EDMG541Wk7

Rawono1

Koenig_and_Schultz_s_Disaster_Medicine_Ch_32-39.pdf

Home >Law homework help >EDMG541Wk7

Floods

Mark E. Keim

Water, water everywhere, and all the boards did shrink; Water, water everywhere, nor any drop to drink.

– Rime of the Ancient Mariner by Samuel Coleridge

OVERVIEW

Definition and Classification

Floods are defined as “the overflow of areas that are not normally submerged with water or stream that has broken its normal confines or has accumulated due to lack of drainage.”1

Engineers studying past floods use statistics to estimate the chance that floods of various sizes will occur. For example, a flood found to occur on the average of 10 times in 100 years would be called the 10% chance flood or the 10-year flood. A flood that only occurs on the average of once every 100 years would have a 1% chance of occurring in any particular year and would be called the 100-year flood or 1% chance flood.2

Floods are classified according to cause (high rainfall, tidal extremes, or structural failure) and nature (e.g., regularity, speed of onset, velocity and depth of water, and spatial and temporal scale). This chapter will discuss impacts according to health out- comes. The influence of flood characteristics on health impacts is discussed where appropriate.

Causes of Floods

Floods may be caused by natural processes that are either fluvial (an abundance of rainfall or melting snow) or coastal (hurricane- related storm surge, coastal inundations, or seismically induced tsunami) in origin.

Human alterations in the environment may also cause flood- ing by alteration of watershed due to deforestation, overgrazing and the failure of dams, embankments and levees,3 channeling

Disclaimer: The material in this chapter reflects solely the views of the author. It does not necessarily reflect the policies or recommendations of the Centers for Disease Control and Prevention or the U.S. Department of Health and Human Services.

of streams, and urbanization of wetlands (which act as a natu- ral flood control by storing water during heavy rains, slowing runoff into streams, and reducing flood peaks). Human alter- ations in the environment affecting global climate change are also predicted to increase the frequency of flooding hazards worldwide.4

Human behaviors can exacerbate flooding severity and impact. Even after prior flooding, human settlement frequently occurs in flood prone areas, thereby increasing a community’s vulnerability to the affects of flooding. Lack of awareness of the dangers posed by fast-moving floodwaters has led to maladap- tive behaviors by people encountering floodwaters. Paradoxically engineering flood controls such as levee and dam construction may contribute to greater human losses and physical damages after a flood disaster (e.g., levee failure).5

Nature of Floods

Fluvial Floods For the purpose of this discussion fluvial (or riverine) flood-

ing will be characterized as either a seasonal flood or a flash flood.

Seasonal floods are typified by a gradual rise to flood stage that may extend across large areas over a long duration. Because seasonal floods are usually caused by a relatively gradual accu- mulation, warning times are generally sufficient to allow safe evacuation of nearby communities. Flash floods are character- ized by a short-duration, high-volume stream flow and usually occur within 6 hours of a rain event, after a dam or levee fails, or after the sudden release of water from an ice or debris jam. Once the flash flood has occurred, it is often accompanied by an extremely short warning and response time with potential for great loss of life.6

Coastal Floods Storm surge, produced by the high winds and vacuum

effect of low-pressure cyclonic storm systems, can produce dramatically high seas that result in coastal flooding. Storm surge–related drownings account for 90% of worldwide deaths

529 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

530 ■ MA R K E. KE I M

related to cyclonic storms.3 Coastal inundations may also occur due to rogue surface waves and cyclonic eddies imparted by weather systems. Seismic events such as earthquakes, landslides, and volcanic eruptions may generate a tsunami pressure wave at sea. Once the tsunami is generated, a series of extremely low frequency, long-wavelength (∼300 km) waves are propagated in an expanding radius from the area of displacement. These waves differ from short-wavelength surface waves (those caused by wind) or storm surges (those caused by cyclones), in that tsunami waves are propagated throughout the entire depth of the ocean. For this reason, tsunamis represent a tremendous amount of potential energy and can travel the speed of a jet air- liner. As the tsunami enters shallow water near coastlines, the enormous kinetic energy previously spread throughout the large volume of deep ocean water becomes concentrated to a much smaller volume of water, resulting in a tremendous destructive potential as it inundates the land.7 This remarkable difference in potential energy imparted over a very large distance by a tsunami compared with other types of floods is unique in char- acter and public health impact. This discussion will therefore focus on floods other than tsunamis, which are the focus of Chapter 36.

Scope of the Problem Flooding is the most common type of disaster worldwide,

accounting for 42% of all disasters during the decade from 1996 to 2005.8 During this same time, floods affected 1.3 billion peo- ple, with more than 90,000 killed. Flooding caused more damage than any other disaster during this time, comprising one-third of all disaster-related costs. Floods have tremendous economic impact worldwide in both high- and low-income nations.8

According to the U.S. National Oceanographic and Atmo- spheric Administration, “In most years, flooding causes more deaths and damage than any other hydro-meteorological phe- nomena. In many years it is common for three-quarters of all federally declared disaster declarations to be due, at least in part, to flooding.”9

Parts or all of more than 20,000 communities in the United States are subject to a substantial risk of flooding. Approximately 7% of the nation’s land area (an area almost as big as the entire state of Texas) is subject to severe flooding.2,10 In the United States, floods cause as much as 90% of the damage from all disasters (excluding droughts).11

Direct economic losses from the 1993 great midwestern U.S. floods surpassed $10 billion.12 Flood damage as a result of the 1998 floods in central Texas was estimated at approximately $900 million, including the costs of damage to 12,000 homes, 700 busi- nesses, and public property.13 In the late summer of 2005, flood- ing brought on by Hurricane Katrina caused more than $200 billion in losses, constituting the costliest disaster in U.S. his- tory.10 Trends toward increasing population density near coasts and in floodplains point to a likely probability of future cata- strophic flood disasters. The current trend of climate change is projected to have an impact on the frequency and severity of floods worldwide.4,14 Munich Re, the world’s largest reinsurance company and a member of the United Nations Environment Pro- gramme’s Finance Initiative, has been compiling annual records on catastrophes and their costs since the 1970s. In 2002, Munich Re reported that rain intensities reached unique values world- wide. The report estimated that during 2002, 42% of worldwide fatalities, 66% of the economic losses and 64% of insured losses were due to floods.14

Table 32.1: Relative Degree of Public Health Impact that May Be Expected After a Flood

Impact Degree

Drowning mortality High-income nations – Few

Low-income nations – Can exceed 100,000/event

Epidemics Can occur in low-income nations

Need for trauma care Rare

Loss of clean water Can be widespread

Loss of shelter Can be widespread

Loss of personal and household goods

Can be widespread

Permanent population migration Rare

Loss of routine hygiene Can be widespread

Loss of sanitation Can be widespread

Disruption of solid waste management

Can be widespread

Public concern for safety High

Increased pests and vectors Can be widespread

Loss and/or damage of healthcare system

Can be widespread

Worsening of existing chronic illnesses

Can be widespread

Toxic exposures Possible

Food insecurity Can occur in low-income nations and remote islands

THE PUBLIC HEALTH IMPACTS OF FLOOD DISASTERS

Public health impacts of flooding include damage to homes and consequent displacement of occupants, infectious disease exac- erbated by crowded living conditions and compromised personal hygiene, contamination of water sources, disruption of sewage service and solid waste collection, increased vector populations, injuries sustained during clean-up, stress-related mental health and substance abuse problems, and death.15

Loss of Safe Water and Adequate Sanitation

During the 1993 great midwestern U.S. flood, 9% of the popu- lation in the state of Iowa suffered a complete loss of the public water system. Twenty-nine counties in Iowa (representing 37% of the population) reported flood damage to water systems, and 31 counties (with 35% of the population) reported flood damage to sewer systems.16

Food Insecurity: Crop Losses and Disruption of Food Distribution

Generalized food shortages severe enough to cause nutritional problems usually do not occur after disasters but may arise among low-income nations in two ways. Food stock destruc- tion within the disaster area may reduce the absolute amount of

Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

FLO O D S ■ 531

food available, or disruption of distribution systems may curtail access to food, even if there is no absolute shortage. Flooding and sea surges can damage household food stocks and crops, disrupt distribution, and cause major local shortages. Food distribution, at least in the short term, is often a major and urgent need, but large-scale importation/donation of food is not usually nec- essary.17 One notable exception occurs when remote low-lying islands are flooded by seawater to an extent that island aquifer becomes brackish and agricultural land becomes salt contami- nated so that it can no longer support gardens for the next several years. Such an inundation event occurred in the U.S.-associated Pacific island nation of the Federated States of Micronesia dur- ing March 2007 and resulted in a loss of food security sufficient enough to warrant a declaration of disaster by the U.S. Federal Emergency Management Agency.18

Loss of Shelter and Population Displacement

Floods displace literally millions of people. Displacement is a key risk factor for morbidity and mortality among disaster-affected populations. During and after displacement, vulnerable popula- tions undergo additional risks for morbidity and mortality.

Misdirected or misguided settlement solutions may increase and prolong the risk of morbidity and mortality among dis- placed populations by providing shelter below internationally accepted standards for space, nutrition, food, clean water, safety and security, sanitation, hygiene, and access to medical care.19

Individual household shelter solutions can be short or long term, subject to the level of assistance provided, landuse rights or ownership, the availability of essential services and social infra- structure, and the opportunities for upgrading and expanding the dwellings. Existing shelter and settlement solutions should be prioritized. Where it is sustainable, members of affected house- holds should be allowed to return to the site of the original dwellings. Affected household members who cannot return to the site of the original dwellings should be able to settle indepen- dently within a host community or with host families, whenever possible. Research has indicated that providing increased social support can significantly lower illness burdens after disasters.19

Household members should be accommodated in temporary camps or mass shelters only as a last resort.

TOXIC CHEMICAL EXPOSURES

The mobilization of chemicals either from storage (e.g., underground fuel tanks, pipelines, hazardous landfill sites, and wastewater lagoons) or by remobilization of chemicals already in the environment (e.g., pesticides, dioxin in river/canal sediment, runoff from roads and bridges, overloaded sewers, and acid mine drainage) has occurred during floods.20 These chemical haz- ards are more likely to mobilize when industrial and agricultural areas are submerged underwater.3 One 2004 review identified epidemiological evidence for flood-related adverse health affects following chemical exposures to carbon monoxide, pesticides, agricultural chemicals, dioxin, volatile organic carbons, heavy metals, cyanide, acid waste water, sulfides, and cadmium.20

According to records kept by the local fire department, 1,200 homes in Grand Forks, North Dakota affected by the 1997 Red River flood reported problems with fuel oil spills ranging from 190 to 985 liters. Experts from the U.S. Environmental Protec- tion Agency conducted a study of 34 homes approximately 1 year after the flood occurred. Six homes (17.6%) still had measur- able hydrocarbon vapors that were considered a serious health

problem. The homeowners were advised to move or undergo major structural work to replace contaminated structures.21

TOXIC MOLD EXPOSURES

In 2004, the U.S. Institute of Medicine reviewed the litera- ture regarding health outcomes related to damp indoor spaces. The findings of this report indicate that indoor environmental conditions and personal practices may provide mold exposures that potentially expose residents and remediation workers to the risk of negative health effects.22

Investigators identified mold as a potential public health problem arising from the 1993 floods in the midwestern U.S.3

Visible mold growth was found in 46% of homes inspected after flooding caused by Hurricane Katrina.22 Predominant fungi indoors and outdoors were Aspergillus and Penicillium species. Although interpreting the significance of measures of airborne mold toxins is complex, indoor air levels were markedly elevated and usual indoor/outdoor ratios for mold were reversed, that is, indoor levels of mold toxin were higher than outdoor.22,23

Among the residents interviewed, two thirds quickly iden- tified particulate respirators as appropriate and necessary res- piratory protection for cleaning of mold. Of those who had cleaned up mold, two thirds did not always use appropriate respirators. Among persons who self-identified as remediation workers, 95% thought mold causes illness and 85% correctly identified particulate filter respirators as the appropriate protec- tion for cleaning up mold; however, 49% of remediation workers had not been fit tested for respirators and 35% of the same group reported that they did not always use respirators.22 These findings suggest that a significant proportion of disaster-affected residents and remediation workers may be exposed to poten- tially hazardous levels of mold contamination by virtue of a lack of understanding or lack of access to personal protection or a lack of compliance when those measures were recognized and made available.22

DISRUPTION OF HEALTHCARE SER V ICES

Flooding may directly damage healthcare facilities or it may hinder public access to these facilities by closing transportation routes. During the great midwestern U.S. floods in 1993, five of the 99 counties representing 14% of Iowa’s population reported closures of primary care physician offices.16

Floods have a substantial impact on the operation of most emergency medical services systems. The primary effect often results from disruption of usual transport routes due to water. Data concerning air transport in flooding collected during Hurricane Floyd demonstrated a nearly 650% increase in heli- copter utilization for emergency medical services transports in the affected areas of the U.S.4 In most flood-related disasters in the U.S. approximately 0.02%–2% of flood survivors require emergency medical attention.24

The Midwest floods of 1993 presented multiple challenges to the six metropolitan medical centers in Des Moines, Iowa when these hospitals lost all public utilities. Healthcare leaders can- celled elective admissions and diverted nonemergency clinical services to alternate facilities. They identified and implemented ancillary resources to maintain essential operations. Modifica- tions were made for alternative methods of infection control, ster- ilization, housekeeping, and food preparation. Planners imple- mented extraordinary measures to maintain adequate amounts of water for laundry, fire protection, cooling, instrument steril- ization, renal dialysis, physical therapy, and dietary services.25

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

532 ■ MA R K E. KE I M

DISRUPTION OF PUBLIC SER V ICES

During the great midwestern U.S. floods in 1993, eight coun- ties in Iowa (24% of the state population) reported interrup- tion in public health services (e.g., supplemental food programs and various clinics such as those for vaccinations and treat- ment of sexually transmitted diseases). Ten counties (15% of the population) reported at least one nonoperational public sewer system.16

Power outages are a common impact of flood disasters. Power outages not related to flooding have been associated with out- breaks of diarrheal illness.26 The disruption of public access to refrigeration may have the potential to impact food safety as well as drug safety. Life sustaining medications such as insulin also require properly controlled refrigeration to remain efficacious.

FLOOD-RELATED MORBIDITY AND MORTALITY

Flood-related Mortality

Floods continue to be the number one nonterrorist-related dis- aster in the United States in terms of lives lost and property damage.27 Over a 25-year period prior to Hurricane Katrina, floods killed approximately 140 Americans and cost $6 billion in property damage each year. In the United States, the most common cause of flood-related deaths is drowning.10

“The number of deaths associated with flooding is closely related to the life-threatening characteristics of the flood (rapidly rising water, deep flood waters, objects carried by the rapidly flowing water) and by the behavior of the victims.”28 The most readily identifiable flood deaths are those that occur acutely from drowning or trauma, such as can occur after being hit by objects in fast flowing waters. The number of such deaths is determined by the characteristics of a flood, including its speed of onset, depth, and the extent of flooded area. Information on risk factors for flood-related death remains limited, but men appear more at risk than women. In high-income countries most deaths are due to drowning and, particularly in the U.S., are vehicle related.4,29 The most likely group to drown in their own homes are the elderly.29

Flash floods are the number one cause of flooding deaths.3

Flash flooding is the leading cause of weather-related mortality in U.S.16 In general, high mortality rates are frequently observed in flash flood incidents, examples of which occurred in Puerto Rico in 1992,30 Missouri in 1993, and Georgia in 1994, and Texas in 2001 when heavy water runoff inundated communities with great immediacy and intensity.5,15,30 The majority of flood- related drownings occur when a vehicle is driven into hazardous floodwaters.4,13

The power of water, especially moving water, is astounding. For example, “Two feet of water will carry away most automo- biles. The lateral force of a foot of water moving at 10 mph is about 500 pounds on the average car. And every foot of water displaces about 1,500 pounds of car weight. So two feet of water moving at 10 mph will float virtually every car.’’31

During the 1998 floods in central Texas, 24 of the 29 deaths directly related to the storm were caused by drowning. Of those 24 drownings, 22 (92%) with known circumstances occurred because the vehicle was driven into high water. These deaths occurred in 16 separate incidents, some with multiple fatalities. Of the 16 water-crossing incidents, 11 (69%) occurred at loca- tions known to reporting authorities to have a history of flooding; 10 (63%) involved trucks and/or sport-utility vehicles.13

Prior to the implementation of early warning, evacuation, and shelter systems, drowning from hurricane storm surge accounted for an estimated 90% of cyclone-attributable mor- tality in both high-income and low-income nations.3 Approxi- mately 8,000 people died as a result of flooding in 1900 after a large hurricane hit Galveston, Texas. In 1928, 1,836 people died as a result of another hurricane storm surge around Lake Okee- chobee in Florida. Most of these deaths were believed to be caused by the large storm surge associated with powerful hurricanes.

Storm surge drowning deaths have decreased markedly in high-income nations due to improvements in population pro- tection measures.32 One notable exception is that more than 1,300 deaths were attributed to Hurricane Katrina, most of which occurred as result of flash flooding caused by catastrophic levee failures, making 2005 the third deadliest year in U.S. history for flood deaths to that date.33

Flood-related Morbidity

POVER TY AND FLOOD-RELATED MORBIDITY

Poverty is a key risk factor for human vulnerability to flood disasters. The correlation between poverty and morbidity is seen clearly during flood disasters. Low-income populations within a given society often dwell in locations at higher risk for flooding and have fewer resources available for response and recovery. Rarely, if at all, are any resources available to prepare for or mitigate flood disasters. High-income populations within a soci- ety possess a much higher level of resilience and availability of resources and can better afford more cost-effective risk reduction measures. Diseases normally endemic to a given population are typically exacerbated as a result of flood disasters. As a result, low-income nations have a higher incidence of flood-related outbreaks of infectious diseases, such as leptospirosis, typhoid, malaria, and cholera. High-income nations tend to suffer few flood-related outbreaks, but instead have a higher proportion of flood-related noncommunicable diseases, such as injuries, mental illness, cardiovascular disease, and chronic obstructive pulmonary disease.

Causes of flood-related morbidity reported in high-income nations during the first 6 weeks after the disaster are frequently equally divided between injuries and illnesses. Injuries com- monly include sprains/strains, lacerations, and abrasions. Many of these injuries occur during the clean-up phase rather than immediately during the flooding. The causes of flood-related ill- nesses in high-income nations are often equally divided between communicable and noncommunicable diseases.16,32

COMMUNICABLE DISEASES AND

FLOOD-RELATED MORBIDITY

The relationship between disasters and communicable diseases is frequently misconstrued. The risk for epidemics is often presumed to be very high in the disaster aftermath. The risk for outbreaks after disasters is often greatly exaggerated by both health officials and the media.17 The risk of infectious diseases after flood-related disasters is often specific to the event itself and is dependent on a number of factors. The risk factors for outbreak after disasters are primarily associated with displace- ment of highly vulnerable populations. Specific risks to these populations are determined by proximity of safe drinking water and functioning latrines, nutritional status of the displaced pop- ulation, level of immunity to vaccine-preventable diseases, and access to healthcare services.17,26,34 Historically, the large-scale,

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

FLO O D S ■ 533

long-term displacement of populations as a result of flood disasters is uncommon. This likely contributes to the low overall risk for outbreaks.34 In high-income countries with adequate public health infrastructure, postimpact surveillance has only occasionally detected increases in life-threatening infectious diseases after disasters and these increases have been relatively small.3,4,34 In comparison, in low-income nations, disaster recovery workers have reported larger outbreaks of infectious diseases including cholera and typhoid, acute respiratory infections, and leptospirosis.23 Despite frequent public concern to the contrary, nonendemic diseases do not spontaneously emerge after flood disasters. Floods exacerbate diseases that are endemic to the affected populations.

DIARRHEAL ILLNESS

In flood conditions there is potential for increased fecal–oral transmission of disease, especially in areas where the population does not have access to clean water and sanitation. Outbreaks of diarrhea occur in both high-income and low-income nations and are associated with locally endemic pathogens. The risk of diarrheal illness appears to be less for high-income countries compared with low-income countries. Diarrheal illness in high- incomes countries most commonly manifests as a self-limiting gastroenteritis without a specifically identified pathogen. Flood- related outbreaks of life-threatening diarrheal diseases (such as paratyphoid and cholera) have been reported in very low-income nations.

Flood-related Diarrheal Illness in High-income Nations In 1983, an outbreak of diarrheal illness was reported in

Utah possibly associated with contaminated water supply that resulted from flooding during the spring snow melt. Five routine bacteriological samples from the source revealed coliform counts to be elevated above acceptable limits.35 A similar period of heavy water runoff associated with unseasonably warm weather and ash fall from the Mount St. Helens volcano eruption in 1980 was also linked to an outbreak of diarrhea due to Giardia lamblia.35

After flooding caused by tropical storm Allison in Hous- ton, Texas during June 2001, 54 (12.9%) surveyed households reported at least one person with illness that occurred after the onset of flooding. Persons living in flooded homes were sig- nificantly more likely than those living in nonflooded homes to report illness; the only specific illnesses significantly associ- ated with residing in a flooded home were diarrhea/stomach conditions.15 In 2002, in a village near Barcelona, Spain an out- break of shigellosis affected more than 10% of the population. The outbreak was linked to consumption of drinking water that may have been contaminated after heavy rain caused floods.36

After Hurricane Katrina in 2006, clusters of diarrheal disease were reported in evacuation centers in four states; gastroenteritis was the most common acute disease complaint among evacuees in Memphis, Tennessee.23 Approximately 6,500 of an estimated 24,000 evacuees in Houston shelters visited Reliant Park medical clinic, and 1,169 (18%) persons reported symptoms of acute gas- troenteritis. In stool samples from 44 patients tested, norovirus was confirmed in 22 (50%) specimens; no other enteropathogen was identified.37

Two cases of toxigenic Vibrio cholerae O1 infection were reported in Louisiana after Hurricanes Katrina and Rita.38–40

There was, however, no epidemic and no evidence to suggest that there was an increased risk of cholera among residents of the gulf coast after these hurricanes.23

Flood-related Diarrheal Illness in Low-income Nations Surveillance data showed an apparent increase of mortal-

ity as a result of diarrhea during the 1988 floods in Khartoum, Sudan,41 but a similar rise was also apparent in the same period of the preceding year.42 Routine surveillance data and hospi- tal admission records showed diarrhea to be the most frequent cause of death following severe flooding in 1988 in Bangladesh, but again, the effect of the flood was not separately quantified from seasonal influences.29 Two devastating monsoon-related floods in Bangladesh in 2004 resulted in very large outbreaks of diarrheal disease reaching epidemic proportions throughout the capital city of Dhaka. Healthcare workers evaluated more than 17,000 patients in one hospital during one of these flood periods. Cholera was the most common cause of admission, and entero- toxigenic Escherichia coli was also an important cause of acute watery diarrhea, particularly in children younger than 2 years of age.23 In a large study conducted in Indonesia in 1992 and 1993, flooding was identified as a significant risk factor for diarrheal illnesses caused by paratyphoid fever.26

RESPIRATOR Y INFECTIONS

Whenever there is a lack of clean water among displaced populations, there is often a concomitant difficulty in main- taining adequate hygiene. This lack of hygiene can lead not only to diarrheal illness, but also to acute respiratory infections. Reported incidence of acute respiratory infections increased fourfold in Nicaragua in the 30 days after Hurricane Mitch in 1998.26

Flooding may also result in episodes of near-drowning and pulmonary aspiration of floodwater. Direct inoculation of the pulmonary system with marine and soil debris may cause seri- ous acute respiratory infections and systemic infections. Flood- related aspiration pneumonia is often polymicrobial.23

VECTOR-BORNE DISEASES

The relation between flooding and vector-borne disease is complex. The predicted impact of severe weather or floods on vector-borne illnesses is less certain than that of enteric infec- tions. Severe weather can either increase or decrease the transmis- sion of vector-borne illness.23,29 Such variance probably mirrors the complexity of a given situation, and partly reflects the preva- lence of vector-borne diseases in the region before the disaster, the identity and ecology of the local vectors (some vectors pre- fer clean water, others prefer organically rich water, some prefer fresh water, and others prefer water containing low amounts of salt), and the impact of control programs or other interventions that minimize vector–human contact (for example, the use of larvicidal or insecticidal agents, access of survivors to nets or window screens, and access of survivors to shelters versus sleep- ing outdoors).23

AR THROPOD-BORNE DISEASES

Flood-related Arthropod-borne Disease in High-income Nations

Floods are often followed by a proliferation of mosquitoes. In the U.S. such disasters are rarely followed by outbreaks of arboviral disease. This is attributable mostly to the relatively low prevalence of vector-borne diseases in the region before the disaster.23,43 Heavy rains and flooding have been associ- ated with outbreaks of St. Louis encephalitis in Florida, believed to be associated with the feeding activities of the responsible mosquito vector.23 In comparison, despite the proliferation of

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

534 ■ MA R K E. KE I M

large populations of mosquitoes known to amplify transmis- sion of arboviruses that cause St. Louis encephalitis and western equine encephalitis after the great midwestern flood of 1993, surveillance data indicated minimal risk for arboviral disease above background levels in the disaster area. During the 1993 floods, 45 counties (53% of the population) in Iowa reported vec- tor problems. Mosquito vectors were found to exist in extremely high levels. Serum conversions were not detected in sentinel chicken flocks nor were there any reported human cases of illness as result of mosquito vectors.16 As a result, contingency plans for large-scale mosquito adulticiding were not implemented, result- ing in an estimated cost savings of more than $10 million. Despite the presumed low risk for mosquito-borne arboviral disease after flood-related disasters in high-income countries, surveillance programs are useful to assist with determining prevalence in large vector populations and prevent unnecessary expenditures associated with the application of insecticides implemented dur- ing prophylactic mosquito control.43

Flood-related Arthropod-borne Disease in Low-income Nations

Malaria outbreaks in the wake of flooding have been reported in several very low-income nations with warm cli- mates.26 Increased numbers of cases of drug-resistant malaria were noted after floods in Sudan44 and an outbreak of more than 75,000 cases of Plasmodium falciparum occurred in Haiti after Hurricane Flora in 1963.26 A four–fivefold increase in malaria incidence also occurred after a flood disaster in 2000 in Mozam- bique.26 After the 2004 Indian Ocean tsunami, no apprecia- ble increase in the number of malaria cases was reported in Indonesia.45

Dengue transmission is influenced by meteorological condi- tions; however, transmission has not been directly attributable to flooding. In Brazil, Indonesia, and Venezuela, rain, temperature, and relative humidity have been associated with patterns of dengue infection.23 Monsoon rains and floods have been asso- ciated with outbreaks of dengue fever in India. In Thailand, dengue was a common cause of fever in children after heavy rain–associated flooding.23

Rodent-borne Diseases Diseases transmitted by rodents may also increase during

heavy rainfall and flooding because of altered contact patterns. Humans usually acquire leptospirosis after exposure to fresh water contaminated with the urine of infected animals, such as rats.

There have been reports of flood-associated outbreaks of lep- tospirosis from a wide range of countries including Argentina, Brazil, Cuba, India, Korea, Mexico, Nicaragua, Philippines, Portugal, Russia, Taiwan and the U.S. (Hawaii and Puerto Rico).23,26,29,46

DERMATOLOGICAL CONDITIONS

Exposure of intact skin to floodwaters does not pose a serious health risk.35

The risk of possible estuary-associated syndrome due to Pfiesteria piscicida is actually thought to decrease during peri- ods of coastal flooding.4 Dermatological conditions, usually in the form of nonspecific rashes, are a commonly reported com- plaint during flood and other disasters.3,17 Among hurricane evacuees from the New Orleans area, a cluster of infections with methicillin-resistant Staphylococcus aureus was reported in

approximately 30 pediatric and adult patients at an evacuee facil- ity in Dallas, Texas. Three of the methicillin-resistant S. aureus infections were confirmed by culture.40

WOUND INFECTIONS

Wound infections are common after disasters. The destruc- tion of the regional health infrastructure, the inability to wash wounds with clean water, and the inability to treat individuals with topical or systemic antimicrobial agents can all lead to severe wound infections, even if the initial trauma was relatively minor. In 2005, after Hurricane Katrina, 24 cases of Vibrio sp. bacterial wound infections were noted among individuals in the affected area. Most patients had associated comorbidities that probably increased their risk of Vibrio wound infection, and many had been wading in floodwaters.23,40

There is no evidence indicating that the risk of tetanus is increased in flood-related lacerations; therefore standard immu- nization practices should be used.4 In comparison, the incidence of tetanus in the disaster setting has been associated with wounds that are highly contaminated with soil as a result of high-energy traumatic inoculation occurring in low-income nations where patients are less likely to have had primary vaccinations and access to postexposure tetanus prophylaxis. Clusters of tetanus cases were reported after the Indian Ocean tsunami as well as the Pakistan earthquake in 2005. These exposures were sustained during the initial trauma on the day of the event and not associ- ated with exposure to floodwater or contaminated water sources. No subsequent cases were reported 2 weeks after the tsunami, indicating an association with the initial traumatic injury on the day of the event and not from wound exposure to flood waters afterward.45

INJURIES

During a flood disaster, injuries are more likely to occur as residents attempt to travel or evacuate through flooded areas and in the aftermath of a flood disaster when residents and workers return to dwellings to clean up damage and debris. Hypothermia with or without submersion injury is seen in some flood casu- alties. Conductive heat loss due to immersion in any water less than 16◦C –21◦C may lead to hypothermia. Wet clothing and water immersion tend to conduct away core body heat even when the ambient air temperature is considerably warm, if it is less than the body temperature. Convective heat losses increase in windy conditions. Electrocutions have occurred as a result of downed power lines, electrical wiring, and improper handling of wet appliances. Injuries from fires and explosions from gas leaks also occur. Musculoskeletal and soft tissue injuries are frequent conditions associated with floods. Lacerations and punctures are common sequelae during postflood clean up and recovery activities.3,4

Although it is a recurrent public concern that animals such as snakes may be forced to seek refuge from rising floodwaters in areas that may be inhabited by humans, public health surveil- lance after floods has not indicated that wild animal bites have been a major problem.3,4 In contrast to wild animals, after Hur- ricane Floyd there were increases in reports of domesticated dog bites.32

In the U.S. state of Missouri after the Midwest floods of 1993, injuries were reported through the routine surveil- lance system. During a 6-week period, a total of 524 flood- related conditions were reported, and of these 250 (48%) were injuries. These injuries were categorized as sprains/strains (34%),

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

FLO O D S ■ 535

lacerations (24%), abrasions (11%), and other injuries (11%).6

Epidemiologists reported similar data from Iowa during the same year.16

MENTAL HEALTH EFFECTS

The long-term effects of flooding on psychological health may be even more significant than for any other illness or injury.47–49 For many people, the emotional trauma continues long after the water has receded.

Factors that appear to make people more vulnerable to the development of psychological problems include the following

■ Objective and subjective characteristics of the disaster – prox- imity of the victim to the disaster site; the duration of the disaster; the degree of physical injury; the witnessing of death or injury47

■ Characteristics of the postdisaster response and recovery environment – community cohesion; secondary victimiza- tion; disruption of social support systems;47 lack of home- ownership and duration of displacement from home19

■ Characteristics of the individual or group – history of psycho- logical problems; elderly; unemployed; single parent; chil- dren separated from their families.47

In one study of the health effects of flooding in 30 locations of the United Kingdom, psychological effects were much more commonly reported after flooding than physical ones. Women also appeared to suffer markedly more frequently than men at the worst time of flooding.19 A number of psychological effects reported by flood victims have been quite strongly associated with reporting physical effects, particularly immediate effects.19

Increased morbidity and mortality following a flood may also result from heightened psychological stress.48 Bennett studied 316 flood respondents and 450 nonflood controls for morbidity and mortality for the year after the Bristol floods in England in 1968. He found higher mortality among residents of the flooded sections, especially the elderly.49 Psychiatric examinations of 224 children 2 years after the 1972 flood in Buffalo Creek, West Virginia showed that 80% of the children were severely emotion- ally impaired by their experiences during the flood.49 Five years after the flood caused by tropical storm Agnes in Pennsylvania, perceived health problems were reported more commonly by flood-affected respondents than nonaffected controls.49

Anxiety and Depression There is considerable evidence for the impacts of anxiety

and depression among flood-affected populations. Most studies are from high- or middle-income countries, including Australia, Poland, the United Kingdom, and the United States, but there is also a study from Bangladesh.29 A few studies have examined flood-related mental health impacts on children. One 1993 study found postflood changes in behavior among children 2–9 years old.29

Posttraumatic Stress Disorder Studies in Europe and North America have revealed post-

flood psychiatric disorders among flood-affected populations that meet the criteria for posttraumatic stress disorder.29 One longitudinal study found 15%–20% of people affected by a disas- ter had symptoms of posttraumatic stress disorder.48 Other stud- ies after the 1997 floods in Poland also suggest long-term negative effects on the well-being of children aged 11–14 and 11–20 years

Table 32.2: Top 10 Medical Conditions Based on Limited Needs Assessments among Persons in Hurricane Katrina Evacuation Centers between September 10 and 12, 2005

Incidence per 1,000 Condition Residents

Hypertension/cardiovascular 108.2

Diabetes 65.3

New psychiatric condition 59.0

Pre-event psychiatric condition 50.0

Rash 27.6

Asthma/Chronic obstructive pulmonary disease

27.5

Flulike illness or pneumonia 26.3

Toxic exposure 16.0

Other infection 15.6

Diarrhea 12.8

Data from Centers for Disease Control and Prevention. Available at: www.cdc.gov/od/katrina/09–19-05.htm.

with increases in posttraumatic stress disorder, depression, and dissatisfaction with life. Six months after Hurricane Floyd similar findings were reported for disaster-affected children aged 9–12 compared with controls.29

Suicides Evidence is limited regarding flood-related suicides. One

study in a high-income nation indicated that suicide rates increased by 13.8% above the predisaster rates.4

EXACERBATION OF CHRONIC ILLNESS

Conditions related to exacerbations of chronic disease may comprise a majority of patient complaints among flood-affected populations in high-income nations, especially in shelter settings. Table 32.2 provides an example of typical patient complaints after flood displacement in a high-income nation.

The development of hypertension by male flood victims was reportedly greater than by male non-flood victims in the 5 years after flooding caused by Hurricane Agnes in Pennsylva- nia.49 Benin also reported a marked increase in hypertension in Russia after the 1964 flood and in Moldavia after two successive floods in 1969.49

CARBON MONOXIDE POISONING

Carbon monoxide poisonings occur, usually in association with loss of electrical power, when flood-affected populations improperly use carbon monoxide–emitting fuel sources within poorly ventilated, enclosed spaces. This happens when people place generators indoors, in garages, or outdoors but near win- dows. Other carbon monoxide–emitting sources include indoor burning of charcoal for cooking and heating and using leaf blow- ers indoors for flood clean up. After the 2004 hurricanes in Florida, 157 persons were treated from 51 exposure incidents, with six reported deaths. Of the 167 cases of carbon monoxide poisonings associated with Hurricane Katrina in 2005, 48.5% were treated and released without undergoing hyperbaric oxygen

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

536 ■ MA R K E. KE I M

Table 32.3: Four Main Strategies and Associated Tools for Floodplain Management2

Strategy Tools

Modifying vulnerability Regulations

Development and redevelopment policies

Disaster preparedness

Flood forecasting, warning, and evacuation plans

Flood proofing and elevation

Modifying hazard Dams and reservoirs

Dikes, levees, and flood walls

Channel alterations

High-flow diversions

Storm water management

Shoreline protection

Land treatment measures

Modifying impact Information and education

Flood insurance

Tax adjustments

Flood emergency measures

Disaster assistance

Postflood recovery

Restoring and preserving the natural and cultural resources of floodplains

Regulations

Development and redevelopment policies

Information and education

Tax adjustments

therapy, 43.7% were released after undergoing hyperbaric oxygen therapy, and 7.8% were hospitalized (most for just 1 day). Among these patients, 80% complained of headache, 51.5% nausea, 51% dizziness, 31.5% vomiting, and 16.4% dyspnea, and 14.5% expe- rienced loss of consciousness. The mean carboxyhemoglobin level was 19.8%, with a range of 0.2%–45.1%.11

CURRENT STATE OF THE ART

Sustainable Development and Disaster Risk Management

A widely used international definition of sustainable develop- ment is “development which meets the needs of the present without compromising the ability of future generations to meet their own needs.”50 The concept of sustainable development was first codified decades ago during the Declaration of the United Nations Conference on the Human Environment at Stockholm in June 1972. Principle 1 of this declaration stated, “Man has the fundamental right to freedom, equality and adequate condi- tions of life, in an environment of a quality that permits a life of dignity and well-being, and he bears a solemn responsibility to protect and improve the environment for present and future generations.”50

Over time the overall approach to emergencies and disas- ters among nations has shifted from ad hoc postimpact activ- ities to a more systematic and comprehensive process of risk management that emphasizes the importance of preimpact risk reduction activities including prevention, mitigation, and pre- paredness.

Building on these core principles of both sustainable devel- opment and disaster risk management, the 2002 World Sum- mit on Sustainable Development plan of implementation stated, “An integrated, multi-hazard, inclusive approach to address vul- nerability, risk assessment and disaster management, including prevention, mitigation, preparedness, response and recovery, is an essential element of a safer world in the twenty-first cen- tury.”51 Sustainable disaster risk management is a comprehen- sive approach to reduce disaster impact on a society over time without transference of additional risk and associated costs to future generations.

Toward Sustainable Flood Risk Management

The risk from disaster occurs when vulnerable populations are exposed to a hazard such as flooding and there are insufficient resources to match the immediate needs. This can lead to mor- bidity and mortality.

Risk assessment is used to quantify environmental health risk. For flooding events, the risk equation has been applied to estimations of the likelihood of disaster impact as follows:

p(flood hazard) × p(vulnerability or “susceptibility”)

− p(absorbing capacity or “resilience”) = p(disaster impact).

The probability of the hazard occurring is based on models that apply historical data from past floods in the same specific location being analyzed for risk.

The term "floodplain management" embodies the effort to manage the waters and lands subject to flooding to

■ Reduce all losses due to flooding ■ Protect and enhance the natural values (inherent social, eco-

nomic, ecological, and cultural) of floodplains

It includes actions by all levels of government and the private sector ranging from constructing massive dams to the zoning decisions of small communities. It involves concerns with wet- lands, water quality, location of new developments, and a host of other considerations. Planning and carrying out a compre- hensive floodplain management program usually requires the cooperative participation of all levels of government and the pri- vate sector. The four main strategies of floodplain management relate to modifying the likelihood of the hazard ever occurring, the vulnerability (or susceptibility) to the hazard, and increas- ing the resilience or capacity of the society to absorb the impact of the hazard without mismatch of needs and resources. See Table 32.3.

The newly emerging paradigm of sustainable flood manage- ment is derived from the concept of sustainable development. A “seismic shift” has been described as taking place in man- aging flood risk. Many countries’ well-established reliance on structural defenses is being questioned and cheaper and more sustainable alternatives are being sought.52

In the past, due to emphasis on structural defense, allevi- ation was prioritized above complimentary strategies such as

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

FLO O D S ■ 537

promoting avoidance, raising awareness, and providing assis- tance. For many decades this paradigm of flood defense has been successful in protecting urban dwellers from riverine inundation and enabling farmers to cultivate or raise livestock right up to the edge of the river or seashore. This strategy, initially ques- tioned because of the threat posed by climate change, is being increasingly abandoned.52 More cost-effective strategies for dis- aster risk reduction through flood prevention, preparedness, and mitigation are now gaining wider application.

The Scotland Water Environment and Water Services Act of 2003 states that public officials have a duty “to promote sus- tainable flood management and act in a way best calculated to contribute to the achievement of sustainable development.” In practice this means that officials must seek to balance social, economic and environmental needs within a framework that incorporates intergenerational equity.52 Intergenerational equity is a concept based on the fundamental principle of sustainable development that calls for a meeting of present population needs without compromising the ability of future generations to meet their own needs. Present development should not add a burden of disaster risk for future generations to pay through expensive and inefficient response and recovery measures. It is the responsibil- ity of each generation to avoid public policies and developmental practices that increase the future risk of disaster.

Sustainable flood risk management also seeks to reduce risk at all stages of the disaster cycle giving preference to more cost- effective disaster risk reduction activities such as prevention, preparedness, and mitigation.

Lessening Public Health Vulnerability to Flood Disasters

Until recently, vulnerability has mostly been measured in terms of flood damage to infrastructure and commerce. The current state of the art is seeking to further define and measure human vulnerability in terms of indicators for health, socioeconomic status, and quality of life. By focusing on vulnerability and the ability of individuals and communities to recover (resilience), sustainable flood management places individuals at risk in the center stage and places the responsibilities for enhancing social equity and promoting community cohesiveness on appropriate authorities. This is coupled with a heightened sense of indi- vidual responsibility. Fundamental to sustainable flood man- agement is a change in attitude in which a willingness to take on greater personal responsibility for mitigating flood losses steadily replaces undue reliance on state intervention when losses occur.52

Flood risk assessments commonly focus on protecting prop- erty and are the basis for financial decision-making. Environ- ments are often protected irrespective of the cost. The health and social impacts suffered by people affected by floods remain mostly ill considered. In this respect, a better understanding of the social effects of individuals involved in floods is needed. Disruption of people and communities cannot be measured in monetary costs. Floods can cause health impacts that are endur- ing, including the stress and trauma created months or years afterward. The psychological effects of flooding can continue for months or even years after the event and are often more pro- nounced than the physical health effects.53 Tapsell et al., have pro- posed a Social Flood Vulnerability Index (SFVI), which measures the impact that floods could have on the communities potentially affected. The SFVI is a composite-added index derived from a review of the existing literature and hundreds of interviews of

disaster victims. It is based on three social characteristics and four financial deprivation indicators. It recognizes that age and financial status of the affected populations are the most impor- tant variables, followed by the prior population health status. The seven indicators used for derivation of the SFVI are

■ Unemployment ■ Overcrowding ■ Nonownership of a car ■ Nonownership of a home ■ Long-term illness ■ Single parent family ■ Elderly

Strengthening Public Health Resilience to Flood Disasters

A 2002 Institute of Medicine report claimed that U.S. govern- mental public health agencies have long suffered “grave under- funding and political neglect.” After Hurricane Katrina, a 2005 editorial also criticized the less than adequate support for a resilient public health system that protects human life when disasters occur.54 Populations that possess a high capacity for access to public health and medical services are more resilient and less vulnerable to flood-related morbidity or mortality. There is growing concern regarding the longer-term impacts of climate change on human health, including flooding.4,14,19 The sustain- able development of cost-effective public health and medical sys- tems serves to strengthen population resilience to flood disasters. Methods that seek to reduce population vulnerability also serve to increase resilience of the public health against flood-related morbidity and mortality.

Preventing the Public Health Impact of Flood Disasters

Since 1926, flood damage has increased in both the U.K. and the U.S. despite local efforts and federal encouragement to mit- igate flood hazards and to regulate development in flood prone areas.52–55 Although flood-related mortality during the past half-century has declined among high-income nations (mostly because of improved warning systems), economic losses have continued to rise due to increased urbanization and coastal devel- opment.56 Many people lack the ability to prevent flood hazards from occurring, leaving the public health and medical sectors to play an important role in mitigating and preventing the public health impact of disasters.

The majority of disaster-related mortality is directly attri- butable to drowning. Therefore, population protection measures such as evacuation are aimed at preventing drowning by warning populations and moving them away from the flood hazard. The incidence of flash floods in the United States is increasing; how- ever, the mortality from these floods is decreasing. This decrease in mortality parallels advances in the U.S. National Weather Sys- tem advanced warning system. Warning has been identified to be a substantial factor in decreasing mortality from flash floods by more than 50%.4

Health communication is a valuable tool in educating the public before and after flood impact regarding protective behav- iors that help to prevent drowning (i.e., cautions regarding driv- ing vehicles in flooded areas). Other injuries and illness may also be prevented through public awareness and education that promotes safe and healthy activities during flood response and

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

538 ■ MA R K E. KE I M

recovery efforts. Injuries such as electrocutions, burns, and car- bon monoxide poisonings are typical examples of flood-related morbidity that are preventable through public awareness and health education campaigns. Chronic psychological and medical illnesses can also be prevented through activities that adequately manage stress in the disaster-affected population.

The appropriate use of personal protective equipment (Chap- ter 13) among disaster-affected populations and recovery work- ers can help to prevent secondary disaster-related morbidity due to toxic exposures from chemicals or mold. Measures to increase awareness of the appropriate respiratory protection among the public are warranted to lessen potential exposures to mold hazards. This information can be provided via public ser- vice announcements on radio and television and by educational sessions for employees of home improvement stores and other commercial entities that sell respirators. To decrease the prob- ability of flood-related illness, flood victims and relief workers should practice proper hand hygiene (wash their hands with soap and water) before preparing or eating food, after toilet use, and after participating in flood clean up or after handling potentially contaminated articles.57

Chemical risk assessments can be used to identify and charac- terize industrial and agricultural hazardous material sites located in flood prone locations to prevent toxic exposures and associ- ated adverse health effects.

Public policy may guide landuse and zoning regulations that prevent population displacement. These efforts subsequently lessen the potential future need for expensive shelter and set- tlement options that must also include access to basic necessities for space, food, water, sanitation, hygiene, security, and medical care. Lessening population displacement markedly decreases the risk for morbidity and the excess demand for services as a result of flooding.

Well-established disease surveillance systems are necessary to monitor health effects among the flood-affected population and direct cost-effective public health interventions that may pre- vent secondary morbidity and mortality. Clinical and laboratory surveillance should be integrated to rapidly detect disease and guide therapy.

Early implementation of environmental health capacity related to toxicology, water, sanitation, waste management, and vector control can prevent disease due to toxic, vector-borne, food-borne, and water-related illnesses.

Flood disaster risk assessments should be used to guide both local and national decisions regarding location/relocation of crit- ical public health and healthcare facilities outside of floodplains whenever possible, so that secondary disasters (i.e., public health and medical facility evacuations) do not occur at healthcare facil- ities.

Public health programs that offer primary prevention of chronic disease (as compared with merely managing existing illness) also serve to lessen the vulnerability of the disaster- affected population to flood-related morbidity and mortality. In this sense, building and maintaining a baseline robust pub- lic health infrastructure also serves to reduce the public health impact of floods. Population reliance on food sources and dis- tribution systems that are not vulnerable to flooding may pre- vent any significant public health impact attributable to mal- nutrition. Reliable and economical access to drugs and medical care also helps to prevent adverse health effects. This is best accomplished through a well-developed and equitable healthcare system.

Mitigating the Public Health Impact of Flood Disasters

The public health impact of floods reflects secondary effects of the disaster such as population displacement and disruption of existing health services. To minimize these secondary effects, public health and healthcare relief efforts must be coordinated within the general emergency management cycle before, during, and after the impact phase and throughout recovery. Strate- gies should seek early reinstatement of normal routine activities of daily living among disaster-affected populations to mitigate ongoing adverse psychological and other health effects.

Mitigation is defined as the reduction of harmful effects of a disaster by limiting the disasters impact on human health and economic infrastructure. In the past, mitigation measures have been used in the traditional fields of engineering and urban planning. Flood-related mitigation activities lessen deaths and injuries by ensuring structural safety through enforcing adequate building codes, promulgating legislation to relocate structures away from flood prone areas, planning appropriate land use, and managing coasts and floodplains.3 Critical public health and medical assets can be identified before flood impact and engineering measures can be taken to mitigate loss of critical health infrastructure and assets during flooding. Key buildings may be designed or strengthened in ways that lessen the risk of flood damage.

Shelter and settlement solutions for displaced populations must adequately address basic human needs related to space, water, sanitation, hygiene, nutrition, and security to mitigate the public health impact of floods. Settlement solutions that do not promote hastening resettlement and minimizing population displacement may contribute to the long-standing psychologi- cal impact of the disaster and its subsequent association with chronic health effects. Therefore, public health workers should be involved in decisions regarding the rapid reinstatement of healthy homes and healthy communities that would have a direct impact on the health of flood-affected populations.

Preparing for the Public Health Impact of Floods

The public health impact of flood disasters is predictable for both high- and low-income nations. It is therefore possible to prepare emergency response and recovery activities that will lower the risk of flood-related morbidity and mortality.

“Public health emergency preparedness is the capability of the public health and health care systems, communities, and individuals, to prevent, protect against, quickly respond to, and recover from health emergencies, particularly those whose scale, timing, or unpredictability threatens to overwhelm routine capa- bilities.”58

An emergency management program has the goal of strengthening the “overall capacity and capability of a country to manage all types of emergencies and bring about an orderly transition from relief through recovery and back to sustained development.”59

Eleven “Es” of emergency preparedness have been described

■ Evaluation and forecasting of hazard ■ Early warning ■ Evacuation ■ Emergency operations planning ■ Education and awareness ■ Exercises and drills

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

FLO O D S ■ 539

■ e-Health and electronic media ■ Epidemiology ■ Equipment and supplies ■ Enforcement of landuse regulations and zoning codes ■ Economic incentive

In recent years, accurate weather forecasts linked to timely warn- ing systems for hazardous flooding have effectively mitigated the affects of floods on the health and well-being of commu- nities. Emergency operations planning for response and tran- sition planning for recovery are core elements of public health preparedness. Once plans are developed, both the public and response communities must be made aware and trained to imple- ment protective behavior. Exercises test the validity of emergency plans and the effectiveness of education and training. Epidemi- ological investigations identify the adverse health effects of flood disasters. Surveillance activities monitor health trends, allowing for early warning and intervention. The availability of equipment and supplies (i.e., personal protective equipment, boats and heli- copters, power generators, water pumps, and water purification units) offers the affected community enhanced absorptive capac- ity and resilience. Enforcement of landuse regulations and zoning codes also helps to break the response cycle by lessening or elim- inating the risk of population exposure to flooding.

Despite the increased level of preparedness, flood-related deaths, diseases, and injuries continue to occur in affected com- munities.3 The development of robust emergency operations plans17,60 and the design of effective messages to prompt desired behavior are key elements for saving lives.

Responding to the Public Health Impact of Floods

Community Needs Assessment Immediately after the flood impact phase, responder teams

conduct rapid needs assessments to identify gaps between health and medical needs of a flood affected community and available resources. A needs assessment typically consists of administering a standardized questionnaire that assesses health and medical and pharmaceutical needs; the status of public health services; and access to basic services such as water, sanitation and hygiene, food, shelter, sewage, and electricity.

Disease Risk Assessment Responding effectively to the needs of the disaster-affected

population requires an accurate disease risk assessment that includes both acute and chronic disease, including injuries as well as illnesses. A systematic and comprehensive postflood evalua- tion should identify: 1) diseases that are common and endemic to the affected area; 2) living conditions of the affected population; 3) availability of safe water and adequate sanitation facilities; 4) underlying nutritional status and immunization coverage among the affected population; and 5) degree of access to healthcare and to effective clinical care.26 Risk assessments can also characterize flood-related toxic exposures.

Fatality Management The public is often concerned about the danger of disease

transmission from decaying corpses. Responsible health author- ities should be aware that health hazards such as epidemics asso- ciated with unburied bodies are minimal, particularly if death resulted from trauma or drowning. It is far more likely that sur- vivors will be a source of disease outbreaks.17 Mass graves are not

necessary when considered solely to prevent the spread of dis- ease caused by mass fatalities. Normal funeral ceremonies and practices should be respected and maintained whenever possible (Chapter 21).

Clinical Diagnosis and Management of Flood-related Morbidity

Health caregivers should anticipate flood-related adverse health effects and be prepared to detect and intervene effectively when they occur. Clinicians should maintain a high index of sus- picion for illness and injuries commonly associated with floods. Flood-related injuries frequently include soft tissue lacerations, contusions, and abrasions. Endemic infectious diseases as well as chronic diseases may worsen. Mental illness and toxic exposures, for example carbon monoxide and mold, are known to increase after floods but may manifest with nonspecific symptoms such as malaise, anxiety, headaches, and nausea.

Studies of flood disasters have shown that outbreaks of vaccine-preventable diseases rarely result.35 Despite the fact that these epidemics have not been reported to occur following flood disasters in the United States, public demand for emergency mass immunization, especially against typhoid fever, hepatitis, and tetanus, is common. Assuring the safety of water and food sup- plies is of paramount importance in preventing enteric disease transmission when water and sewage systems have been com- promised. Active surveillance data should be used to justify con- sideration of an immunization campaign. Basic rules of hygiene and sanitation are far more important than immunizations in preventing infectious disease that floodwaters could potentially spread.57 Mass immunization in the absence of a documented outbreak is usually counterproductive during flood disasters and diverts limited human resources and materials from other more effective and urgent measures.3,17,35 Mass vaccination would be justified only when the recommended sanitary measures do not have a preventive effect and if there is evidence of the progressive increase in the number of cases of illness with the risk of an epidemic.17

Mass Care and Shelter of Flood-displaced Populations Public health officials should be involved in decisions regard-

ing the mass care and settlement of displaced populations to ensure a safe and healthy environment. Public health workers assist in performing food safety and water quality inspections, and assessment of sanitation and hygiene during mass evacuation and in shelters.

Environmental Health Services for Flood-affected Populations

The demands for environmental health services and consul- tation are high during flood disasters. The following are common considerations6

■ Purification of drinking and cooking water ■ Disinfection of wells ■ Food safety ■ Sanitation and personal hygiene ■ Mosquito control

Surveillance for Flood-related Morbidity and Mortality Although communicable disease outbreaks worldwide are

rare after flooding, some potential does exist for disease trans- mission; therefore, flood-affected communities should be under

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

540 ■ MA R K E. KE I M

close surveillance.57 Mortality surveillance is performed to deter- mine the nature and circumstances surrounding flood-related deaths so that appropriate preventive actions can be taken to reduce further mortality. Morbidity surveillance is conducted to determine: 1) any increases in diseases that are endemic to the area; 2) any cases of infectious disease that must be con- tained or controlled; and 3) any cases of injuries that may require public advisories. Flood-specific surveillance systems are often used to determine any increases in vector populations such as mosquitoes. In addition, public health officials should conduct laboratory-based surveillance of drinking water sources such as public and private wells.3

Chemical Emergency Response The role of public health in the investigation of a flooding

event that results in a chemical release should include the follow- ing activities20

■ Hazard identification ■ Risk communication ■ Liaison with other relevant responder agencies ■ Technical advice including aspects of toxicology, decontam-

ination, antidotes, and personal protective equipment ■ Registry and follow-up of exposed casualties ■ Emergency plan development for flooding in other high-risk

areas

The Role of Social Support in Improving Behavioral and Physical Health Outcomes

A complex set of social and other factors define flood victims’ susceptibility to health and stress effects. There is some evidence that effective community and professional agency management of the flooding aftermath can mitigate deleterious mental health outcomes, which in turn are strongly associated with physical health effects.19 Research from the United States has indicated that providing increased social support can significantly lower illness burdens after disasters.61

Recovering from the Public Health Impact of Floods

Long-term recovery from the public health impact of flood dis- asters can take years. Additional financial, health, and emotional costs may continue long after basic utilities and shelters have been reinstated. The disaster recovery phase may also offer a window of opportunity for improving risk reduction strategies, such as preparedness and mitigation efforts. Recovery strategies that promote future reduction of risk should be prioritized. Per- manent migrations of populations are rare after disaster floods. Disease surveillance systems are necessary to identify the long- term adverse health effects of flood disasters. Public health and medical services must recover to normal levels to detect and help patients to manage the long-term risks of flood-associated illness, such as mental illness and cardiovascular disease.

RECOMMENDATIONS FOR FURTHER RESEARCH

Flooding is one of the most widespread disaster hazards, posing multiple risks to human health. Despite this, there has been only limited systematic research on the health outcomes of flooding.19

There is surprisingly limited evidence about the health effects of floods, particularly in relation to morbidity. There are virtually

no studies available on the effectiveness of public health mea- sures, other than flood warning systems. A wide range of health risks have been well documented, although there remains scien- tific uncertainty regarding the strength of association and public health burden for specific health effects.29 Overall, there are few data on the long-term health impact of flooding.48

Although some studies conducted during and after floods have provided important information on factors contributing to the risk of morbidity and mortality, unresolved inconsistencies remain nearly 20 years after being posed as research recommen- dations.

Research recommendations originally offered by Jean French and Kenneth Holt in 1989 include49

■ Factors influencing actions people take in the face of flash flood warning and evacuation notices should be studied fur- ther

■ Study should be done to assess the circumstances under which there is sufficient time to permit evacuation by car or when it is safer to abandon vehicles and escape to higher ground on foot

■ A cohort of flood victims should be followed over time to determine whether they are at higher risk than a comparable group of nonflood victims of having adverse physical and mental health effects

■ Systematic study should be undertaken taken to determine whether an increase in certain biological agents results from disrupted water supplies and sewage systems after floods and whether this is related to geographical location

■ Systematic studies should be undertaken to examine the release of chemical agents during flooding and the poten- tial for contamination of human pathways from such events

■ A reporting system should be established to more accurately assess the number of deaths and injuries associated with each flood and the circumstances surrounding each flood death and injury

There is a need for standardized criteria for estimating flood damage and public health impact. Perception of flood dam- age is influenced by historical experience. For example, in low- vulnerability U.S. states, floods causing more than $1 million of damage are notable events and are likely to be reported. Con- versely, in high-vulnerability states, damage of $5 million or more occurs frequently, so smaller damages might seem unremarkable and be underreported.55

The effectiveness of detection and warning systems should be evaluated and researchers should make recommendations for appropriate standards for such systems in ensuring greater warn- ing sensitivity.3 More research into the behavioral outcomes of health risk communication is important. This is especially signif- icant to help prevent the continued cases of drowning related to vehicle use, as well as other preventable morbidity and mor- tality such as mold-related exposures and carbon monoxide intoxication.

In 2005, Ahern and colleagues identified the following knowl- edge gaps with respect to managing the public health impact of flood disasters29

■ Mental health impact of flooding, especially the long-term impacts, and their principal causes

■ Nature and magnitude of mortality risk in the period after flooding

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

FLO O D S ■ 541

■ Quantification of the risk of infectious and vector-borne diseases following floods

■ Effectiveness of warning systems and public health measures in reducing flood-related health burdens

■ Defining health-related costs of flooding in terms of how they influence decisions about specific interventions

■ Quantification of the degree to which climate and landuse change will contribute to flood risk and associated health burdens in different settings

The World Health Organization also recognizes that the mental health consequences of floods “have not been fully addressed by those in the field of disaster preparedness or service deliv- ery,” although it is generally accepted that disasters, such as earthquakes, floods, and hurricanes, “take a heavy toll on the mental health of the people involved, most of whom live in developing countries, where capacity to take care of these prob- lems is extremely limited.” Areas of particular concern relate to anxiety and depression, posttraumatic stress disorder, and sui- cide.29 In summary, much work remains to be done to define evidence-based approaches to decrease morbidity and mortality from floods, a disaster with one of the greatest overall public health impacts.

REFERENCES

1. Gunn SWA. Multilingual Dictionary of Disaster Medicine and International Relief. Dordrecht: Kluwer Academic Publishers, 1990.

2. Floodplain Management Association. Overview of flood- plain management. Available at: http://www.floodplain.org/ overview of floods.htm. Accessed September 2, 2007.

3. Malilay J. Floods. In: Noji ER, ed. The Public Health Consequences of Disasters. New York: Oxford; 1997:287–300.

4. Bey T, van Weizsaecker E, Koenig KL. Global warming: polar bears and people – implications for public health preparedness and disaster medicine: a call to action. Prehosp Disaster Med. 2008;23(2):101–102.

5. Poole J, Hogan D. Floods. In. Hogan D, Burstein J, eds. Dis- aster Medicine. Philadelphia: Lippincott Williams & Wilkins; 2007:214–4.

6. Centers for Disease Control and Prevention. Morbidity surveil- lance following the Midwest flood – Missouri, 1993. MMWR. 1993;42(41):797–798.

7. Keim M. Cyclones, tsunamis and human health. Oceanography. 2006;19(2):40–49.

8. International Federation of Red Cross and Red Crescent Soci- eties. World Disaster Report 2006. Bloomfield, CT: Kumarian Press; 2006:211–218.

9. National Weather Service. Natural hazard statistics. Available at: http://www.nws.noaa.gov/om/hazstats.shtml. Accessed Novem- ber 26, 2008.

10. U.S. Department of the Interior, U.S. Geological Survey. Flood Hazards – A National Threat. U.S. Geological Survey Fact Sheet 2006–3026. Available at: http://pubs.usgs.gov/fs/2006/3026/. Accessed November 26, 2008.

11. Llewellyn M. Floods and tsunamis. Surg Clin Am. 2006;86 :557– 578.

12. Parret C, Melcher NB, James RW. The discharges in the upper Mississippi River basin. US Geological Survey Circular 1120-a. Denver, CO: US Government Printing Office. 1993.

13. Centers for Disease Control and Prevention. Storm-related mor- tality – Central Texas, October 17–31, 1998. MMWR. 2000; 49(07):133–135.

14. Munich Reassurance Company. Annual Review: Natural Catas- trophes 2002. Available at: http://www.unep.org/download file .multilingual.asp?FileID=96. Accessed November 26, 2008.

15. Centers for Disease Control and Prevention. Tropical Storm Alli- son rapid needs assessment Houston, Texas, June 2001. MMWR. 2002;51(17):365–369.

16. Centers for Disease Control and Prevention. Public health conse- quences of a flood disaster – Iowa, 1993. MMWR. 1993;42:653– 656.

17. Noji E. Public health issues in disasters. Crit Care Med. 2005;33(1):S29–33.

18. Federal Emergency Management Agency. Federated States of Micronesia Drought Emergency Declaration, July 31, 2007. Available at: http://www.fema.gov/news/event.fema?id=8465. Accessed November 26, 2008.

19. Tunstall S, Tapsell S, Green C, et al. The health effects of flooding: social research results from England and Wales. J Water Health. 2006;(04)3: 365–380.

20. Euripidou E, Murray V. Public health impacts of floods and chemical contamination. J Pub Health. 2004;26(4):376– 383.

21. Potera C. Fuel damage from flooding: finding a fix. Env Health Persp. 2003;111(4). Available at: http://www.ehponline .org/members/2003/111-4/innovations.html. Accessed Novem- ber 26, 2008.

22. Centers for Disease Control and Prevention. Health concerns associated with mold in water damaged homes after hurricanes Katrina and Rita – New Orleans area, Louisiana, October 2005. MMWR. 2006;55(02):41–45.

23. Ivers, LC, Ryan ET. Infectious diseases of severe weather- related and flood-related natural disasters. Curr Opin Infect Dis. 2006;19(5): 408–414.

24. Noji E. Natural disaster management. In: Auerbach P. ed. Wilderness Medicine: Management of Wilderness in Environmen- tal Emergencies. 4th ed. St. Louis: Mosby; 2001:1603–1621.

25. Peters M. Hospitals respond to water lost during the Midwest floods in 1993: preparedness and improvisation. J Emerg Med. 1996;14(3):345–350.

26. Watson JT, Gayer M, Connolly MA. Epidemics after nat- ural disasters. Emerg Infect Dis. 2007;13(1). Available at: http://www.cdc.gov/ncidod/EID/13/1/1.htm. Accessed Novem- ber 26, 2008.

27. Kim SH. Flood. In Ciottone G. ed. Disaster Medicine. 3rd ed. Philadelphia: Mosby; 2006:489–491.

28. WHO. Europe fact sheet 05/02 Flooding: health effects and preventive measures. Available at: www.euro.who.int. Accessed August 20, 2007.

29. Ahern M, Kovats RS, Wilkinson P, et al. Global health impacts of floods: epidemiologic evidence. Epidemiol Rev. 2005;27:36– 46.

30. Staes C, Orengo JC, Malilay J, Rullan J, Noji E. Deaths due to flash floods in Puerto Rico, January 1992: implications for prevention. Intl J Epidemiol. 1994;23:968–975.

31. American National Red Cross. Flood and flash flood. Available at: http://www.redcross.org/services/disaster/keepsafe/flood .html. Accessed November 26, 2008.

32. Centers for Disease Control and Prevention. Morbidity and mortality associated with Hurricane Floyd–North Carolina, September–October 1999. MMWR. 2000;49(17):369–372.

33. Schultz J. Epidemiology of tropical cyclones: the dynamic of disaster, disease, and development. Epidemiol Rev. 2005;27: 21–35.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

542 ■ MA R K E. KE I M

34. Noji E. The nature of disasters. In: Noji ER, ed. The Pub- lic Health Consequences of Disasters. New York: Oxford; 1997: 3–20.

35. Centers for Disease Control and Prevention. Current trends flood disasters and immunization – California. MMWR. 1983; 32(13):171–178.

36. Tuffs A, Bosch X. Health authorities on alert after extensive flooding in Europe. BMJ. 2002;325:405.

37. Centers for Disease Control and Prevention. Norovirus out- break among evacuees from hurricane Katrina – Houston, Texas, September 2005. MMWR. 2005;54:1016–1018.

38. Centers for Disease Control and Prevention. Two cases of tox- igenic Vibrio cholerae O1 infection after Hurricanes Katrina and Rita – Louisiana, October 2005. MMWR. 2006;55:31– 32.

39. Centers for Disease Control and Prevention. Vibrio illnesses after hurricane Katrina – multiple states, August–September 2005. MMWR. Morb Mortal Wkly Rep. Available at: http:// www.cdc.gov.mmwr/preview/mmwrhtml/mm5437a5.htm. Accessed August 20, 2007.

40. Centers for Disease Control and Prevention. Infectious disease and dermatologic conditions in evacuees and rescue workers after Hurricane Katrina – multiple states, August–September, 2005. MMWR. 2005;54:961–964.

41. McCarthy MC, He J, Hyams KC, et al. Acute hepatitis E infection during the 1988 floods in Khartoum, Sudan. Trans R Soc Trop Med Hyg. 1994;88:177.

42. Woodruff BA, Toole JM, Rodriguez DC, et al. Disease surveil- lance and control after a flood in Khartoum, Sudan, 1988. Dis- asters. 1990;14:151–163.

43. Centers for Disease Control and Prevention. Rapid assessment of vectorborne diseases during the Midwest flood United States, 1993. MMWR. 1994;43(26):481–483.

44. Centers for Disease Control and Prevention. Report: Interna- tional notes health assessment of the population affected by flood conditions – Khartoum Sudan. MMWR. 1989;37(51&52): 785–788.

45. Guha-Sapir D, van Panhuis W. The Andaman Nicobar earth- quake and tsunami 2004: impact on diseases in Indonesia. Center for Research on the Epidemiology of Disasters (CRED), Brussels Belgium 2005. Available at: http://www.tsunami- evaluation.org/NR/rdonlyres/10D559A4–0FB9–4F1D-A666– 8ACCD173CAC3/0/cred Tsunami report health disease1.pdf. Accessed September 3, 2007.

46. Centers for Disease Control and Prevention. Report: leptospiro- sis after flooding of a university campus – Hawaii, 2004. MMWR. 2006;55(05):125–127.

47. Gerrity E, Flynn B. Mental health consequences of disasters. In: Noji ER, ed. The Public Health Consequences of Disasters. New York: Oxford; 1997:101–121.

48. Ohl C, Tapsell S. Flooding and human health: the dangers posed are not always obvious. BMJ. 2000;321:1167–1168. (Editorial)

49. French JG, Holt KW. Floods. In Gregg MB, ed. The Public Health Consequences of Disasters. Atlanta: US Department of Health and Human Services, Public Health Service, CDC; 1989:69–78.

50. United Nations Environment Programme. Declaration of the United Nations Conference on the Human Environment. Available at: http://www.unep.org/Documents.Multilingual/ Default.asp?DocumentID=97&ArticleID=1503. Accessed Nov- ember 26, 2008.

51. World Summit on Sustainable Development Plan of implemen- tation, Johannesburg, South Africa, Find at url: http://www .johannesburgsummit.org/html/documents/summit docs/ 2309 planfial.htm. Accessed August 20, 2007.

52. Werrity A. Sustainable flood management: oxymoron or new paradigm? Area. 2006;38(1):16–23.

53. Tapsell S, Penning-Rowsell E, Tunstall S, et al. Vulnerability to flooding: health and social dimensions. Phil Trans R Soc Lond. 2002;360:1511–1525.

54. Anonymous. Katrina reveals fatal weaknesses in US public health. Lancet. 2005;366:867. (Editorial)

55. Pielke RA Jr, Downton MW, Barnard Miller JZ. Flood Dam- age in the United States, 1926–2000: A Reanalysis of National Weather Service Estimates. Boulder, CO: UCAR. Available at: http://www.flooddamagedata.org./. Accessed November 26, 2008.

56. US Department of Commerce, National Oceanic and Atmo- spheric Administration, National Weather Service. Floods: The Awesome Power. Available at: http://www.nws.noaa.gov/ om/brochures/floodbrochure 9 04 low.pdf. Accessed Novem- ber 26, 2008.

57. Centers for Disease Control and Prevention. Outbreak of diar- rheal illness associated with a natural disaster – Utah. MMWR. 1983;32(50):662–664.

58. Nelson C, Lurie N, Wasserman J, et al. Conceptualizing and defining public health emergency preparedness. Am J Public Health. 2007;97(S1):S9–11.

59. deBoer J, Dubouloz M, eds. Handbook of Disaster Medicine. The Netherlands: International Society of Disaster Medicine; 2000.

60. Keim M. Developing a public health emergency operations plan: a primer. Pac Health Dialog. 2002;9:124–129.

61. Lutgendorf SK, Antoni MH, Ironson G, et al. Physical symptoms of chronic fatigue syndrome are exacerbated by the stress of Hurricane Andrew. Psychosom Med. 1995;57:310–323.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Cyclones/Hurricanes/Typhoons

Kelly R. Klein and Frank Fuh-Yuan Shih

OVERVIEW

Introduction

In the history of all ancient civilizations, there are fantastic tales in which a country or kingdom is saved by divine intervention. For example, when the Mongols sought to conquer Japan and complete their control of all Asia, a divine wind known as the Kamikaze saved the Japanese people from Kublai Khan by sinking the invasion fleet. This powerful storm, which saved Japan, is known today as a tropical cyclone.

Storms such as these have caused deaths in the hundreds of thousands and billions of dollars in property loss in the past 100 years on coastlines around the world. The risk seems to be increasing as more and more people decide to live in vulnerable coastal areas. As discussed in the 1999 Hangzhou Declaration in China, more than half the world’s population lives in coastal areas with several of the fastest growing cities, Jakarta, Shanghai, and Miami, all projected to have 20–30 million in population by the year 2025. In the U.S., it is estimated that by 2010, 60% of the population will live on a coast.1 With the projected increase in number and intensity of tropical cyclones, this becomes a signif- icant threat. The focus of this chapter is on the impact of tropical cyclones on human societies. This includes public health; the mortality and morbidity resulting from these events; interven- tion measures such as evacuation; medical preparedness for the affected population; and mitigation, prevention, and response strategies for the medical community drawn from a global per- spective.

Tropical Cyclones

Tropical cyclones, often referred to as hurricanes, cyclones, and typhoons, are given different names depending on their partic- ular geographical locations (see Table 33.1). All are capable of producing large-scale devastation.2 In the northern hemisphere, from the International Date Line to the Greenwich meridian, they are known as hurricanes. In the Pacific, north of the equa- tor and west of the International Date Line, they are known as typhoons. In the Indian Ocean, they are known as cyclones. The tropical cyclone, among the most destructive of weather systems,

is defined meteorologically as a storm system characterized by a low-pressure center with surrounding thunderstorms that pro- duce strong winds and flooding rain.3 These storms are capable of producing up to 20 billion tons of rainwater per day. A fully developed hurricane contains the equivalent energy of 2 million Hiroshima-sized atomic bombs.4,5

These massive and deadly storms originate over tropical or subtropical oceans and derive their storm energy from these warm waters. They begin as fragile meteorological entities that require several factors to ensure their formation.

1 Warm oceanic waters that are at least 26.5◦C 2 An atmosphere that cools rapidly with moist layers at mid-

troposphere elevations to enhance thunderstorm formations 3 The Coriolis effect to rotate the winds and the near-surface

disturbance to create a vortex with minimal vertical wind shear

Cyclonic systems, which may last over open waters for more than 2 weeks, rotate counterclockwise in the northern hemi- sphere and clockwise in the southern hemisphere with gen- eral storm movement of east to west. As the storm develops, it progresses through successive meteorological stages: tropical wave, tropical disturbance, tropical depression, tropical storm, and cyclone (see Table 33.2). Storms strengthen over water by the energy released from differences between water and upper

Table 33.1: Classification of Cyclones Based on Geographical Location

Hurricane: the North Atlantic Ocean, the Northeast Pacific Ocean east of the dateline, or the South Pacific Ocean east of 160E

Typhoon: the Northwest Pacific Ocean west of the dateline

Severe tropical cyclone: the Southwest Pacific Ocean west of 160E or Southeast Indian Ocean east of 90E

Severe cyclonic storm: the North Indian Ocean

Tropical cyclone: the Southwest Indian Ocean

543 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

544 ■ KE L LY R. KL E I N A N D FR A N K FU H-YUA N SH I H

Table 33.2: Storm Progression

Storm Type Wind Speeds Duration Metrological Features

Tropical/Easterly wave variable 24 h Low pressure moving westward through the trade wind easterlies. Associated with extensive cloudiness and showers

Tropical disturbance variable >24 h Area of organized convection. Often the first developmental stage of any subsequent tropical depression, storm, or cyclone

Tropical depression <38 mph (16 m/s) Having 1 or more closed isobars (line drawn on the weather map of equal barometric pressures)

Tropical storm >39 mph (17 m/s) No classic developed eye; rain bands form outward from the center; given a name and it is tracked

Cyclone >74 mph (33m/s)

atmospheric temperatures and are further defined by degrees of barometric pressure, precipitation counts, and the radius of their cloud mass. A well-developed storm has a highly organized warm center of low barometric pressure and definite cyclonic or cir- cular surface wind movement. Due to reduction of temperature disparity, they weaken upon landfall. Occasionally there are two mature-stage tropical cyclones that directly interact with each other in a phenomenon known as the Fujiwhara effect. If con- ditions are right, the Fujiwhara effect occurs when two cyclones come within 300–700 nautical miles of each other and begin to rotate around each other (Figures 33.1 and 33.2).

When a tropical cyclone develops, it is graded on a 1–5 rating scale based on wind speed, using the Saffir–Simpson

scale and/or the Australian Tropical Cyclone Intensity scale (Table 33.3). These scales are used to estimate potential prop- erty damage and the degree of flooding expected along the coast after the tropical cyclone has made landfall. It is important to remember that although the intensity of the winds is predictive of damage, the speed with which a storm moves through an area also has significant impact. Tropical cyclones cause loss of life and property damage primarily due to their strong winds, flood- ing from the inundating rains, and storm surges. In addition, secondary events are often induced by these cyclones, creating or exacerbating new or existing hazards. Such problems include tornadoes, landslides, mudslides, and increased flooding due to levee breaches (see Chapter 32).

Figure 33.1. Fujiwhara effect: Typhoons Ione and Kirsten, August 24, 1974. Image ID: wea00481, NOAA’s National Weather Service (NWS) Collection Source: NOAA Photo Library. See color plate.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

CYC LO N E S/HU R R I C A N E S/T Y P H O O N S ■ 545

Table 33.3: Saffir–Simpson Hurricane Scale

Storm Surge Feet Category MPH (km/h) (Meters) Damage Caused

Tropical depression 0–38 (0–62) 0 (0) None

Tropical storm 39–73 (63–117) 0–3 (0–0.9) None

1 Minimal 74–95 (119–153) 4–5 (1.2–1.5) Minimal to buildings structures; primarily to unanchored mobile homes, shrubbery and trees; some coastal road flooding and minor pier damage

2 Moderate 96–110 (154–177) 6–8 (1.8–2.4) Some roofing material, door and window damage; considerable damage to vegetation, mobile homes and piers; small craft in unprotected anchorages break moorings

3 Extensive 111–130 (178–209) 9–12 (2.7–3.7) Structural damage to small residences and utility buildings with a minor amount of curtain wall failures; mobile homes are destroyed; flooding near the coast destroys smaller structures with larger structures damaged by floating debris

4 Extreme 131–55 (210–249) 13–18 (4–5.5) More extensive curtain wall failures with some complete roof structure failure on small residences; major beach erosions; major damage to lower floors of structures near the shore

5 Catastrophic 156> (>250) >18 (>5.5) Complete roof failure on many residences and industrial buildings; some complete building failures with small utility buildings blown over or away; major damage to lower floors of all structures located less than 15 ft (3 m) above sea level

One of the most devastating metrological manifestations of the tropical cyclone is its storm surge. Here, water accu- mulates along the coast as the storm approaches, pushed for- ward by the wind and speed of the storm. Depending on the slope of the continental shelf, the storm surge can be quite mas- sive and destructive. Its effects are exacerbated by topographi- cal changes due to deforestation, topsoil erosion, and increased coastal construction. Without natural barriers to block the water and wind, the cyclone’s effects are carried much farther inland, increasing secondary effects from landslides and building col- lapses.

The severe effects of storm surge are augmented by the local tide and the storm’s extremely low barometric pressures. This barometric impact causes the water surface to rise 1 cm for every millibar reduction of air pressure; hence, the more intense the

Figure 33.2. Fujiwhara effect: Hurricanes Emmy and Frances, August 31, 1976. Image ID: wea00489, NOAA’s National Weather Service (NWS) Collection. Source: NOAA Photo Library. See color plate.

storm (lower barometric pressure), the greater the storm surge.6

The Galveston hurricane that killed 8,000 people in the year 1900 is an example of the effects of storm surge. It produced a 5-m storm surge that flooded the island of Galveston, Texas, which has a maximum height of only 3 m. In 2005, during Hurricane Katrina, a recorded surge wave averaging 2 meters traveled as far inland as 19 km. During that same hurricane, a documented storm surge of 8.5 m was recorded in the city of Pass Christian, Mississippi.7

STATE OF THE ART

Consequences for Public Health Recent large cyclonic storms such as Hurricanes Katrina and

Rita in the U.S. and Typhoon Nari that caused devastation in Taiwan reaffirmed the need to meet the complex challenge of public health planning, especially for those with special needs. Burkle and Rupp stated that disasters “keep governments and planners honest by defining public health and exposing its vul- nerabilities.”8 Noji noted that a variety of public health emergen- cies share a common thread by “adversely impacting the pub- lic health system and its protective infrastructure (e.g., water, sanitation, shelter, food, and basic health).” Poverty and social inequality, environmental degradation from inappropriate land use, and rapid population growth all contribute to the pub- lic health effects of a tropical cyclone’s landfall.9 Following a disaster, levels of resilience to both chronic stress and any sub- sequent catastrophic shock are typically lowered for individuals and groups. An increased risk of susceptibility exists to diseases, both physical and mental. Depending on its magnitude, the great- est potential for loss of life does not come from the actual event, but instead comes from everyday health risks such as reduced access to potable water, break down of sanitation systems, lack of medical care for chronic medical and psychiatric conditions, and exposure to entomological vectors. These conditions create

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

546 ■ KE L LY R. KL E I N A N D FR A N K FU H-YUA N SH I H

Table 33.4: Impact of Cyclones and Storm Surges on the Community

Categories of Impact Components Involved Indicators of Impact

Physical Inadequate physical protection; poor-quality housing and infrastructure; disruptions of communication, roads, utilities, public works infrastructure

Trauma-related death tolls; damage/loss physical properties such as infrastructure, homes, industry, animal, and crops; disruption of normal life, migration to safe places, lack of electricity, potable water, food sources, waste build-up

Economic Loss of livelihood and income opportunities; loss of assets and savings; need for recurrent aid, lower socioeconomic stratification

Low income, poverty, unemployment, landlessness, unequal land distribution, lack of relief and rehabilitation, and forced movement of lower-income populations

Agricultural Land degradation; intrusion of salt water for irrigation increasing seasonal, unplanted fields

Low productivity, frequent crop loss, outbreak of migration among the owners of small farms and farm laborers; lack of money for purchasing seed

Social Disintegration of social organization, increased incidence of female-headed households and resource-poor communities; poor education services

Social/ethnic crisis; social marginalization, violence and crime; apathetic attitude; identity crisis; plight of people for safety and survival

Environmental Land and environmental degradation; deforestation, loss of biodiversity and marine resources, increase in salinity, intrusion of salt water, lowering of water table, dams

Deforestation; loss of soil fertility; limiting of biodiversity; increase refugees, migrants, and homelessness; rising disaster-related deaths

Public Health Disruptions of healthcare and utility services, inadequate sanitation, lack of qualified physician and clinical services

Increase mortality and morbidity; poor health and malnutrition; disease epidemics; exacerbation of chronic diseases, increased PTSD

long-term effects, depending on the socioeconomic stratifica- tion of the area involved, with lower baseline status most often associated with poor outcomes.9–11

The public is generally aware that tropical cyclones are capa- ble of causing devastating damage, severely crippling if not destroying society and its infrastructure. Yet, it is expected by the effected community that established public health and health- care systems will continue to provide services not only in the days leading up to the storm but during and after the event as well.12 Because of these demands, it is imperative that the medical and public heath care community be prepared to not only man- age injuries created by the storm, but also to provide continued care for patients with chronic medical conditions and those with special medical needs (see Chapter 8). Typically, these include victims with hypertension, diabetes, renal failure needing dialy- sis, mental illness, and physical disabilities. This is in addition to ensuring safe public drinking water; appropriate sewage disposal; control of disease vectors such as mosquitoes and rats; food dis- tribution; and protection of food supplies from contamination (Table 33.4).

Evacuations

Ideally, no one would be physically present to suffer death or injury during the landfall of a devastating tropical cyclone. In fact, due to their well-defined paths of travel and the use of modern meteorological tracking systems, 70% of hurricanes will be forecasted 24 hours in advance of their approach to land based on their speed and direction during the previous 24–36 hours.13

Computer-based models, such as those that use the Sea, Lake and Overland Surges from Hurricanes (SLOSH) software, can assist emergency planners in predicting storm surge heights. Based on these predictions, people in vulnerable areas are often asked to evacuate voluntarily while local officials assist by changing traffic flow patterns. Often, the contra flow technique is used, in which

both lanes of a roadway are used for outgoing traffic. Researchers have observed that, depending on local and personal experiences, citizens will engage in two opposing types of behavior prior to an evacuation order. They will either spontaneously self-evacuate or, despite storm warnings and subsequent evacuation orders, refuse to comply and remain in their homes sheltered in the same way they have done in previous years during storms.14

Evacuation is a very complex undertaking requiring the coor- dination of a multitude of factors, not the least of which is main- taining basic public health services for evacuees. In addition to the logistics, cost is an issue. This includes not only the expenses related to the evacuation itself, but also of lost revenue to dis- placed individuals and to local industry. Therefore, the decision to evacuate an area has significant ramifications.

Hospitals also face the threat of hurricane-induced evac- uations. From the years 1971–1999 in the U.S., hurricanes prompted over 38 hospital evacuations.15 Mathematical model- ing of hospital evacuations predicts that, depending on resources available, it could take between 30 and 60 hours to evacuate 50 patients.16 Evidence from hospital evacuations after Hurricane Rita suggests the actual time frame may be somewhat shorter, with seven hospitals evacuating in an average of 29 hours.17

Many countries other than the U.S. do not undergo mass evacuations in the face of a storm. For example, in Taiwan, a community that is frequently subjected to tropical cyclones, most of the buildings are wind resistant, and only people who live in flood plains, mudslide prone areas, or who are physically disabled or dependent are considered for evacuation.18

The decision to evacuate is based on available resources and an estimate of the resulting economic impact as well as the poten- tial loss of life. Because storm landfall predictions can lead to expensive preparations and subsequent disruptive population movements, individuals involved in making such decisions must consider the potential negative impacts of an evacuation on com- mercial, healthcare, and other public health activities. Making

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

CYC LO N E S/HU R R I C A N E S/T Y P H O O N S ■ 547

Table 33.5: Time Considerations for Phases of Population Evacuation

Evacuation Phase Time Needed

Mobilizing community evacuation resources

Hours

Communicating appropriate protective action instructions to the public

Hours

Individual mobilization of resources to leave the area at risk

Hours to days

Completing the physical evacuation of people occupying the affected area

Days

the decision to evacuate is a challenge for administrators and community leaders and involves determining who should be evacuated, when the evacuation should start, and the logistics for the evacuation, that is, mass transportation, traffic patterns, and provision of control/security. There are four major event scenarios for which evacuation decisions may affect credibility of the policy makers and economic losses.19

■ Evacuation with direct damage to the area or structure evac- uated: no lives lost due to the damage nor credibility lost but large economic costs through loss of revenue and expenses incurred

■ Evacuation with no damage to the area: no lives lost due to the storm but a loss of credibility with large economic costs through loss of revenue and expenses incurred

■ No evacuation with damage to structures and the area: even if no lives lost due to the damage there is a loss of credibility and a large economic cost due to repair and loss of revenue

■ No evacuation with no damage to the area in the absence of a direct impact: no lives lost, no credibility lost, and no economic effects

The time required to accomplish an evacuation once the physical movement of people is underway depends on the char- acteristics of the area and on the availability of public transporta- tion and large highways. Research indicates that, once a decision to evacuate is made, up to 2 hours may elapse before most people in the affected area hear, absorb, and decide to respond to the instructions (Table 33.5).20 It is intuitive to think that a larger population warrants a longer evacuation time, but warning and evacuation times do not necessarily increase with population size and density. This is true, in part, because the infrastructure capac- ity (e.g., street system, public transportation resources) necessary for moving people out of the area is generally more extensive in regions with greater population.21 In areas where there is pub- lic reluctance to evacuate or the evacuation routes are limited, repeated warnings may be necessary. Characteristics of a good evacuation plan include

■ Identification of available resources such as community faith- based organizations and voluntary medical and fire assets

■ Knowledge of vulnerable populations, which would include those who are elderly, ventilator dependent, and have lan- guage barriers

■ Awareness of hazardous sites in the area: flood zones, refiner- ies, and hazardous material sites

■ Knowledge about main transportation assets: highways, trains, buses, and airports

■ Shelter locations and staging areas for evacuation

Mortality

In violent tropical cyclones, almost all primary weather-related deaths are attributed to the storm surge. Examples include the cyclones that impacted Bangladesh in 1970 and 1991, the Indian coastal states of Andhra Pradesh in 1977 and Orissa in 1999, the Indian state of Gujarat along its coast facing the Arabian Sea in 1998, and the state of Mississippi in 2005 (Hurricane Katrina). Other immediate deaths result from tornadoes, flying debris, and collapsing structures.14,22–23 In Taiwan and many other Asian/Pacific basin countries, mortality numbers due to storm surge and mudslides remain quite high despite warnings. This has been attributed to deforestation, which allows for mud- slides and debris to flow through farming communities dur- ing the torrential rainstorms that accompany tropical cyclones (Figure 33.3). Many countries prone to damage from storm surge have installed early warning systems, which if used in conjunc- tion with timely evacuations and storm-resistant sheltering, can achieve a decrease in mortality rates. Prediction and warning systems have been credited with protecting lives in the Missis- sippi counties of Mobile and Baldwin. Here, computer models predicted a large storm surge 2 days in advance of the hurricane’s arrival, allowing for evacuation and adequate preparation.24 In the countries of Bangladesh and Cuba, storm-related mortality rates are lower despite the lack of sophisticated electronic warn- ing systems. This is due to the use of trained volunteers who implement a well-known and easily recognizable flag and siren signal system. Conversely, simply having the technology will not guarantee a decrease in mortality. This was seen in Haiti during tropical storm Jeanne in 2004. The area had a good electronic warning system, but due to a coup earlier that year, there were no emergency managers available to utilize the system and more than 1,000 people died.25

Morbidity

During a tropical cyclone disaster, there are phase-specific mor- bidity patterns. Understanding these patterns may help with medical and emergency planning. In the pre-landfall phase, injuries result from storm preparation and evacuation activi- ties. Typical problems include car crashes and falls from ladders. When the tropical cyclone makes landfall, injuries occur from nonreinforced structural collapses, wind-borne debris, falling trees, drowning, and downed power lines. In the immediate postimpact phase, traumatic injuries result from electrocution by downed power lines, falling objects and trees, and severe lac- erations from chain saws as people clear debris from houses and roadways. As a result of power outages, burn injuries and carbon monoxide poisoning may occur from cooking, lighting equipment, or improperly ventilated gas-powered generators.22

It is during the postimpact phase that a surge occurs in the demand for healthcare by patients with chronic medical condi- tions. Interventions such as dialysis and refills of medications for hypertension, diabetes, psychiatric illnesses, and chronic pain are required. During the recovery phase of the disaster, acute care

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

548 ■ KE L LY R. KL E I N A N D FR A N K FU H-YUA N SH I H

Figure 33.3. Area of mudslide resulting from deforestation. Personal photo, Hawaii, 2007. See color plate.

needs transition to chronic care. Within weeks, the increased need for generalists, pediatricians, obstetricians, nephrologists, psychiatrists, and cancer specialists replaces the need for surgeons and emergency medicine providers.

The provision of mental health services for victims and rescuers is an important component of any disaster recovery process (see Chapter 7). Therefore, the recovery plan should include provisions for the involvement of psychiatrists and men- tal health workers. Patients will require continued treatment of their addictions, depression, and schizophrenia. In addition, ser- vices are needed for victims suffering from acute stress syndromes and posttraumatic stress disorder (PTSD).26

Many people initially experience fear and distress at the time of a tropical cyclone’s impact, but the majority of them quickly return to normal. Some people may experience persistent dis- tress that affects functional capability, and a subset of these people will progress to PTSD. In the U.S., surveillance systems detected increases in rates of psychological disorders after hurricanes. Disorders occurring after the storms that victims had never pre- viously experienced included PTSD, major depression, and anx- iety. Risk factors attributable to adverse mental health outcomes included: 1) the severity of an individual’s exposure to a family member’s injury or death; 2) experiencing extensive property loss or displacement; 3) belonging to a vulnerable group such as women, children, the elderly, and the poor; and 4) existing psy- chopathology. Social support, self-efficacy, and positive coping strategies can ease the severity of mental health consequences. Early intervention allows mental health professionals to triage people with an increased risk for more severe mental illness.27

Infectious and Environmental Diseases

The likelihood of infectious outbreaks in a community following a tropical cyclone may increase for a multitude of causes: disrup- tion of public health services and healthcare infrastructure, dam- age to water and sanitation networks, population displacement, and crowded conditions in temporary shelters. Fecal–oral routes of infection are often the cause and result in outbreaks of diseases

such as cholera, hepatitis, shigella, and other diarrheal illnesses. In crowded shelters, outbreaks of measles and meningitis have been reported, but not in epidemic proportions.

In regions of the globe where infectious disease vectors such as mosquitoes or fleas exist, outbreaks of diseases such as typhoid fever, encephalitis, or plague may occur under certain conditions and when these vectors are already present in the ecosystem. For example, following flood inundations in tropical areas, ecologi- cal conditions are frequently optimal for mosquito reproduction. In 1963, following Hurricane Flora, 75,000 cases of Plasmodium falciparum, a potentially deadly form of malaria, were recorded in Haiti. This is much higher than the number of infections normally observed.28 Despite a popular disaster myth, if a dis- ease pathogen is not normally present in the affected area, then that particular disease cannot occur in that region despite ideal environmental conditions.29

Special Needs Populations

Philosophers have said that a true measure of a society’s greatness is how it protects those least able to care for themselves. Large cyclonic storms in modern times, such as Hurricanes Katrina and Rita in the U.S. and Typhoon Nari in western Pacific, underscore the necessity to meet the complex challenges of public health and disaster planning for those individuals with special needs by making sure that emergency management plans address this population (see Chapter 8). Studies in the U.S. indicate that up to 19% of the general population is disabled. Often, people with special needs have difficulty receiving and understanding public emergency broadcasts (for example, due to language barriers or physical limitations). They also have difficulty taking protective actions such as moving to a “special needs” shelter or complying with evacuation orders. In addition to persons with physical disabilities, other individuals that belong to the special needs population would include

■ People without access to transportation or who lack financial resources

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

CYC LO N E S/HU R R I C A N E S/T Y P H O O N S ■ 549

■ People who do not speak English or communicate differently (the hearing impaired)

■ Migrant workers, homeless persons, visitors, and tourists ■ People who are in confined facilities (e.g., schools, hospitals,

nursing homes, and prisons) ■ Children

The special needs population is an important but challenging one to include in the disaster planning process. It is clear they will require much in the way of resources during a tropical cyclone disaster. It is essential that emergency managers anticipate the needs of this population and create appropriate plans. These written plans should be practiced and revised through exercises and drills to ensure assistance will be available to the special needs population.

Hospital Mitigation, Preparedness, Response, and Recovery

Preimpact Phase Pre-event planning for a devastating tropical cyclone is essen-

tial for continued hospital operations during and after the dis- aster, and it starts years before it is needed. As previously stated, the public expects that hospitals and the healthcare system will continue treating current patients and also providing care to those seeking emergency medical attention regardless of what disaster has just occurred.12 This means that hospitals, clinics, and medical personnel must be prepared and have planned in advance to deal with the large number of issues associated with a tropical cyclone: loss of electricity, failure to deliver fuel or food, emergency generator failure, tainted municipal drinking water supplies, hospital flooding, and personnel shortages. In addition, planning should include how to feed the hospital’s patients, their families, staff, and the staff ’s families.30

Prior to the topical cyclone season, it is imperative that emer- gency management programs are created to address the needs of the hospital and the community it serves. After plans to support these comprehensive programs have been developed, they must be practiced and modified before the need to use them arises. They should incorporate a list of volunteers that includes medi- cal personnel, environmental workers, and social workers along with their current contact information. Plans should also include an up-to-date list of hospital assets such as ventilators and auto- claves and possible hospital hazards such as liquid oxygen tanks. Additional plan components should address 1) morgue capa- bilities and contingency plans for managing the deceased when the morgue is full or power to cool the area is lost, 2) provi- sion of staff emergency information kits, and 3) memoranda of understanding (MOU) with other hospitals, vendors, and emer- gency medical services for services. In addition, plans should include what pharmaceuticals, if any, to give patients when they are discharged. Issues include whether the hospital gives a week’s supply of essential medications or just discharges patients with a prescription and a list of possibly open pharmacies (Table 33.6).

An often-overlooked point in hospital disaster planning is addressing the needs of vulnerable community members such as the elderly, the infirmed, and those who are ventilator or oxygen dependent. These individuals will frequently use the hospital for shelter and basic care when surrounding infrastructure fails and floodwaters rise. In this situation, social workers are incredibly important. They can compile a list of 24-hour pharmacies, oxy- gen companies that will deliver canisters, and shelters the can

accept patients when they are discharged postevent. Social work- ers often have this information readily available as they manage these issues on a daily basis.

Impact Phase and Immediately Thereafter

Typically, hospitals experience a lull in emergency department visits around the time the tropical cyclone makes landfall and in the storm’s immediate aftermath. Once this respite is over, there is generally a rapid increase in visits to the hospital, mostly for emergency trauma care. In planning for this phase of the disaster, managers should be aware that most patients will not require advanced life support. In fact, medical data from three tropical cyclones affecting Taiwan indicate that only one fifth of the patients needed an ambulance for transportation to the hospital and that 90% of the patients seen for emergency care did not require hospitalization. The most common injuries recorded were soft tissue injures followed by head injuries and orthopedic problems such as sprains and fractures.18

From a staffing perspective, the hospital will need environ- mental crews to assist with cleanup and engineers to assess build- ings for signs of damage. Other staff required to support medical care needs include nurses to staff floor beds; operating room staff to care for trauma victims as well as nondisaster related surgi- cal cases such as appendicitis; intensive care unit personnel; and extra emergency department staff to assist with storm-related injuries.

Postimpact Phase

This phase can last days to years depending on the magnitude of the tropical cyclone and the devastation that it has brought to the area. Hospital effects can vary from minor flooding damage to permanent closure, as was seen in New Orleans after Hurri- cane Katrina. The functional status of hospitals will influence what actions are necessary to return a community to baseline health and medical status. In many countries, disaster response teams exist that respond quickly and offer assistance provided by medical professionals including surgeons, pediatricians, mid- wives, and emergency medicine providers. The assignment of such teams to the disaster area is temporary, deployments lasting a few weeks but generally not for months.

If there is widespread devastation, establishing alternate care sites may be an option. This would help hospitals manage the initial onslaught of victims requesting emergency medical care and allow institutions to more effectively distribute the patient load. “Surge” facilities, staffed by community providers or gov- ernment assets delivering healthcare to patients with nonacute to moderately acute medical needs, should be available for at least a week after the cyclone makes landfall.31 Conversely, if the hos- pital has sustained only minor damage, then the need for a surge facility would be reduced. Planning for this phase is challenging because it is difficult to anticipate how events will unfold after each tropical cyclone. An example is the impact of Hurricane Katrina on the city of New Orleans. There would not have been the devastation that occurred if the levees had not broken and flooded parts of the city.

In this recovery phase, hospitals must be ready to support the community. If there has been a drop in municipal water pressure, water sources must be tested before hospital admin- istrators permit use for drinking and equipment sterilization.30

In the weeks following a devastating tropical cyclone, trauma

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

550 ■ KE L LY R. KL E I N A N D FR A N K FU H-YUA N SH I H

Table 33.6: Personal and Hospital Needs

Personal A week’s supply of prescription medication

An extra pair of glasses or contact lenses plus lens solution

A week’s supply of potable water (2.5–3 L/d - climate dependent)

Full tank of fuel for generators and vehicles

Clothing appropriate to weather conditions

Flashlight

Family evacuation plan with understood rendezvous points

Hospital Considerations

Infrastructure Generators that are biannually tested

Generators in a location that will not flood

Fuel tanks for generators in a location that will not flood and taint the fuel

Fuel tanks located for easy access to refueling

Fuel tanks filled pre-event

Potable water stored

Morgue on the generator circuit

MOU with fuel vendors to ensure resupply

MOU with emergency generator companies to provide support in case of generator failure

Sandbags and lumber for windows and doors

MOU for liquid oxygen supplier

Operations Plan for emergency staffing and postevent cleanup staff

List of essential jobs to include housekeeping, nursing, social work, laboratory technicians, respiratory therapy, cafeteria support staff, and cooks

Accurate contact list for all personnel

MOU with food vendors

MOU with other hospitals in case of evacuation

MOU with transportation vendors

MOU with ventilator companies

Downtime areas for staff and their families

Pharmacy supplies for staff and families

Pharmacy: essential medications for discharged patients

Evacuation plans

Operating room supplies, e.g., autoclaves, sterile instruments

Intensive care operations supplies

and acute care needs gradually evolve to chronic illnesses and psychological needs. Patients with renal failure will need routine dialysis, patients with cancer will require continued chemother- apy or radiation treatments, people will exhaust their supplies of medications and need refills, patients with chronic conditions will suffer acute manifestations, and victims with depression and other debilitating psychological illnesses will present to hospitals requesting assistance. Staffing will be stretched thin and many people might not return to work as their personal life issues will take precedent. Others might move out of the area entirely and seek new employment elsewhere. Within a few weeks to months after the event, the emergency assistance personnel originally dispatched to the disaster zone from outside the area will return to their communities. Those living in the devastated areas must begin rebuilding their community. There are no easy answers or simple templates to support this reconstruction effort. Hard work and communication within the community and with the local and regional governments will facilitate movement towards the goal of return to normalcy.

Pharmaceutical Needs

Part of disaster planning is deciding which medical supplies should be stockpiled for a tropical cyclone and its aftermath.

Many of the lists and recommendations available are based on hearsay and personal experiences. There is a paucity of evidence- based literature available for use by hospitals, clinics or disaster medical teams. Published studies that examined the needs of patients following multiple hurricanes found that wounds, mus- culoskeletal pain, medication refills, upper respiratory infections, rashes, and abdominal complaints were the most common con- ditions in people seeking emergency medical care.32 Analysis of Hurricane Andrew medical treatment data gathered by the U.S. Centers for Disease Control and Prevention demonstrated that 16% of households were unable to obtain prescription drugs for chronic and acute conditions between 3 and 10 days after the hurricane hit Florida and Louisiana.33

Based on this observational literature, there seems to be a need for the following medications in the immediate aftermath of a cyclone: tetanus toxoid, oral and parental antibiotics, hypo- glycemics, cardiac medications, respiratory agents, antiepilep- tics, analgesics, gastrointestinal drugs, and psychotropics.34 Med- ications needed after an event can be modeled after an average community hospital’s normal emergency department pharmacy usage because chronic diseases will still require treatment.35 From this information, medications could be stockpiled for both emer- gency treatment and for medication refills until normal phar- macy services are restored or outside assistance is available.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

CYC LO N E S/HU R R I C A N E S/T Y P H O O N S ■ 551

For catastrophic disasters, which result in a total collapse of a region’s medical infrastructure, the World Health Organization, the High Commissioner for Refugees, UNICEF (United Nations Children’s Fund, formerly United Nations International Chil- dren’s Emergency Fund), the United Nations Population Fund, Médecins Sans Frontières, the International Committee of the Red Cross, and the International Federation of Red Cross and Red Crescent Societies designed the New Emergency Health Kit. It was created to meet the primary health needs of a displaced population of 100,000 people for three months. The kit includes medicines, disposable items, sterilizable instruments, and basic sterilization equipment. The primary unit is intended for use by basic health workers; the supplementary unit is designed for physicians and advanced practitioners and is used to augment the Basic Unit if there are advanced providers available.36

RECOMMENDATIONS FOR FUR THER RESEARCH

Tropical cyclones will have a disproportionally greater impact on areas that are socioeconomically depressed; the worse the poverty, the more devastating will be the disruption to infra- structure, public health, and medical care in that region. In developing countries or poverty stricken areas of wealthy nations, tropical cyclone mortality continues to be significant, with the majority of deaths occurring from storm surge. In situations where the local and regional infrastructure has been severely damaged, morbidity is more evenly distributed throughout the population, causing devastating and long-term consequences to the affected region. Governments and nongovernmental agencies throughout the world have been working diligently to improve mitigation and preparedness efforts in areas prone to these dev- astating storms. These regions seem to be increasing in num- bers and they are experiencing increasingly severe storms. There have been successes in decreasing mortality in many developing and wealthy nations by using early warning systems, improving building codes so structures are better able to withstand tropical cyclone winds, creating storm-safe shelters, and implementing early evacuation in areas at risk from flooding, landslides, and deadly storm surges. These projects work well as long as the gov- ernment remains stable and provides philosophic and economic support to these mitigation projects.

Further studies are needed regarding hospital design and construction techniques to minimize flooding of critical areas and generator failures like those seen in the Houston floods in 2001 following tropical storm Allison. Hospitals must examine where critical patient care areas are located within their facili- ties, develop designs that permit expeditious and safe evacua- tions, and eliminate elements that could make evacuation diffi- cult should it become necessary. Additionally, better studies are needed to create appropriate recommendations for healthcare response teams. Multidisciplinary study groups should investi- gate alternative methods of population evacuation such as the use of boats, buses and trains where feasible.

The process of mitigation, preparation, response, and recov- ery from the effects of a tropical cyclone is a very complex and expensive process for public health and hospitals. These entities are expected to remain open and functioning regardless of the disaster’s intensity. Plans should address infrastructure, staffing, psychological, and community needs. Further research is needed in these areas to ensure these entities remain functional and can provide appropriate services after a cyclone.

REFERENCES

1. Davis I. The planning and maintenance of urban settlements to resist extreme climatic forces. In: Oke T, ed. Urban Clima- tology and its Applications with Special Regard to Tropical Areas. World Climate Programme – Proceedings of the Technical Con- ference, Mexico. Geneva: World Meteorological Organization; 1986:277–312.

2. Holland GJ. Ready reckoner. In: Global Guide to Tropical Cyclone Forecasting. WMO/TC-No. 560, Report No. TCP-31. Geneva: World Meteorological Organization; 1993:Chapter 9.

3. International Federation of Red Cross and Red Crescent Soci- eties. World Disasters Report 1993. Norwell, MS: Kluwer Aca- demic Publishers; 1993.

4. Alvarez R. Tropical cyclone. In: Ingleton J, ed. Natural Disaster Management. Leicester: Tudor Rose; 1999:34–36.

5. Gray W. General characteristics of tropical cyclones. In: Pielke R, Pielke R, eds. Storms. Vol 1. London: Routledge; 2000.

6. McGuyire B, Mason I, Kilburn C. Natural Hazards and Environ- mental Change. London: Arnold; 2002.

7. Graumann A, Houston T, Lawrimore J, et al. Hurricane Kat- rina: A Climatological Perspective. Preliminary report NOAA’s National Climatic Data Center, Technical Report 2005–01.

8. Burkle FM Jr, Rupp G. Hurricane Katrina: disasters keep us honest (Commentary). Monday Develop. September 26, 2005; 23(17):5.

9. Noji EK. The nature of disaster: general characteristics and public health effects. In Noji EK ed. The Public Health Conse- quences of Disasters. New York: Oxford University Press; 1997:3– 20.

10. Pelling M. The Vulnerability of Cities: Natural Disasters and Social Resilience. London: Earth Scan Publications; 2003.

11. Norris FH. Psychosocial Consequences of Natural Disasters in Developing Countries: What Does Past Research Tell Us About the Potential Effects of the 2004 Tsunami? Available at: http://www.redmh.org/research/specialized/tsunami summary. pdf. Accessed November 21, 2008.

12. Pan American Health Organization. Mitigation of Disasters in Health Facilities: Evaluation and Reduction of Physical and Func- tional Vulnerability. Volume II: Administrative Issues. Washing- ton, DC: Pan American Health Organization. 1993.

13. Alexander D. Natural Disasters. New York: Chapman & Hall, Inc.; 1993.

14. Perry RW. Population evacuation in volcanic eruptions, floods and nuclear power plant accidents: some elementary compar- isons. J Commun Psychol. 1983;11(1):36–47.

15. Sternberg E, Lee GC, Huard D: Counting crises: US hospital evacuations, 1971–1999. Preshosp Disast Med. 2004;19(2):150– 157.

16. Taaffe K, Johnson M, Stenmann D. Improving hospital evacu- ation planning using simulation. In: Perrone LF, Wieland FP, Liu J Lawson BG, Nicol DM, Fujimotos RM, eds. Proceedings for the 2006 Winter Simulation Conference, Monterey, California, 2006.

17. Andress K, Downey E, Schultz CH: Implications of hospital evac- uation after Hurricane Rita. Acad Emerg Med. 2007;14(Suppl 1):S190.

18. Shih, FY. Risk Analysis of Disasters and Preventive Strategies Implication in Taiwan. Doctoral dissertation. 2007.

19. Lindell MK, Prater CS. A hurricane evacuation management decision support system (EMDSS). Nat Hazards. 2007;40:627– 634.

20. Sorensen JH, Vogt BM, Mileti DS. Evacuation: An Assessment of Planning and Research. Report prepared for US Federal Emer- gency Management Agency, RR-9. 1987.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

552 ■ KE L LY R. KL E I N A N D FR A N K FU H-YUA N SH I H

21. Vogt BM, Sorensen JH. Evacuation Research: A Reassessment, ORNL/TM-11908. Oak Ridge, TN: Oak Ridge National Labo- ratory; 1992.

22. Malilay J. Tropical cyclones. In: Noji E, ed. The Public Health Consequences of Disasters. New York: Oxford University Press; 1997:207–227.

23. Wisner B, Blaikie P, Cannon T, Davis I. At Risk: Natural Haz- ards, People’s Vulnerability and Disasters. 2nd ed. New York: Routledge; 1994:243–273.

24. NOAA website. Available at: http://www.srh.noaa.gov/mob/ 0805Katrina/. Accessed November 21, 2008.

25. Reuters website. Talking Point: Why is Haiti so prone to disaster? Available at: http://www.alertnet.org/thefacts/reliefresources/ 109655418734.htm. Accessed November 21, 2008.

26. North CS, King RV, Polatin P, et al. Psychiatric illness among transported hurricane evacuees: acute phase findings in a large receiving shelter site. Psychiatric Ann. 2008;38(2):104– 114.

27. Rodriguez SR, Tocco JS, Mallonee S, Smithee L, Cathey T, Bradley K. Rapid needs assessment of Hurricane Katrina evacuees – Oklahoma,September 2005. Prehosp Disaster Med. 2006;21(6):390–295.

28. Toole MJ. Communicable disease and disease control. In: Noji EK, eds. The Public Health Consequences of Disasters. New York: Oxford University Press; 1997:79–100.

29. Committee on Research Priorities for Earth Science and Pub- lic Health and National Research Council. Earth Materials and Health: Research Priorities for Earth Science and Public Health. 1st ed. Washington, DC: National Academies Press; 2007.

30. Klein KR, Rosenthal MS, Klausner HA. Blackout 2003: prepared- ness and lessons learned from the perspectives of four hospitals. Prehosp Disaster Med. 2005;20(5):343–349.

31. Meredith JT. Hurricanes. In: Hogan DE, Burstein JL, eds. Dis- aster Medicine. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2007:205–213.

32. Nufer KE, Wilson-Ramirez G. A comparison of patient needs fol- lowing two hurricanes. Prehosp Disaster Med. 2004;19(1):146– 149.

33. Rapid health needs assessment following Hurricane Andrew – Florida and Louisiana,1992. MMWR. 1992;41:687–688.

34. Sepehri G, Meimandi MS: The pattern of drug prescrip- tion and utilization among Bam residents during the first six months after the 2003 Bam Earthquake. Prehosp Disaster Med. 2006;21(6):396–402.

35. Rosenthal MS, Klein K, Cowling K, Grzybowski M, Dunne R. Disaster modeling: medication resources required for disaster team response. Prehosp Disaster Med. 2005;20(5):309–315.

36. WHO website: The New Emergency Health Kit. Available at: http://www.who.int/medicinedocs/pdf/whozip31e/whozip31e. pdf. Accessed November 21, 2008.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Tornadoes

Arthur G. Wallace Jr.

OVERVIEW

Tornadoes occur worldwide, with the greatest incidence in North America.1 Australia ranks second in incidence to the United States while countries such as Italy, New Zealand, and the United Kingdom rival this incidence if it is expressed as tornadoes per area rather than in absolute numbers. Tornadoes sometimes occur with very little or no advanced warning, causing consid- erable structural damage, traumatic injury, and death. Although meteorologists are capable of advanced forecasting of weather conditions that favor tornado development, the exact touchdown location and ground track of a tornado is not yet predictable. Even with advanced warning, there are still the challenges of notifying the population at risk and communicating the correct response to minimize injury or death.

Historical data reveal similar patterns of injury and death in all tornado disasters.1 Community healthcare systems face significant challenges in attempting to manage the influx of tor- nado casualties. The sudden surge in patient volume, added to the routine daily challenge of providing medical care in hospi- tals functioning near 100% capacity, can overwhelm an already stressed healthcare system.

The track of destruction and injury produced by tornadoes is relatively small in proportion to the exposed population. The occurrence of a “worse-case scenario,” the impact of a large tornado on a densely populated event located at a fairgrounds or sports arena, could produce significant casualty numbers. Fort Worth, Texas had such a near miss when a sudden massive hailstorm caught 10,000 spectators in an open area.2 A sudden onset storm with little advance warning over a populated location could create a mass casualty event that exceeds the capabilities of any medical response system. This chapter will review risk factors for injury and death, injury patterns, and mass casualty scenarios by using modeling and real event reports.

STATE OF THE ART

Terms

Understanding that many terms necessary for discussing torna- does are somewhat technical and not universally understood, it

is prudent to define and explain some of these expressions so their meaning is clear. In addition, as current weather broadcasts and warnings are becoming widely available via television and wireless services, there are certain terms that should be familiar to emergency system managers, planners, and responders.

Super cell – A thunderstorm with a persistent rotating updraft. Super cells are rare, but are responsible for a remarkably high percentage of severe weather events – especially tornadoes, extremely large hail, and damaging straight-line winds.3

Mesocyclone – A storm-scale region of rotation, typically approximately 2–6 miles (3.3–10 km) in diameter and often found in the right rear flank of a super cell.3

Hook (or Hook Echo) – A radar reflective pattern char- acterized by a hook-shaped extension of a thunderstorm echo, usually in the right-rear part of the storm (relative to its direction of motion). A hook often is associated with a mesocyclone and indicates favorable conditions for tornado development.3

Fujita Scale

Dr. T. Theodore Fujita developed a tornado rating scale that has been in use for the past 30 years (Table 34.1).4 Essentially, the scale rates tornadoes on wind speed and the most intense damage cre- ated in the path of the storm based on direct observation. By Dr. Fujita’s own admission, this scale is somewhat arbitrary and sub- ject to observational bias. Furthermore, no objective measure- ments exist comparing wind speed and damage relationships.5

Beginning in February 2007, the United States began using the enhanced Fujita scale (EF Scale) to rate tornadoes (Table 34.2).6

This new scale takes into account 28 different indicators evaluat- ing the impact of a storm on buildings constructed of gradually stronger material and applies weighted values to the degree of structural damage. This method of rating a tornado is performed after the storm has passed and represents a set of wind estimates (not measurements) based on an assessment of structural dam- age. Although it is not a contemporaneous reading, the esti- mated wind speed is used in analysis of the predicted morbidity and mortality as they relate to storm strength. A significant tor- nado is considered one rated at F2 or greater.7 Some researchers use the F2 rating and aforementioned criteria as indicators for

553 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

554 ■ ART H U R G. WA L L AC E JR.

Table 34.1: Fujita Tornado Damage Scale∗

Scale Wind Estimate (mph) Typical Damage

F0 <73 Light damage. Some damage to chimneys; branches broken off trees; shallow-rooted trees pushed over; signboards damaged.

F1 73–112 Moderate damage. Peels surface off roofs; mobile homes pushed off foundations or overturned; moving autos blown off roads.

F2 113–157 Considerable damage. Roofs torn off frame houses; mobile homes demolished; boxcars overturned; large trees snapped or uprooted; light-object missiles generated; cars lifted off ground.

F3 158–206 Severe damage. Roofs and some walls torn off well-constructed houses; trains overturned; most trees in forest uprooted; heavy cars lifted off the ground and thrown.

F4 207–260 Devastating damage. Well-constructed houses leveled; structures with weak foundations blown away some distance; cars thrown and large missiles generated.

F5 261–318 Incredible damage. Strong frame houses leveled off foundations and swept away; automobile-sized missiles fly through the air in excess of 100 m (109 yards); trees debarked; incredible phenomena will occur.

∗ Developed in 1971 by T. Theodore Fujita.4

Table 34.2: Enhanced Fujita Scale 20066

Little Damage

Minor Damage

Roof Gone

Walls Collapse

Blown Down

Blown Away

Damage f scale

Windspeed F scale F0

f 0–

– – f0

f 2

f5f5f5f5

f5 f5 f5

17 m/s 32 50 70 92 116 142

3192612071581137340 mph

To convert f scale into F scale, add the appropriate number

Weak Outbuilding

Strong Outbuilding

Weak Framehouse

Strong Framehouse

Brick Structure

Concrete Building

–3

–2

–1

The Fujita tornado scale (F scale) pegged to damage-causing windspeeds. The extent of damage expressed by the damage scale (f scale) varies with both windspeed and the strength of structures.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TO R NA D O E S ■ 555

Figure 34.1. Funnel cloud in Ardmore, Oklahoma, 1985.9 See color plate.

examining mortality data.5 Tom Grazulis, an accomplished mete- orologist and tornado researcher, considers any tornado that results in death as a significant tornado.8 If future development of tornado analysis could generate a product that described tor- nado strength over a storm track in real time, it would be a significant response-planning asset for emergency planners.

Tornado Epidemiology

Tornadoes are generated when warm moist air from the Gulf of Mexico, moving as a warm front, collides with a cold front traveling down from the Northwest. This creates an unstable environment where moist warm air rises rapidly through colder dry air. As the warm air rises with increasing intensity, thunder- storms are formed. The increase in energy potential within these storms sometimes results in the development of super cells. As super cells continue to grow, a rotation of wind begins within the storm, creating a mesocyclone. As the mesocyclone spins, clouds descend toward the earth forming a funnel-shaped struc- ture with a visible rotation (Figure 34.1).9 The rotating funnel cloud is officially called a tornado when it makes contact with the ground (Figure 34.2).10

The United States has an average of 1,000 tornadoes each year. Ten percent of these tornadoes are rated at F2 or greater. Violent tornadoes, rated F4 or F5, only occur 2% of the time but account for 67% of tornado-associated deaths.5

The annual occurrence of tornadoes appears to be increas- ing (Figure 34.3).11 The increasing incidence of tornadoes may simply reflect advances in detection rather than a true increase in events. In addition, more frequent observation of storms that have gone unreported in the earlier years might explain this change. Perhaps more important than the number of storms is where they occur. Population density in relation to storm occurrence may be more predictive of potential disaster-creating events.12 With a population shift toward the southeast coastline and a subsequent population reduction in parts of the Great Plains, there may be a shifting of tornado hazard areas.12 A geo- graphical area known as “Tornado Alley” that is infamous for 25% of significant tornadoes had only 9% of major killer storms from 1980–2000.5

Significant tornadoes, when rated by storm strength, occur most frequently in high-prevalence areas; however, when tor- nado significance is rated by associated deaths, both killer torna- does (one to seven fatalities) and major killer tornadoes (eight

or more fatalities) have consistently occurred more frequently outside of the high-prevalence “Tornado Alley” area over the 54 years between 1950 and 2004 (Figure 34.4).13 In Figure 34.4, note that the “Tornado Alley” area lies just to the left of the significant tornado regions. Reviewing National Climate Data Center records for injury and death related to F2 or higher tor- nadoes from 2001 through September 2006 revealed only 32% of reported injuries and 35% of deaths were in high-occurrence areas (Figure 34.5).14,15 Possible explanations include increased population awareness and response to tornado warnings in high- occurrence areas compared with less awareness and less protec- tive response behavior in low-occurrence locations.5 Along with a direct storm strike in a high-population area, another poten- tial catastrophic event would be multiple storms similar to the “Super Outbreak” event in 1974. Over a 2-day period, 148 tor- nadoes affected 13 states causing 330 deaths and 5,484 injuries (Figure 34.6).7

Risk of Injury and Death

Although a tornado is inherently dangerous, it only becomes a serious risk if it threatens a populated area with damage and injury. The vulnerability of the affected population defines the risk of the storm. Many variables characterize a given popula- tion’s vulnerability. Advanced warning and the subsequent pro- tective actions taken by individuals affect health outcomes during the preimpact and impact phase of the storm. Tornado warnings issued by the National Weather Service provide, on average, 11 minutes advanced notice. This would require an action plan to seek appropriate shelter immediately.

Risk of injury or death has been associated with storm strength, victim age, type of material used for dwelling construc- tion, shelter location, income level, and time of day.16 Elderly persons may have reduced physical response capability, impaired sensory function, or reduced access to appropriate shelter. Pre- fabricated and frame houses may not provide protection, evi- denced by the fact that a significant number of deaths and injuries occur in individuals sheltering within these structures. Seeking shelter in a basement has been associated with injury and death if the house is shifted off the foundation or the basement walls collapse on victims.17–20 Storms after dark may not be visible or may occur when victims are asleep and unaware or unable to hear a warning siren. People with lower incomes may reside in dwellings that cannot withstand high wind stress or in areas of a community without public warning sirens. Locations at

Figure 34.2. Tornado in Mayfield, Oklahoma, 1977.10 See color plate.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

556 ■ ART H U R G. WA L L AC E JR.

Figure 34.3. Incidence of tornadoes in the U.S.11

significant risk included schools, churches, and restaurants where multiple potential victims are present at the time of a direct tor- nado strike.

Most tornado deaths occur at the time of impact. The most common mechanisms of death result from rapid movement of victims through the air, accelerated by storm winds, who then strike a stationary object. The reverse is also true; wind-driven projectiles can strike people or collapsing structures can crush them.21 Traumatic injuries to the head, thorax, and abdominal organs are the most common causes of immediate mortality in tornado victims.21 According to National Climatic Data Center data from 2001–September 2006, overall mortality rates range from 5% to 10% of all those injured. Hospital mortality rates are significantly lower. The overall mortality trend from 1940 to 2005 is shown in Figure 34.7.22

At first glance, it appears one could make the assumption that the emergency medical system is performing adequately and no further critical review is necessary. The overall tornado

Figure 34.4. Relative frequency of killer tornado events around “Tor- nado Alley.”13

mortality is decreasing and is accompanied by an even smaller hospital mortality rate. The low overall mortality rate may not, however, adequately reflect trauma system performance. This rate becomes deceptively diluted because it is analyzed as a per- centage of total injuries, regardless of how minor the wounds are. By considering even minor wounds in the calculation, this inflates the size of the denominator and makes the mortality rate appear smaller. Perhaps a better marker is the critical mortality rate.23 This has been explored in the trauma literature and exam- ines mortality rates only in the victims with significant injuries, defined as an Injury Severity Score greater than 15. It does not include those with minor injuries. Using the critical mortality rate, a more accurate assessment of system performance is pos- sible. One such measurement is time to surgical consultation, as a delay in surgical referral has been associated with preventable trauma deaths estimated to range from 2% to 50%.25 Another situation that can generate an increased critical mortality rate is when a tornado strike produces large numbers of patients who seek medical treatment simultaneously. Other factors that can influence this rate are reviewed below.

Injury Patterns

Injury patterns can be categorized by the time at which they occur in relation to the storm. This approach may affect treat- ment strategies. Preimpact injuries can include traffic collisions as people flee ahead of the storm, and falls that occur while running down stairs into basements or storm shelters. Impact phase injuries result from victims being thrown against objects by strong storm winds or being struck by wind blown projec- tiles. The latter phenomenon can predispose to wound contami- nation by organisms that become airborne in high-wind events. Wound infection resulting in septicemia has been responsible for some hospital mortality.26 Recommendations have been made that wounds occurring during this phase should be considered for delayed primary closure. For unclear reasons, this approach is often avoided in clinical practice.21,27 Postimpact injuries include puncture wounds and lacerations associated with debris

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TO R NA D O E S ■ 557

Figure 34.5. 2001–2004 Killer tornadoes.15

Figure 34.6. Super outbreak April 3–4, 1974.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

558 ■ ART H U R G. WA L L AC E JR.

Table 34.3: Mechanisms of Blast Injury∗

Category Characteristics Types of Injuries

Secondary Results from flying debris and bomb fragments – Penetrating ballistic (fragmentation) or blunt injuries

– Eye penetration (can be occult)

Tertiary Results from individuals being thrown by the blast wind – Fracture and traumatic amputation

– Closed and open brain injury

Quaternary – All explosion-related injuries, illnesses, or diseases not due to primary, secondary, or tertiary mechanisms

– Includes exacerbation or complications of existing conditions

– Burns (flash, partial, and full thickness)

– Crush injuries

– Closed and open brain injuries

– Asthma, COPD, or other breathing problems from dust, smoke, or toxic fumes

– Angina

– Hyperglycemia, hypertension

Notes 1. The Primary Category is not included in the chart as it does not apply to tornadoes. 2. Any body part may be affected by the secondary, tertiary, or quaternary mechanisms. COPD, chronic obstructive pulmonary disease. ∗ Department of Health and Human Services.24

removal and electrocution while individuals are working around downed power lines erroneously thought to be inactive.

Approximately 50% of injuries seen in hospital emergency care areas are soft tissue wounds and include lacerations, contu- sions, and punctures.21 Fractures occur in 30% of victims and are the most common cause for hospital admission.21 Head injuries including extra- and intracranial trauma, accounts for 7% of injuries.21

Figure 34.7. U.S. tornado fatality rates.22

Injury patterns caused by tornadoes can be contrasted and compared to injuries from blasts or explosions. With the excep- tion of primary blast injury, the other categories are similar in terms of mechanisms of injury and event-associated complica- tions. Treatment options for blast injuries are discussed in more detail in Chapter 26. Table 34.3 outlines blast categories; tor- nado injuries are similar.24 One difference between blast and tornado injury patterns may be the body part affected, because

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TO R NA D O E S ■ 559

blasts typically occur without warning, while tornado victims may become aware of the approaching threat at the last moment and take a protective posture that could influence the injury pattern.

Immediate Medical Considerations

Prehospital Impact Tornadoes have both seasonal and geographical variability.1

Awareness by the local population at high risk provides a starting point to begin understanding what vulnerabilities exist. Because tornadoes usually strike with little advanced warning; the medi- cal system response will consist of whatever the local capabilities are at that time. State and federal assistance, if needed, may not be readily available in the early phase of the response. If the closest medical center is in a rural or suburban area, established trans- fer or referral arrangements with a trauma center should be in place for those victims requiring such services. If there are mul- tiple medical facilities in the response area, a coordinating center should assist with victim distribution to avoid overloading any particular facility. This approach was effective in the April 8, 1998 F5 tornado response in the U.S. state of Alabama.28 These coor- dinating centers take the form of medical emergency response centers in association with metropolitan medical response sys- tems and become activated during events requiring a coordinated response.

A strategy to deal with the initial victim surge is the “First- Wave” protocol reviewed by Auf der Heide.29 This is a casualty distribution plan in which hospitals are identified by capability to treat victims of a given severity category (immediate, delayed, minor) and includes how many patients in each category they can accommodate. Initial emergency medical services (EMS) transports could then attempt even distribution of the victims to appropriate medical centers, avoiding overloading one facility as much as possible. The impact of an incident management sys- tem on resource distribution and casualty convergence cannot be overemphasized. It is often not just a paucity of resources, but their maldistribution that adversely impacts the delivery of medical care. An effective incident management system can sub- stantially improve utilization of existing resources.

Triage Overtriage is assigning a patient or victim to a higher level

of urgency than is actually necessary. Undertriage is underesti- mating the true nature of the victim’s condition and assigning an inappropriately lower acuity status. Acceptable rates exist for over- and undertriage. Triage officers try to minimize underes- timating a patient’s medical condition, but they also realize that overtriage may create an additional burden to the system. Dur- ing routine medical operations, overtriage happens daily and is usually acceptable because it is a temporary situation and has no negative impact on outcomes. Although unproven, in a disaster or mass casualty situation, overtriage could theoret- ically reduce ability to rapidly identify and prioritize victims that require immediate treatment. If the majority of victims are transported either to the closest facility or to a tertiary referral center, those centers will become overburdened to the point that system efficiency deteriorates. Arrival of victims from a mass casualty event usually occurs in a distinct pattern. The first wave is characterized by those that self-refer and transport themselves to a hospital, usually the closest one. As the first victims arrive, the triage team may begin assigning emergency beds, ordering

Table 34.4: Recommendations for Hospital Workers in Preparation for and Response to an Imminent Tornado Event

1. Protect yourself, staff, and current patients (inside hallways away from windows and glass)

2. Consider a safe alternative triage and treatment site

3. Assess trauma team availability, regional referral support, and local community support agencies

4. Appropriately scrutinize triage decisions on early ambulatory arrivals, anticipating that later arrivals by ambulance may be more critical

5. Inspect wounds for foreign bodies and contamination

6. Provide tetanus prophylaxis if indicated

7. Limit laboratory investigations to those critical to manage patients from the acute event

radiographic and laboratory studies, and requesting consulta- tions from specialty services. (This first wave references ambula- tory victims, as opposed to the “First-Wave” protocol mentioned previously referring to EMS-transported victims.)

The second wave of patients usually arrives approximately an hour after the event begins. These victims are usually transported by EMS and may be the more seriously injured than the self-presenting patients that arrived earlier. It has been postulated that the consumption of medical resources by victims with less severe conditions that arrive in the first wave via self- transport may negatively impact the care of those more seriously ill that arrive later via the EMS system. In addition, it has been hypothesized that overtriage may have a direct linear relationship to critical mortality rate.30 No studies published to date have proven either of these assertions. In fact, the data demonstrate, at best, an association of higher mortality with overtriage. Data also support the opposite interpretation that severe events with high mortality rates induce overtriage behavior in responders. An analysis of the London subway bombing in July 2005 reported significant overtriage but no increase in critical mortality.31

Resolution of this controversy must await further studies. The hospital’s ability to increase its surge capacity to care for

this patient population may be affected by the rate of victim entry into the trauma care area, rather than by the number of avail- able beds or level of emergency department staffing.23 Results of modeling using an actual event reenacted at a trauma facility examined casualty load and its impact on surge capacity. Using the optimal level of care, defined as the resources committed to a single trauma patient on a normal working day, alterations in care levels were noted to occur with increasing numbers of critical victims.32

Hospital Impact In the situation in which a tornado strike at or near a hos-

pital is imminent, personnel must respond quickly. Although no evidence-based recommendations exist, the actions listed in Table 34.4 are reasonable and prudent. These will help protect personnel and prepare the hospital for incoming casualties.

Admission rates of tornado victims are typically less than 25%.21 Estimating the overall number of victims a facility can manage would depend on the resources available, knowing that serious injuries occur 10%–15% of the time.23 For example, if three trauma teams were available, 30 victims could theoretically

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

560 ■ ART H U R G. WA L L AC E JR.

be processed without compromising the level of care for those needing urgent intervention.

Recommendations for Further Research

Special Medical Needs Populations Tornado damage to a retirement center or nursing home

may force evacuation of the residents. If arrangements for relo- cation are delayed, even without acute injuries from the tornado event, this population of individuals with special medical needs may be referred to the emergency response system and be trans- ported to local medical facilities. In addition, special medical needs populations may not be permitted in general community shelters. Alternatives are needed to improve the management of these displaced individuals (see Chapter 8). Research is necessary to develop best practices that guide strategies for designation and management of off-site facilities that can rapidly be converted to temporary shelters for this group of victims.

Rapidly Establishing Alternate Care Sites If a medical facility is threatened or damaged by a storm, a

number of issues arise. Top concerns include protection of inpa- tients should the decision be made to shelter in place, evacuation of patients to another facility, and establishing an alternate care site for victim reception. Furthermore, if a densely populated area is struck by a significant tornado generating an exceedingly high number of injured victims, local healthcare facilities could be overwhelmed. In these scenarios establishment of an alternate triage and treatment area may serve as one remedy to system overload. Research examining this issue would be useful.

Triage Concepts Earlier discussion reviewed issues related to the triage pro-

cess. Questions remain regarding the effectiveness of triage methodology and its impact on patient outcomes, including whether overtriage adversely affects critical mortality rates and surge capacity. From past accounts, it appears that less severely injured tornado victims arrive early, are frequently ambulatory, and usually not transported by EMS. The majority of injuries sus- tained by this group will not be life threatening. Those requiring admission, typically less than 25% of all victims, arrive later and are usually transported by EMS. Could triage accuracy and speed improve if those arriving physiologically intact and ambulatory were assigned to a delayed care category? The reliability of the motor component in the Glasgow Coma Scale for predicting increased risk of morbidity and mortality from trauma has been demonstrated in previous studies.33 Military and civilian pro- grams have used this concept in field triage strategies. Such algo- rithms as Simple Triage And Rapid Treatment (START) 34 and the triage concept taught in the Advanced Disaster Life Support course direct victims to specific acuity categories based, in part, on their response to verbal commands.35 Using such methods that are easily taught and applied could be helpful in improving triage accuracy and the assignment of only appropriate high- risk victims into the treatment areas first. Although not without potential error, comparison of outcomes within these current methods when applied to tornado victims would be helpful.

Injury Classification Currently there is no standardized classification scheme for

tornado injuries. Soft tissue injury, muscle strain, orthopedic injury, trunk, and head injuries are all broad categories men-

tioned. Using a system similar to that for blast injury categoriza- tion may help to corroborate other reviews of tornado injury patterns. Timing of injury occurrence as pre-event, intraevent, or postevent may be important predictors of wound contami- nation and infectious complications. Investigation of such issues may help define whether primary or delayed wound closure is appropriate.

Standards of Care Typical daily trauma care utilizes significant resources and

personnel for individual patients. In a mass casualty event, the system may need to operate in a population-based mode. A great deal of controversy currently exists regarding how this should be achieved, including debate on altering care. Further investigation is needed on how standard of care issues are addressed, including development of templates that can recommend a time stratifi- cation strategy for medical care based on available resources, referral sources, injury patterns, and patient volume.

REFERENCES

1. Greenough G, McGee M, Bernard S. The potential impacts of climate variability and change on health impacts of extreme weather events in the United States. Environ Health Perspect. 2001;109(2):191–199.

2. Calianese E, Jordan J. The Mayfest High-Precipitation Super- cell of 5 May 1995 – A Case Study NWS Forecast Office Lub- bock, Texas. Available at http://ams.confex.com/ams/pdfpapers/ 47099.pdf. Accessed January 9, 2009.

3. Branick M. A Comprehensive Glossary of Weather Terms for Storm Spotters. NOAA Technical Memorandum NWS SR- 145. 2006. Available at: http://www.srh.noaa.gov/oun/severewx/ glossary2.php. Accessed November 17, 2008.

4. University of Wisconsin-Madison’s (UW) Cooperate Insti- tute for Meteorological Satellite Studies (CIMSS) Available at: http://cimss.ssec.wisc.edu/oakfield/Fscale.htm. Accessed Jan- uary 9, 2009.

5. Concannon P, Brooks H, Doswell C. Climatological Risk of Strong and Violent Tornadoes in the United States. Second Con- ference on Environmental Applications. American Meteorlogi- cal Society. Long Beach, CA, January, 2000; paper 9.4.

6. Edwards R. Storm Prediction Center. Available at: http:// www.spc.noaa.gov/efscale./. Accessed January 9, 2009.

7. Edwards R. Storm Prediction Center. Significant Torna- does. 2006. Available at: http://www.spc.noaa.gov/faq/tornado/. Accessed November 17, 2008.

8. Grazulis T. The Tornado: Natures Ultimate Windstorm. Norman, OK: University of Oklahoma Press; 2001.

9. Funnel Cloud (1985) Ardmore, Oklahoma. NOAA photo library. NOAA Central Library; OAR/ERL/National Severe Storms Laboratory. Available at: http://www.photolib.noaa.gov/ htmls/nssl0132.htm. Accessed December 4, 2008.

10. Tornado (1977) Mayfield, Oklahoma. NOAA photo library. NOAA Central Library; OAR /ERL/National Severe Storms Laboratory. Available at: http://www.photolib.noaa.gov/htmls/ nssl0107.htm. Accessed December 4, 2008.

11. Tornado occurrence. Available at: http://www.hprcc.unl.edu/ nebraska/US-tornadoes-1950-to-2006-bar.html. Accessed Dec- ember 4, 2008.

12. Boruff B, Easoz J, Jones S. Tornado hazards in the United States. Climate Res. 2003;24:103–107.

13. Ashley W. Meteorology Program, Department of Geography, Northern Illinois University. Available at: http://www.niu.edu/

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TO R NA D O E S ■ 561

PubAffairs/photos/opt/releases/tornado-alley-large.jpg. Acces- sed December 4, 2008.

14. National Climatic Data Center. Available at: http://www4.ncdc. noaa.gov/cgi-win/wwcgi.dll?wwevent∼storms. This is an inter- active site the reader may enter a query for a particular year then specify tornado as the event and the injury and death statistics will be displayed for any or all states. Accessed January 9, 2009.

15. Killer tornadoes. Available at: http://www.hprcc.unl.edu/ nebraska/U S SEVERE.html. Accessed November 17, 2008.

16. Simmons K, Sutter D. Protection from nature’s fury: an analysis of fatalities and injuries from F5 tornadoes. Nat Hazards Rev. 2005;6(2):82–87.

17. Centers for Disease Control. Tornado associated fatalities– Arkansas 1997, MMWR. 1997;46(19):412–416.

18. Centers for Disease Control. Texas disaster–Texas, May 1997. MMWR. 1997;46(45):1069–1072.

19. Oho Y. Risk factors for death in the 8 April 1998 Alabama tor- nado. Quick Response Report # 145. Boulder, CO: Natural Haz- ards Response Applications Center. 2002. Available at: http:// www.colorado.edu/hazards/research/qr/qr145/qr145.html. Accessed January 9, 2009.

20. Centers for Disease Control. Tornado disaster –Illinois 1990. MMWR. 1991;40(2):33–36.

21. Bohonos J, Hogan D. The medical impact of tornadoes in North America. J Emerg Med. 1999;17(1):67–73.

22. Available at: http://www.hprcc.unl.edu/nebraska/us-tornado- deaths1940-2005.html. Accessed December 4, 2008.

23. Hirshberg A, Holcomb J, Mattox K. Hospital trauma care in multiple-casualty incidents: a critical view. Ann Emerg Med. 2001;37(6):647–652.

24. Centers for Disease Control and Prevention. Available at: http://www.bt.cdc.gov/masscasualties/explosions.asp. Accessed December 4, 2008.

25. Gruen R, Jurkovich G, McIntyre L. Patterns of errors contribut- ing to trauma mortality: lessons learned from 2594 deaths. Ann Surg. 2006;244(3):371–380.

26. Millie M, Senkowski C, Stuart L. Tornado disaster in rural Geor- gia: triage response, injury patterns, lessons learned. Am Surg. 2000;66(3):223–228.

27. May B, Hogan D, Feighnor K. Impact of a tornado on a community hospital. J Am Osteopath Assoc. 2002;102:225– 228.

28. May A, McGwin G Jr, Lancaster L. The April 8,1998 Tornado assessment of the trauma system response and the resulting injuries. J Trauma. 2000;48(4):666–672.

29. Auf der Heide E. Disaster Response: Principles of Preparation and Coordination. St. Louis: C.V. Mosby; 1999:188–192.

30. Frykberg ER. Disaster Planning and Management Princi- ples. Available at: http://www.facs.org/education/congress2002/ gs30frykberg.pdf. Accessed December 4, 2008.

31. Aylwin C, Konig T, Brennan N, et al. Reduction in critical mor- tality in urban mass casualty incidents: analysis of triage, surge, and resource use after the London bombings on July 7, 2005. Lancet. 2006;368(9554):2219–2225.

32. Hirshberg A, Scott B, Granchi T. How does casualty load affect trauma care in urban bombing incidents? A quantitative analy- sis. J Trauma. 2005;58(4):686–695.

33. Meredith W, Rutledge R, Hansen A. Field triage of trauma patients based upon the ability to follow commands: a study in 29,573 injured patients. J Trauma. 1995;38(1):129–135.

34. Super G, Groth S, Hook R, et al. START: Simple Triage and Rapid Treatment Plan. Newport Beach, CA: Hoag Memorial Presbyterian Hospital; 1994.

35. American Medical Association MASS Triage. Advanced Disaster Life Support: Provider Manual. Version 2.0. Chicago, IL: Ameri- can Medical Association; 2004.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Earthquakes

Carl H. Schultz and Solisis Deynes

OVERVIEW

Disasters have posed a significant threat to human lives and prop- erty throughout history. Such events include hurricanes, floods, tornadoes, tsunamis, and earthquakes, among others. During the past 40 years, disasters have caused more than 3 million deaths worldwide, including more than 1 million deaths from seismic events.1–3 Earthquakes are considered one of the most destruc- tive disasters. An average of 16 earthquakes leading to death occur throughout the world each year with many more leading to injury and property damage.2 The Hanshin-Awaji earthquake in Japan produced at least 100 billion USD in damage and killed more than 6,000 people.6 The 1994 Northridge earthquake in Cali- fornia caused an estimated 20–30 billion USD in damages, and preliminary damage estimates for Hawaii’s 2006 temblor reached 100 million USD.4,5 These events occurred in nations that have sophisticated, modern seismic building codes. Outcomes can be even more destructive in less developed countries (Table 35.1).

Society’s continued vulnerability to the devastating effects of earthquakes is due to several factors. Earthquakes are within the category of sudden-impact disasters that strike quickly and without warning, making mitigation and evacuation efforts dif- ficult. The quantity of property damage, loss of life, disruption of economic activity, and interference in the provision of important services associated with the effects of an earthquake vary depend- ing on its magnitude and on the degree of earthquake prepared- ness and mitigation measures implemented in the region affected by the temblor.7,8 Other factors influencing the severity of the earthquake’s impact include the day of the week, time of day, population density, location, distance from the epicenter (to an extent), local geological conditions, and building design.9 Inad- equate building materials, structural design deficiencies, and the absence of laws regulating building construction will increase susceptibility to seismic damage. In addition, earthquakes can cause long-term disruption to transportation, communication, and financial infrastructures.

Worldwide demographics also play a role. Many large pop- ulation concentrations exist along major fault lines. These pop- ulations are at higher risk of earthquake-related morbidity and mortality.10 Despite this high risk, population density continues

to increase in many of these regions, exacerbating the potential for future injuries and deaths after a seismic event.

STATE OF THE ART

Earthquake Characteristics

To understand the issues involved with managing the earthquake threat, it is necessary to explore the basic concepts related to seis- mic events in some detail. Several theories exist that attempt to explain earthquake behavior. The concept most widely accepted by seismologists is the Tectonics Plate Theory. This theory is based on the structure of the earth’s crust, and asserts that in the initial formation of continents, all of the earth’s land mass was aggregated in a single unit. This unit subsequently fragmented, and the fragmented sections known as tectonic plates, began moving against each other (Figure 35.1).8 These land sections remain in constant motion. Where the edges of these tectonic plates meet is referred to as a major fault line. Surrounding the major faults are minor ones that also give rise to seismic

Table 35.1: Earthquake Mortality

Year Location Deaths

1923 Japan 143,000

1927 China, Tsinghai 200,000

1948 USSR 110,000

1970 Peru 67,000

1976 China, Tangshan 255,000

1985 Mexico 10,000

1990 Iran 40,000

1993 India 10,000

2003 Iran 31,000

2004 Indonesia (with tsunami) 283,000

2005 Pakistan 80,361

562 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 563

Figure 35.1. Major world faults. From: pubs.usgs.gov/gip/volc/fig37.gif.

events. A variable degree of deformation associated with increas- ing stress accumulates along fault lines as the land masses move past each other. Even though the Tectonic Plate Theory success- fully explains most earthquakes, there are seismic events that it cannot adequately clarify, such as the activity in the New Madrid zone and around Charleston, North Carolina in the United States. The New Madrid zone is located along the Mississippi River Val- ley in the central United States, approximately 1,000 miles away from the nearest plate boundary. More information is needed to understand how earthquakes occur in these locations.4,8

left-lateral strike-slip fault

right-lateral strike-slip fault

Normal fault

Hangingwall

Footwall Footwall

Hangingwall

Reverse (thrust) fault

~30°~60°

Figure 35.2. Types of fault motion characterizing strike-slip and dip- slip faults.

Tectonic plates move in relation to one another in three spe- cific patterns described as strike-slip, dip-slip, and oblique-slip (Figure 35.2). Strike-slip faults occur when the plates slide hori- zontally past one another. Dip-slip faults occur when the plates slide over or under each other.8 These are further characterized as normal faults (where the underlying segment moves upward) and reverse faults (where the underlying segment moves down- ward). Oblique-slip faults exhibit both types of motion when they rupture. The dip-slip faults are associated with the develop- ment of tsunamis.

Regarding terminology, the location where the fault rupture begins is known as the hypocenter (or focus) and is located below the earth’s surface. The point that is located directly above the hypocenter on the earth’s surface is called the epicenter (Figure 35.3). When an earthquake takes place, stress is relieved

Figure 35.3. Relationship between hypocenter and epicenter. Modified from http://www.minerals.nsw.gov. au/ data/page/264/ 21 2.gif.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

564 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

P-Wave

S-Wave

Surface Wave

Figure 35.4. Earthquake shock waves. Modified from home.hiroshima-u.ac.jp/er/Resources/ Image195.gif.

along the fault lines as the land masses shift and release energy. The energy is released in the form of seismic waves. Temblors produce three types of seismic waves: primary (P), secondary (S), and surface (L) waves (Figure 35.4). P and S waves are referred to as body waves, meaning that they develop at the hypocen- ter and radiate in all directions through the earth’s interior.8

The L waves are surface waves and can only move through the crust.

The P waves move in a longitudinal direction and are the fastest of the seismic waves, traveling at 4.8 km/second. The S or shear waves travel at 3.2 km/second and cause the earth to move at right angles from the direction of the P waves.8 It is this difference in velocity between the P and S waves that allows determination of the epicenter. The different rates of travel between the P and S waves also produce two separate perceptions by individuals. The P wave generates an acoustic signal that sounds like an approaching train whereas the S wave produces a sharp jolt. The L or Love wave is a slow surface disturbance that causes swaying of tall buildings and large swells in bodies of water.8 It is the major cause of damage and injury resulting from earthquakes.

The frequency and amplitude of the vibrations produced at the surface and the subsequent earthquake severity depend on the amount of mechanical energy released, the distance/depth of the focus, and the structural properties of the soil near the surface.8,11 Distance from the epicenter is a poorer predictor of

severity because the transmission strength of earthquake shock waves is influenced by ground composition, liquefaction, and landslide susceptibility. Soils with high water content transmit energy waves that cause significant mortality, morbidity, and structural damage even when located at great distances from the epicenter. In contrast, solid rock transmits earthquake energy with a minimum of vibration. This explains why areas situated far from the epicenter (in a liquefaction zone) can be potentially more severely impacted by seismic intensities than locations near the epicenter (on bedrock).8

Earthquakes are characterized by intensity and magnitude. Magnitude is the total energy generated by the temblor. This energy is measured with a seismograph and then converted using the Richter scale (Table 35.2). The Richter magnitude scale is a logarithmic scale that estimates the total energy released by an earthquake.11 A change of 1 unit in the Richter scale corresponds to a 10-fold change in ground motion and a 32-fold change in radiated energy.11 Measurements on the Richter scale below 2.0 are not usually felt and measurements on the Richter scale above 5.0 can cause damage. A major earthquake consists of a magnitude of 7.0 or greater on the Richter scale. A major earth- quake can be preceded by less severe preliminary tremors known as foreshocks. There can also be smaller events after the main earthquake known as aftershocks, which can produce further damage and may necessitate evacuation of the area.8,11

Table 35.2: Richter Scale. (Data from Table 10.1, Carla W. Montgomery, “Fundamentals of Geology,” Wm. C. Brown, 1993 – Original Data taken from Gutenberg and Richter, “Seismicity of the Earth and Associated Phenomena,” Princeton University Press, 1954.)

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 565

Modified Mercalli Scale Richter Scale Instrumental Intensity Scale

Perceived Shaking

Potential Damage

Peak ACC (%g)

Peak Vel (cm/s)

Instrumental Intensity

I Felt by almost no one Not felt None <.17 < 0.1 I

II Felt by very few people

2.5

Generally not felt, but recorded on seismometers.

III many, but they often do not realize it is an earthquake.

Weak None .17–1.4 0.1–1.1 II–III

IV Felt indoors by many. Feels like a truck has struck the building.

Light None 1.4–3.9 1.1–3.4 IV

V everyone; many Felt by nearly

people awakened. Swaying trees and poles may be observed

3.5 Felt by many people.

Moderate Very light 3.9–9.2 3.4–8.1 V

VI people run outdoors. Felt by all; many

Furniture moved, slight damage occurs.

Strong Light 9.2–18 8.1–16 VI

VII Everyone runs outdoors. Poorly built structures considerably damaged; slight damage elsewhere

4.5 Some local damage may occur

Very strong

Moderate 18–34 16–34 VII

VIII Specially designed structures damaged slightly, others collapse.

Severe Moderate/Heavy 34–65 31–60 VIII

IX considerably All buildings

damaged, many shift off foundations. Noticeable cracks in ground.

6.0 A destructive earthquake

Violent Heavy 65–124 60–116 IX

X Many structures destroyed. Ground is badly cracked.

7.0 A major earthquake

XI Almost all structures fall. Very wide cracks in ground

XII Total destruction. Waves seen on ground surfaces, objects are tumbled and tossed.

8.0 and up

Great earthquake

Extreme Very heavy >124 >116 X+

Tremor noticed by

Figure 35.5. Comparison between the Modified Mercalli Scale, the Richter scale, and the instrumental intensities. Data were compiled and table was created by Schultz.

Measurement of the earthquake’s intensity determines the degree of ground shaking in a particular location. Intensity is calculated using two separate methods: one is an objective approach using instrumentation and the other is subjective eval- uation based on human observations and perceptions. The first method uses motion detectors to record peak ground velocity (PGV) and peak ground acceleration (PGA). The greater the degree of ground motion, the higher is the recorded velocity and acceleration. This information is referred to as the instrumental

intensity. The second method, the modified Mercalli intensity (MMI) scale, relies on observing the extent of damage to prop- erty and from the perceived degree of shaking reported by people experiencing the earthquake.7

The MMI is a 12-point subjective scale (Figure 35.5). Using this scale, intensity determinations are made after an earthquake when local U.S. Postal Service employees are interviewed regard- ing their perceived shaking experience and the visible structural damage they observed. Different governmental workers may

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

566 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

perform this task in other countries. Using one of the 12 cat- egories on the scale, a value is selected that most accurately represents the degree of shaking and damage. This value is then assigned to the pertinent zip or postal code. Recording the MMI values for all zip codes in the earthquake zone yields a representa- tion of the overall intensity. Although MMI scale measurements are subjective, they are generally valid. Researchers have found that MMI values correlate with structural damage, deaths, and traumatic injuries.8,10

PGV and PGA determinations are based on the degree of ground velocity and acceleration measured at localized points obtained by ground sensors placed in earthquake prone areas. Because these sensors are placed in only a few locations, the use of instrumental intensities as a worldwide measure is lim- ited. If available, they provide precise estimates of intensity not influenced by the subjective perception of motion or damaged structures. Comparisons between the Richter scale, the MMI scale, and Instrumental Intensities are depicted in Figure 35.5. Instrumental intensities recorded during an earthquake are bet- ter predictors of injury and lethal outcomes than the rate of building collapse.9,10,12

Seismology, the study of earthquakes and the propagation of seismic waves, has made progress in estimating the probability of a strong earthquake occurring during any 24-hour period in certain parts of the United States, such as California. These pre- dictions are based on evaluations of previous events and analysis of potential relationships between events. Seismologists cannot currently predict with certainty when or where the next earth- quake will occur or its intensity level.13

MANAGEMENT ISSUES

The persistent threat of seismic events and the difficulties involved with mitigating their effects highlight the importance of disaster preparedness. When planning, the following should be considered: 1) modify the initial response of prehospital and hospital care, 2) maximize the effective and efficient use of com- munity resources, 3) awareness of the most common clinical conditions seen after an earthquake and how to treat them, 4) acknowledge that outside help would probably take more than 24–48 hours to arrive.

Incident Command

In the initial period after an earthquake, a certain degree of uncertainty arises not only for the people in a community, but also in the healthcare system. Once it is clear that the demand on resources exceeds what is available under standard operating pro- cedures, establishing a system of command and control, known as an Incident Command System (ICS), becomes indispensable to effective management.

ICS is a structure that can provide direction and rapidly establish control of the event. It is based on a concept initially used by California firefighters in the 1970s for effective coor- dination and resource control when battling wildfires. The ICS directs activity during the response through decisions made by a single individual known as the incident commander and imple- mented by a formal chain of command involving others in an organized fashion. By providing federal and state government agencies with a standardized system of command and control, complex situations can be managed more efficiently during a

disaster, protecting lives and property. In cases in which multiple jurisdictions have a role in event management, the ICS becomes a Unified Command System and may include representatives from the state and federal levels in addition to local entities.4,14

A detailed explanation of ICS can be found in Chapter 9. An incident management system is used at each responder

entities’ Emergency Operations Centers (EOC). This is the loca- tion where personnel representing various organizations from the public and private sectors meet during an emergency event to: 1) coordinate response and recovery actions, 2) conduct strategic decision making, and 3) manage resource allocation.4,14

Prehospital Response

After an earthquake, victims’ lives may depend on how rapidly they are extricated from collapsed buildings and how expedi- tiously they receive medical treatment. In urban areas of devel- oped countries, it is common that paramedics are responsible for this type of initial prehospital care; however, in a large-scale earthquake they may be unavailable for this activity, particularly in systems in which they have primary responsibilities as fire- fighters. In systems not using paramedics, first responders can be police, firefighters, or other ambulance personnel. In some locations, such as California, firefighters receive cross-training as paramedics. Wherever fires occur immediately after an earth- quake, such as occurred in the 1989 Loma Prieta temblor in northern California, the priority is to minimize the amount of damage resulting from these blazes.4 Therefore human resources are initially directed to fire suppression. This can leave the affected area without paramedic support. Even though it was only a moderate-sized earthquake, during the Loma Prieta tem- blor, sections of San Francisco went without paramedic support in the critical early hours following the initial shaking because of the prioritization of fire suppression over rescue efforts.

Complicating the threat to the population, secondary events that increase the lethality of the initial earthquake can occur after seismic activity. Such events include fires, landslides, floods, and tsunamis, among others.2,8 During the 1994 Northridge, California earthquake, fires and burns accounted for 6.1% of the fatalities and 7.3% of the hospitalized injuries, even though nonresidential buildings were the most affected and the fires were quickly controlled.2

Another common problem faced by industrialized societies after an earthquake is exposure of the population to released toxic materials stored in such facilities as chemical plants. Following the 1989 Loma Prieta earthquake, approximately 20% of the post earthquake injuries were caused by toxic materials.7 Under these circumstances, chemical decontamination must be addressed to limit the number of exposed victims and property damage, and improve community safety. When dealing with the threat of chemical contamination, it is reasonable to consider directing human resources to assist with the decontamination processes instead of the rescue effort.

Communication between paramedics in the field and receiv- ing hospitals can be disrupted after a seismic event. Notifying hospitals of victims’ impending arrival becomes difficult, increas- ing the risk that healthcare resources cannot be effectively man- aged. Many individuals will not wait for the arrival of paramedics or other providers. Family and friends will frequently drive vic- tims to the nearest hospital, overwhelming the facility’s capacity. During the Puerto Limon, Costa Rica earthquake in 1999, most victims were transported by survivors.15 During the Gujarat,

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 567

India earthquake in 2001, most of the victims used private trans- portation to reach hospitals.16 Major trauma victims will be brought to nontrauma centers and many patients without life- threatening injuries will arrive at trauma centers. This mismatch of medical needs and available resources can potentially over- whelm capacities and result in inefficient utilization. Even though this phenomenon is undesirable, it will be difficult to avoid dur- ing the first hours after an earthquake.

Even in well-developed systems in industrialized societies, breakdown of communication systems makes coordination of field care and disposition more difficult and jeopardizes the direction and coordination of all prehospital activities. Para- medic radios frequently rely on repeaters that can fail, making transmission of radio signals to the dispatchers or base stations providing contact with paramedics in the field difficult. Ambu- lances from different jurisdictions tend to use different frequen- cies, making communication between field units and the central coordinating body problematic. The creation of a globally recog- nized disaster frequency would make a significant contribution to the improvement of communication under these circumstances. Satellite or cellular telephones also have limitations.

Many systems have been developed to address the communi- cation problems that arise after disaster events. A wireless trans- mission system for disaster patient care (WISTA) was devel- oped to assist emergency personnel in treating disaster victims and coordinating medical resources.17 Another communication innovation, the Wireless Internet Information System for Medi- cal Response in Disasters (WIISARD) is a project by the Univer- sity of California, San Diego designed to support disaster relief operations in the field.17 Although these systems show promise, at the time of this writing, none has actually been implemented after an earthquake.

Movement in and out of a disaster zone is difficult follow- ing an earthquake. Significant damage occurs to transportation infrastructures such as roads, bridges, traffic signals and road lighting. During the 1994 Northridge earthquake, 15% of the fatal injuries were motor vehicle related, primarily as a result of the disruptions in traffic control devices.10 Earthquake-related motor vehicle injuries were 5.23 times more likely to result in fatality than in a hospitalized injury.2 Eighty one percent of deaths on public roadways following the Loma Prieta event were asso- ciated with the collapse of freeway structures.10 Landslides, soil settlement, and slope failures, among others, increase damage to the highways. Ill or injured survivors who require transportation to medical centers will experience delays until the safest routes to hospitals are identified. The early use of law enforcement person- nel (including State National Guard assets in the U.S.) to control important transportation areas can improve this situation.4

The initial prehospital response should be directed toward the provision of emergency medical assistance, and then fol- lowed by search and rescue activities.7,15 In some studies, death and injury rates were 67-fold and more than 11-fold higher, respectively, for trapped victims in comparison to those not requiring extrication.7,18 Injuries caused by collapsing structures or falling building components were 8.36 times more likely to result in fatalities than in hospitalizations.2 Approximately 90% of the deaths in all earthquakes are the result of structural col- lapse.8 There is a well-documented decrease in survival for vic- tims trapped longer than 24–48 hours after an earthquake, as seen in the Campania-Irpinia earthquake (1980) in Italy and the Tangshan earthquake (1976) in China.3,11 In Italy, a survey of 3,169 survivors showed that 93% of those who were trapped and

survived were extricated within the first 24 hours and 95% of those who died expired before extrication.18 Estimates of surviv- ability among entrapped victims in Turkey and China indicate that within 2–6 hours, fewer than 50% of those buried were still alive.3,12,18 In a study of the 1980 Italian earthquake, investiga- tors concluded that 25%–50% of the victims who were injured and died slowly could have been saved if they had received initial life-saving treatment immediately.3,7,12,18

After an earthquake, an increased requirement for surgical services is expected during the first 72 hours; acknowledging this requirement is important for effective resource utilization.8 Early rapid assessment of the extent of damage and injuries is neces- sary to help mobilize resources and direct them where they are most needed.10 Rescue workers frequently perform triage, using a system similar to Simple Triage and Rapid Treatment (START). This methodology can theoretically sort patients into groups of increasing acuity; however, controversy exists regarding the efficacy of such systems and data supporting their use are lim- ited. Two studies evaluating START triage have been published and suggest it may be useful.19,20 The START system emphasizes basic life-saving measures such as opening an airway and apply- ing direct pressure for external bleeding control but rescuers are instructed not to provide definitive care procedures at the scene. More information about the START system can be found in Chapter 12.

A significant number of victims who survive the initial impact subsequently die because of a delay in arrival of life-saving emergency medical care. One study investigating the 1988 Arme- nia earthquake reported that deaths might have been prevented if victims had received appropriate medical attention within the first 6 hours after the event.3,18 A theoretical increase in survival would have been possible if trained urban search and rescue (USAR) teams had been available during the first hours after the disaster. USAR teams can help local leaders with structural assessment, advanced search and rescue techniques, and special- ized medical training helping to limit the period of entrapment for victims.21 The utility of USAR teams has, however, been questioned. It requires nearly 24 hours for a team to deploy from an unaffected region and begin search and rescue activities in the United States and even more time to provide interna- tional assistance, thereby limiting its usefulness for saving lives after an earthquake.21 In the Northridge earthquake, 19 hours passed before the Riverside USAR team began operations at the Northridge Meadows Apartments.4 The first U.S. USAR team required 48 hours to begin operations in Turkey after the 1999 earthquake.4 A careful investigation of the organizational bar- riers preventing a rapid response by medical rescue teams is required to improve their ability to save the lives of the most seri- ously injured disaster victims.15 Reports show that up to 90% of survivors are rescued by civilian volunteers within the first 24 hours after an earthquake event.18

Hospital and Community Response

Hospitals are the traditional source of medical care for victims after seismic events; however, hospitals can also be impacted by earthquakes, aggravating the initial imbalance between demand for medical care and surge capacity. Healthcare facility closure and evacuation are mandated if management staff identify envi- ronmental or structural factors that put patients’ safety at risk. An early structural assessment performed by expert engineers or, if not available, on-site personnel is critical. Therefore, institutional

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

568 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

disaster plans must include a hospital functional status evalua- tion.22 Guidance for structural assessments is provided by the ATC-20 and ATC-20–2 documents, which were developed with the support of the U.S. Federal Emergency Management Agency, the National Science Foundation, and the State of California’s Office of Emergency Services and Office of Statewide Health Planning and Development.3 Structure-dependent collapse pat- terns are documented in the Federal Emergency Management Agency’s Structural Engineer Training Manual.21 During the Northridge earthquake, eight (9%) of 91 acute care hospitals were evacuated.23 Six hospitals, two of which complied with cur- rent building codes, evacuated patients within 24 hours. Four of these institutions ordered complete evacuations, including the two institutions that met current building standards. The remaining two facilities ordered partial evacuations.4,23 Two hos- pitals whose initial inspections revealed no critical damage and therefore continued providing patient care were subsequently evacuated and condemned. The first facility completely evacu- ated patients 3 days after the temblor and the second one after 14 days due to identification of structural damage that required demolition.23 It appears that hospitals reported to be secure may later prove vulnerable. After the 2005 Pakistan earthquake, 65% of the hospitals in the affected area were destroyed or badly damaged.24

Hospitals that are damaged and cannot continue providing inpatient care need information to facilitate the transfer and evacuation of their inpatients. All else being equal, evacuation should be coordinated by the EOC if one exists to ensure effec- tive resource utilization; however, hospitals can safely and effec- tively evacuate patients with or without assistance from an EOC. Although coordination with the EOC is desirable, if time is crit- ical and communications are intact, hospitals can successfully evacuate patients directly. Such activity is facilitated if mutual aid agreements with other hospitals are already in place. During the Northridge earthquake, both strategies were equally effec- tive.23 Evacuation transport problems can often be solved by collaboration between military and civilian groups. In the 1999 Marmara earthquake in Turkey, military boats and helicopters were used to transfer patients to remote major cities.24

Hospitalized individuals in critical condition require exten- sive resources, which are limited after a seismic event. Evacu- ating the sickest patients first from hospitals appears to work best, lessening the burden on healthcare facilities while improv- ing the chances for better care of remaining patients or arriv- ing victims.4,23 When time is critical and structural collapse is imminent, evacuating the healthiest patients first permits move- ment of the greatest number of patients in the least amount of time.4 During the Northridge earthquake, one evacuating institution believed that patients were in immediate danger and chose to evacuate the healthiest patients first, successfully evac- uating all patients (a total of 334) to nearby open areas in 2 hours.23 Methods to evacuate patients from a hospital may vary. In the Northridge temblor, supervisors transported patients by using available equipment such as backboards, wheelchairs and blankets, among others.23 Vertical evacuation of patients (mov- ing patients from one floor of the hospital to another) should generally be accomplished using stairwells and not involve ele- vators until they have been inspected.

Functioning hospitals must activate their disaster plans to prepare for the influx of victims. Effective command and control of the facility’s response requires the implementation of a sys- tem flexible enough to be used in medical institutions and not

dependent on the presence of any specific individual. A proposed model used by many hospitals is known as the Hospital Incident Command System (HICS) and is based on general ICS princi- ples. HICS has been effectively used during earthquake events.4

More information on HICS is available in Chapter 20. After establishing a command post and implementing HICS,

hospital administration is primarily concerned with evaluat- ing personnel availability, communications, and resources. Hos- pital staff, like the general population, use telephones as a major method of communication. Because standard telephones frequently fail, secondary communication methods should be available. Examples of such methods include: alphanumeric pagers, priority phones, fax machines, the Internet/e-mail, cel- lular phones, pay phones, wireless systems, ham radio sys- tems, portable two-way radios, satellite telephone systems, and runners.

Hospital personnel who are at home after a disaster can be difficult to contact. Therefore, implementation of a disaster callback policy is necessary. Because communications may be disrupted, such a policy could state that personnel should report to work in a major disaster unless they are notified to stay at home. Disaster planners can expect that most hospital staff will remain at work and not abandon their responsibilities. After the Northridge earthquake, as in other earthquakes, the majority of staff remained on duty. Most workers who did not immediately report to the hospital had problems with either communica- tion or transportation.4 To maintain the flow of medical goods, notifying suppliers of an abrupt increase in demand for certain products may be difficult secondary to communication limita- tions. An agreement requiring suppliers to deliver a set amount of supplies after a disaster, such as a seismic event, should be established a priori.

In actuality, it is difficult for hospitals to meet the entire demand for medical assistance because they typically operate at a high census with little surge capacity for accepting additional patients. The solution to providing an effective medical response is optimizing the use of all available resources. This requires hos- pitals to establish plans for improving surge capacity after an earthquake, even if some structures are damaged.7 Typical plans include use of additional space not traditionally designated for patient care, such as cafeterias, auditoriums and parking lots. After a temblor in Puerto Limon, Costa Rica (1999), the only treatment facility was declared structurally unsafe. This resulted not only in patient evacuations but establishment of treatment areas in the parking lot.15 After the Northridge and Loma Prieta earthquakes, treatment areas were established outside the hospi- tals as well.15

In situations in which receiving hospitals become nonfunc- tional, community-based solutions to augmenting surge capacity can be used. One such model is the Medical Disaster Response (MDR) program. The main advantage of the MDR program is that victims receive rapid advanced medical care even if hospi- tals are damaged or destroyed. This project, which was developed by emergency physicians in southern California, focuses on the training of healthcare personnel in the management of unique medical problems encountered after earthquakes. It utilizes supplies stored in the community before an earthquake occurs, similar to the concept behind the Strategic National Stock- pile, but at the local level. Under austere conditions, the MDR model directs the initial management of casualties by specially trained local personnel who use medical supplies at designated sites within the community.3 Using this model, local healthcare

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 569

providers in or near the disaster zone can respond immediately and deliver advanced patient care. Detailed discussion of the MDR project can be found in the literature.3,25,26

Information about the status of surrounding hospitals is valuable. Some will sustain damage and be unable to accept victims from the field. Others will be in the process of evac- uation. Remaining functional hospitals can expect an increase in emergency department volume as well as requests to accept patient transfers from damaged institutions. Therefore, it would be extremely useful to establish a hospital communication system that could provide a rapid estimate of the number of remaining functional facilities. Implementation of such a system would per- mit these undamaged hospitals to estimate the potential demand for their services in the immediate postdisaster period. Health- care institutions could then decide whether to cancel elective surgeries, discharge stable patients earlier than planned, and implement other components of their surge capacity plans.27

Several communications systems currently exist but have not always performed optimally after seismic events.

It is tempting to assume that hospitals located near the epi- center will have a higher probability of structural damage. As such, incident managers may incorrectly assume that hospitals located farther away will remain functional and will preferentially direct patients to these facilities. Although this is true for insti- tutions located a large distances from the epicenter, it does not appear these assumptions are correct for facilities located close to the epicenter. A study examining the association between dis- tance from the epicenter and hospital evacuation found no rela- tionship between these two variables (Figure 35.6).28 All facilities studied were located within approximately 32 km of the epicen- ter. In contrast, a strong association existed between PGA and hospital evacuation, regardless of the institution’s location. It may be appropriate for incident managers to consult earthquake shake maps before making patient transport decisions. In regions with ground motion sensors, computers can generate shake maps identifying the areas of greatest shaking within minutes after a seismic event occurs.

Surge capacity is critical for a health system’s disaster pre- paredness. Deficient surge capacity limits the ability of health- care systems to respond to disasters successfully.29 Surge capacity plans require multidisciplinary collaboration to be successful. With regards to the personnel component of surge capacity, Schultz and Stratton proposed the creation of a database to provide rapid emergency credentialing of volunteers. If imple- mented, the strategy would offer an accurate, inexpensive, effi- cient, sustainable and U.S.-based Joint Commission–compliant tool for rapidly expanding healthcare personnel support for hos- pital patient care. This database is created from a list provided by each hospital of personnel and physicians with unrestricted privileges currently credentialed at the facility. It is shared with all participating institutions and a county healthcare agency. Volun- teers who appear in the database could obtain hospital privileges at any hospital within the county for the first 72 hours after a disaster event.29 Should the database be shared between coun- ties, such individuals could receive emergency privileges in other jurisdictions as well.

Patient tracking and the creation of medical records can be problematic after earthquakes. Many patients will not have access to typical items that document personal information and iden- tity (e.g., driver’s license, medical insurance cards) or will have a health condition that makes gathering information from them impossible (e.g., altered mental status). Therefore, creating a

5 miles210

405

105

118

101

170

134

91 47

Burbank

Glendale

West HollywoodBeverly

Hills Wilshire Blvd

Culver City

Inglewood Los Angeles International

Airport

Matina Del Rey

Pacific Ocean

Los Angeles

Alhambra

Pasadena

G ie

ne ga

B lv

d La

ur el

B lv

Epicenter location Evacuated hospitals Evacuated and condemned hospitals Control hospital

Figure 35.6. Epicenter and hospital locations, Los Angeles County, CA. This map shows the geographical locations of the study and control hospitals, as well as the epicenter of the 1994 Northridge earthquake. From Schultz CH, et al., Ann Emerg Med. 2007;50:320– 326.

medical record and tracking a patient’s movement through the healthcare system can be challenging. Potential solutions include documenting descriptive information such as sex, estimated age, height, color of skin, unique skin marking (mole, tattoo, or scar), eye color, and possibly fingerprints in the medical record.8 Using records created with generic fictitious names before an event can address registration issues. In settings in which patient vol- ume overwhelms hospitals, documenting patient care may not be possible. After the Costa Rica earthquake, hundreds of patients received medical care even though documentation of these inter- actions was absent in many instances.15 No widely accepted solu- tion to the patient-tracking problem has yet been found, although some have suggested using bar codes or radiofrequency identifi- cation devices (see Chapter 25).

Building damage and structural collapse resulting in loss of life are predictable consequences of earthquakes. This seri- ous problem is due, in large part, to lack of appropriate seis- mic construction codes in many countries, including wealthy nations.18,30 A study of the 1976 Guatemala earthquake con- cluded that deaths and injury are critically dependent on housing damage and construction materials used.18 Mitigation and pre- paredness are the only ways to save lives and preserve resources at the same time.30 Engineering and epidemiological evaluations are fundamental for understanding the effects that seismic forces have on the different types of building structures.31 Data collected on these seismic forces and subsequently analyzed will help to improve building design, casualty estimates, and disaster plan- ning and preparedness.15 Building codes developed for seismic areas are intended to prevent catastrophic structural failure and

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

570 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

reduce mortality with the formation of large void spaces in any structure that does fail.21 Implementation and enforcement of building codes will help prevent deaths and improve the safety of the population. Even though building codes can reduce morbid- ity and mortality from an earthquake, they are not 100% effective. During the Northridge earthquake, two of the hospitals that suf- fered significant nonstructural damage were constructed under the most updated building codes.23 During the retrofit process, there was the unexpected discovery of failed welds in steel-frame buildings that engineers had incorrectly believed could toler- ate earthquake forces.4 In addition to improving building design components, other interventions that reduce injury involve non- structural mitigation. Such activities address the most common cause of injuries in mild to moderate temblors. These include bracing bookcases to the wall, securing heavy electronic equip- ment, and using earthquake hooks to hang picture frames con- taining glass to prevent them from falling and shattering.

The first influx of patients is expected to arrive within 30– 60 minutes after the temblor, probably suffering from minor injuries such as uncomplicated lacerations, contusions, or frac- tures. Because supplies may be limited in a disaster situation, hospital personnel should initially allocate resources to the more critical patients. Therefore, victims with uncomplicated condi- tions should be quickly evaluated and directed to an observation area. Treatment should be postponed until more detailed infor- mation regarding the overall demand for medical care is available. Many victims suffering from minor injures may not require hos- pital care at all. After the 1989 Loma Prieta earthquake, as many as 60% of those with earthquake-related injuries either treated themselves or received treatment in nonhospital settings.7 It is noteworthy that neither age nor sex is considered a risk fac- tor for injury. No consistent relationship between injuries and demographic factors has been reported in the literature.9,10,18,32

When treating uncomplicated lacerations during a disaster, providers should be concerned about an increased risk of infec- tion and missed foreign bodies. To decrease these risks, some experts recommend allowing uncomplicated lacerations to close by second intention or to use the technique of delayed primarily closure. There is insufficient evidence to justify any particular treatment recommendation at this time.

The second group of patients to arrive is composed of those victims with more serious conditions such as crush injury. In developed nations, these victims are more typically transported by paramedics via ambulances because they were more diffi- cult to extricate or transport by lay individuals. They may also suffer from other medical conditions, making their initial treat- ment more challenging. Depending on the scenario, the number and acuity of patients presenting for acute care at local emer- gency facilities may overwhelm available resources. Therefore, implementation of a triage process is necessary. Triage decisions must focus not only on patient acuity but, more importantly, on which individuals should be prioritized for medical or sur- gical treatment. START triage does not discriminate between those who will consume large amounts of limited available resources or those whose prognosis will remain poor despite aggressive treatment. To allocate resources more appropriately, a proposed algorithm called the Secondary Assessment of Victim Endpoint (SAVE) was created. The goal of this algorithm is to optimize reduction of victim mortality and morbidity by allocat- ing resources only to those who will benefit. This means that not all victims will receive the same degree of care they would under normal circumstances. This treatment approach contradicts the

usual medical philosophy of providing potentially unlimited care to each patient.22

The SAVE triage algorithm, which was designed for mass casualty events, is a triage pathway driven by estimated patient outcomes. It is based on outcomes data from trauma patients and those with other medical conditions who receive standard treat- ment. It arbitrarily recommends withholding aggressive medi- cal care from those with less than a 50% chance of survival or from those who will essentially deplete all available resources.26

Implementation of SAVE triage should only occur in situations in which the time for return to standard operations is unknown and so optimal utilization of limited resources is necessary.26 A detailed discussion of SAVE triage can be found in the litera- ture.26

The need for rapid screening tools to prioritize patient care in disasters has received a great deal of interest. Ultrasonogra- phy holds promise as a technology for evaluating mass casualty victims and has many benefits. It is 1) noninvasive, 2) portable, 3) easily repeatable, and 4) highly sensitive for intraperitoneal blood and other conditions. Newer indications for this tech- nique in the evaluation of trauma patients include pleural effu- sions, pneumothorax/hemothorax, and fractures. The practical utilization of sonography has been studied in disaster settings (e.g., following Hurricane Katrina in the U.S. in 2005) and in the prehospital setting. Technologies have progressed to the point that ultrasound devices are available in handheld form. During the 1988 Armenia earthquake, physicians with two ultrasound machines were able to triage 400 blunt trauma patients in 48 hours.33 Ultrasound has also been used in Kosovo, Afghanistan, and Iraq by the deployed military forces.34

Medical Issues

After an earthquake, victims experience major injuries and ill- nesses that will require treatment in hospitals. Typical exam- ples include skull fractures with intracranial hemorrhage, spinal cord injuries, and intrathoracic, intraabdominal, and intrapelvic organ trauma.7 Major medical complications described in pre- vious earthquakes include hypothermia, wound infections, gan- grene requiring amputations, sepsis, adult respiratory distress syndrome, exacerbations of chronic pulmonary disease such as asthma, myocardial infarctions, multiple organ failure, and crush syndrome.7 Variable degrees of hypothermia were documented in patients evacuated from the Armenia earthquake of 1988.18,21

Although not strictly a complication of seismic activity, many investigators have also noted an increase in childbirths after an event.

Crush injuries are associated with the development of com- partment and crush syndrome, all of which are commonly found in earthquake victims. Crush syndrome, one of the most com- mon causes of death in a seismic event, usually results from excessive pressure on limbs, which damages muscle tissues.35 The life-threatening effects of crush syndrome include hypovolemic shock, acidosis, rhabdomyolysis, acute renal failure (ARF), elec- trolyte disturbances (hypocalcemia, hyperkalemia), acute respi- ratory distress syndrome, disseminated intravascular coagula- tion, and fatal cardiac arrhythmias.7,36,37 Calculated mortality rates for patients with rhabdomyolysis and/or crush syndrome who required renal dialysis reach up to 40% or more.38,39 In the 1988 Armenia earthquake, more than 1,000 victims trapped in collapsed buildings developed crush syndrome and 323 devel- oped secondary acute renal failure requiring renal dialysis.7

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 571

During the Kobe earthquake, crush syndrome was observed in 13.8% of hospitalized patients and acute renal failure developed in half of these individuals.35,37 After the 1991 Costa Rica earth- quake, autopsies performed on a sample of victims showed crush injury as the predominant mechanism of injury and cause of death.15

Standard therapy for crush syndrome includes large amounts of intravenous fluids (more than 10 L/d), bicarbonate, manni- tol (10 mL/h of a 15% solution), monitoring of cardiac rhythm and urine output, and dialysis.35,37 Because there is a high risk for ARF and hyperkalemia, intravenous fluids must be potas- sium free. In the 1993 Turkey earthquake, researchers noted that some of the victims transferred from the field received solu- tions containing potassium.36 In patients who require dialysis, implementing a protocol using intermittent hemodialysis dur- ing short sessions supervised by experienced personnel allows the treatment of several patients per day with a single machine.35,37

Continuous renal replacement therapy and peritoneal dialysis are less efficient in removing potassium in comparison to inter- mittent hemodialysis.35,37 Lower mortality rates are seen when patients with crush injury–associated ARF are treated with ade- quate dialysis at intensive care facilities and under the supervision of healthcare professionals.39

In the austere environment that frequently follows an earth- quake, prevention of crush syndrome may be difficult. Large volumes of intravenous fluid are limited. In these situations, amputation or fasciotomy may be the only treatment available to prevent the otherwise inevitable morbidity and mortality result- ing from uncontrolled crush syndrome. A victim can deterio- rate quickly to the point of death as the crush-injured area is extricated from the collapsed structure.21,37 During the 1976 Tangshan earthquake in China, patients died suddenly of cardiac arrest soon after extrication, presumably due to hyperkalemia.15

When a high degree of suspicion exists that crush syndrome and associated hyperkalemia will develop during or immediately after rescue, it is recommended to delay extrication until personnel can institute therapy to threat metabolic derangements (several liters of intravenous isotonic saline).15,35,36 If this treatment is not available, amputation of the involved extremity is an option. Another suggested approach to this situation is application of a tourniquet to the compromised limb before the patient’s extri- cation. Theoretically, this could avoid sudden death due to the ensuing reperfusion of the crushed extremity.40 No data currently exist to support this intervention.

When treating compartment syndrome, it is recommended that fasciotomies should be done only when clear indications are present, such as with intracompartmental pressure measure- ments above 30–35 mm Hg.37 This suggestion is based on the risk of infection related to the procedure itself, which increases the morbidity and mortality of the victim. If, however, the pro- cedure is performed at a stage when all muscle is still viable, and the wound is kept open with sterile dressings to allow healing by secondary intention, infection is rarely seen. Administration of antibiotics may also improve the prognosis, although this is con- troversial.40 Medical management of compartment syndrome by using mannitol has been proposed as an effective and less haz- ardous alternative to surgical intervention.38

Amputations also play a role in facilitating extrications and in the removal of severely mangled extremities. The mangled extremity severity score is a tool that can assist medical providers in making the decision about whether to amputate an injured lower extremity in cases in which survival of that limb is ques-

Table 35.3: Mangled Extremity Severity Score (MESS) Score

Skeletal/Soft Tissue Injury

Low energy (stab; simple fracture; civilian gunshot wound) 1

Medium energy (open or multiple fractures, dislocation) 2

High energy (shotgun; military gunshot wound, crush injury) 3

Very high energy (as above, plus gross contamination, soft tissue avulsion)

Limb Ischemia

Pulse reduced or absent but normal perfusion 1

Pulseless; paresthesia; reduced capillary refill 2

Cool, paralyzed, insensate, numb 3

(Score doubled for ischemia longer than 6 h)

Shock

Systolic blood pressure maintained >90 mm Hg 0

Transient hypotension 1

Persistent hypotension 2

Age (y)

<30 0

30–50 1

>50 2

Adapted from: Robertson PA. Prediction of amputation after severe lower limb trauma. J Bone Joint Surg Br. 1991;73(5):816.

tionable (Table 35.3).41 A total score of 7 or greater indicates that there is a high probability the lower extremity will be unsalvage- able despite aggressive treatment and that amputation should be considered.26,40,41 This assessment has a specificity between 90% and 100%, making it acceptable for use in disaster settings. In this situation, a guillotine amputation performed at the most distal site possible is best and has a low incidence of infection even without antibiotics.40

Painful procedures such as amputation, fasciotomy, and frac- ture reduction require effective anesthesia and analgesia. Avail- able options are the use of regional anesthetics (nerve blocks) and administration of systemic medications. Even though they may be an attractive option, regional blocks can produce inconsistent results. Ketamine administration is another reliable method for achieving effective procedural anesthesia and analgesia.

Ketamine, a dissociative agent, has been categorized as a fast, short-acting, safe, and effective drug that preserves airway reflexes and supports the cardiovascular system. In a review of 11,589 patients treated with ketamine, only two healthy patients required intubation.3 Ketamine has been used safely and effec- tively by nonmedical persons in unmonitored, austere environ- ments in Afghanistan to facilitate field amputations.3 In the absence of contraindications, ketamine can be administered safely via the oral, intravenous, rectal, and intramuscular routes. The recommended dosage for intravenous administration is 2 mg/kg and for intramuscular injection, it is 4–6 mg/kg.3,40

Intravenous fluids may be limited in the first 48 hours after a seismic event. Therefore, common patterns of practice must be modified regarding the type and amount of fluids used in a disaster setting. There are limited data that support the use of

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

572 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

hypertonic saline for initial resuscitation and normal saline for early maintenance.3,42 Doses of 4 mL/kg of hypertonic saline seem to be safe and effective in trauma patients, burn victims, and children.42,43 After patients improve with hypertonic saline challenges, they can be supported with normal saline. Concerns about the development of hypernatremia and a hyperosmolar state limit the amount of hypertonic saline that can be used. The main objective in the switch of resuscitation fluids is reducing the volume of normal saline that must be kept in inventory for initial patient stabilization. Liter bags of saline are bulky, heavy, and require significant space for storage. Medical personnel can initially resuscitate two–four times the number of patients with a given volume of hypertonic saline as they can with the same volume of normal saline. Solutions of 5% dextrose in quarter normal saline could be given to patients unable to ingest fluids or who require hypotonic hydration.3 More information is needed to know the optimal intravenous solution for resuscitation and stabilization of disaster victims.

Patients suffering from abdominal trauma should be resus- citated with hypertonic saline boluses. Some experts recom- mend maintaining blood pressure in the moderate hypoten- sive range, targeting a systolic of 90 mm Hg or mean arterial pressure of 60 mm Hg. They suggest that normalizing blood pressure may be associated with increased internal hemorrhage and exacerbated morbidity and mortality.26,42 Definitive evi- dence for or against this approach is lacking. Adults who do not respond to fluid resuscitation should be reevaluated with regard to the continuation of aggressive care. The use of ultrasound can quickly confirm the presence of significant intraabdominal hem- orrhage. In these situations, if intraperitoneal bleeding is iden- tified, current resources are limited, and surgical intervention is unavailable, consideration for the provision of comfort care only is warranted. Children with abdominal trauma should be treated aggressively because many will survive without operative intervention.26

After a catastrophic earthquake, by the time substantial out- side medical resources arrive to the disaster area, most victims have either received initial treatment or are dead; however, the need for maintaining medical care, supporting the damaged healthcare infrastructure, and supplementing limited resources persists. Groups that arrive after the immediate response phase is over are critical to preserving the lives of survivors. Disas- ter Medical Assistance Teams are available in some countries with variable missions and response times. One important func- tion they can provide is to care for ambulatory patients in any location. These teams can decrease the healthcare burden on emergency departments by offering another option for med- ical care in the days to weeks after the earthquake. Also, the Renal Disaster Relief Task Force (RDRTF), created by the Inter- national Society of Nephrology in 1989, can increase surge capacity for renal dialysis. These teams can provide additional dialysis capacity for existing renal patients as well as victims suffering from crush syndrome. The positive impact from these task forces was demonstrated following the Marmara earthquake in Turkey in 1999 and the Yogyakarta earthquake in Indone- sia in 2006.35,44,45 The RDRTF is composed of three divisions: the American, European, and Pacific groups.35 Disaster Medical Assistance Teams and the RDRTF are meant to be self-sufficient and to establish patient care capability with limited consumption of community resources. The time required for these teams to arrive at the disaster zone is often more than 48 hours (sometimes

weeks in remote areas), limiting their effectiveness in reducing the burden of acute patient care in many cases.44 These teams will help in the reconstitution of the basic medical care system because hospitals, clinics, and medical offices may be destroyed or incapacitated.

An important area to consider is access to information by the public. In an event like an earthquake, locating individuals quickly is extremely difficult. In the immediate aftermath, many victims will leave the area, die, or be hospitalized. Patient transfers are common and victims will often receive care in facilities far from where they live. The creation of a public information center will decrease the time and effort people will spend looking for relatives and improve family reunification.

Nontrauma-related Medical Conditions

A review of pertinent literature reveals that up to 20% of the patients hospitalized after earthquakes suffer from nontrauma- related conditions.12 Unlike the direct health effects of disasters, indirect health effects are predicable and preventable.31 The con- ditions most commonly seen are acute exacerbations of chronic conditions such as diabetes, chronic obstructive pulmonary dis- ease, acute coronary syndromes and myocardial infarctions, end- stage renal disease, anxiety, hypertension, and the onset of spon- taneous abortions, among others.7,45 Beginning on the third day following the 1999 Chi-Chi earthquake in Taiwan, medical diseases became the most common cause for hospitalization.12

After the 1985 Mexico City earthquake, there was an increase in the number of spontaneous abortions, premature births, and normal deliveries, making those conditions the primary reasons for admission in chronic care facilities.7 Healthcare professionals should be aware of the possibility that an increased incidence in these medical conditions will occur following earthquakes, and they should be prepared to treat them.

Myocardial Infarctions

Increased numbers of patients presenting with conditions related to cardiovascular disease such as myocardial infraction and car- diac arrest have been reported after earthquakes.4,31 A 50% increase in cardiac deaths was reported in the first 3 days follow- ing the 1981 Athens 6.7-magnitude earthquake.4,7 The existence of this phenomenon suggests that it may be caused by psychologi- cal stress resulting in an increase in catecholamines, vasoconstric- tion, or a hypercoagulable state and not to an increase in physical exertion.4,7,31 The incidence of sudden cardiac death rose dra- matically in the first 24 hours after the Northridge earthquake, and then decreased in the following 6 days. This suggests that those who died immediately after the earthquake were already susceptible to cardiac events and would probably have died in the next several days. It appears the earthquake may have accelerated the process.

In most studies, increases in cardiac mortality were reported to occur within a few days following earthquakes; however, after the 7.2-magnitude 1995 Great Hanshin-Awaji earthquake, aug- mented cardiac mortality was reported over a period of weeks.31

Earthquakes that cause chronic stress among the population dur- ing the recovery and reconstruction stage may result in prolonged cardiac events. In addition, increased mortality and worsening prognosis from myocardial infarction can be due to loss of stan- dard treatment resources because of hospital damage. Further

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 573

investigation is needed to clarify the factors affecting regional variations in terms of the extent and duration of increased car- diac mortality after an earthquake.31

Chronic Obstructive Pulmonary Disease

Respiratory injury is a major cause of death among victims of earthquakes. Early mortality caused by airway obstruction, asphyxiation, and dust-induced fulminant pulmonary edema is one of the possible scenarios. Part of the increased incidence in respiratory diseases after earthquakes is attributable to the inhalation of dust produced by collapsed structures during early search and rescue activities.7

End-stage Renal Disease

After a seismic event, there is an increase in the number of patients with end-stage renal disease whose conditions deterio- rate. Dialysis-dependent patients cannot receive their treatments on schedule because the water and electricity required for this process are often unavailable after earthquakes. In addition, dial- ysis centers and dialysis material may be destroyed and dialysis center personnel may be lacking either because of transporta- tion issues or becoming victims themselves. In addition to the paralysis of communication and transportation systems, centers with the capabilities to perform dialysis can be overwhelmed by new patients with ARF secondary to crush syndrome as well as by patients with chronic renal disease whose usual site of care is not functional. The immediate options for increasing surge capacity for dialysis patients are either to decrease the number of weekly dialysis sessions for each patient or to shorten the dura- tion of each session, as was done after the Marmara earthquake in 1999.45 In the aftermath of this seismic event, officials not only implemented shortened periods for dialysis treatment but also issued specific patient instructions to follow strict dietary and fluid restrictions. Compliance with these recommendations was high. As a result, despite less frequent dialysis, weight gain and blood pressure among these patients did not differ when com- pared with the predisaster period.45 As new temporary centers for dialysis open or alternate facilities are made available, patients can receive their treatments at nearby satellite outpatient units until damaged facilities are restored.37

In the first days after a disaster, dialysis-dependent patients should strictly follow recommendations for fluid restriction and should avoid foods rich in potassium. In the situation in which dialysis treatment centers are inoperative or cannot be reached, potassium exchange resins kept in the home can serve as a useful temporizing measure to prevent life-threatening hyperkalemia.45

In an extreme situation in which the physician needs to prioritize treatment between patients with acute or chronic renal failure, dialysis should be administered first to those with ARF. These individuals, if they survive their critical state, are more likely to live longer and healthier lives.45 Disaster plans should consider management strategies for chronic renal patients who are depen- dent on dialysis, because they will consume a large amount of resources that are needed in the aftermath of a large temblor.

Management of Psychological Distress

Experiencing the impact of a disaster can negatively affect responders and victims, resulting in enormous emotional con-

sequences. Responders may need supportive psychological ser- vices in addition to the general population. Community needs assessments must include all potential patients. Specific analysis should include information about age, race, cultural background, socioeconomic status, and special needs populations so that pro- grams can be developed to meet the requirements of these diverse groups. The Red Cross is often the first agency to begin coordi- nating mental health programs and provides crisis counseling for extended periods of time.46 Skills that are most often needed include training and experience in the prevention and manage- ment of posttraumatic stress disorder, trauma and bereavement counseling, debriefing, and crisis intervention.46 Mental health workers should be available 24 hours a day. In addition, it is crucial to provide individuals who can assist with the debriefing of the mental health teams at the end of their shifts. Community groups that desire a role in the mental health response must be identified and trained before an event to be effective.

Epidemics

After an earthquake, the affected area is potentially predisposed to epidemics of diseases normally endemic in the population. The most important factors that contribute to an area’s vulnerabil- ity to infectious diseases are the loss of adequate water supplies and sanitation systems. Failure of these systems increases the risk for development of water-borne diseases. In addition, earth- quakes lead to population displacement and crowding, both of which have been proven to increase infectious disease trans- mission.47 Despite these theoretical risks for such epidemics, it appears unlikely that they will actually occur.4,8 Only two arti- cles in the literature reported outbreaks after an earthquake. A malaria outbreak was reported after the 1999 Costa Rica earth- quake.47 In addition, after the 1994 Northridge earthquake, a coccidiomycosis outbreak was reported.47 It appears that mass vaccination campaigns, based solely on the fear of possible epi- demics are inappropriate. An epidemiological surveillance sys- tem should direct such interventions based on measured disease activity.

Disposal of Bodies

Decaying corpses represent a concern to the public and to health authorities; however, the belief that an increased risk of disease transmission exists due to decomposing dead bodies in the after- math of major disasters is a myth. Authorities should be aware that the health hazards related to unburied human remains, prin- cipally those of trauma victims, are negligible. Mass burials or cremations, which require the use of enormous amounts of fuel, destroy any evidence for future identification, and do not respect some religious rituals, are not justified on the basis of public health concerns.48,49 One situation in which handling human remains can represent a health risk is during epidemics of trans- mittable infectious diseases. Even in these situations, there is no reason to deprive families of their wishes to manage their dead relatives according to their customs, as long as they follow cer- tain safety measures.47 In addition, it is important to identify victims. This has relevance not only to regional governments but also to provide family members a greater degree of stability and closure in reference to their losses. Fingerprints, photographs, dental records, imaging, and DNA analysis can be used to help identify victims.49

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

574 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

Prevention

The most important factors required for reducing earthquake mortality reside in the institution of local and federal pre-event disaster mitigation and preparedness plans.15 These plans must include preparedness for early extrication of trapped victims and early treatment of immediate medical conditions, with effec- tive use of available resources. Implementation of an effective incident management system along with previously negotiated mutual aid agreements and transportation networks, including airlifts, can also influence the prognosis of victims.

Disaster preparedness plans should focus on ways to augment prehospital response to minimize the lethal impact of earth- quakes and to maximize life-saving potential within the first 24 hours after the disaster event. Prehospital providers must ini- tiate abbreviated patient assessment protocols designed for disas- ters rather than those used during everyday responses. Immediate needs assessments that are associated with an efficient organiza- tional structure are required to optimize the disaster response. This must be done with the coordination of state, local, national, and international agencies working together to achieve a com- mon goal. It is of great importance to incorporate data obtained from formal training exercises into the disaster planning process and to develop interorganizational relationships among entities that will participate in a disaster response.

The Joint Commission requires hospitals seeking their accreditation to include an evacuation strategy in their emer- gency management programs. Institutions must periodically test their evacuation plans by using drills that model the disruption of emergency departments’ and hospitals’ normal activities.50

Although the Joint Commission also requires that these types of disaster drills include the regional responders needed in an earth- quake, hospitals often do so in an ineffective manner, failing to realistically assess overall preparedness.

Critical actions taken before, during, and after a disaster may save lives and minimize property damage.10 Communities that are at risk for earthquakes must adopt effective response strate- gies to decrease losses after an event. Substantial numbers of victims are rescued by community members in the first hours after an earthquake. Therefore, to decrease morbidity and mor- tality, planners should train local residents in search and rescue activities. To reduce the number of deaths suffered by extricated victims, care should be provided for life-threatening injuries within the first 6 hours.10 Because insufficient and uncoordi- nated medical responses are often reported in the aftermath of earthquake events, more attention to a local medical response plan, such as the MDR program, is warranted. In addition, indi- viduals must secure items such as bookshelves and computers as an injury-prevention technique. Anchoring loose objects can substantially reduce injuries. After the Northridge earthquake, 16 hospitals were individually reviewed and reported that most of the injuries resulting in hospital admission were caused by falls or by being hit by objects.36 Many tools and devices are currently available for securing both heavy and smaller objects.9

Education in earthquake risk reduction should be a major focus for healthcare and community members. Educational pro- grams must include 1) the training of teams for search and res- cue operations and needs assessments; 2) how to identify safe sites where people can be relocated after the event; 3) train- ing healthcare professionals about conditions commonly seen in earthquake victims and treatments for them; 4) how to main- tain pharmaceuticals and supplies needed for the most common

medical conditions after an earthquake; 5) how to assess the structural safety of facilities that are essential in the operation of disaster responses (e.g., hospitals) and upgrading them as nec- essary; 6) how to plan for alternate water supplies; 7) preparing plans for the maintenance of viable vehicle transportation cor- ridors; 8) how to access and operate emergency communication systems; and 9) training teams to assess nonstructural damage and determine whether buildings are safe for reoccupancy.8

RECOMMENDATIONS FOR FUR THER RESEARCH

Although investigators examining the medical consequences of seismic events have made significant progress, additional work is necessary to further improve the care and outcomes of earth- quake victims. These endeavors will require a true transdisci- plinary approach involving participation by individuals from multiple medical, health, and nonmedical specialties. A few projects with the potential for achieving this goal follow.

A rapid and accurate estimate of casualty numbers and the extent of their injuries is necessary to coordinate the size and type of healthcare resources required to provide care to earth- quake victims. Delays in obtaining this information will lead to an inappropriate or inadequate medical and health response. Current estimates are based on generalized information gathered from previous seismic events and do not reflect individual varia- tions in building construction or population density. In addition, current casualty estimation computer programs do not provide estimates in real time and are not specific for earthquakes because they do not fit the model on which the software is based. New models are needed that can utilize population data for individ- ual locations and rapidly generate estimates based on observed building damage. Designing such models is possible using data- mining techniques combining new engineering building dam- age classification data and population injury information. Such innovative research has yet to be funded.

If a large-magnitude earthquake occurs in a densely popu- lated area, implementing a system for triaging medical casualties will be necessary. There are several triage systems available but data on their effectiveness are limited. Most research to date has focused on triage tool assessment using drills or models thought to approximate the types of victims caused by seismic activ- ity, such as individual trauma patients. More investigations are required of mass casualty triage system performance based on meaningful outcomes data in actual earthquakes to clarify which of these tools is most appropriate.

The field management of victims suffering from crush injuries remains controversial. If resources are plentiful, stan- dard treatment protocols apply. Under austere conditions, how- ever, the most appropriate approach to such victims is unknown, especially when supplies of intravenous fluids are limited. The roles for field fasciotomies, amputations, tourniquets applied to crushed extremities before extrication, and the use of hypertonic saline require additional study. Animal models for crush injury exist but have not been widely used to investigate these aspects of crush injury management.

The large number of victims generated by a powerful earth- quake will require hospitals to quickly increase surge capacity. Research has identified the major components of this process and programs exist to support some aspects of surge capacity, includ- ing the U.S. National Pharmaceutical Stockpile, community- based programs such as the MDR Project, the Emergency System

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 575

for Advanced Registration of Volunteer Health Professionals, and the creation of Federal Medical Stations in the U.S. It remains unknown, however, how effective these programs will be. Fur- ther investigation is needed to examine what interventions will be successful under actual disaster conditions. If current programs fail to adequately address surge capacity, disaster experts must identify new approaches that can meet the rapidly expanding demand for care.

Identifying the areas most heavily damaged by an earth- quake is important. These locations are not evenly distributed near the epicenter but are scattered over a wide area affected by the temblor. Because most deaths and injuries result from build- ing damage, quickly obtaining such information can provide responders with the locations where most victims will be found. In addition, it can suggest which hospitals remain functional and which are evacuating patients. The current technology using ground motion sensors that measure PGA and PGV can gener- ate shake maps within minutes of a seismic event. These maps depict the intensity of ground motion over a wide area. Although this information suggests areas of higher and lower probability for structural failure, it is insufficient to reliably predict building damage. More sophisticated software that incorporates ground motion along with soil conditions and types of building con- struction could yield a more accurate view of building damage and subsequent injury potential.

Disaster management focuses on needs assessments of the affected population, the efficient utilization of resources, the prevention of further adverse health effects, and the evalua- tion of relief program effectiveness to plan for future disasters.48

Improvements in disaster mitigation and responses will require a carefully organized, multidisciplinary evaluation of the conse- quences associated with these events. Therefore, a government- sponsored national or international research center for the mul- tidisciplinary evaluation and study of disasters is needed.

SUMMAR Y

Earthquakes are spontaneous events whose consequences are associated with an increase in morbidity, mortality, and costly property damage. Understanding the challenges that arise after seismic events can assist with planning, improving strategies for mitigating their effects, and assisting with the coordination of local resources in more effective ways. The medical management of earthquake victims remains a difficult challenge; however, the amount of data on this subject continues to increase, allowing the development of new recommendations. Implementation of field triage protocols (e.g., START and SAVE) and the use of a unified command system are vitally important in support- ing early organizational efforts in the chaotic environment that surrounds disasters. Acknowledging that hospitals will be over- whelmed during the first 24–48 hours helps responders prepare for that reality, supporting plans to augment surge capacity and decreasing victim mortality and morbidity. Hospital structures remain susceptible to seismic damage and must prepare and test reliable evacuation plans that incorporate effective strategies used successfully in previous events. Knowledge of common medical conditions that arise after an earthquake, such as crush syn- drome, respiratory symptoms, renal failure, fractures, and lac- erations can help healthcare professionals more effectively man- age these medical problems. In addition, refining plans for the management of chronic medical conditions and the appropriate

disposition of dead bodies is important to improve earthquake response.

REFERENCES

1. Kazzi AA, Langdorf MI, Handly N, White K, Ellis K. Earth- quake epidemiology: the 1994 Los Angeles Earthquake emer- gency department experience at a community hospital. Prehosp Disaster Med. 2000;15(1):12–19.

2. Peek-Asa C, Kraus JF, Bourque LB, Vimalachandran D, Yu J, Abrams J. Fatal and hospitalized injuries resulting from the 1994 Northridge earthquake. Int J Epidemiol. 1998;27(3):459– 465.

3. Schultz CH, Koenig KL, Noji EK. A medical disaster response to reduce immediate mortality after an earthquake. N Engl J Med. 1996;334(7):438–444.

4. Schultz CH. Earthquakes. In: Hogan DE, Burstein JL, eds. Dis- aster Medicine. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2007:185–193.

5. Risk Management Solutions. 2006 Kiholo Bay Hawaii Earth- quake RMS Event Report. 2006:1–16. Available at: http://www .eeri.org/lfe/pdf/usa Kona Hawaii EQ ReconReport.pdf. Ac- cessed March 1, 2009.

6. Shibata H. Disaster in relation to the Hanshin-Awaji earthquake: a comment on risk-informed design. Technol Law Insurance. 1999;4(1):31–36.

7. Naghii MR. Public health impact and medical consequences of earthquakes. Rev Panam Salud Publica. 2005;18(3):216– 221.

8. Perez E, Thompson P. Natural hazards: causes and effects. Lesson 2 earthquakes. Prehosp. Disaster Med. 1994;9(4):260–269.

9. Shoaf KI, Sareen HR, Nguyen LH, Bourque LB. Injuries as a result of California earthquakes in the past decade. Disasters. 1998;22(3):218–235.

10. Ramirez M, Peek-Asa C. Epidemiology of traumatic injuries from earthquakes. Epidemiol Rev. 2005;27:47–55.

11. Adams RD. Earthquake occurrence and effects. Injury. 1990;21(1):17–20.

12. Chan YF, Alagappan K, Gandhi A, et al. Disaster management following the Chi-Chi earthquake in Taiwan. Prehosp Disaster Med. 2006;21(3):196–202.

13. Gerstenberger MC, Wiemer S, Jones LM, Reasenberg PA: Real- time forecasts of tomorrow’s earthquakes in California. Nature. 2005;435:328–331.

14. Mignone AT Jr, Davidson R. Public health response actions and the use of emergency operations centers. Prehosp Disaster Med. 2003;18(3):217–219.

15. Pretto EA, Angus DC, Abrams JI, et al. An analysis of prehos- pital mortality in an earthquake. Disaster Reanimatology Study Group. Prehosp Disaster Med. 1994;9(2):107–117.

16. Roy N, Shah H, Patel V, Coughlin RR. The Gujarat earthquake (2001) experience in a seismically unprepared area: commu- nity hospital medical response. Prehosp Disaster Med. 2002; 17(4):186–195.

17. Yuechum C, Xin H, et al. WISTA: A Wireless Transmission Sys- tem for Disaster Patient Care. 2nd International Conference on Broadband Networks. 2005;2:1041–1045.

18. Noji EK, Kelen GD, Armenian HK, et al. The 1988 earthquake in Soviet Armenia: a case study. Ann Emerg Med. 1990;19(8): 891–897.

19. Garner A, Lee A, Harrison K, Schultz CH: Comparative analysis of multiple-casualty incident triage algorithms. Ann Emerg Med. 2001;38:541–548.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

576 ■ CA R L H. SC H U LT Z A N D SO L I S I S DEY NE S

20. Kahn C, Schultz CH, Miller K, Anderson C: Does START triage work? An outcomes-level assessment of use at a mass casualty event. Acad Emerg Med. 2007;14(Suppl 1):S12–S13.

21. Macintyre AG, Barbera JA, Smith ER. Surviving collapsed struc- ture entrapment after earthquakes: a “time-to-rescue” analysis. Prehosp Disaster Med. 2006;21(1):4–17.

22. Socna J, Sella T, Shaham D, et al. Facing the new threats of ter- rorism: Radiologists’ perspectives based on experience in Israel. Radiology. 2005;237:28–36.

23. Schultz CH, Koenig KL, Lewis RJ. Implications of hospital evac- uation after the Northridge, California, earthquake. N Engl J Med. 2003;348(14):1349–1355.

24. Surviving the Pakistan Earthquake: Perception of the affected one year later. Fritz Institute. 2006. Available at: http://www .fritzinstitute.org/prsrmPR-PakistanEarthquakeSurvey.htm. Accessed March 1, 2009.

25. Schultz CH, Koenig KL. Earthquakes and the practicing physi- cian. West J Med. 1992;157(5):591.

26. Benson M, Koenig KL, Schultz CH. Disaster triage: START, then SAVE – a new method of dynamic triage for victims of a catas- trophic earthquake. Prehosp Disaster Med. 1996;11(2):117–124.

27. Einav S, Feigenberg Z, Weissman C, et al. Evacuation priorities in mass casualty terror-related events: implications for contingency planning. Ann Surg. 2004;239(3):304–310.

28. Schultz CH, Koenig KL, Lewis RJ: Decision-making in hospital earthquake evacuation: does distance from the epicenter matter? Ann Emerg Med. 2007;50:320–326.

29. Schultz CH, Stratton SJ. Improving hospital surge capacity: a new concept for emergency credentialing of volunteers. Ann Emerg Med. 2007;49:602–609.

30. Mas Bermejo P. Preparation and response in case of natural disasters: Cuban programs and experience. J Public Health Policy. 2006;27(1):13–21.

31. Angus DC, Pretto EA, Abrams JI, et al. Epidemiologic assessment of mortality, building collapse pattern, and medical response after the 1992 earthquake in Turkey. Disaster Reanimatology Study Group (DRSG). Prehosp Disaster Med. 1997;12(3):222– 231.

32. Osaki Y, Minowa M. Factors associated with earthquake deaths in the great Hanshin-Awaji earthquake, 1995. Am J Epidemiol. 2001;153(2):153–156.

33. Vanholder R, van der Tol A, De Smet M, et al. Earthquakes and crush syndrome casualties: Lessons learned from the Kashmir disaster. Kidney Int. 2007;71(1):17–23.

34. Blaivas M, Kuhn W, Reynolds B, Brannam L. Change in dif- ferential diagnosis and patient management with the use of portable ultrasound in a remote setting. Wilderness Environ Med. 2005;16(1):38–41.

35. Brooks AJ, Price V, Simms M. FAST on operational military deployment. Emerg Med J. 2005;22(4):263–265.

36. Vanholder R, Sever MS, Erek E, Lameire N. Acute renal fail- ure related to the crush syndrome: towards an era of seismo- nephrology? Nephrol Dial Transplant. 2000;15(10):1517–1521.

37. Sever MS, Erek E, Vanholder R, et al. The Marmara earthquake: epidemiological analysis of the victims with nephrological prob- lems. Kidney Int. 2001;60(3):1114–1123.

38. Sever MS, Vanholder R, Lameire N. Management of crush- related injuries after disasters. N Engl J Med. 2006;354(10):1052– 1063.

39. Sever MS, Erek E, Vanholder R, et al. Clinical findings in the renal victims of a catastrophic disaster: the Marmara earthquake. Nephrol Dial Transplant. 2002;17(11):1942–1949.

40. Erek E, Sever MS, Serdengecti K, et al. An overview of morbidity and mortality in patients with acute renal failure due to crush

syndrome: the Marmara earthquake experience. Nephrol Dial Transplant. 2002;17(1):33–40.

41. Schultz CH, Koenig KL. Preventing Crush Syndrome: Assisting with field amputation and fasciotomy. JEMS. 1997; Feb:30–37.

42. Robertson PA. Prediction of amputation after severe lower limb trauma. J Bone Joint Surg Br. 1991 Sep;73(5):816–818.

43. Driessen B, Brainard B. Fluid therapy for the traumatized patient. J Vet Emerg Crit Care. 2006;16(4):1–24.

44. Klein D, Millo Y, Shuvurum A, Tzur H. The use of alkaline hyper- tonic saline solution for resuscitation of severe thermally injured patients (our experience). Ann Medit Burns Club. 1994;7(4): 194.

45. Sever MS, Erek E, Vanholder R, et al. Features of chronic hemodialysis practice after the Marmara earthquake. J Am Soc Nephrol. 2004;15(4):1071–1076.

46. Bowenkamp C. Coordination of mental health and commu- nity agencies in disaster response. Int J Emerg Ment Health. 2000;2(3):159–165.

47. Floret N, Viel JF, Mauny F, Hoen B, Piarroux R. Negligible risk for epidemics after geophysical disasters. Emerg Infect Dis. 2006;12(4):543–548.

48. Noji EK. The public health consequences of disasters. Prehosp Disaster Med. 2000;15(4):147–157.

49. Management of Death Bodies after Disaster: A Field Manual for First Responders, Washington, DC: PAHO;2006. Available at: www.icrc.org/Web/Eng/siteeng0.nsf/htmlall/p0880/$File/ICRC 002 0880.PDF!Open. Accessed March 1, 2009.

50. Mattox K. The World Trade Center attack. Disaster prepared- ness: health care is ready, but is the bureaucracy? Crit Care. 2001;5(6):323–325.

ADDITIONAL READING

Auf der Heide E. The importance of evidence-based disaster planning. Ann Emerg Med. 2006;47(1):34–49.

Bourque LB, Siegel JM, Shoaf KI. Psychological distress follow- ing urban earthquakes in California. Prehosp Disaster Med. 2002;17(2):81–90.

Conover WA. Earthquakes and the office-based surgeon. West J Med. 1992;157(1):79–82.

Delaney JS, Drummond R. Mass casualties and triage at a sporting event. Br J Sports Med. 2002;36(2):85–88.

Ellidokuz H, Ucku R, Aydin Y, Ellidokuz E. Risk factors for death and injuries in earthquake: cross-sectional study form Afyon, Turkey. Croat Med J. 2005;46(4):613–618.

Guha-Sapir D, Carballo M. Medical relief in earthquakes. J R Soc Med. 2000;93(2):59–61.

Gutierrez E, Taucer F, De Groeve T, Al-Khudhairy DH, Zaldivar JM. Analysis of worldwide earthquake mortality using multi- variate demographic and seismic data. Am J Epidemiol. 2005; 161(12):1151–1158.

Liang NJ, Shih YT, Shih FY, Wu HM, et al. Disaster epidemiology and medical response in the Chi-Chi earthquake in Taiwan. Ann Emerg Med. 2001;38(5):549–555.

Mahue-Giangreco M, Mack W, Seligson H, Bourque LB. Risk factors associated with moderate and serious injuries attributable to the 1994 Northridge earthquake, Los Angeles, California. Ann Epi- demiol. 2001;11(5):347–357.

Miyamoto M, Sako M, Kimura M, Kanno T, et al. Great earth- quakes and medical information systems, with special reference to telecommunications. J Am Med Inform Assoc. 1999;6(3):252– 258.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EA RT H Q UA K E S ■ 577

Ogawa K, Tsuji I, Shiono K, Hisamichi S. Increased acute myocar- dial infarction mortality following the 1995 Great Hanshin- Awaji earthquake in Japan. Int J Epidemiol. 2000;29(3):449– 455.

Peek-Asa C, Ramirez M, Seligson H, Shoaf K. Seismic, structural, and individual factors associated with earthquake related injury. Inj Prevent. 2003;9(1):62–66.

Pho RW. Management of the severely traumatised limb. Singapore Med J. 2003;44(6):277–279.

Porter K, Shoaf K, Seligson H. Value of Injuries in the Northridge earthquake. Earthquake Spectra. 2006;22(2):555–563.

Rodgers J, Foushee R, Terndrup TE, Gaddis GM. Research methods of inquiry. Acad Emerg Med. 2006;13(11):1183–1192.

Schultz CH, Koenig KL, Auf der Heide E, Olson R. Benchmarking for hospital evacuation: a critical data collection tool. Prehosp Disaster Med. 2005;20(5):331–342.

Shoaf KI, Peek-Asa C. Survey research in disaster public health. Prehosp Disaster Med. 2000;15(1):57–63.

Shoaf K, Sauter C, Bourque LB, Giangreco C, Weiss B. Suicides in Los Angeles County in relation to the Northridge earthquake. Prehosp Disaster Med. 2004;19(4):307–310.

Sieh K. Sumatran megathrust earthquakes: from science to saving lives. Philos Transact A Math Phys Eng Sci. 2006;364(1845):1947–1963.

Wooding S, Raphael B. Psychological impact of disasters and terrorism on children and adolescents: experiences from Australia. Prehosp Disaster Med. 2004;19(1):10–20.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Tsunamis

Samuel J. Stratton

OVERVIEW

A tsunami is a series of waves created when a massive volume of ocean water is rapidly displaced. Tsunami waves are created as the mass of displaced water radiates relative to gravitational forces across an ocean or sea. Commonly, submarine earthquakes are associated with tsunamis, but other geophysical events causing mass displacement of water will generate tsunami waves. These may consist of underwater landslides, volcanic eruptions, mete- orite impacts, and submarine explosions, including nuclear det- onations.1 Tsunami events are best classified as sudden-impact disasters.

The term tsunami is of Japanese origin from the words “tsu” meaning harbor and “nami” meaning wave.1 Tsunamis usually occur in a series of nonrhythmic waves as opposed to a single wave. The first tsunami wave to approach a shore is often not the largest in the series. In open ocean, tsunami waves can have a wavelength of up to 700 km (435 miles) and propagate at speeds of 640 km (400 miles) per hour.2 In open water, tsunamis may have a wave height (amplitude) of only a few centimeters. Upon reaching shallow water, however, the waves slow and build to heights with inertial energy well beyond those of wind-generated waves (Figure 36.1).

Tsunamis can cause severe damage to coastal areas as they “run-up” onshore and dissipate wave energy caused by the mas- sive displacement of ocean water. The destructive effect of a tsunami is controlled by the submarine topography in front of the land area that the tsunami approaches. A sloping beach or land positioned on a submarine ridge will sustain damage from the direct impact of high waves, whereas a wide and shallow con- tinental shelf will absorb most of the wave energy and protect a land mass behind it.2 Tsunamis are different from wind waves and tidal movements because of the large amount of energy they contain and the long, wide character of the waves.

Tsunamis present in two different forms: local and ocean- wide waves. Local tsunami waves arise when earthquakes or undersea disruptions occur near a shore. Local tsunamis can be occur after an earthquake or subsurface event has been detected by residents of the shoreline that is at risk. Local tsunami waves

run-up to shore with little warning other than the preceding event causing displacement of water. In the 2004 Indian Ocean Tsunami, 130,000 persons in the coastal area of Aceh Province, Indonesia near the originating earthquake were killed by direct effects of the tsunami and earthquake.3 The same earthquake generated an ocean-wide tsunami that killed 145,000 persons on distant shores throughout the Indian Ocean in Thailand, the Maldives, India, and Sri Lanka.3 Ocean-wide tsunamis are gen- erated by distant earthquakes or submarine events that may or may not be felt by affected shoreline residents. A classic exam- ple of an ocean-wide tsunami is the 1960 Chilean tsunami that was generated by a magnitude 9.5 earthquake off the coast of southern Chile, causing devastation in Chile. The tsunami waves spread across the Pacific striking Hilo Hawaii 14.8 hours after the initial quake, killing 61 persons with the highest wave measuring 10.5 m (35 ft); it continued on to run-up on the coastal area of Sanriku, Japan killing 142 persons.4

Historically, tsunamis have occurred in all the oceans of the world, with the coast of Maine struck in 1926 and ancient reports of tsunamis in the Mediterranean Sea.5 Most tsunamis of con- sequence strike in the Pacific Ocean. Here they cause significant destruction because ocean topography frequently includes land masses on the edges of ocean canyons rather than on a continen- tal shelf, as is the case with the east coast of the U.S.6 The Pacific Rim is also highly active with earthquake activity, raising the risk for tsunamis.7

STATE OF THE ART

On-shore Tsunami Effects

As tsunami waves run-up onto coastal areas, there is churning of silt, sand, and organic matter that is then thrown onto the shore. Onrushing seawater, sand, and debris cause direct dam- age to structures and roads (Figure 36.2). In addition, persons in the path of the tsunami waves are swept into on-rushing waves. They suffer drowning and aspiration of seawater with suspended material as well as blunt injury from heavy debris captured by the waves. Tsunami waves destroy and damage essential service areas,

578 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TS U NA M I S ■ 579

Normal situation

Disturbance

Coastal flooding

Fault

Flood tide

Ebbing tide

Earthquake hypocenter

Overlapping primary waves

Raising

Sinking

Figure 36.1. Diagram showing the generation of tsunami waves by a submarine earthquake. With disruption of the ocean floor, energy is transferred to the water mass causing displacement of a large mass of water. The energy of the ocean floor disruption is then dispersed in the form of a water mass or wave. When the tsunami strikes a shoreline, the energy stored in the form of water mass is dissipated on the shoreline. See color plate.

waste management systems, flood control systems, and struc- tures. Because tsunamis are high-energy waves, boats and other offshore objects can be torn from moorings. These objects, along with anchors and other fittings, are thrown violently against the shore causing direct damage to sea walls, buildings, and other objects in the path of the wave. Once tsunami waves run-up on shorelines, it is not uncommon for automobiles, buses, refriger- ators, trees, large rocks, and other debris to be thrown against any object in the path of the waves. Many coastal areas have electrical lines close to shore, raising the risk for disruption of these lines and secondary electrical injury to humans and ani- mals. The rolling, crushing nature of tsunami waves can cause release of biological toxins from storage containers, automo- biles, electrical devices, gasoline stations, and other damaged sources.8

As tsunami waves recoil, debris and toxins are pulled into the ocean with forces that are close to that of the force of the oncoming waves (Figures 36.3, 36.4). Because tsunami waves present in series, debris, chemicals, and suspended material are churned and thrown back and forth to the shoreline and ocean. In addition to damage from heavy objects, toxins, and silt, tsunami run-ups are associated with on-shore fires as natural gas lines and flammable materials and liquids are exposed.8

Figure 36.2. During tsunami events heavy debris is churned onto and off shore as shown by this large coral rock thrown on shore in the Solomon Islands during the 2007 tsunami. (Source: United States National Oceanic and Atmospheric Administration, by John Beba, Woodlark Mining Limited.) See color plate.

Tsunami waves have little effect on deep-water structures as they traverse open waters. They dissipate large amounts of energy as they run-up to shore. Near shore coral reefs and ecosystems can be severely affected by tsunami strikes, causing long-term loss of near shore marine life and habitats that can take cen- turies to replace.8 Loss of coastal marine ecosystems can lead to loss of fishing grounds and affect tourism and other human industries. Tsunami run-ups in industrialized coastal areas can cause release of environmental toxins including paints, oils, gaso- line, detergents, and solvents. Release of these toxins can affect both marine and human populations for extended periods. Fur- thermore, release of human and animal biological waste can contaminate and damage food and water supplies for extended periods.

Tsunami waves striking industrialized areas often break nat- ural gas and electrical supply lines. As water recedes, the risk of fire is high from ignition of natural gas supply lines and other combustible material left in the wake of the waves. Many of the

Figure 36.3. Sri Lanka 2004 tsunami wave striking shore, submersing bases of trees and low-lying buildings. Note the forceful churning of water as the wave is striking. (Source: United States National Oceanic and Atmospheric Administration, by Chris Chapman, Cambridge, UK.) See color plate.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

580 ■ SA M U E L J. ST R AT TO N

Figure 36.4. A Sri Lanka 2004 tsunami wave receding and pulling debris and objects into the ocean with forces that nearly equal the energy of the incoming wave. (Source: United States National Oceanic and Atmospheric Administration, by Chris Chapman, Cambridge, UK.) See color plate.

secondary fires are caused by ignition of oil products, including gasoline, diesel fuel, plastics, and solvents. Because water alone will not extinguish these types of fires and these burning materi- als are lighter than water, subsequent on-rushing tsunami waves can spread fire and burning debris throughout the affected area. Both thermal and chemical burns of survivors are to be expected when tsunami waves strike industrialized shorelines.

IMMEDIATE TSUNAMI RISKS FOR HUMANS AND ANIMALS

Drowning is the obvious risk for humans and animals when a tsunami strikes an inhabited shoreline. Many persons survive the initial effects of an oncoming rolling wave but are swept to sea to drown in open ocean waters that are churning as tsunami waves strike. Because of this phenomenon, children, women, and the disabled are more likely to drown during a tsunami strike because they may lack the strength to hold onto stationary objects and

Figure 36.5. This picture taken in Karaikal, India after the 2004 Indian Ocean tsunami shows the debris field left by the tsunami waves. High- force movement of heavy debris and sharp objects within a tsunami wave cause major injuries to exposed humans and animals. (Source: United States National Oceanic and Atmospheric Administration, by Joseph Trainor, University of Delaware, Disaster Relief Center.) See color plate.

Table 36.1: Immediate Tsunami Injury Risks

Submersion Drowning Aspiration lung injury Tympanic membrane rupture

Blunt force injury Crush injury Closed head injury Solid organ blunt force injury Spinal injury

Cervical spine injury Compression fracture thoracic and lumbar spine

Orthopedic injury Long bone fracture and contusion Dislocation of shoulder, elbow, knee, digits Amputation of digits, hands, feet Pelvic fracture

Eye injury, both blunt and penetrating Soft tissue injury

Laceration Contusion/abrasion

Penetrating trauma Foot injury from sharp debris Penetrating thorax/abdominal injury

Burn Thermal Chemical

Dental trauma

avoid being swept out to the open ocean.3 Attempts to survive by clinging to floating debris are often futile because the waves strike and recoil from the shoreline in a churning, mixing motion.

In addition to drowning, death occurs by blunt force injury as heavy objects are thrown against persons and structures as the advancing waves hit the shore. Fatal crush injury and blunt trauma are reported as the second most common cause for death related to tsunamis.9 Mortality is also associated with impale- ment by large pieces of glass and sharp, protruding metal, wood, and vegetable matter formed and exposed by the forces of the tsunami waves (Figure 36.5). Descriptions of persons found impaled through the torso on bamboo shoots, metal rods and pipes, and splintered tree branches and trunks are not uncom- mon. Closed head injury secondary to blunt trauma and blunt force solid organ injury with subsequent hemorrhage are also recognized causes for fatal injury related to tsunami events.9

Building collapse from tsunami waves undermining structural foundations is another cause for crush injury.9

Prolonged serious medical conditions resulting directly from tsunami events are uncommon and most persons either survive to be “ambulatory” or are killed suddenly during the event.3,10

Table 36.1 lists the injuries that can be anticipated in survivors. Lacerations, soft tissue injuries, and orthopedic injuries are most often encountered in the aftermath of a tsunami strike. Because those suffering these injuries are in a contaminated environment, tetanus prophylaxis is a primary concern. In opposition to stan- dard practice in which most soft tissue wounds and lacerations would be immediately sutured or closed, wounds presenting dur- ing the acute disaster phase of a tsunami event are often left open to heal by secondary intent or delayed primary closure after thor- ough irrigation and debridement because of the contaminated nature of the injuries and concern for retained foreign bodies.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TS U NA M I S ■ 581

Eye injuries secondary to flying debris and direct injury are common during tsunami events. Frequent episodes of tympanic membrane rupture are also reported and are most likely caused by pressure gradients when one is submersed in the water of a rolling wave. Facial injuries and dental trauma are also to be expected.

Orthopedic injuries are predominately long bone fractures but also include pelvic fracture and spinal injury (including cer- vical spine injury and compression fractures of thoracic and lumbar spine regions). Amputation of digits and partial hand and foot amputations are also commonly reported and are the result of crushing and severing caused by heavy floating debris churned within tsunami waves. Dislocations of the shoul- der, elbow, and knee have also occurred often during tsunami events.

Pulmonary injury due to aspiration of contaminated water is frequent in relation to tsunamis.3 Although pneumothorax due to blunt chest injury and barotrauma would be expected, it has not been reported as a predominate injury. Thermal and chemical burns are reported and expected because caustic toxins can be released from containers and fires occur in the aftermath of a tsunami strike.

PUBLIC HEALTH ASPECTS OF TSUNAMI EVENTS

The immediate public health concerns with tsunami events include loss of shelter, food, water, and clothing supplies. Starva- tion and hypothermia or sunburn and sun exposure are common public health concerns immediately after a tsunami strike.9 Sur- vivors are often left partially clothed without access to potable water or food. Because debris, including broken glass, splin- tered trees, and destroyed buildings, are distributed throughout the immediate environment by the waves, mobility is limited and access to food, water, and shelter is difficult. After tsunami events, initial public health activity generally include environ- mental health actions to develop safe, protected shelter and ready access to drinking water and clothing.3 Providing food supplies appropriate to the cultural norms for the area struck by the tsunami is an important public health priority.

When the immediate needs of the surviving population have been addressed and survivors have moved to safe locations, orga- nized assessment of health-related needs should be conducted. Initial assessment may be by survey of shelters and survivor collection points, but random-cluster analysis of the affected community is often preferred as soon as feasible.11 Knowledge of local health risks and challenges is essential in planning and con- ducting a rapid health analysis. For example, if it is known that malaria is endemic in an area struck by a tsunami, those doing a health assessment would survey for accessibility to protective mosquito netting for sleeping quarters and access to DEET or other appropriate mosquito repellents. An interesting illustra- tion of knowing important local health hazards occurred during the recovery phase of the 2004 Indian Ocean tsunami. Outside experts made unnecessary strong recommendations for inten- sively complicated cholera vaccination programs in areas struck by the tsunami when cholera was not a threat and had not been present in the area for decades.12 Information obtained during rapid public health assessments of an affected population is used to plan and implement immediate health responses. Table 36.2 lists the elements most often addressed during a rapid health assessment following a sudden onset event such as a tsunami.

Table 36.2: Elements of a Rapid Epidemiologic Assessment

1. Determine the overall impact of the event Geographical extent Number of affected persons Estimated duration

2. Assess the impact on health Number of casualties Number injured Number with illness Number well and unaffected

3. Determine the integrity of the healthcare system 4. Determine the specific health needs of the survivors 5. Assess the disruption of essential services that contribute to public

health Water Power Sanitation Communication Shelter Food

6. Determine the extent of resources needed by local authorities for adequate response and recovery to the disaster

The focus of these health assessments is to make prompt esti- mates of the health needs of an affected population.11 Visual inspection, interviews with key personnel, and surveys are the mainstay tools for rapid health assessments.11

Examples of the importance of health assessments in tsunami events is effectively illustrated by the World Health Organiza- tion evaluations done during the December 2004 Indian Ocean tsunami. Initial rapid health assessments done immediately after the event showed that burial of the dead, sheltering survivors, waste and debris removal, and provision of water and food were crucial first efforts to prevent public health problems within the surviving populations. Later health assessments showed that malnutrition, childhood diarrhea, hazardous waste manage- ment, and disposal of rotting animal carcasses were a prior- ity. Importantly, rapid health assessments detected measles cases among children in eastern areas of the affected area, allowing for an intensive measles vaccination program that prevented a measles outbreak among a generally unvaccinated pediatric pop- ulation.13

TSUNAMI WARNING SYSTEMS

Although tsunamis can be classified as sudden onset events, there are often periods of warning before a tsunami wave impact. Local tsunami run-up to a coastal area is often preceded by a submarine earthquake or volcanic eruption that can be felt by those occupying hazardous areas of the shoreline. Understanding the association of tsunamis with earthquakes and other under- water disturbances allows for escape to higher ground or inland areas.

As ocean-wide tsunami waves approach a land mass there is often a profound receding of the water along a shoreline as the tsunami rolls in. Understanding an association between pro- found receding of shore waters and incoming tsunami waves can allow for initial movement of those at risk away from the shore (Figure 36.6). There have been many anecdotal reports of ani- mals unexplainably rushing to higher ground minutes before the

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

582 ■ SA M U E L J. ST R AT TO N

Figure 36.6. The initial receding of water from shore before the run- up of a tsunami wave during the 2004 Indian Ocean tsunami. During this “warning” phase of a tsunami run-up, rocks and sand that are normally submerged along the outer shore become exposed. This receding of the ocean serves as a warning for those along the shore to move to higher ground to avoid a possible incoming tsunami wave. (Source: United States National Oceanic and Atmospheric Adminis- tration, by Chris Chapman, Cambridge, UK.) See color plate.

approach of a tsunami as if they are able to sense the incoming wave.

In 1949, the United States created an organized official tsunami warning system with headquarters located in Hon- olulu.14 In 1960, an earthquake occurring off the southern shore of Chile caused severe local damage and generated an ocean- wide tsunami that impacted Hilo, Hawaii, resulting in signif- icant destruction and killing approximately 60 residents. The same tsunami waves proceeded to Japan where they inflicted sub- stantial damage and killed 200 people. After this 1960 tsunami event, a coalition of Pacific nations was formed to develop a warning system to prevent future damage by ocean-crossing tsunami waves. With support of the United Nations, the inter- national Pacific Tsunami Warning Center (PTWC) was devel- oped.14 Until 1967, the PTWC had responsibility for warning Pacific nations of ocean-wide tsunamis. In 1967, following a severe 1964 Alaskan earthquake and tsunami, the West Coast and Alaska Tsunami Warning Center (WC/ATWC) was estab- lished. Today, the WC/ATWC has responsibility for issuing tsunami warnings for Alaska, British Columbia, Washington State, Oregon, and California. The PTWC monitors the rest of the Pacific.15,16

When the 2004 Indian Ocean tsunami occurred, an orga- nized warning system did not exist for the Indian Ocean area. Subsequently, the PTWC has added the Indian Ocean, South China Sea, and Caribbean to those areas monitored and warned of tsunami risks.15

The PTWC currently has 26 participating international member states.16 Multiple forms of technology are used to deter- mine tsunami risks and events. These technologies include earth- quake sensing and information devices, drifting and moored ocean data buoys, satellite observation equipment, Argo float- ing buoys, integrated on-land observation sites, high-frequency coastal radar, and the Shore Coastal-Marine Automated Net- work. Equipment supporting this network is located in light- houses and shore stations that monitor weather and seismic activity.16,17

A standard terminology has been developed for tsunami alerts that are released from tsunami warning centers. A “tsunami

advisory” indicates that a threat exists but that no tsunami has been detected. A “tsunami watch” indicates that an earthquake or other high-risk tsunami-generating event has been detected and coastal areas are to standby for further information and alerts. A “tsunami warning” indicates that a tsunami has been generated or that conditions are serious enough for coastal communities to take responsive actions. Information generated with a tsunami warning will include earthquake magnitude, originating loca- tion, and arrival times of waves.18

MITIGATING THE EFFECTS OF TSUNAMIS

Tsunami waves are hazards unique to coastal areas of the world. They are a predictable hazard because the origins of tsunamis are known to be earthquakes, submarine volcanoes, landslides, and other sources for displacement of large volumes of water. Human inhabited coastal regions that are in areas of geophysical activity places persons at direct risk for tsunami disaster effects. The coastal region known as the Pacific Rim is a well-recognized tsunami hazard zone.

Human habitation and activity in low-lying coastal lands puts people at direct risk for tsunami injury and loss. Further- more, recreational use of low-lying coastal lands places visitors to these areas at risk for injury and death from tsunamis. Build- ing structures for human habitation near low-lying shorelines in tsunami hazard areas raises the risk for death and injury as well as economic loss from tsunamis. As with many hazards, ignorance of the threat or ignoring the risks posed by tsunamis leads to inevitable calamity. Such consequences can be avoided by simple community planning and education.

Japan, a country that has been repeatedly struck by tsunamis, has been environmentally proactive in protecting human popu- lations from the waves’ effects by building sea walls along shore- lines facing open ocean. Tsunami sea walls range up to 4.5 m (13.5 ft) in height and afford an initial break for potential incom- ing tsunami waves19 (Figure 36.7). Using the same rationale, open low-lying beaches in areas of southern California in the United States utilize earthen barriers behind open beach to dissipate the energy of incoming storm waves and potential tsunamis. Maintenance of natural wet lands, forest, and vegetation that often lie immediately adjacent to open beach areas helps dissi- pate the energy of incoming tsunami waves and further affords protection for inland structures and habitats. Building standards and laws that limit the use of low-lying coastal lands for con- struction of houses, factories, schools, airports, energy plants (including nuclear power plants), and other essential human activity is a proactive method for decreasing the risk posed by tsunamis. Despite this, human encroachment into coastal zones has placed much of the world’s population at risk for injury and death due to tsunamis.20–22 Such decisions are characterized by the need to satisfy political and short-term economic goals rather than the need to implement rational decisions regarding housing and industrial zoning. Individuals responsible for these decisions often lack an understanding of the potential risks.

Community planning for the emergency response to poten- tial tsunami events is important for decreasing risks. As with planning for other disasters, initial community efforts should include a hazards assessment of the affected area. Low-lying coastal lands facing open ocean are at obvious risk for tsunami run-up. In addition, harbors and waterways must also be

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TS U NA M I S ■ 583

Figure 36.7. A protective sea wall constructed to repel and limit poten- tial tsunami damage along the shoreline of Nice, France. This tsunami sea wall was constructed after Nice was struck by a tsunami triggered by a construction-generated shoreline landslide. (Source: author.) See color plate.

considered as these entities have the potential to transmit and sometimes funnel the energy of a tsunami inland. Topographical mapping of shorelines is particularly helpful in identifying poten- tial areas at risk. Generally, areas that are less than 10 m (32.8 ft) above sea level are at highest risk for tsunami damage.19

The key element of tsunami disaster planning is providing a strategy for rapidly evacuating persons in the path of an incom- ing wave to locations inland and at higher elevation (>10 m). As noted earlier, tsunamis are sudden-impact events, but usu- ally can be predicted immediately prior to occurrence. A system for warning those at risk of an incoming tsunami after a trig- gering event (example, an offshore earthquake) will allow for protective evacuation to higher ground inland. In many areas of the Pacific Rim, warning sirens have been installed to rapidly alert those in an area of immediate risk. Warning posters and signs identifying the most efficient route to higher ground are also placed in high-risk areas to facilitate evacuation. Placement of evacuation route direction signs are important interventions. It has been repeatedly noted that many people become direc- tionally disoriented when incoming tsunami waves strike. Total evacuation of a tsunami run-up zone is important as anyone in low-lying areas is at risk for injury or death. This includes emer- gency first responders and security-law enforcement personnel. More sophisticated plans in developed countries not only include evacuation of emergency first responders and security person- nel to higher ground, but simultaneous movement of essential portable equipment with the personnel.21,23

Key to the success of potential emergency evacuation plans is pre-event education of those at risk. The resident population must know basic information about tsunamis. Knowledge of the meaning of warning sirens allows persons to respond more quickly when the alarm is sounded. In addition, it is important to educate those visiting, inhabiting, or working in coastal areas that extreme and sudden raising or receding of the ocean shoreline can be a clue to an incoming tsunami. Persons in tsunami haz- ard areas should be aware of signage that directs people from the risk zone. Other important tsunami information includes under-

standing that the waves come in series and that the first tsunami wave may not be the largest. Furthermore, intervals between waves are not constant and can be markedly asynchronous. After arrival of the initial tsunami wave, further waves in the series can be spread out and may not strike the shore for more than 2 hours. Because of the predictable delay in arrival of the com- plete tsunami wave series, rescue personnel and equipment are generally held back 2 hours from the onset of the first wave to prevent their becoming victims.

It has been noted that tsunami warnings sometimes attract sightseers to the coastal area, causing impairment in attempts to evacuate. Control of evacuation perimeters is important. Communication by standard landlines and wireless (cellular) telephones is often ineffective due to system overload and not available to emergency operations and rescue personnel. Alter- nate communication methods using radio systems in redundant configurations should be considered for potential responders. Organized amateur radio operators have been valuable in help- ing communicate and facilitate coordination of response in past tsunami events.

Recovery efforts are often prolonged following tsunamis and, after injured survivors have received medical care, mimic public health emergencies. Mortality rates have been high in proportion to nonevacuated survivors among those directly in the path of a tsunami waves. Management of the dead and accounting for those lost and probably washed out to sea is an early recovery concern along with providing care to survivors.24 For the man- agement of survivors, ambulatory care supplies such as tetanus vaccine, sterile irrigation solutions, fracture splinting materials, and dressings are often the first to be depleted.10 The disas- ter’s impact on the mental health of victims are a particular concern.

Populations that must be considered when planning for tsunamis include visitors and tourists. Many coastal areas are popular vacation locations and those visiting may not be famil- iar with the risk. This was particularly true with the 2004 Indian Ocean tsunami event. In some areas impacted by this disaster, approximately half of the survivors requiring medical services were tourists and visitors.3 Basic response information for visi- tors and tourists in tsunami hazard areas should be considered part of comprehensive planning.

RECOMMENDATIONS FOR FUR THER RESEARCH

Although standard plans have been developed for emergency evacuation of areas threatened by incoming tsunamis, there has been little research into the quickest and safest evacuation methods. General recommendations suggest those threatened by tsunamis should seek higher ground, but time taken in moving to higher ground may not be as effective as moving into upper stories of solidly constructed multistory buildings. The develop- ment of safe and effective maneuvers for avoiding personal injury from tsunamis represents a significant future research oppor- tunity.

The local and international management of large-scale tsunami events is an area that research can help improve. In January 2007, Claude de Ville de Goyet published a comprehen- sive article describing health lessons from the 2004 Indian Ocean tsunami. His observations are summarized in the following para- graphs. This tsunami event provides a wealth of information.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

584 ■ SA M U E L J. ST R AT TO N

Between 227,000 and 275,000 persons were lost or died as a result of the tsunami, with 1.7 million displaced.3,12 The tsunami affected several countries and varying cultures, with Indonesia and Sri Lanka in states of government instability at the time of the tsunami strike.12 Although thousands of reports and obser- vations of the event are available in journals or on the Internet, few scientifically rigorous studies have been published at the time of this writing. The tsunami-related issues noted by de Ville de Goyet form a basis for a core research agenda for future tsunami disaster research.

De Ville de Goyet provides the following observations in his paper on the 2004 Indian Ocean tsunami.

1. Funding was not a primary obstacle to an effective relief response. At a global average, $7,300 was committed per affected survivor. Yet external responses to the event fell short of being effective, suggesting that abundant finan- cial resources and technology do not guarantee a successful recovery effort.12

2. Few decisions were made based on needs assessments. Accountability of many nongovernmental organizations and United Nations agencies was to their donors rather than to the survivors and local governments. Decisions by outside agencies regarding the types of donations and aid offered were based on political pressures and media influences as opposed to basic epidemiological needs assessments and evaluations.12

3. The national public health capacities of the affected coun- tries were minimally impacted by the disaster except for Aceh Province, Indonesia. In Banda Aceh, which was also impacted by earthquake forces, there was loss of healthcare deliv- ery resources. Otherwise, infrastructure damage and human injury and loss occurred in the coastal areas, with inland areas remaining intact. The tsunami did not damage hospi- tals and public health resources directly. During a tsunami event, people who are in the coastal impact areas drown, die of trauma caused by loose debris, or survive with injuries but remain ambulatory.10,12 Local, organized responses to tsunami events using inland resources that withstand or are not affected by the tsunami waves have been more successful than medical assistance organized from outside a commu- nity.

4. For the Indian Ocean tsunami event, international human- itarian standards were not adapted to local contexts. The “Sphere Handbook” is an internationally accepted stan- dard for disaster response that uses a needs-based approach to compensate for disaster losses.25 International standards published in the Sphere Handbook rely on a strong rights- based approach, which was not a predominate norm in many of the countries affected by the tsunami.

The rigid application of lofty international standards to local situations without adapting to local norms during the Indian Ocean tsunami event caused negative consequences. Respond- ing international organizations targeted populations they could access easily for which the standards could be met, rather than seeking out those populations in locations more difficult to reach. This resulted in an over concentration of resources in urban areas. Furthermore, tourists and refugees became a primary focus as opposed to those locals trying to survive in rural regions and who experienced a more primitive existence prior to the event. There was also an overextension of the emergency phase and delay in

the recovery phase. The continued influx of donated resources provided an incentive to focus on obtaining more material and volunteer aid rather than moving to the more difficult recov- ery actions required. The international response resulted in a dilemma for local health departments and providers because the medical standards provided during the international response could not be sustained after relief efforts ceased. The disaster response medical resources donated and supplied by outside organizations exceeded the local standards available before the event.12

The emergency construction of duplicate health centers and medical clinics by international responders resulted in competi- tion between these agencies and was an example of inappropri- ate international aid.12 Those providing international aid quickly overwhelmed Aceh and Sri Lanka, causing confusion and frustra- tion for victims of the disaster. Further, the incoming aid work- ers and equipment added to the stress of managing the event for local emergency operations administrators. This increased demand for logistical support resulted in a second disaster, as local managers struggled to coordinate the invading relief work- ers and their equipment.12

Marginalization of local health authorities by international responders was another problem that occurred. On-site access to foreign assets, such as air transportation, equipment, and communication systems was limited to international responders. Those in the local response groups who were familiar with and culturally aware of the affected populations were left to struggle using pre-event methods and resources. As already noted, local responders not only had to manage the disaster with existing limited resources, but had the added burden of attempting to coordinate and direct incoming responders from the interna- tional sector.12

An additional factor compounding the difficulties caused by international response groups in the aftermath of this event was the overstating of epidemic risks. The occurrence of major secondary epidemics following sudden-impact disaster events is not the norm.26 Although there was no evidence for epidemic risks during the tsunami event, humanitarian agencies stimu- lated fear of epidemics and diverted attention from recovery efforts. Rather than focusing on basic disaster medical science using surveillance techniques and health education, the interna- tional effort concentrated on immunizations for cholera, which is logistically and technically complex. This cholera immuniza- tion campaign resulted in the loss of scare operational resources for a nonexistent threat, ignoring more obvious threats such as the large numbers of children who were not immunized for measles. This very real hazard was minimized by some inter- national “experts.” The little-publicized but substantial threat from measles was identified and immunization was successfully conducted using standard field surveillance and immunization programs for high-risk populations.27 As demonstrated here, health priorities should be based on sound public health hazard and risk assessments that usually require field work as opposed to theoretical pontification. For the affected population, repairing the environmental (water and sanitation) and economic (fishing and food production) infrastructure proved to be an immediate health priority.12

Medical and public health problems encountered following tsunamis have been described in a general sense, but primarily for the acute recovery or response phases. The long-term effects on health have been minimally described and this area represents a great opportunity for further research.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

TS U NA M I S ■ 585

SUMMARY

Tsunami waves represent high-energy massive movements of water generated when enormous volumes of ocean are displaced by events such as earthquakes and submarine volcanic eruptions. The coastal regions of the Pacific Ocean are the areas of high- est risk for tsunamis, but these disasters can strike anywhere that an ocean or sea meets a coastline. Injury and destruction from tsunamis occur as the waves strike the shoreline. The most common cause of death from tsunamis is drowning, followed by blunt injury resulting from loose debris thrown inland by tsunami wave run-up.

The essential elements for prevention of tsunami injury and damage are limiting construction in exposed coastal areas, devel- oping sea walls and coastal protection barriers, and educating those at risk to seek higher ground immediately when the threat of tsunami is high. Recognizable warning systems and planned evacuation routes are keys to tsunami-response planning. It is important to understand that tsunami waves come in series and that the first wave to strike shore is not necessarily the largest of the series.

As with other sudden-impact disasters, most of the mortality related to the event occurs immediately with the onset of the event. The majority of survivors requiring medical resources are not critically ill. Most are in need of ambulatory medical care and attention to existing chronic diseases (such as diabetes) that were present prior to the event. Predominate immediate public health needs are for shelter, water, food, clothing, and mental health support. Epidemics are uncommon in relation to tsunami events. Local health agencies and governments are best at managing the response to a tsunami event and international aid organizations should remain in the background and coordinate their responses with local authorities.

REFERENCES

1. National Oceanic and Atmospheric Administration. Tsunami Vocabulary and Terminology. Available at: http://www.tsunami. noaa.gov/terminology.html. Accessed November 14, 2008.

2. Ross DA. Tsunami Primer. Woods Hole Oceanographic Institution. Available at: http://www.whoi.edu/page.do?pid= 12462&tid=282&cid=7258&print=this. Accessed November 14, 2008.

3. World Health Organization. TRIAMS Final Report. Available at: http://www.ifrc.org/docs/pubs/disasters/triams-bangkok-en. pdf . Acessed November 23, 2008.

4. World Health Organization. Emergency and Disasters Data Base. Available at: http://www.em-dat.net/. Accessed November 14, 2008.

5. National Geophysical Data Center. Tsunami Events Full Search. Available at: http://www.ngdc.noaa.gov/nndc/. Accessed November 14, 2008.

6. National Oceanic and Atmospheric Administration. Tsunamis. Available at: http://www.noaa.gov/tsunamis.html. Accessed November 14, 2008.

7. U.S. Geological Survey. National Earthquake Information Center-NEIC. Available at: http://earthquake.usgs.gov/regional/ neic/ Accessed November 14, 2008.

8. National Oceanic and Atmospheric Administration. Potential Ecological Impacts of Indian Ocean Tsunami on Nearshore

Marine Ecosystems. Available at: http://www.noaanews.noaa. gov/stories2005/s2362.htm. Accessed November 14, 2008.

9. World Health Organization. Injuries and Disability: Priorities and Management for Populations Affected by the Earthquake and Tsunami in Asia. Available at: http://www.who.int/violence injury prevention/other injury/tsunami/en/index.html. Accessed November 14, 2008.

10. Stratton SJ, Tyler RD. Characteristics of medical surge capac- ity demand for sudden-impact disasters. Acad Emerg Med. 2006;13:1193–1197.

11. Wetterhall SF. Noji EK. Surveillance and epidemiology. In: Noji EK, ed. The Public Health Consequences of Disasters. New York: Oxford University Press; 1997:37–64.

12. de Ville de Goyet C. Health lessons learned from the recent earth- quakes and tsunami in Asia. Prehosp Disaster Med. 2007;22:15– 21.

13. World Health Organization South Asia Tsunami Situa- tion Reports. Available at: http://www.who.int/hac/crises/ international/asia tsunami/sitrep/en/. Accessed November 14, 2008.

14. National Oceanic and Atmospheric Administration. PTWC His- tory. Available at: http://www.prh.noaa.gov/ptwc/history.php. Accessed November 14, 2008.

15. National Oceanic and Atmospheric Administration. PTWC Responsibilities. Available at: http://www.prh.noaa.gov/ptwc/ responsibilities.php. Accessed November 14, 2008.

16. University of Washington. The Tsunami Warning System. Available at: http://www.ees.washington.edu/tsunami/general/ warning.html. Accessed November 14, 2008.

17. National Oceanic and Atmospheric Administration. Tsunami Warning Systems. Available at: http://www.ndbc.noaa.gov/ Accessed November 14, 2008.

18. National Weather Service. Tsunami Safety Advice. Avail- able at: http://wcatwc.arh.noaa.gov/tsunamiready/safetya.pdf . Accessed November 14, 2008.

19. U.S. National Research Council. Preventing Earthquake Disas- ters: The Grand Challenge in Earthquake Engineering a Research Agenda. Washington, DC: National Academies Press; 2003:12– 25.

20. Dudley WC, Lee M. Tsunami! Honolulu: University of Hawaii Press; 1998.

21. National Weather Service. Tsunami Ready. Available at: http://www.tsunamiready.noaa.gov/. Accessed November 14, 2008.

22. Centers for Disease Control and Prevention. Tsunamis. Avail- able at: http://www.bt.cdc.gov/disasters/tsunamis/. Accessed November 14, 2008.

23. American Red Cross. Tsunami. Available at: http://www. redcross.org/services/disaster/0,1082,0 592 ,00.html. Accessed November 14, 2008.

24. Centers for Disease Control and Prevention. Rapid health response, assessment, and surveillance after a tsunami – Thai- land, 2004–2005. MMWR. 2005;54:61–64.

25. Sphere Humanitarian Charter and Minimum Standards in Disaster Response Handbook. Revised 2004 ed. Available at: http://www.sphereproject.org/. Accessed November 14, 2008.

26. Pan-American Health Organization/World Health Organization (PAHO/WHO). Natural Disasters Myths and Realities. 2001. Available at: http://www.paho.org/English/DD/PED/myths.htm. Accessed November 14, 2008.

27. Centers for Disease Control and Prevention. Assessment of health-related needs after tsunami and earthquake – three dis- tricts, Aceh Province Indonesia, July-August 2005. MMWR. 2006;55:93–97.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Winter Storms

John M. Wightman, James A. Fenno,

and William H. Dice

OVERVIEW

Severe winter storms can be life-changing events that isolate and disrupt families; shut down schools, businesses, and government; prevent air, ground, and water transportation; and destroy large components of agricultural and service industries. Public safety can be threatened when roads are impassable, power grids are down, and telecommunications are inoperable.

Humans can exist in extremely cold environments, but their physiology remains geared mostly for the tropics from which the species originated. Adaptation to cold is most importantly behavioral. Heat loss is slowed or prevented by avoiding contact with cold surfaces and ingestion of cold substances, sheltering from wind and precipitation, wearing of protective clothing, and physically moving to a warmer location. Peripheral vasocon- striction and shivering only modestly and temporarily protect humans exposed to significantly cold temperatures.1 The conse- quences of many winter storms make it difficult or impossible for unprepared or unassisted humans to achieve the behavioral modifications necessary to mitigate the direct effects of cold or to obtain the resources needed to maintain health.

This chapter discusses the problems winter storms pose, how they fit into the continuum from minor annoyance to major disaster, and what can be done to prepare for future events that threaten the welfare of those in their paths. The disaster life cycle – mitigation, preparedness, response, and recovery – used here includes a continuous loop of planning and pre- paredness; warning, if any; the event itself; immediate response, which is almost always local; rapid assessment to identify needed resources; definitive response with ongoing surveillance and repeated assessments; recovery to baseline; and system improve- ments to increase preparedness for the next event.

Scope

Winter storms are relatively uncommon causes of disasters in Canada, the United Kingdom (U.K.), and the United States (U.S.). This is especially true in the U.S. when compared to flooding and other severe storms, which individually and col- lectively have constituted the bulk of Presidential disaster dec-

larations.2 In the 5 decades prior to the 2007–2008 winter sea- son, the Federal Emergency Management Agency (FEMA) had listed 99 winter storms resulting in federal emergency declara- tions and 148 resulting in major federal disaster declarations. These represented 3% and 8.5% of the totals in each category, respectively.3 Since 1980, the U.S. National Climactic Data Cen- ter documented 70 weather-related disasters that each cost over 1 billion (normalized to 2002 U.S.) dollars in economic damages.4

Four involved catastrophic winter storms in the 1990s with a combined mortality of almost 500 people.5 On the other hand, the scope of any individual potential injury/illness-creating event (PICE) in any country does not need to be as great as these to significantly alter baseline societal patterns and have a major negative regional or national economic impact.6 These smaller events can still cause patients to access the healthcare system. The 2007 winter season witnessed two major storms affecting dense population centers in Europe and North America, result- ing in increased human morbidity and mortality, and significant damage to societal infrastructures.

From January 15–19, 2007, a cyclone over the Netherlands generated a “European windstorm” with high, sustained winds and gusts up to 202 km/h (120 mph). Power was severed to more than 50,000 homes in the U.K. alone. Several major high- ways across Europe were forced to close, commuter rail traffic was slowed, and hundreds of commercial flights were cancelled. Ferries were halted on navigable waterways, one freighter ran aground, and another carrying hazardous cargo had to be aban- doned. Overall, this storm named “Kyrill” caused widespread damage across the British Isles and Western Europe and resulted in 47 deaths. Falling objects and motor vehicle collisions (MVCs) seemed to dominate as mechanisms of fatal injury.7

Three major winter cyclones also made their way across North America the same month: 1) from Texas to southeastern Canada, January 11–16; 2) from Texas to the Carolinas, January 16–19; and 3) across the U.S.–Canadian border, January 19–24. More than 1 million people were without power for days during some portion of the 2-week period. Large portions of several U.S. states, plus the entire state of Oklahoma, were declared dis- aster areas. MVCs accounted for the majority of the 87 deaths attributed to the storms.8

586 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 587

Table 37.1: Winter Weather Definitions According to the U.S. National Weather Service Glossary

Avalanche. A mass of snow, rock, or ice falling down a mountain or incline. In practice, it usually refers to a snow avalanche.

Blizzard. The following conditions are expected to prevail for a period of 3 hours or longer: sustained wind or frequent gusts to 56 km/h or greater and considerable falling or blowing snow reducing visibility frequently to less than 402 m.

Blowing Snow. Wind-driven snow that reduces surface visibility. Blowing snow can be falling snow or snow that has already accumulated but is picked up and blown by strong winds. Blowing snow is usually accompanied by drifting snow.

Coastal Flooding. The inundation of land areas adjacent to bodies of salt water connected to the Atlantic Ocean, Pacific Ocean, or Gulf of Mexico caused by sea waters over and above normal tidal action. This flooding may impact the immediate oceanfront, gulfs, bays, back bays, sounds, and tidal portions of river mouths and inland tidal waterways.

Cyclone. A large-scale circulation of winds around a central region of low atmospheric pressure: counterclockwise in the northern hemisphere, clockwise in the southern hemisphere.

Drifting Ice. In hydrological terms, pieces of floating ice moving under the action of wind and/or currents.

Drifting Snow. Drifting snow is an uneven distribution of snowfall/snow depth caused by strong surface winds. Drifting snow may occur during or after a snowfall. Drifting snow is usually associated with blowing snow.

Drizzle. Precipitation consisting of numerous minute droplets of water less than 0.5 mm in diameter.

Flood. Any high flow, overflow, or inundation by water that causes or threatens damage.

Freeze. A freeze is when the surface air temperature is expected to be 0◦C or below over a widespread area for a climatologically significant period of time.

Freezing Rain. Rain that falls as a liquid but freezes into glaze upon contact with the ground. Freezing drizzle, fog, and boat/ship spray also occur.

Heavy Snow. Snowfall accumulating to 10.2 cm or more in depth in 12 h or less, or snowfall accumulating to 15.2 cm or more in depth in 24 h or less.

Ice Fog. A suspension of numerous minute ice crystals in the air or water droplets at temperatures below 0◦C, based at the earth’s surface, which reduces horizontal visibility. Also called freezing fog.

Ice Jam. In hydrological terms, a stationary accumulation that restricts or blocks stream flow.

Ice Storm. Describes occasions when damaging accumulations of ice are expected during freezing rain situations. Significant accumulations of ice pull down trees and utility lines resulting in losses of power and communication. These accumulations of ice make walking and driving extremely dangerous. Significant ice accumulations are usually accumulations of 6.4 mm or greater.

Lake-effect Snow. Snow showers that are created when cold, dry air passes over a large warmer lake, such as one of the U.S. Great Lakes, and picks up moisture and heat.

Nor’easter. A strong low-pressure system that affects the Mid-Atlantic and New England states in the U.S. It can form over land or over the coastal waters. These winter weather events are notorious for producing heavy snow, rain, and tremendous waves that crash onto Atlantic beaches, often causing beach erosion and structural damage. Wind gusts associated with these storms can exceed hurricane force in intensity. A nor’easter gets its name from the continuously strong northeasterly winds blowing in from the ocean ahead of the storm and over the coastal areas.

Rain. Precipitation that falls to earth in drops more than 0.5 mm in diameter.

Sleet. Pellets of ice composed of frozen or mostly frozen raindrops or refrozen partially melted snowflakes. These pellets of ice usually bounce after hitting the ground or other hard surfaces. Heavy sleet is a relatively rare event defined as an accumulation of ice pellets covering the ground to a depth of 12.7 mm or more.

Snow. Precipitation in the form of ice crystals, mainly of intricately branched, hexagonal form and often agglomerated into snowflakes, formed directly from the freezing (deposition) of the water vapor in the air.

Snow Flurries. An intermittent light snowfall of short duration (generally light snow showers) with no measurable accumulation (trace category).

Snow Shower. A short duration of moderate snowfall. Some accumulation is possible.

Snow Squall. An intense, but limited duration, period of moderate to heavy snowfall, accompanied by strong, gusty surface winds and possibly lightning (generally moderate to heavy snow showers). Snow accumulation may be significant.

Wind Chill. Increased wind speeds accelerate heat loss from exposed skin, and the wind chill is a measure of this effect.

This information is in the public domain. It was retrieved from the National Weather Service Glossary.9

Winter Storms

Definitions used in this chapter are listed in Table 37.1.9 Win- ter precipitation comes in the form of rain, freezing rain, sleet, and snow.10 Colder temperatures and strong winds may magnify the ruinous effects of each on the environment. They also have direct individual and combined effects as wind chill on exposed

humans and animals. Predominately warm climates may not experience freezing precipitation at all, but heavy winter rains and wind can place people and property in danger, especially through flooding. In colder climates, flooding may also occur secondary to ice jams obstructing flowing bodies of water, or as a consequence of melting ice and snow, although these events may not necessarily be related to a storm. Snow avalanches

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

588 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

Figure 37.1. Power lines can be downed by the direct weight of ice or by trees and limbs felled by ice, as in this example from Spring- field, Missouri, following a January 2007 ice storm. This photo- graph is in the public domain. It was retrieved from Wikipedia at: http://en.wikipedia.org/wiki/Image:Icestorm 003.jpg. See color plate.

and snow runoff flooding are discussed in other sections of this chapter.

Winter storms are generally categorized by the type of pre- cipitation (in the preceding paragraph)10 or by storm type (i.e., blizzards, ice storms, lake-effect storms, and nor’easters).10,11

Each has its own unique features with common elements that include the following:

■ Cold has direct effects on people, property, and electri- cal and mechanical systems. Damage can be temporary or permanent. Frostbite and hypothermia are the two medical conditions most commonly associated with cold environ- ments. They can occur indoors when power failure limits heat generation or outdoors when people are stranded in a storm, conducting necessary activities, or during winter recreation.

■ Frozen precipitation creates wet conditions that can accel- erate heat loss from humans and animals. Unlike liquid water that runs off surfaces, ice and snow accumulate on objects adding significant weight to structures that may not be designed to withstand the added stress. Ice can pull down power and telecommunication lines (Figure 37.1), crush roofs of buildings, and collapse bridges. Ice flowing on rivers can also damage bridges and watercraft. Ice jams may cause upstream flooding.

■ Dangerous movement results from slick ground conditions, degraded visibility, and icing conditions that affect air and water transportation. These conditions inhibit the ability of people to obtain supplies and resources, as well as make it difficult to obtain assistance when help is needed. They also may indirectly contribute to falls, MVCs, recreational accidents, and other injuries related to snow removal and cleanup operations.

All these features of winter storms create hazardous situa- tions, which slow emergency response and increase the risks for responders.

At lower elevations from October to April, populations of American, Asian, and European countries in extratropical north-

ern latitudes are at risk from the effects of winter weather. How- ever, regional probabilities for specific types of storms differ. As illustrated in Figure 37.2 for the U.S., winter storms are more frequent in northern states, mountainous regions, and east of the Great Lakes. They rarely occur in the southern U.S., but ice storms are especially treacherous when they do.

Human Impact

As with all PICEs, the impact winter storms have on society is what defines their magnitude. Regions having little experi- ence with winter storms often have the least prepared popu- lations and therefore are at even greater risk. These areas may also have local and state governments with the least capacity to respond rapidly and effectively. Unprepared communities can exponentially increase the human and economic impact of an event. For instance, 1 m of snow blanketing rural areas of America’s Great Plains may minimally disrupt populations and their associated agricultural and ranching industries. In dis- tinction, 1 cm of ice from New York City to Washington DC could paralyze major commercial, financial, and governmental centers.

Understanding the human impact of catastrophic events, so that society can be better prepared for future challenges, is the primary mission in the fields of emergency management and disaster medicine. With regard to winter storms, the medical literature mostly covers the last 3 decades. The health effects of storms that move up the Ohio River Valley from the Texas Gulf Coast and the so-called nor’easters have been the types most extensively reported in the English-language medical literature. This is due to their relative frequency and great potential impact on large population and economic centers in the north-central U.S. and northeastern seaboard.

Perhaps one of the most studied was an ice storm that affected an area of North America centered over the St. Lawrence River and extending east toward Nova Scotia, occurring on January 4–10, 1998. It left more than 4 million people without elec- tricity – some for up to 33 days – mostly in southern portions of Ontario, Quebec, and Nova Scotia in Canada, and northern New England in the U.S.12 Those persons displaced from their homes in Ontario, who could not move in with a relative or friend, constituted almost 5% of the affected population, plac- ing approximately 140,000 persons in 454 emergency shelters.13

Some Canadian hospitals were without power for 3 weeks.14

Estimated total damages for the two countries amounted to 4–6 billion U.S. dollars.12 One of this chapter’s authors participated in the U.S. federal disaster response to northern New York State in the aftermath of this storm.

Mortality Determining that an individual’s death resulted from extreme

temperatures is a diagnosis of exclusion and difficult to make without awareness of risk factors and circumstances. Databases likely underestimate the problem, as some types of reporting are not required, no precise case definitions exist, and there is little quality control over death certificates.15 Nonetheless, U.S. data suggest that mortality from excessive cold is less common than from excessive heat.16 However, this might not be true on a regional basis, or in countries with different climates.

Deaths from all causes in the U.S. are more common in January than in any other month. This is especially true for the elderly, and this higher mortality rate has been specifically

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 589

Figure 37.2. Winter storm hazards in the United States. This figure is in the public domain. It was reproduced from the National Weather Service’s pamphlet entitled Winter Storms: The Deceptive Killers.23 See color plate.

linked to colder periods.17 A study of four consecutive winters in Minnesota showed a slight increase in mortality during cold days, but a greater increase in cardiovascular mortality during periods following snowfall.18 One report covering six consec- utive January months in Pennsylvania found that there was a 1.27-times (95% confidence interval [CI] = 1.12–1.44) increased relative risk of dying during “extreme climatic con- ditions” – defined as when the temperature was less than −7◦C or more than 3 cm of snow had fallen.19 A British study noted a statistically higher risk of dying in winter months during the years from 1986 to 1996 (1.5% higher for every 1.5◦C decrease in temperature), especially if no central heating was used in the household (odds ratio 95% CI = 1.009–1.022).20

The exact risk of storm-related mortality from all causes is not well known. No universal requirement exists to report that an individual’s death is, or is not, directly related to a weather phe- nomenon. Because of this, any generalized data must be viewed skeptically, unless the dataset from which it is derived is speci- fied. For instance, textbooks published in 2006 and 200721,22 have stated that 25% of deaths are in people who are outdoors in the storm, and most of the rest die in automobiles.23 However, these numbers come from a U.S. governmental organization that had no mechanism by which to obtain representative data. Therefore, planners cannot rely on these statistics to predict where resources will be needed to mitigate mortality rates before, during, or after a storm.

A massive blizzard in New England on February 6, 1978, was followed by two public health reports regarding mortal- ity. Twenty-seven storm-related fatalities were identified in Mas-

sachusetts, but no overall increase in total mortality was appreci- ated.24 In Rhode Island, on the other hand, researchers concluded that there was an increase in total mortality in the first 5 days fol- lowing the storm,25 although statistical methods were not used to compare the study group with an unexposed cohort.

Morbidity During and immediately after a winter storm, the emergency

department (ED) will be the first functional area affected. It is the location where ambulances bring patients from the commu- nity and where the public is accustomed to seeking unscheduled care. This assumes any given healthcare system can maintain its capabilities, or rapidly implement its preexisting plans or ad hoc processes to increase surge capacity.

A Massachusetts study of 15 hospitals found a significant decrease in ED visits on the day of the blizzard in February 1978, but daily census rapidly returned to baseline.24 A survey of five major hospitals in northeastern New York State noted a similar trend after a January 1996 blizzard, but this was followed by a marked increase in ED volume the next day.26 Only one liter- ature article specifically reported the effects of a winter storm on a pediatric ED. The author found that the total daily cen- sus increased 35% in the 36 hours before a blizzard hit eastern Pennsylvania and Delaware in January 1996, decreased to very low levels during the storm, then slowly returned to normal over the next 4 days.27 A postevent spike frequently reported in adult and combined adult/pediatric EDs was not seen, although there was an increase in the percentages of higher-acuity problems and near tripling of the admission rate.27

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

590 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

Following a particularly heavy snow storm in a region of the U.S. where snow is expected, common storm-related mecha- nisms of injury (in descending order of frequency) included slips and falls, MVCs, being struck by falling objects, carbon monox- ide (CO) poisoning, and those related to equipment use such as chainsaws and snowblowers. A number of additional prob- lems resulted from deficiencies in access to customary care, lack of home heating due to power loss, and inability to discharge patients to their existing home situation.26

Similar findings were reported from a single university med- ical center in the aftermath of an ice storm that hit North Car- olina in December 2002, and interrupted power to 1.3 million homes.28 Ice more commonly snaps power lines and fells trees than does heavy snow. Therefore, it was not surprising that indi- viduals struck by falling objects during assessment and cleanup operations was the most common patient presentation. An epi- demic of CO poisoning was also seen at the same institution. These mechanisms accounted for all the life-threatening injuries, except one elderly patient who likely became hypothermic after a stroke. Slips and falls, injuries associated with darkness, and burns rounded out the other causes of injury. In this study, inves- tigators could not reliably determine that MVCs resulted from storm-related conditions in many instances, so they were not examined specifically.28

One of the first epidemiological reports documenting an increased incidence of fractures associated with a winter storm was published by Ráliš. The storm in question produced 5 days of snow and ice over a 1 week period around the 1978–1979 New Year.29 In a U.K. ED that saw more than 93,000 patients per year, the number of patients with fractures increased by 2.85 times normal, peaking at a rate of 1 in every 5 patients. In descending order of frequency, the locations of fractures were the wrist and forearm, foot and ankle, hand, hip, leg, chest and spine, and head.29 Others in the U.K. noted similar patterns when they subsequently reviewed their own experiences in the early 1980s.30–32

In the midwestern U.S., ice storms have led to numerous orthopedic injuries in patients presenting to EDs. Falls on ice, particularly in the elderly, more commonly resulted in extremity fractures than falls during equivalent periods of snow cover in St. Louis.33 In the 9 days after a winter storm passed through the city of Indianapolis in February 1994, 327 injuries in 259 individuals who slipped on ice were seen at just one central hospital. Most were back injuries of various types, but more than one-third of patients were diagnosed with fractures of the nonaxial skeleton.34 Following the 13-day ice storm of January 1998, Canadian EDs reported more than one-third of the injuries seen were directly storm related.35 This was also true in both adult and pediatric populations in Montreal.14 Montreal General Hospital alone performed 60 emergency orthopedic operations for storm-related injuries.14

CO poisoning is generally associated with winter months and its incidence can be increased by a storm that causes widespread power losses. Production of CO usually results from burning fuel indoors for heat, electricity generation, or cooking when alternative means are not readily available.36 Exhaust fumes filling automobiles from snow-obstructed tailpipes is another mechanism of CO poisoning following winter storms.37

The rate of CO cases increased after a winter storm dis- rupted power to a large portion of the Seattle–Tacoma area of the northwestern U.S.38,39 Thirty incidents resulted in 81 cases at 13 hospitals through the 3-day storm in January 1993. Another

spike in CO cases occurred after two winter storms in late 1996.39

Following the North American ice storm of January 1998, more than 1,000 cases of CO poisoning from at least 700 individual incidents were reported in Quebec alone, and this was likely fewer than actually occurred.14 Four hospitals in rural Maine reported 42 incidents causing 100 cases with up to 8 patients arriving from a single scene.40 Approximately half that number was reported for two EDs in just one Ontario city.41 After a December 2002 ice storm in North Carolina, a single university hospital saw 200 cases of CO poisoning in 1 week,42 and another one in a different city saw 48 after the same storm.28

An increased incidence of fatal and nonfatal acute coronary syndromes (ACS) has been associated with winter storms in sev- eral reports.18,43,44 Specific risk factors could not be identified in a small cohort of patients surviving myocardial infarctions after a January 1979 blizzard in Chicago.45 Short-term cold exposure was not believed to be the cause of increased ACS when a Cana- dian study examined 15 years of epidemiological data.46 On the other hand, a Dutch study concluded that windchill might have a greater effect on cardiovascular risk than cold air itself.47

Heavy snow shoveling has been shown to elicit sustained heart rates near age-calculated maximums with aerobic oxygen demands similar to arm-crank ergometry.48 Although a Cana- dian study concluded that the incidence of “heart attacks” was independent of snowfall, the ones that did occur were more likely to follow shoveling.49 In distinction, an American study of ten EDs in one New York county after a January 1996 blizzard found a 6.6-times increased relative risk (RR) of ACS events. These occurred primarily following snow shoveling, and often in per- sons without histories of coronary artery disease.50 Comparing consecutive January months from 1991 to 1996, a Pennsylva- nia study noted increased RRs of cardiac mortality in males. The risk increased for progressively younger age groups: 1.28–2.21 for those > 65 years old; 1.32–2.38 for those 50–64 years old; and 2.35–5.35 for those 35–49 years old (RR 95% CIs).19 In another study, the rate of noncardiovascular illnesses was not noted to change.50

With regard to pregnant women, onset of term labor was statistically more common with the low atmospheric pressures associated with weather fronts and nor’easters in one central Massachusetts study. However, the difference was interpreted as not clinically significant.51

Accidental hypothermia can occur at virtually any ambient temperature, but it is more strongly associated with cold air and cold water submersion.52,53 Death rates in the U.S. are 0.08–1.99 per 100,000 population in the contiguous 48 states, although, as expected, they are higher in Alaska.54,55 Therefore, accidental hypothermia is not a common cause of death but it is a potentially preventable one.

CURRENT STATE OF THE ART

Many paradigms for disaster response have been proposed. The American Medical Association has promulgated the acronym DISASTER in its Basic and Advanced Disaster Life Support train- ing courses,56,57 but efficacy data are lacking and other systems may be equally effective. Regardless of the approach, decision makers must know a problem exists before they can imple- ment preparedness plans and devote resources to a targeted and coordinated response. It is usually easy to detect that a win- ter storm has dropped precipitation on a given region, but its

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 591

Table 37.2: Potential Impact of Winter Storms59

Index Category

Maximum Snowfall Rate

Maximum Snowfall Amounts

Potential Wind Speeds

Maximum Snow Depth Drifting Nature of Societal Disruption

1 <2.5 cm/h <25 cm Weak <50 cm Minimal (h) to Nuisance (up to 1 d)

2 2.5 cm/h 50 cm Strong 100 cm Nuisance (up to 1 d) to Inconvenience (a few days)

3 5.0 cm/h 75 cm Gale 200 cm Inconvenience (a few days) to Crippling (several days)

4 7.5 cm/h 100 cm Gale or hurricane 300 cm Crippling (several days) to Paralyzing (up to 1 wk)

5 >7.5 cm/h >125 cm Gale or hurricane >500 cm Paralyzing (up to 1 wk)

Reproduced with permission from the American Meteorological Society with modifications for clarity.

human impact is much more difficult to determine, especially when aerial overflights and on-the-ground access to affected areas are limited. Establishing an incident command structure (see Chapters 9 and 19), regardless of size, should bring together the resources officials need to determine the security and safety of affected areas, identify hazards to responders, and coordinate the support necessary to begin rescue and recovery efforts.

Winter Hazards

Storms and their aftermaths affect both populations and societal systems, the latter of which includes responding to community emergencies and delivering healthcare outside and inside hos- pitals. Public health surveillance and interventions, emergency medical services (EMS) systems, and regional hospital capacity can all be adversely affected by power losses, hazardous driving conditions, and cold environments.

Weather Information One of the purposes of government is to protect the pubic

welfare, particularly with resources not available to individuals or private groups. Most developed nations have one or more methods of notifying their populaces of important weather conditions that may adversely impact people or property. In the U.S., this responsibility begins with the National Weather Service (NWS).

The NWS may issue a “winter storm watch” when “the risk of a hazardous weather or hydrologic event has increased signif- icantly, but its occurrence, location, and/or timing is still uncer- tain.”9 The purpose of a watch is intended to alert the population at risk to initiate protective actions. Once its weather predictions are more certain, a “winter storm advisory” or “winter storm warning” may be issued – or the message may be more specific, such as “blizzard warning.”

■ The term “advisory” highlights special weather conditions that are less serious than a warning. They are for events that may cause significant inconvenience, and if caution is not exercised, could lead to situations that may threaten life or property.9

■ The term “warning” is issued when a hazardous weather or hydrologic event is occurring, is imminent, or has a very high probability of occurring. A warning is used for conditions posing a threat to life or property.9

Although print media is useful for long-range forecasts (i.e., > 24 h), broadcast media and the Internet are the most common methods used for more immediate notifications of approach- ing winter storms. The U.S. National Oceanic and Atmospheric Administration (NOAA) All-Hazards Weather Radio broadcasts important information directly from the nearest NWS office around the clock.58

Storm severity categories may help populations at risk, emer- gency response organizations, and other public and private facilities (e.g., government centers and hospitals) prepare for potential effects. Hurricanes and tornadoes each have a well- known severity index associated with them: 1–5 for the former and 0–5 for the latter (see Chapters 33 and 34). A five-level cate- gorization scheme for winter storms has also been proposed.59 It is based on intensity – directly proportional to atmospheric pres- sure gradients – and duration – inversely proportional to forward speed, when slower speed equates to more precipitation over any given area. Although many factors may modify the characteris- tics of a specific event, based on storm and population features, the overall index category may help predict a storm’s impact (Table 37.2). While not in as widespread use as the categoriza- tion schemes for hurricanes, impact predictions can similarly be modified moment-to-moment as constantly updated weather data are received and analyzed.

Responder Risks Working in cold environments can be risky for the ill pre-

pared.60 Emergency responders who venture into these condi- tions must be ready for hazardous driving or flying conditions; the effects of cold, wind, and wet conditions on themselves, their patients, and their vehicles; and on potential problems with communications created primarily by atmospheric conditions or secondarily by damage to power and telecommunications systems.

Responders must also ensure that their equipment and sup- plies are functional in cold environments. Vehicles must be well maintained and prepared for operations in cold weather or under slick surface conditions. This is true for all ground- based vehicles, which are subject to maintenance standards at the local and national levels if the equipment is government- owned. These same standards may not apply to vehicles owned by individuals or businesses (e.g., utility companies). The same applies to rotary-wing aircraft used as ambulances or for other public service applications, the maintenance and operations of

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

592 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

which are even more stringently controlled, especially as regards weather.

Winter environments are harsh on vehicles. Engine and transmission oils become more viscous at lower tempera- tures and vehicles require longer warm-up times when kept in unheated locations. Any water in fuel lines may freeze, thus obstructing flow to the engine. Block heaters can help keep flu- ids somewhat warmer to facilitate prompt starting. Windshield wipers can clear the windshield of liquid water droplets, but as rain freezes, or turns to sleet or heavy wet snow, buildup can hamper visibility. Headlamps can also become covered with wet snow, reducing the light output reaching roadways to almost nothing. Snow may obscure directional signals, parking lights, and emergency flashers, making it difficult for other motorists to see vehicles at sufficient distances to avoid collisions on slick roadways.

Responding to emergency scenes is hazardous, even with- out the dangerous road conditions and limited visibility caused by winter storms, as well as the increased risk of becoming stranded far from shelter. EMS, fire, and law enforcement per- sonnel usually receive formal instruction on emergency driving and gain important experience on the job. On the other hand, driving instruction for winter weather is often only didactic in nature. Few response organizations provide hands-on training that allows drivers to push their limits under real conditions in controlled situations where the potential for personal injury or property damage is minimized. Similarly, few utility compa- nies and other public service organizations – including those delivering healthcare in fixed facilities such as clinics and hospi- tals – provide any driver training to those who must respond to an outdoor or building worksite. When road conditions exceed driver abilities, the number of emergency scenes may increase, and thereby compound the overall demand for rescues and out- of-hospital medical responses.

When driving during or after a winter storm, speed must be reduced for both traction and visibility concerns. Heavy snowfall and windswept snow can cause near “white out” conditions with visibility less than 50 m. Slick conditions on road surfaces will increase stopping distances. Negotiating hills and curves may be difficult. Even stepping out of vehicles can be hazardous, as slips and falls can injure responders. Parking or exiting a vehicle on a roadway can be very dangerous when other vehicles are moving around it. Distracted by flashing emergency lights, other drivers may lose control of their own vehicles causing impacts with emergency response equipment and their crews or with other vehicles, causing more injuries that will consume even more automobiles.

Black ice is a term referring to a thin layer of ice that cannot be seen when it is adherent to darkly colored road surfaces. Yet it is just as treacherous as an icy surface several centimeters thick. This invisible ice layer can develop in minutes when a wet roadway at the freezing point suddenly becomes colder due to evaporation via wind or with the loss of the sun’s heat (e.g., clouds, sunset, or shade).

Crashes and vehicle incapacitation can result in response personnel being stranded while driving in inclement weather, ice, or heavy snow. Radio communications may be degraded and cellular telephones may be nonfunctional during or immediately after a winter storm. Response personnel must be prepared for the inability to communicate with their own dispatch authority, other emergency response agencies, and an Emergency Opera- tions Center (EOC). Medical control for patient care advice or

authorizations beyond an EMS crew’s standing orders or scope of practice might be limited. Communications problems may also make it difficult to call for assistance, if stranded.

All emergency response personnel venturing into the winter environment must have operational guidelines for a commu- nications outage or cold weather vehicle failure. Most author- ities recommend stranded motorists stay with their vehicles – although caution to guard against cold injury, if not heating the interior, and CO poisoning, if running the engine or using some other kind of heat source.

Many emergency response vehicles, which operate in areas likely to receive ice and snow, are equipped with chains. These can be attached either directly to their tires or to a deployable apparatus under the chassis. In the latter situation, at the touch of a button, chains are rotated on a horizontal axis near the tires resulting in enhanced traction equivalent to chained tires.

Response personnel must be properly attired, equipped, and trained for work in cold, windy, and wet scenarios. Typical cloth- ing worn by EMS responders on a day-to-day basis may not ade- quately protect them from the elements, as winter storms may force them into prolonged scene times in outdoor environments, to which they may be unaccustomed.

Care should be taken to select clothing that will minimize the risk of injury yet maximize dexterity and function. The over- arching goals would be preservation of core body heat and mini- mizing exposure of skin to cold air and surfaces. Such garments, while important, are insufficient alone to protect responders. Rescue personnel must be knowledgeable about environmen- tal conditions so that proactive and reactive behaviors prevent problems before damage occurs or mental status is affected to the point that behavior is no longer compensatory. The follow- ing are recommendations for the prevention of cold-induced challenges.52,61,62

■ Maintain adequate hydration, nutrition, and get sufficient rest. Being in excellent physical condition is a definite advan- tage. Use of tobacco products is detrimental to good circu- lation. Clean, healthy skin is protective for cold injury, but washing too frequently can dry and damage the skin’s pro- tective barrier.

■ Wear garments that provide thermal insulation and trap air between fabric layers. The mnemonic for prevention of COLD stands for clean fabrics; opening for ventilation during exercise, to avoid wetting perspiration; loose layers to retain insulating air pockets, and allow for donning and doffing lay- ers as conditions change; and dry garments, which must be changed if they get wet. Make sure the head is well insulated, because significant body heat can escape from the exposed scalp.

■ Avoid tight-fitting clothing and restrictive gloves or boots. Liners improve the insulation of both. When dexterity is not needed, mittens are more effective at retaining heat than are gloves.

■ A balance must exist between external water impermeability, and the risk of retaining sweat inside protective garments that cannot breathe. Particularly with regard to the feet, wet conditions soften the skin, which can lead to debilitating nonfreezing injuries like trench foot. Even damp socks should be exchanged for dry ones, so extras must be available.

■ Protect exposed skin surfaces from cold and windy air, as well as from cold liquids and surfaces. Avoid damaging skin through ultraviolet radiation directly from the sun or

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 593

indirectly from reflection off ice, snow, water, windows, or light-colored building surfaces.

■ Use the buddy system. Trained individuals, who consciously and frequently check each other, will decrease the likelihood of insidious frostbite and hypothermia.

A discussion of the precautions needed for specialized responses, such as cold water or mountain rescue, is beyond the scope of this chapter.

Local Medical Responders

Winter storms may severely limit the ability of first responders to reach callers in a timely manner or to find victims during a community needs assessment or deliberate search. Once located, the delivery of care to these victims may occur in a relatively austere medical environment for periods of time that may far exceed those to which EMS personnel are accustomed. Trans- portation to a permanent or temporary medical facility, or even to a heated shelter with power and food, may be so treacher- ous that the risk/benefit ratio is higher than staying in place. Response organizations and systems must consider how to best plan, train, operate, and recover in these environments.

Human access, care, and evacuation (HACE) is a term coined by Mark Gebhart and James Gruenberg of the National Center for Medical Readiness (personal communication). The concept describes the situations faced by emergency responders in per- forming their out-of-hospital duties after a call for help has been initiated and will serve as an outline for the next discussion.

Access The austere environment associated with storms can signifi-

cantly limit responders’ ability to access known victims, or search areas to locate missing victims. Much has been written regard- ing tactics, techniques, and procedures for searching for victims who cannot call for help after an area has been hit with freezing precipitation. However, virtually nothing has been published on best practices for EMS personnel, who must access persons who have requested urgent assistance.

Because such efforts would certainly require coordination with other public safety agencies (e.g., fire and law enforcement) the effort should be managed through a local or regional EOC functioning under an incident command structure. Utility com- panies also participate when downed power lines or ruptured natural-gas or water pipelines threaten the safety of respon- ders. In advance of a winter storm, a local EOC could prohibit ambulances from traveling off snow-plowed roadways, yet still be responsive to requests for assistance. Response assets could be reorganized into task forces, such that a call for medical assis- tance would result in dispatch of a fire engine with personnel, a snowplow, and a four-wheel-drive vehicle with a command offi- cer and a medic. These resource groupings could be kept intact between calls. Public health representation could also facilitate collection of data for a rapid needs assessment.

Care Despite governmental regulations and industry guidelines on

vehicle operations, there are fewer standards regarding storage and use of onboard medical equipment and pharmaceuticals in cold weather. Ground-based ambulances can often be parked in a shelter or garage while awaiting calls. This is rarely an option for helicopters that must respond rapidly, especially when they

are hospital based. Three studies have examined the tempera- ture inside medication containers stored on rotary-wing ambu- lances, two from North America63,64 and one from Europe.65

Both identified temperatures far outside the range recommended for pharmaceutical storage, although the clinical impact of this is unknown. An important exception would be cold intravenous fluids, which could be potentially harmful to many patients if given in any significant quantity.

Individual out-of-hospital patient care should remain essen- tially unchanged, as long as responders are aware of the increased occurrence of cold weather conditions and consideration is given to potential delays in access and evacuation. Illnesses may be more advanced or complications of injuries may be more man- ifest when access is delayed. Additionally, the duration of care may be extended for longer than EMS and rescue personnel are otherwise accustomed. Depending on limitations in evacuation options or extensions of transportation times, the duration of on-scene medical care may last for hours, or perhaps even days, in extreme circumstances.

Evacuation delays can necessitate hemorrhage control meth- ods less familiar to civilian EMS responders. Application of clot enhancing agents such as fibrin, microporous polysaccharide microsphere, mineral zeolite, or poly-N-acetylglucosamine (chi- tosan) may be useful adjuncts.66,67 Exsanguinating extremity hemorrhage may require control with a proximal tourniquet, either by inflating a blood-pressure cuff or applying a prefabri- cated or field-expedient device.

Delayed response to patients with advanced disease is less concerning for the responder community because people often postpone seeking medical care, even in the absence of a storm. On the other hand, injured patients are likely to seek care immedi- ately, yet medically trained providers may have difficulty access- ing them rapidly. Delayed complications (e.g., established wound infections, gangrene or tetanus, and compartment syndromes) or progressive conditions (e.g., increased intracranial pressure, pulmonary contusions, and slow intracavitary hemorrhage) may be less familiar to EMS personnel. Their scope of practice may only allow for supportive care, even when delays in evacuation to definitive care could result in loss of life or limb.

With regard to cold weather conditions, frostbite and hypothermia are the most common. The out-of-hospital man- agement of these conditions should focus on removing the patient from exposure to the cold insult and supporting the patient during evacuation to a higher level of care. Active rewarm- ing is difficult in the field. In situations of widespread power loss, the only source of added heat may be the interior of an evacu- ation vehicle. Other techniques may be available, if specialized equipment is positioned in the response vehicle and functional. In general, though, active rewarming is normally conducted at a medical facility.52

Cold injuries can be divided into nonfreezing and freezing categories. The former includes pernio, trench foot, and immer- sion foot. Pernio results from the combined effects of wet and cold skin, but may occur in dry nonfreezing conditions. Trench foot follows chronically wet feet exposed to near-freezing temper- atures and made relatively ischemic by vasospasm and increased tissue pressure induced by standing. Immersion foot results from skin that has become waterlogged from prolonged cold water immersion.61 Injuries that actually freeze tissue include frostnip and frostbite.62

Pernio, also called chilblains, is a localized inflammatory condition of the skin, which results from an abnormal tissue

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

594 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

response that develops over 12–24 hours after cold exposure. It most commonly manifests as tender, bluish or purplish, sub- cutaneous nodules in exposed areas. These lesions are often associated with edema or blister formation. A history of Ray- naud’s phenomenon is not uncommon. Treatment consists of local massage to stimulate blood flow and slow rewarming at normal room temperatures. Active rewarming with higher tem- peratures significantly increases the intense burning and itching associated with resolution.61,68,69 Nifedipine, a dihydropyridine calcium channel blocker, may have a role in hastening clear- ance of the lesions with less discomfort and may also reduce the likelihood of recurrence.70

Trench foot and immersion foot are clinically indistinguish- able. They both progress insidiously through three phases: pre- hyperemic, hyperemic, and posthyperemic. Intense vasospasm causes skin blanching and mottling in the prehyperemic phase. Peripheral pulses may be diminished and capillary refill is usu- ally prolonged. Continued exposure results in anesthesia and gait disturbances from damage to sensory and proprioceptive nerves.61,68 Rewarming creates a hyperemic condition with ery- thema, petechiae, swelling, pain, and hypesthesia – yet is still associated with prolonged capillary refill.71 Epidermal slough- ing may occur.61 Nerves controlling voluntary muscular action and vibratory sensation may be adversely affected.61 Therefore, the goal of treatment is rewarming core body temperature with- out directly warming the affected body parts, so as to keep the metabolic demands of the injured tissue low.61,68 The posthyper- emic phase is not normally seen in the field, unless evacuation is significantly delayed after rewarming.

Frostnip, which heralds the beginning of tissue ice-crystal formation, is a warning sign that frostbite is imminent. Vasocon- striction leads to pallor, localized pain, and sensory numbness. Clinical manifestations can be readily reversed by preventing additional cooling and by rewarming the affected body parts.61,68

Frostbite represents freezing of extra- and intracellular water with cell injury or destruction leading to tissue damage, which can be compounded by microvascular stasis and ischemia. Typ- ical symptoms progress from feeling local cold to loss of sensa- tion.62 Pain may occur in a “watershed” area between frostnip and frostbite, typically more proximal to fully involved areas of distal tissue damage. Frozen skin may appear a waxy yellowish white or translucent bluish color. It may be frozen solid.

Prevention of further cooling should be the primary goal prior to in-hospital management of frostbite.62,72 Wet, or possi- bly frozen, clothing should be gently removed. If stuck to the skin, other portions of any garments can be cut away, leaving frozen bits of clothing adherent to the skin.73 Out-of-hospital rewarm- ing of frozen tissue is generally discouraged, unless evacuation to definitive medical care will be significantly delayed.72

A 10-year Canadian study found that delay to medical care was one factor associated with poor frostbite outcome, so delayed access or prolonged evacuation times may have implications for prognosis.74 One author has suggested 2 hours of field time as a cutoff to begin thawing in the field, but only if no chance of refreezing is possible.75

In these settings, the position of the International Commis- sion for Alpine Rescue might provide the most useful extrap- olation from mountaineering to victims stranded by a winter storm.76 Guidelines are divided into whether the victim is out in the open or inside shelter but commonalities include: removal of wet clothing; orally administered warm fluids; and aspirin up to

1 g or ibuprofen up to 800 mg, if available. In shelter, if a warm 37◦C bath is considered for active rewarming, the patient should not be allowed to subsequently use the effected body part that has been warmed, which includes walking if the feet are involved, until after definitive care has been rendered. Because edema will ensue as the part is warmed, the area should be elevated and dry dressings loosely applied.76

Prevention of further cooling is also the primary goal for the out-of-hospital treatment of hypothermia.52,53,72,77 How- ever, other field management options are limited. Patients with mild hypothermia (i.e., core body temperature 32–35◦C and still capable of shivering) may warm themselves by being cov- ered with dry, heat-retaining clothing, blankets, or other items. These interventions allow endogenous heat production through metabolism and the shivering reflex. These constitute the pas- sive external rewarming technique. Active external rewarming involves the addition of exogenous heat to the body through a warm environment or radiant heaters. These can be supple- mented by warm oral or intravenous fluids. These methods of active internal rewarming have only a minor temperature-raising effect, although an added benefit results because many victims of hypothermia are also dehydrated.

Bradycardia is frequently observed at core temperatures less than 28◦C. In severe hypothermia, when the temperature is less than 25◦C, more life-threatening cardiac dysrhythmias occur. Development of new atrial fibrillation is ominous, as it may herald the risk for subsequent ventricular fibrillation. When dys- rhythmias complicate hypothermia, the treatment of choice is rewarming. Medications are not likely to be helpful until the heart is warm.78 Transthoracic pacing is possible, but not required in most cases.78,79

Evacuation Just as winter weather can affect the ability of first respon-

ders to access victims, it can also make it hazardous or difficult to transport them from out-of-hospital locations to medical facili- ties. The usual assets may not be available. Standard ambulances may be incapable of driving to or from the scene. Few have four- wheel drive, so responders may need to use nonmedical vehicles of expediency. Moreover, helicopters are frequently incapable of flying. In a Canadian study, 30% of requested helicopter EMS missions were aborted due to weather. This was the next most common reason behind the 42% combination of the helicopter not being the appropriate vehicle and cancellation by medical control.80

Determining the most appropriate destination for patients should be guided by local protocols, but the environment may force additional considerations in a disaster setting. Some med- ical facilities may be directly affected by the weather. Impaired ground and air accessibility, power losses, and staff absenteeism are just a few of the reasons a hospital or other facility can lack capacity to receive new patients. Other facilities may be over- whelmed by high patient volume and higher patient acuity in a storm’s aftermath. Routing of transportation assets directly to centers offering specialty services may be required for pediatrics, major trauma, serious burns, and hyperbaric oxygen therapy; however, transportation to these locations can be dangerous in hazardous winter conditions.

Finally, the possibility of being stranded during evacua- tion can complicate patient care, forcing paramedics to treat patients for much longer durations than medical training, supply

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 595

quantities, and vehicle power were designed. The availability of medical control telecommunications to assist medics in these unfamiliar situations is frequently uncertain.

Local Medical Receivers

The most significant problems for hospitals often relate to inade- quate staff (e.g., inability to surge with off-duty personnel due to weather, on-duty staff overworked due to lack of relief) or capa- bility degradation (e.g., interruption of facility water and power and supply depletion). Nonetheless, patients will arrive seeking care for a variety of baseline and storm-related problems. As in the out-of-hospital setting, the care delivered to individual patients is similar to what is required in other situations, except that cold-induced injury, hypothermia, CO poisoning, and other storm-associated issues may complicate management. Establish- ing disaster triage protocols to screen victims for these potentially occult problems may enhance ED operations.

Primary Triage Because the ED is the focal point for victim reception, it is also

the site for initial triage of persons requesting medical care. All individuals must be screened for emergency conditions whether or not their problems are attributable to the storm. In addition to standard triage questions for potentially serious illnesses and injuries, high indexes of suspicion for frostbite, hypothermia, and CO poisoning are required. Triage personnel should specifically ask about exposure to cold, wet, or windy conditions, as well as the possibility of exposure to burning fuels. Such diligence is necessary to limit the risk of missing these potentially occult conditions in the face of an unrelated chief complaint.

If sufficient resources are available for the volume and acuity of conditions, little or no adjustment to routine operations is necessary. However, any excessive patient need or limitation in healthcare capacity may require a different approach to triage and activation of facility disaster plans. Protocols should address the probability of various conditions, establish reasonable diag- nostic parameters recognizing the potential limitations in labo- ratory and radiology services, provide management guidelines, and identify admission and discharge criteria.

Emergency Care Traumatic injuries directly related to the inclement weather

include those sustained during outdoor movement (e.g., slips and falls and MVCs), cleanup efforts (e.g., snow clearing, tree removal, and utility repair), and recreational activities (e.g., sledding and snowmobiling). As most of these mechanisms are familiar to emergency professionals, only those unique to winter weather will be covered in this chapter.

Snowmobiles can be important methods of transportation in some northern areas. Snowmobile collisions involve drivers, passengers, pedestrians, and people being pulled on sleds. Most injuries are musculoskeletal, but head, chest, and abdominal injuries occur frequently. Injuries around the knee are common, and fractures of the femur and tibia represent almost half of the injuries. In a study of snowmobile collisions in Canada, 74% of victims required surgery with a mean of 1.6 operative procedures per patient.81

Patients with complex wounds caused by a variety of mecha- nisms may present to the ED. Many follow the use of power equip- ment for clearing debris, fallen trees, and snow. Chainsaw injuries

Figure 37.3. Typical snowblower injury. The patient placed his hand into the running auger to remove a chunk of ice with resulting open fractures of the index and long fingers. The avascular, denervated index finger was amputated at the metocarpalphalyngeal joint. Photograph courtesy of William H. Dice, MD. See color plate.

occur throughout the year,82 but snowblower injuries can be expected following a heavy snowfall.83 In a 1997 storm in Rhode Island, seven of 11 patients injured by snowblowers indicated that they placed their hands into a running machine, although three of the remaining patients said that the machine was off at the time of injury. Amputation was common as were fractures and ten- don injuries. Ten hand injuries were managed in the ED by hand surgeons and one required inpatient treatment. The majority of cases involved the index, long, and ring fingers (Figure 37.3).84

Management of open wounds should follow traditional wound care guidelines. Wounds should be thoroughly described with emphasis on size, shape, depth, foreign material, local perfusion, distal circulation, tendon function, neuromuscular strength, and two-point discrimination prior to induction of anesthesia. Plain radiography should be routine. Computed tomography scanning or ultrasound may be used to further eval- uate for deep foreign bodies.

Bleeding and pain must be controlled prior to imaging stud- ies. Resuscitation may be required, and significant blood loss might necessitate transfusion. Administration of blood prod- ucts in the face of a winter weather emergency should be care- fully considered because of the limited ability to replace banked blood. Both regional anesthesia and procedural sedation are use- ful adjuncts to ED evaluation and management of large complex wounds.

An important prerequisite for complete assessment is a bloodless field. Arterial bleeding can be controlled by ligating small vessels or using a blood-pressure cuff inflated proximally. This tourniquet technique can be applied for up to 2 hours while vascular control is achieved, the wound is explored, and copious irrigation is accomplished.

All open injuries from power tools should be considered crushed and contaminated, and thus require thorough clean- ing to remove gross debris and reduce bacterial load. A pulsatile lavage system using 130 kPa (19 psi) and high flow rates is recom- mended. Irrigation should be performed within 6 hours of injury to get the best results.85 Use of a 19-gauge catheter attached to a 30- or 60-mL syringe can achieve an adequate stream pressure of

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

596 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

35–55 kPa (5–8 psi). Clean tap water is an acceptable alternative to saline for irrigation of wounds.86 Debridement of devitalized tissue is important.

Because a winter storm might delay presentation of patients with complex wounds and availability of consultants may be limited, emergency physicians and others performing primary patient evaluations should consider delayed primary closure in some cases, especially in those whose wounds are more than 6 hours old. If not limited by restrictions on travel or personal resources, the patient can return for daily dressing changes prior to definitive debridement and closure in 3–5 days. Nonoperative ED management of fingertip amputations is accepted practice, although telephone consultation with a hand specialist is advised prior to repair.87

Tetanus prophylaxis is an important consideration for com- plex wounds. Every patient with a wound is susceptible to tetanus, even though there are few reported cases in developed nations. Public health authorities recommend a tetanus toxoid booster when there is evidence of a complete initial vaccination series. In the absence of primary immunization, tetanus immune glob- ulin plus tetanus toxoid is recommended with referral for the second and third doses of tetanus toxoid to complete primary immunization at a later time.88

The use of antibiotics by any route in the management of sim- ple wounds offers no clear advantage in preventing infection.89

Physicians should consider the potentially limited availability of pharmacies and the ability of a patient to return for complica- tions when recommending antibiotics or other medications. Use of antibiotics for noninfected simple wounds is a questionable practice and is generally not supported in the literature. How- ever, wounds that involve tendons, bones, nerves, or large vessels generally include antibiotics in ED management.

In addition to trauma, the winter environment is further associated with cold injuries and hypothermia in individuals representing a wide demographic range.73 Disabled persons, the elderly, and the socially disadvantaged are at particular risk.52,74

Electrical failures lead to use of alternative power and heat sources that are associated with CO poisoning.40 Patients can succumb to permanent disability or even death if healthcare providers do not actively assess for clinically occult primary or complicating conditions secondary to CO exposure.

Freezing (e.g., frostbite) and nonfreezing (e.g., trench foot and immersion foot) cold injuries are debilitating and should be anticipated. The risk for cold injury is markedly higher in both African American men and women than in whites of either sex.90 Additional risk factors include alcohol or drug use, psy- chiatric disease, motor vehicle trauma, and motor vehicle break- down.54,55 Patients with prior cold injury appear to be susceptible to recurrence.74

Symptoms and signs of cold injury begin when skin tem- perature falls below 15◦C, and vasoconstriction reduces blood flow. Skin changes occur when extracellular ice crystals form at temperatures below 0◦C. Intravascular sludging begins, endo- thelial leakage occurs, and edema develops. Findings associated with cold injury are based on the changes associated with tissue ischemia and the freezing process.73

Frostbite is definitively managed by active rewarming with- out massage. A practical, three-step approach is required to manage frostbite efficiently: 1) remove any constricting or wet clothing, gently drying the involved area, and protecting it from evaporative cooling and direct pressure with loosely applied dry dressings and padding to halt ongoing insults; 2) examine

affected skin areas for signs of ischemia, change in texture (e.g., waxy, inflexible, solid), presence of blisters, and loss of sensa- tion; and 3) rewarm affected parts rapidly until all skin out to the most distal portions appears perfused and pliable.

In the ED, frostbite is treated with rewarming, hydration, wound care, and pain control. Rapid rewarming is the immediate objective once frostbite is discovered. Dry external heat is not appropriate for frostbite. Rewarming should take place in a water bath at 40–42◦C, usually for 15–30 minutes until signs of skin reperfusion are evident (i.e., red or purple color and pliable texture). Systemic fluid resuscitation is usually not required, as it is in thermal burns. Once the part is rewarmed it should be elevated, splinted, and the toes and fingers separated with cotton. Rewarming is painful and parenteral opioid analgesics are usually required. Debridement of the blisters that form after rewarming is controversial and is generally not performed in the ED.

Use of thrombolytic agents has shown some promise in reducing tissue loss when given within 24 hours of rewarm- ing.91 Other attempts to use agents to enhance blood flow into and through thawed tissue have not been successful in clinical studies. Heparin, dextran, intra-arterial reserpine, hyperbaric oxygen, and surgical sympathectomy have been used in frostbite treatment without dramatic changes in tissue salvage.91

Frostbite may be categorized as deep or superficial, although some clinicians use a grading system of 1–4. Classification is made after rewarming. Favorable prognostic signs in superficial frostbite include sensation to pinprick, blisters containing clear fluid, and normal skin color. Poor prognostic signs in deep frost- bite include nondeforming hard skin, loss of sensation, blisters filled with dark or bloody fluid, and nonblanching cyanotic skin color.68

Tools are not available in the ED to define the full extent of cold injury and provide a prognosis for tissue loss. The standard of care includes a “wait and see” approach to amputation that can be quite prolonged because mummification may take up to 3 months. Researchers have studied, with limited success, the value of magnetic resonance imaging and magnetic resonance angiography in predicting tissue viability.92 Technetium-99m (99mTc) scanning has also been used to identify tissue that will require amputation.93 One study in France of severely frostbit- ten hand injuries suggested that 99mTc scanning in the first few days after rewarming predicts the level of amputation for 84% of cases.94 However, other studies suggest that 99mTc images do not accurately identify eventual levels of gangrene.92 Generally, surgeons are reluctant to decide on an amputation level until complete demarcation occurs several months later.

Patients can experience long-term symptoms from their cold injuries. Deep frostbite is associated with cold sensitivity, increased sweating, pain, hypersensitivity, and skin color change. Superficial frostbite is associated with persistent cold sensitivity, numbness, and loss of sensation.62 Nonfreezing trench foot can result in muscle atrophy and contractures.61 Cold injuries may result in chronic occupational impairment.

The incidence of accidental hypothermia is a public health issue year-round, especially as it relates to socially disadvantaged persons. These groups are at risk for being disproportionately affected by a winter storm. Other groups may also be affected if they are displaced from their residences by power failures or structural damage. Hypothermia can masquerade as other ill- nesses, especially in people with risk factors for altered mental sta- tus, syncope, and dysrhythmias. It can also be missed in trauma

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 597

victims, who have more obvious challenges. Hypothermia can be overlooked in patients who are homeless, use alcohol or drugs, or have chronic conditions such as diabetes, hypothyroidism, and psychiatric illness.95 Triage procedures should include core temperature measurement during cold environment disasters.

Hypothermia coupled with infection is associated with increased mortality. Patients with slow rewarming rates (0.5– 1◦C/h) and low serum albumin are more likely to harbor infections. Hypotension, slow rewarming rates, and bradycardia might be predictors of death during or soon after rewarming.96

Sepsis should always be included in the differential diagnosis of all degrees of hypothermia.

Victims of trauma, particularly vehicle crashes, are at risk for hypothermia as a result of delays in discovery, extrication, or evacuation. Trauma patients with a core temperature less than 34◦C might have a 35% increase in mortality.97 Hypothermia is associated with electrolyte abnormalities, acid-base distur- bances,98 coagulopathy, and thrombocytopenia.99 Hypothermic coagulopathy appears to be similar to disseminated intravascu- lar coagulation and might require management with plasma, clotting factors, and platelets. Partial thromboplastin and pro- thrombin times should be measured early in the evaluation of trauma patients who might be hypothermic.

Complete blood count, basic metabolic profile, and arte- rial blood gas analysis may be helpful in assessing hypothermic patients. Temperature correction of blood gas results is unnec- essary during initial management. Severe hypothermia can be associated with lower cardiac output from loss of plasma volume and correlates with a rise in hematocrit of approximately 2%/◦C decrease in core temperature.52 The loss of intravascular fluid can potentially decrease renal blood flow by 50%, which may result in renal failure. In the absence of significant electrocardiographic changes, correction of low potassium is generally not required because warming reverses the abnormality.

Endotracheal intubation and mechanical ventilation are nec- essary for patients with respiratory failure or cardiac arrest. Traditional rapid-sequence induction drugs are acceptable for hypothermic patients; care is needed to minimize airway and cardiac stimulation because of the risk of ventricular fibrillation. Life-threatening arrhythmias associated with hypothermia are notoriously difficult to manage.78

Preventing additional heat loss from wet clothing and skin, and exposure to ambient air is the first step in the ED man- agement of hypothermia. Attention is then directed to core rewarming, which is the major focus of activity to reverse the untoward effects from lowered core temperature. Passive and active external rewarming techniques are recommended for mild hypothermia, when core temperatures are greater than 32◦C. Covering the patient with blankets accomplishes rewarming by containing and reflecting heat generated by the patient’s own metabolism. As in the out-of-hospital setting, active external rewarming with radiant warmers or heated air can supplement this approach. These techniques can raise body temperature 0.5– 0.8◦C/h.73

Moderate and severe hypothermia require more aggres- sive management with active core rewarming – methods being mostly determined by physician expertise and hospital capa- bilities. Although core temperature “afterdrop” is an anecdotal observation when active external rewarming is used, the clini- cal significance of central pH and temperature changes reported when the periphery and core are warmed concurrently remains elusive.100

Indicators of irreversible hypothermic cardiac arrest include serum potassium concentrations of greater than 10 mEq/L or fib- rinogen less than 50 mg/dL.95 When potentially reversible cardiac arrest complicates hypothermia, cardiopulmonary bypass (CPB) is the treatment of choice when available.101,102 The warming rate for CPB is 1–2◦C every 3–5 minutes. A mechanical chest com- pression device might be useful when the availability of CPB is delayed.103 Other indications for CPB are solidly frozen extremi- ties, rhabdomyolysis, or failure of other rewarming techniques.95

An alternative to CPB is hemodialysis, which achieves similar rewarming rates, and offers some advantage when electrolyte abnormalities, metabolic acidosis, or renal failure complicate hypothermia.104

Most hospitals are not equipped to perform emergency CPB or dialysis. Active internal rewarming by aerosol inhalation, gas- tric lavage, bladder lavage, and warm enemas may be appropriate for moderate hypothermia but are mostly ineffective for victims with severely depressed core temperatures.

However, thoracic and peritoneal lavage are alternative ther- apeutic options in the ED. In either technique, two intracavitary tubes are placed under aseptic conditions. Sterile water or crys- talloid fluid warmed to 40–42◦C is allowed to circulate through either the chest or peritoneal cavities from one tube to the other and drain by gravity. Thoracic lavage is an effective method of active internal rewarming105 and can achieve warming rates of 3– 6◦C/h.106 Thoracic lavage might be preferable because the heart is warmed directly. Peritoneal lavage is an additional effective method that can achieve warming rates of 2–4◦C/h.107 They can be performed together.

CO poisoning is another winter-related condition, which can be confused with nonspecific viral syndromes and other common causes of headache because patients most commonly present with headache, nausea, and dizziness.108 People with other medical conditions such as heart or lung disease might be more susceptible to the effects of CO and present to the ED for treatment of chest pain or shortness of breath.109 Severe CO poisoning contributes to ischemic myocardial injury and can double mortality rates.110

Depression, dementia, and psychosis have been reported between 2 and 28 days after CO poisoning. Carboxyhemoglobin concentration does not, however, reliably predict delayed neu- rological sequelae such as an inability to concentrate, learning impairment, memory loss, and abnormal motor function.111

The standard measurement for the presence or absence of CO is blood carboxyhemoglobin concentration. Pulse oximetry is unreliable when CO is present because artificially high saturation readings are associated with carboxyhemoglobin.

The treatment for CO poisoning is a high concentration of inspired oxygen. The half-life of CO breathing room air is approximately 5 hours. The half-life decreases to about 1.5 hours on 100% oxygen by either face mask or ventilator and to 0.7 hours using hyperbaric oxygen therapy.108,112 Hyperbaric cham- bers traditionally used to treat CO poisoning are located at fixed facilities but success has also been reported with portable chambers.113

Generally, hyperbaric oxygen therapy is indicated for any history or persistence of syncope, altered mental status, neu- rological impairment, or myocardial ischemia, as well as for most pregnant patients.108 Hyperbaric treatment might also be indicated in patients found to be at higher risk for long-term cognitive impairment (i.e., patients >36 years old, exposure >24 h duration, or “higher” carboxyhemoglobin levels).114 One

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

598 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

study suggested that hyperbaric treatment reduced one in six cases of delayed neurological conditions when treatment was begun as late as 24 hours after poisoning.115 Others studies have found little benefit of hyperbaric treatment when delayed longer than 6 hours.116 In any case, hyperbaric facilities may be scarce resources that are difficult to access during the aftermath of a winter storm.

Because hyperoxia is associated with hypocapnia, which in turn decreases cerebral blood flow, one study of healthy volunteers indicated that ideal treatment of CO poisoning might include maintenance of normocapnia.117 An animal model suggested that hyperventilated intubated patients breathing higher concentrations of CO2 to maintain an arterial paCO2 of 5.33 kPa (40 mm Hg) might eliminate CO twice as fast,118

but the clinical significance of this in human CO poisoning is unknown.

In terms of preparedness for epidemics of CO poisoning, one report of an epidemiological spike during a winter storm found that almost four of every five patients arrived between 18:00 and 06:00 hours.39 In this study, the 81 patients were distributed over 3 days and 13 hospitals,39 so the impact on individual EDs in that urban area was not too great. Even at a large university medical center in North Carolina, an epidemic of 200 cases in 1 week resulted in the need for hyperbaric oxygen therapy that outstripped available chambers in the area.42 The chamber at Hôpital du Sacre-Cœur de Montréal was used to treat 45 patients in the first 9 days after the Ice Storm of ’98.14

Secondary Triage Secondary triage to specialty centers can be interrupted by

degraded conditions at sending and receiving facilities in the midst of a weather emergency. Interfacility patient transporta- tion in inclement weather or over dangerous roads can be a serious challenge. Regional referral centers have been created for services not available at all hospitals. Examples include such spe- cialty services as trauma, burns, and hyperbaric therapy. Rural and suburban hospitals could require all of these services in asso- ciation with a winter storm event. In situations where resource needs exceed availability, these facilities can be overwhelmed.

A scarce specialty resource may be unavailable. Therefore, contingency plans must be in place for community hospitals to manage cases on a temporary basis that are normally transferred under less constrained circumstances. Increased needs and evac- uation delays may require all healthcare professionals to adjust procedures. Tertiary care centers should be prepared to increase capacity for those they can receive and provide enhanced con- sultative services by telecommunications to facilities unable to transfer patients (see Chapter 23).

Disposition ED and hospital discharge planning may require coordina-

tion in ways that are atypical and challenging. Displaced per- sons frequently find it difficult to return home. Physicians must consider several issues when discharging such patients. Is the patient’s home still intact? If so, will the patient go to a home without power, heat, or communications? If not, will the patient go to a shelter that may limit their ability to care for them- selves? Other options include patients remaining at the hospital for longer periods. However, this can potentially limit the num- ber of additional admissions. Moving patients to less crowded hospitals is difficult due to existing road conditions or weather affecting or aeromedical transportation. Even when patients can

be discharged, home healthcare services (e.g., oxygen, pharma- ceuticals, and nursing care) or the ability to return for addi- tional treatment (e.g., chemotherapy or dialysis), on which they depend, may not be available. Follow-up clinics may not be open for days or weeks.

Healthcare Systems

Winter storms may limit the capacity of first receivers to manage patient flow to their institutions. Patients often find their own transportation or are transported by others to community Eds, regardless of their capacity to accept new victims. During these mass casualty situations, the standards of care are often modified, so the most good can be done for the most people. To achieve this goal, providers should shift to a “minimal acceptable care” standard. These and related concepts are discussed in more detail in Chapter 3.

Permanent Facilities The most critical systems for hospital operations are119

■ Physical plant ■ Utilities ■ Personnel ■ Supplies and equipment ■ Internal and external communications ■ Transportation ■ Supervisory and managerial support

These can all be adversely impacted by winter storms. Power loss may affect lighting, medical equipment, and safety systems. Frozen pipes may block ready access to clean drinking water, water for personal hygiene, and sewage outflow. Hazardous road conditions may make it difficult for the next shift or surge per- sonnel to report for work. Those already on duty at a facility may have to work extended shifts, or additional days, without relief. Supplies become depleted and are not replaced and mal- functioning medical equipment is not repaired due to lack of technicians.

Information on winter storm impact on individual health- care facilities can be found in articles on medical care, but health- care systems studies have not been published. Reports from insti- tutions that have chosen to share their experiences are strictly anecdotal, but are illustrative of potential problems others may face in the future.

A November 1996 ice storm occurring in Washington State interrupted power to approximately one-third of its population and a number of hospitals. One trauma center required six diesel generators to maintain operations. Had it not received power from a secondary utility feed, it might have been without elec- tricity for 12 days.120

In the aftermath of the Ice Storm of ’98, many Canadian hospitals operated on generator power for almost 3 weeks.14

Even in the absence of a disaster situation, one report identified significant facility power losses on the day of a mid-autumn heavy snowstorm. At one hospital, a complete power loss forced nurses to hand-ventilate patients in its intensive care unit for 45 minutes.26 Facilities must know what equipment does not receive power from emergency generators.120

During a 2006 October snowstorm in western New York, power failure at a county water treatment plant threatened the potable water supply to the regional trauma center and children’s

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 599

hospital. The hospitals’ supply of water was placed in jeopardy and delivery of bottled water was difficult or impossible due to icy roads.121,122

Telecommunications may be interrupted for a variety of rea- sons. This may affect telephones and pager systems. One hospital had to arrange for emergency delivery of numerous cellular tele- phones just to communicate within the facility itself.122

Public and private transportation may be dangerous, diffi- cult, or impossible. Therefore, staff absenteeism can be a signif- icant concern in the aftermath of winter storms. Managers have employed emergency vehicles to transport personnel to work and relieve those on duty at the time of a storm.26

Transportation can also be problematic in three other situa- tions: moving victims from the community to hospitals or other healthcare facilities; accomplishing interfacility hospital trans- fers for access to specialty care; and discharging some patients to home or to skilled nursing facilities. After the Ice Storm of ’98, some hospitals opened additional ward beds for discharged patients who could not return home.121 Providing home health- care services was a challenge noted after an ice storm occurring in southeastern Oklahoma in December 2000.122

Sheltering employees and their families (including pets in some cases) often becomes an additional function of a fixed facility. During the prolonged crisis following the Ice Storm of ‘98, Montreal General Hospital set up shelters for employees and their families in unused portions of the facility. It provided free food service and child care in order to ensure it had enough staff to meet patient needs.14 Ottawa Civic Hospital opened two oncology floors to house staff for up to 3 weeks.121 Oth- ers have reported implementing similar strategies, or utilizing nearby compassionate-care facilities.120

Healthcare facilities have also taken in displaced persons.122

One psychiatric hospital in Ontario sheltered many of the need- iest people in its community but then found that its personnel had to care for the medical needs of several elderly boarders.123

Providing shelter services in addition to supporting an increased ED or inpatient census may require additional security assets.121

Temporary Facilities A variety of temporary facilities might be established by gov-

ernmental and nongovernmental organizations to mitigate the human impact of severe winter storms. Most of these would be for the purposes of providing shelter from the elements and acting as distribution centers for water and food, although lim- ited first aid services might be available at some. In the context of a community’s emergency operations plan, the likelihood of limited ground movement in a storm’s aftermath would argue for staffing these facilities and notifying the public of their loca- tions in advance of expected severe weather. The media could use public service announcements to direct persons in need to the shelters that remain functional after the event.

Temporary shelters can house and care for those who are dis- placed from their homes due to structural damage or from the lack of power or heat. Community centers, schools, churches, and government buildings are commonly used for sheltering. Some of these buildings may remain connected to a functioning power grid or have emergency generators facilitating the provi- sion of heat, water, and amenities such as microwave ovens and televisions – all of which can make the disrupting experience of sheltering more tolerable. However, the number of toilets, the availability of shower facilities and food preparation areas, and the need for extra security for large numbers of occupants might

stress these facilities. Loud background noise, lack of privacy, and poor sleep cycles in such conditions lead to mental health issues. These include anxiety and depression, substance abuse, psychosis, suicidal ideation or attempts, and long-term sequelae such as posttraumatic stress disorder.

Infectious diseases can be difficult to control in shelters. Bac- terial and viral upper respiratory infections, gastrointestinal dis- turbances, and other maladies can be spread rapidly between per- sons living under these conditions. Sheltered populations may be at greater risk if shelter officials have limited medical or public health training or the supply of sanitation items such as anti- bacterial hand wipes, alcohol-based hand sanitizers, and surface disinfectants is limited.

If a potentially contagious disease is identified, measures should be undertaken to isolate infected and exposed persons. This can be exceedingly difficult in an austere shelter environ- ment as separate rooms in which to segregate individuals or families are frequently lacking. Curtains or hanging sheets and arranging cots to prevent face-to-face relationships are useful techniques to mitigate transmission risk. Infected individuals should be restricted from group events (e.g., meals, social activ- ities, and gatherings for announcements) to prevent cross con- tamination. In the event of gastrointestinal problems, strict san- itation must be enforced at toilet facilities, and these should be cleaned and disinfected more often than usual.

The availability, storage, preparation, and distribution of water and food are a concern in any sheltering situation. Dur- ing disasters, food is often donated by nongovernmental orga- nizations and local ad hoc community-based groups. It is also provided by institutions such as hospitals, nursing homes, jails, and other facilities accustomed to preparing large quantities of food on a regular basis. Proper storage is also an issue for perish- able items. Any leftover foods must be stored properly to avoid spoilage.

Care should be taken to ensure that personnel are adhering to health codes. Prepared food must be kept heated or cooled to proper temperatures during preparation, storage, and service to avoid the potential for large-scale food-borne illness, which can also disable rescuers and healthcare workers if they ingest the sup- plied food. Local, regional, state, or federal public health officials can be consulted before an event for assistance with planning and early after a storm for assistance with implementation of food management programs.

Pots, pans, and utensils used to cook food must be thoroughly cleaned. If a shelter is without running water but has access to delivered water, a three-bucket system can be used. Hot, soapy water in the first bucket is used for washing and subsequent con- tainers are used for a two-stage rinse. Alternatively, personnel can take dishes and food preparation equipment to another location for washing. The use of disposable plates, bowls, and utensils can make it feasible to feed hundreds of people and reduce the need for washing of dishes, but this approach comes with additional logistics issues of supply replenishment and waste disposal. The same would apply to prepackaged meals.

In disaster situations requiring additional healthcare capac- ity, consideration should be given to using resources developed for other purposes. The Modular Emergency Medical System (MEMS) created primarily for a response to biological terrorism is an excellent example of this dual-use approach in the U.S. If prepared and integrated prior to an event, the concept provides a framework for expanding community healthcare capacity as required. Under MEMS, one or more high-volume reception

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

600 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

Table 37.3: Some Possible Survey Questions During Modified Cluster Sampling of Households in Areas Affected by a Winter Storm

■ Not enough water or food ■ No home (displaced or difficulty returning) ■ No running water ■ No electricity ■ No heat ■ No functioning toilet ■ No telephone ■ No personal vehicle ■ No access to commercial or public transportation ■ Unable to obtain needed medications or home health services ■ Injuries ■ Illnesses (acute or chronic) ■ Need for medical care ■ Need for counseling ■ Special needs (disabilities, extremes of age, mental health issues, pets,

other)

Homes and shelters can also be inspected by assessment teams for haz- ardous conditions. Public service information, education, and limited medical care may also be provided.

This information is in the public domain. It was modified from the CDC.128

and triage facilities, known as Neighborhood Emergency Help Centers, can be established directly within affected areas. These provide initial community healthcare that is more easily accessi- ble when victims do not have ready access to transportation.124

When hospitals are overwhelmed, managers can establish one or more off-site inpatient facilities known as Acute Care Cen- ters. These are staffed and equipped to manage large volumes of patients with less serious problems, thus allowing hospitals to concentrate on more seriously ill and injured individuals.125

Rapid Needs Assessments

Rapid assessments126 for identifying immediate challenges cre- ated by a winter storm and determining potential resources required to mount an effective response are necessary to avoid a dysfunctional approach. The fundamental challenge is rapidly gathering reliable and useful data when operation of ground and air transportation vehicles might be difficult or dangerous. Snowmobile drivers and Nordic skiers may be useful in these situations.

Modified cluster sampling is an epidemiological method often advocated by the U.S. Centers for Disease Control and Pre- vention (CDC). This process gauges the impact on communities that have sustained severe property damage that make egress and ingress difficult in the initial aftermath of an event.127,128

The technique involves sampling 30 randomly selected clusters of land in the affected area. Assessment teams attempt to inter- view persons still present at these locations. Typical questions are listed in Table 37.3. Data are collected and analyzed to estimate rates, which are then extrapolated to total population numbers based on pre-event census information.

This sampling method was used with success 3 days after Hur- ricane Andrew hit Florida in August 1992.129 It was also used in rural Maine following the Ice Storm of ‘98, but the severe limita- tions on travel delayed the survey 10 days.130 Nonetheless, results

of the assessment revealed that 14% of the affected population was still without electricity, 18% were using gasoline-powered generators, and 68% had their utility power restored in that time frame. Many had no telephone service but most had access to public service broadcasts through either radio or television.

Needs assessments can also identify victims with injuries and illnesses, either related or unrelated to the storm. Illnesses may be new or exacerbations of chronic conditions from exposure to the cold environment or inability to access customary care. Healthcare access – in the forms of EMS availability or capabilities of individuals and families to travel to medical offices, clinics, and hospitals – can also be assessed. Of particular note in the CDC’s assessment after the 1998 ice storm, potentially hazardous sources of CO were identified in many homes without electrical power. Only 8% had working CO detectors.130

External Response

Major disasters disrupt or overwhelm local response capacity to such a degree that assistance from outside the affected area may be required to help mitigate the human impact of the event. Regional resources may be held at the county, province, or state levels – the exact terminology differs between countries. As such, specifics are beyond the scope of this chapter. The important concept is that any response must be coordinated through an incident management system capable of linking all of the neces- sary resources.

When local and regional resources do not have the capacity to significantly mitigate the effects of a PICE, national or inter- national resources are often necessary. In the U.S., the President must declare a federal disaster after formal requests by the gov- ernor(s) of the state(s) affected (see Chapter 9). With regard to winter storm response, FEMA’s snow-assistance policy focuses on record or near-record snowfalls for an area to qualify for aid.131

Therefore, local and state governmental and nongovernmental organizations must be prepared for storms of lesser magnitude. Since inherent delays in mobilizing any federal governmental or military response are inevitable – unless these assets are posi- tioned pre-event in advance of an approaching storm – planners must acknowledge that local resources will be the only response in the first 3–4 days after an event.

A Disaster Medical Assistance Team (DMAT) is a commu- nity-based federal government asset designed to provide medi- cal care during a disaster or other local, regional, or state event. A team can also be federalized in support of a national event outside its jurisdiction of attachment. There are approximately 50 DMATs in the U.S. Each team is sponsored by an organi- zation such as a public safety agency, hospital, or nonprofit public or private group. DMATs are composed of physicians, physician assistants, nurses, pharmacists, respiratory therapists, paramedics, emergency medical technicians, and a variety of healthcare, logistical, and administrative personnel. DMATs con- sist of 50–125 members to ensure that a sufficient number of people can deploy from their day-to-day jobs on short notice. Approximately 35 are typically chosen for a particular mission.

DMATs are designed to function as rapid-response elements to supplement local medical care until other resources can be mobilized or the situation is resolved. They deploy as a self- sufficient unit with medical equipment and supplies, climate- controlled tents, electrical generators, and other support equip- ment necessary to establish a base of operations. They can treat up to 250 patients per day in a fixed or temporary site. They can

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 601

Table 37.4: Internet Resources for Public Individual and Group Preparedness Information

Canada Public Health Agency http://www.phac-aspc.gc.ca/cepr-cmiu/index.html

United Kingdom Department of Health http://www.dh.gov.uk/keepwarmkeepwell

United States American Red Cross http://www.redcross.org/services/prepare/0,1082,0 252 ,00.html Centers for Disease Control and Prevention http://www.bt.cdc.gov/disasters/winter/factsheet.asp DisasterCenter.com http://www.disastercenter.com/guide/winter.html Emergency Management http://www.fema.gov/plan/index.shtm Health and Human Services http://www.hhs.gov/disasters/emergency/naturaldisasters/cold/index.html National Weather Service http://www.nws.noaa.gov/om/brochures/wntrstm.htm Occupational Safety & Health http://www.osha.gov/SLTC/emergencypreparedness/guides/winterstorms.html Ready.gov http://www.ready.gov/america/publications/allpubs.html

Many local and state governments, nongovernmental organizations, and universities also have useful information.

also potentially supplement the staff at existing local healthcare facilities, although credentialing issues have limited this role to date. Responsibilities of DMATs may include triage, provision of high-quality medical care in austere settings, and preparation for evacuation to more appropriate healthcare facilities. DMAT personnel may also be deployed to more distant facilities to assist in receiving large numbers of patients. DMATs are discussed in more detail in Chapter 9.

Multiple DMATs have deployed in response to northeastern U.S. ice storms. They have provided needed medical care to com- munities or when local doctor’s offices, clinics, and pharmacies had closed due to lack of power or inability of staff to travel to their places of employment. The majority of DMATs supported local hospital staffs, allowing them to rest, tend to their families, and address damage to their property. Some members assisted local nursing home staff in caring for residents while others con- ducted shelter clinics on a rotating basis.

Following the Ice Storm of ’98, a mobile mission was devised. A utility vehicle was stocked with equipment, supplies, and medi- cations. Support staff accompanying the vehicle included a physi- cian, a nurse, and a paramedic. This traveling clinic visited each shelter in the area daily. Several house calls were also made to check on families who remained in their homes. Some were treated in place for pneumonia, viral syndromes, Streptococcal pharyngitis, and other minor illness and injuries. Many people displaced from their homes and staying in shelters had forgot- ten to bring or exhausted their supply of regular medications. DMAT teams provided small amounts of these medicines until local pharmacies were functional and people could safely travel to pick up prescriptions.

Public Information

The ultimate goal of any public educational effort is to prevent problems before they occur. To achieve this endpoint, the CDC in the U.S. has called for standardization of public health messages issued for a variety of events.132 One purpose of this initiative will be “to receive, manage and disseminate alerts, protocols, procedures and other information for public health workers, primary care providers, and public health partners in emergency response.”133 This program is supported by a publication from the National Disaster Education Coalition entitled a Guide for Standard Messages specific to winter storms.134

Medically related information disseminated to the public and to healthcare personnel should be based on evidence where it exists. Areas of focus should be the causes of morbidity and mortality associated with winter storms: cold exposure; CO poi- soning; and injuries such as MVCs, slips and falls, snow removal, and those involved in recreational activities.

Cold Exposure A number of risk factors are associated with accidental

hypothermia. These include extremes of age (particularly <1 y and ≥60 y), male sex (likely behavioral), ethanol ingestion, treat- ment with neuroleptic medications (often used in patients with psychiatric conditions), hypothyroidism, and malnutrition. Two populations at particularly high risk are: 1) relatively young par- ticipants in outdoor sports and recreational activities, who sus- tain overwhelming cold stress, and 2) “vulnerable populations” exposed to moderate indoor cold stress, most specifically the elderly.17 In a longitudinal study of 47 elderly people in the U.K., a progressive decrease in thermoregulatory capacity associated with increasing age was noted.135 Numerous reports have docu- mented greater morbidity and mortality in the elderly following winter storms.17,33,52,74 Similarly, ethanol ingestion and psychi- atric illness were found to be risk factors for frostbite. In one Canadian study, 77% of cases presented with one of these as the cause of impaired mental capacity.74

Little has been written and almost no research performed on the impact of public information campaigns in preventing frostbite and hypothermia. The U.K. has been conducting a cam- paign called “Keep Warm – Keep Well” for a number of years.136

However, publications of any public health research regarding its impact on reducing cold-associated morbidity and mortality could not be found in open sources.

Assuming an intact power supply and Internet access, some public resources may be found in Table 37.4. Alternatively, doc- uments could be printed in advance of a disaster in preparation for possible power loss. Local organizations could also develop similar educational products for their own public awareness pro- grams.

Carbon Monoxide Poisoning As always, prevention is the key to mitigating the impact of

CO poisoning. Because CO is colorless and has no detectable odor or noxious properties, preventive measures should involve

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

602 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

a combination of public information campaigns and detec- tor/alarm devices.

Efforts to increase public awareness must consider uneven ethnic distributions in the epidemiology of CO poisoning. There- fore, awareness programs must be individualized to each spe- cific region or jurisdiction.137,138 In the 1993 storm affecting the largest urban population center of Washington State, the majority of patients poisoned with CO from indoor cooking with charcoal briquettes were ethnic minorities. Asians repre- sented the largest group, most of whom did not speak English. By comparison, all cases caused by gasoline-powered electricity generators involved non-Hispanic whites.39 Newspaper, radio, and television were extensively utilized in an educational cam- paign and one fire department even distributed 2,000 leaflets door-to-door. However, all warnings were communicated only in English.38 In Rochester, New York, public education may have contributed to the modest reduction in absolute numbers of CO poisoning cases seen during an ice storm in 2003 compared with a similar storm in 1991. On the other hand, there was a sig- nificant decrease in indoor cooking as a source mechanism.139

Others have reported difficulty disseminating information when telecommunications were disrupted by the storm.13

CO detectors with an audible alarm have the potential to alert exposed individuals if appropriately installed and maintained.140

One study concluded that detectors might prevent up to half of unintentional deaths.141 The same study also found that 42% of those who were likely asleep at the time of fatal poisoning had alcohol in their system. This proportion was similar for victims whose blood alcohol level was either above or below 100 mg/dL.141 In a report of unintentional nonfire-related CO poisonings following the December 2002 ice storm in North Carolina, researchers noted that the severity of poisonings in households with functioning alarms was much less than in those without them.142

Vehicle Incapacitation The use of any vehicle carries with it the risk of mechanical

breakdown or inability to negotiate winter terrain. When vehi- cles are incapacitated in noninclement weather or when help is readily accessible, it usually causes more frustration and aggra- vation than illness or injury. However, when driving in an inhos- pitable and dangerous environment, such as the aftermath of a severe winter storm, or when rescue is delayed by environmental conditions or shear EMS call volume, immobility could rapidly increase the risk to health.

Once stranded, an automobile’s occupants must decide whether to remain with the vehicle or walk to better shelter. History suggests staying with the vehicle is the best course of action in most circumstances. Travel across terrain can be dif- ficult due to snow and ice, especially with no knowledge of the area. Blowing snow can quickly cause disorientation. Land navi- gation is a learned skill, for which many motorists and emergency responders have no training. A detailed map or portable global positioning system device may help, but individuals may still be at risk from lack of sufficiently protective clothing. Remaining in a vehicle can retain some heat for a short period and will block any wind that can hasten the onset of frostbite and hypother- mia. During any overland trek, a fall resulting in an injury to an ankle, knee, or back could incapacitate an individual and increase the risk from cold exposure. Unwittingly treading on thin ice over a body of water could lead to drowning or immersion injury.

Table 37.5: Suggested Items for a Winter Storm Survival Kit to Be Carried in Motor Vehicles for the Contingency of Being Stranded

■ Blankets or sleeping bags ■ Flashlight with extra batteries∗ ■ First aid kit ■ Knife ■ High-calorie food that is nonperishable ■ Extra clothing, hat, mittens or gloves, boots ■ Large empty can and plastic cover with tissues/paper towels for san-

itary purposes ■ Smaller metal can and waterproof matches to melt snow for drinking

water ■ Sand or cat litter for tire traction ■ Tire chains ■ Shovel ■ Windshield scraper and brush ■ Tool kit ■ Emergency tire repair equipment (canned compressed air and

sealant) ■ Tow rope ■ Booster cables ■ Water container ■ Road maps and compass or global positioning system device ■ Signal devices (e.g., flares, light-emitting diode lights, pylons) ■ Cellular telephone with charger or citizens band (CB) radio with

extra batteries∗ ■ AM/FM and National Oceanic and Atmospheric Administration

radios with extra batteries∗

∗ Lithium batteries have longer life and work better in cold temperatures.

This information is in the public domain. It was modified from the CDC.144

The chances of being seen and rescued are much higher when staying with a vehicle, as passing cars may render aid in short order if the road is frequently traveled. If others familiar with the motorists’ travel plans note that the stranded individuals are overdue, rescue assets will inevitably be dispatched. Attaching a brightly colored object to the antenna, putting up the hood, activating emergency flashers, and deploying road flares all signal distress. A dark unlit car on the edge of the road does not beckon help nearly as well as one with multiple active methods indicating the need for assistance. Care should be taken to prevent the buildup of snow against the exhaust system of the vehicle, which could cause CO poisoning without warning.37

Gasoline tanks should be kept near full to avoid ice forma- tion inside the tank and fuel lines and to have enough fuel to idle the car 10 minutes each hour for heat if stranded. Motorists should not travel alone whenever possible, and someone should be made aware of their timetable and estimated routes of travel. The U.S. National Oceanic and Atmospheric Administration143

and the CDC144 suggest carrying a storm-survival kit when driv- ing during the winter season (Table 37.5).

Other Injuries The CDC has stated that “officials should consider the fol-

lowing recommendations during blizzards: 1) early in the storm, warn against nonessential driving; 2) announce publicly that persons who must drive should have extra clothes and food with them and remain in their vehicle if stranded; and 3) advise extreme caution if the heating system is used while the vehicle is

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 603

stopped [even for short periods], because exhaust systems may become blocked with snow and ventilation adequacy is difficult to determine.”145

Heavy snow can collapse the roofs of buildings not designed for the added weight. Family, friends, and the media often advise homeowners to clean snow from their roofs, but this can be a haz- ardous undertaking for people who do not have the agility and balance, strength and dexterity, or safety equipment to undertake the task.146 Official broadcasts should warn the public about the potential dangers of clearing rooftop snow, and offer alternative resources for those who do not feel safe or capable of doing it. Media announcements regarding the cardiovascular risks of shoveling and the safe use of snowblowers and other power equipment might also be considered, but the potential impact on the epidemiology of injuries is unknown at this time. The same is true for mechanisms of injury related to winter recreational activities.

Fall prevention has been part of the geriatric medicine research agenda for some time.147,148 With regard to slippery conditions caused by a winter storm, the elderly remain the pop- ulation at highest risk.33 Some prevention strategies may have application to other populations as well. Additional strategies may require development but none can be expected to have sig- nificant impacts on incidences and outcomes without ongoing surveillance programs to elucidate the epidemiology of winter slips and falls.

Perhaps the best public awareness campaigns involve sim- ply extending existing community injury prevention and con- trol measures. This would require continuous research efforts in injury epidemiology and clinical management to determine best practices for incidence reduction and improved outcomes.149

Information on subpopulations related to winter storms might be extracted from larger population-based data collection and analysis programs.

RECOMMENDATIONS FOR FURTHER RESEARCH

As with many research questions in public health, it is difficult to measure prevention. Particularly in the field of disaster medicine, comparisons can often be made only to baseline rates of mor- bidity and mortality before a PICE. Evaluating any data between different events, even those caused by the same insult such as a winter storm, is made extremely difficult by the heterogeneity of multiple characteristics of the incident itself and the region affected. Determining the effectiveness of any change in medi- cal interventions or response tactics, techniques, or procedures often lacks any valid cohort to which new data can be compared. Numbers are often small, especially when trying to examine sub- sets of populations that may derive the most benefit from any change in standard practices.

Although numerous disaster medicine centers have sprung up in the last decade, no nationally or internationally coordi- nated research agenda has been disseminated. The U.S. National EMS Research Agenda150 made no specific call for outcomes research in disaster medicine. With specific regard to winter storms, no recommendation was made to study the potential occupational safety and health effects of driving in hazardous winter conditions or working in cold environments, better meth- ods for managing cold-induced problems in the field, or out- of-hospital management of patients for prolonged periods of time.

Most disaster research must be multidisciplinary and col- laborative,151 with relationships made and responsibilities delin- eated before an event. Until more consistent data are collected for more events over a greater period of time, winter storm research will need to focus primarily on population effects. The goal is to develop better methods of rapidly identifying populations at risk for adverse health impacts caused by difficult outdoor movement, widespread power losses, and limited access to health care. Spe- cific areas should include risk mitigation; system preparedness; out-of-hospital access, care, and evacuation; rapid needs assess- ments; and infrastructure recovery.

Risk Mitigation No known force can stop a winter storm. However, the

risks to potentially vulnerable populations are subject to mitiga- tion. Research should focus on making society less susceptible to disruption. Developed solutions would necessarily require a cost/benefit analysis based on sound evidence. For instance, one method for potentially decreasing the incidence of accidental hypothermia and CO poisoning might be to equip every build- ing at risk with a safe gasoline-powered generator and plenty of fuel. However, supplying millions of homes with this emergency capacity in order to save at most a few hundred lives every year might be too expensive. Perhaps a pre-event public education campaign would be more cost effective. Neither option, nor any of a myriad of other programs in the spectrum of cost versus benefit, should be implemented without research in one or more representative geographical regions or vulnerable populations.

One key feature to any pre-event risk mitigation policy is updating quantitative and qualitative data on vulnerable pop- ulations as often as practicable. Census data are useful, but so much more is necessary following a PICE. Are there buildings at higher risk for power loss than others? Which of those have safe alternate energy sources? Which households have infants, elderly, or disabled persons? Which have persons requiring home-health services such as oxygen, antibiotics, chemotherapy, or dialysis that could not be delivered over hazardous road conditions?

System Preparedness Protection of the public safety and healthcare infrastruc-

ture against physical hazards (e.g., power loss, and inaccessi- bility) and system breakdown (e.g., victim needing unavailable resources) should also be researched. How can rescue capacity, EMS response, and hospital capabilities be preserved during and after a major winter storm? How fast can roads be cleared? How well can response vehicles move around the affected area? Other research priorities involve organization of the healthcare sys- tem. Should EMS personnel be more dispersed throughout the community prior to storm arrival? Should hospital staffing be bolstered before some storms to decrease absenteeism and allow for those on duty to maintain more normal work–rest cycles? Emergency management tactics, techniques, and procedures for planning, operations, safety, and overall effectiveness should also be scrutinized through research. To date, it is unknown whether these or any other interventions will improve medical or mental health outcomes after an event. Only proactive research will be able to answer these questions.

Human Access, Care, and Evacuation True research into evidence-based best practices for out-

of-hospital medical response to victims of a winter storm is virtually nonexistent. As such, there are many questions that

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

604 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

require answers before significant funds should be invested in unproven equipment and technologies.

What balance of weather conditions and population needs make it unsafe to attempt or continue access efforts? What are the best search techniques for victims who cannot call for help? What types of vehicles or vehicle equipment should be used to access victims who are able to call for help? Which agencies are needed to facilitate access to a neighborhood with ice or snow cover, felled trees, and downed power lines? How best can rescue and EMS personnel be protected when working in cold environmental conditions? What are the caloric and fluid requirements of response personnel? Should work–rest cycles be different under cold stress?

In a mass casualty situation, what triage methodologies result in the best outcomes for populations affected by a winter storm? Preventing additional heat loss is most likely beneficial; however, should active-external, or even active-internal, rewarming be undertaken in the out-of-hospital setting or in an ambulance en route to the hospital? If so, what are the best methods of accomplishing this? How much rewarming should be achieved before hospital arrival? Is the care of a cold water immersion victim any different from that for a patient suffering from “dry” hypothermia? How does frostbite or hypothermia complicate other medical conditions for which EMS may have primarily been called? Should medic training include the diagnosis and treatment of advanced complications not commonly seen when healthcare access is not significantly delayed? Should medics be trained to manage patients for longer periods of time than they are accustomed when evacuation times are not limited by weather or road conditions? What balance of those conditions and patient medical requirements make it unsafe to attempt evacuation? Which patients should simply receive care but not be evacuated? Can protocols be used for these decisions or must medical control be consulted by radio or telephone to achieve safe dispositions for patients?

What types of vehicles or vehicle equipment should be used to evacuate victims? Is there any benefit to air versus ground versus water evacuation? Are field-expedient nonmedical vehicles safe for some patients? If so, which ones, and who makes the decision regarding vehicle selection? What is the best destination for a severely hypothermic or frostbitten patient? If it is a tertiary referral center, is there degradation of benefit over time? Should an intervening stop at a less-capable ED be made for immediate care before transfer to the referral center?

Rapid Needs Assessments The CDC has posted suggestions for community studies in

the early aftermath of any event that causes widespread property damage.128 The first step would be to identify neighborhoods or other specific areas that are the most severely affected. This could be accomplished by satellite data, aerial overflights, or by reports from reliable observers within the distressed area. Once ground access is possible, surveyors would collect data (Table 37.3) from members of households, occupants of busi- nesses and public buildings, and emergency responders. Disaster managers can use this information to determine: 1) the demo- graphics of affected persons and families, 2) their physical and mental health; 3) their medical needs; and 4) their living condi- tions and additional resources required. Disaster response orga- nizations should use the current information provided by the modified cluster sampling technique to target subsequent relief

efforts, but even this method of data collection and analysis are worthy of study for accuracy and utility.

Repeated assessments throughout the response and recov- ery phases should be used to verify the effectiveness of measures taken. This last piece is frequently missed, which results in esti- mating the impact of the event but not the impact of various aspects of emergency management. Comparing prior commu- nity assessments within the same event would have more validity than comparing them with a different historical event. If relief efforts or other interventions are deemed unsuccessful, this could focus interevent research on new methods or technologies that may be more useful in similar circumstances following a future PICE.

Infrastructure Recovery Although recovery should begin at the same time as the ini-

tial response, the latter will dominate early and the former will dominate late in the continuum of the disaster life cycle. System- atic and targeted data collection must be conducted throughout all four phases of the disaster (see Chapter 9). What worked and what did not, costs and benefits of different approaches, and proposed plans for the next event must be shared with other communities facing similar problems. As yet, there is no single repository for best practices in response and recovery, except in limited regional contexts.

Analysis of recovery data from the last event in a given region, or a critical review of past disasters in other regions, should produce better preparedness for the next occurrence. Not all PICEs are the same, but similar events cause similar challenges. By studying such incidents, knowledge can be learned and solutions proposed and tested before they are needed to save lives and protect property and systems. Implementation of these research priorities has the potential to significantly decrease morbidity and mortality from winter storms.

REFERENCES

1. Crawshaw LI, Wallace HL, Dasgupta A. Thermoregulation. In: Auerbach PS, ed. Wilderness Medicine. 5th ed. Philadelphia: Mosby-Elsevier; 2007:110–124.

2. Federal Emergency Management Agency. Presidential Disaster Declarations: December 24, 1964 to March 3, 2007. Available at: http://www.fema.gov/pdf/hazard/map/declarationsmap1964 07.pdf . Accessed November 25, 2008.

3. Federal Emergency Management Agency. Major Disaster Decla- rations, Emergency Declarations, and Fire Management Assis- tance Declarations. Available at: http://www.fema.gov/news/ disasters.fema. Accessed November 25, 2008.

4. National Climactic Data Center/National Oceanic & Atmo- spheric Administration. 1980–2006 Billion Dollar US Weather Disasters. Available at: http://www.ncdc.noaa.gov/img/reports/ billion/disasters2006.pdf . Accessed November 25, 2008.

5. National Climactic Data Center/National Oceanic & Atmo- spheric Administration. Billion Dollar Storms. Available at: http://www.ncdc.noaa.gov/oa/reports/billionz.html#chron. Accessed November 25, 2008.

6. Koenig KL, Dinerman N, Kuehl AE. Disaster nomenclature – a functional impact approach: the PICE system. Acad Emerg Med. 1996;3:723–727.

7. Wikipedia: The Free Encyclopedia. Kyrill. Available at: http://en. wikipedia.org/wiki/Kyrill /storm. Accessed November 25, 2008.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 605

8. Wikipedia: The Free Encyclopedia. January 2007 North Ameri- can Ice Storm. Available at: http://en.wikipedia.org/wiki/North American ice storm of 2007. November 25, 2008.

9. National Weather Service/National Oceanic & Atmospheric Administration. National Weather Service Glossary. Available at: http://www.nws.noaa.gov/glossary. Accessed November 25, 2008.

10. The Weather Channel. Types of Winter Precipitation. Avail- able at: http://www.weather.com/encyclopedia/winter/precip. html. Accessed November 25, 2008.

11. The Weather Channel. Types of Winter Storms. Available at: http://www.weather.com/encyclopedia/winter/types.html. Accessed November 25, 2008.

12. Wikipedia: The Free Encyclopedia. North American Ice Storm of 1998. Available at: http://en.wikipedia.org/wiki/North American ice storm of 1998. Accessed November 25, 2008.

13. Riddex L, Dellgar, U. The ice storm in eastern Canada 1998: KAMEDO-Report No 74. Prehosp Disaster Med. 2001;16:50–52.

14. Hamilton J. Quebec’s Ice Storm ’98: “all cards wild, all rules bro- ken” in Quebec’s shell-shocked hospitals. CMAJ. 1998;158:520– 524.

15. Nixdorf-Miller A, Hunsaker DM, Hunsaker JC. Hypother- mia and hyperthermia medicolegal investigation of morbid- ity and mortality from exposure to environmental temperature extremes. Arch Pathol Lab Med. 2006;130:1297–1304.

16. Dixon PG, Brommer DM, Hedquist BC, et al. Heat mortality versus cold mortality: a study of conflicting databases in the United States. Bull Am Meteorol Soc. 2005;86:937–943.

17. Kilbourne EM. Cold environments. In: Gregg MB, ed. The Public Health Consequences of Disasters. Atlanta: Centers for Disease Control; 1989:63–68.

18. Baker-Blocker A. Winter weather and cardiovascular mortality in Minneapolis–St Paul. Am J Public Health.1982;72:261–265.

19. Gorjanc ML, Flanders WD, VanDerslice J, Malilay J. Effects of temperature and snowfall on mortality in Pennsylvania. Am J Epidemiol. 1999;149:1152–1160.

20. Aylin P, Morris S, Wakefield J, et al. Temperature, housing, deprivation and their relationship to excess winter mortality in Great Britain, 1986–1996. Int J Epidemiol. 2001;30:1100–1108.

21. Sisitsky A. Winter storm. In: Ciottone GR, Anderson PD, Auf der Heide E, et al., eds. Disaster Medicine. Philadelphia: Elsevier-Mosby; 2006:499–501.

22. Lane AQ, Barnes SB. Winter storms. In: Hogan DE, Burstein JI, eds. Disaster Medicine.2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2007:236–247.

23. National Weather Service. Winter Storms: The Deceptive Killers. Washington DC: National Oceanic & Atmospheric Administra- tion; 2001.

24. Glass RI, O’Hare P, Conrad JL. Health consequences of the snow disaster in Massachusetts, February 6, 1978. Am J Public Health. 1979;69:1047–1049.

25. Faich G, Rose R. Blizzard morbidity and mortality: Rhode Island, 1978. Am J Public Health. 1979;69:1050–1052.

26. Geehr EC, Salluzzo R, Bosco S, et al. Emergency health impact of a severe storm. Am J Emerg Med. 1989;7:598–604.

27. Attia MW. The blizzard of 1996: a pediatric emergency depart- ment. Prehosp Emerg Care. 1998;2:285–288.

28. Broder JA, Mehrotra A, Tintinalli J. Injuries from the 2002 North Carolina ice storm, and strategies for prevention. Injury. 2005; 36:21–26.

29. Ráliš ZA. Epidemic of fractures during periods of snow and ice. BMJ. 1981;282:603–605.

30. Avery JG. Fractures during ice and snow. BMJ. 1982;284:270. (Letter)

31. Wood L. Fractures during ice and snow. BMJ. 1982;284:425. (Letter)

32. Manning DB, Jones C, Bruce M. Fractures during ice and snow. BMJ. 1982;284:508–509. (Letter)

33. Lewis LM, Lasater LC. Frequency, distribution, and management of injuries due to an ice storm in a large metropolitan area. South Med J. 1994;87:174–178.

34. Smith RW, Nelson DR. Fractures and other injuries from falls after an ice storm. Am J Emerg Med. 1998;16:52–55.

35. Hartling L, Pickett W, Brison RJ. The injury experience observed in two emergency departments in Kingston, Ontario during “Ice Storm ’98.” Can J Public Health. 1999;90:95–98.

36. Ernst A, Zibrak JD. Carbon monoxide poisoning. N Engl J Med. 1998;399:1603–1608.

37. Centers for Disease Control and Prevention. Carbon monoxide poisonings associated with snow-obstructed vehicle exhaust sys- tems – Philadelphia and New York City, January 1996. MMWR. 1996;45:1–3.

38. Centers for Disease Control and Prevention. Unintentional car- bon monoxide poisoning following a winter storm – Washing- ton, January 1993. MMWR. 1993;42:109–111.

39. Houck PM, Hampson NB. Epidemic carbon monoxide poison- ing following a winter storm. J Emerg Med. 1997;15:469–473.

40. Daley WR, Smith A, Paz-Argandona E, et al. An outbreak of carbon monoxide poisoning after a major ice storm in Maine. J Emerg Med. 2000;18:87–93.

41. Hartling L, Brison RJ, Pickett W. Cluster of unintentional carbon monoxide poisonings presenting to the emergency departments in Kingston, Ontario during “Ice Storm ’98.” Can J Public Health. 1998;89:388–390.

42. Ghim M, Severance HW. Ice storm-related carbon monox- ide poisonings in North Carolina: a reminder. South Med J. 2004;97:1060–1065.

43. Glass RI, Zack MM. Increase in deaths from ischaemic heart disease after blizzards. Lancet. 1979;1:485–487.

44. Spitalnic SJ, Jagminas L, Cox J. An association between snow- fall and ED presentation of cardiac arrest. Am J Emerg Med. 1996;14:572–573.

45. Glass RI, Wiesenthal AM, Zack MM, Preston M. Risk factors for myocardial infarction associated with the Chicago snow storm of Jan 13–15, 1979. JAMA. 1981;245:164–165.

46. Anderson TW, Rochard C. Cold snaps, snowfall and sud- den death from ischemic heart disease. CMAJ. 1979;121:1580– 1583.

47. Kunst AE, Groenhof F, Mackenbach JP. The association between two windchill indices and daily mortality variation in the Netherlands. Am J Public Health. 1994;84:1738–1742.

48. Franklin BA, Hogan P, Bonzheim K, et al. Cardiac demands of heavy snow shoveling. JAMA. 1995;273:880–882.

49. Persinger MA, Ballannce SE. Snow fall and heart attacks. J Psy- chol. 1993;127:243–252.

50. Blindauer KM, Rubin C, Morse DL, McGeehin M. The 1996 New York blizzard: impact on noninjury emergency visits. Am J Emerg Med. 1999;17:23–27.

51. Noller KL, Resseguie LJ, Voss V. The effect of changes in atmo- spheric pressure on the occurrence of spontaneous onset of labor in term pregnancies. Am J Obstet Gynecol. 1996;174:1192– 1199.

52. Danzl DF. Accidental hypothermia. In: Auerbach PS, ed. Wilder- ness Medicine. 5th ed. Philadelphia: Mosby-Elsevier; 2007: 125–160.

53. Giesbrecht GG, Steinman AM. Immersion in cold water. In: Auerbach PS, ed. Wilderness Medicine. 5th ed. Philadelphia: Mosby-Elsevier; 2007:160–188.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

606 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

54. Centers for Disease Control and Prevention. Hypothermia- related deaths – Utah, 2000, and United States 1979–1998. MMWR. 2002;51(4):76–78.

55. Centers for Disease Control and Prevention. Hypothermia- related deaths – United States 1999–2002 and 2005. MMWR. 2006;55:282–284.

56. American Medical Association. All-hazards course overview and DISASTER paradigm. In: Basic Disaster Life Support Provider Manual. version 2.5. Chicago, IL: American Medical Association; 2004:1–27.

57. American Medical Association. Application of the DISASTER paradigm and RED survey. In: Advanced Disaster Life Support Provider Manual. version 2.0. Chicago, IL: American Medical Association; 2004:1–19.

58. National Weather Service/National Oceanic and Atmospheric Administration. NOAA Weather Radio All Hazards. Available at: http://www.weather.gov/nwr/. Accessed November 25, 2008.

59. Zielinski GA. A classification scheme for winter storms in the eastern and central United States with an emphasis on nor’easters. Bull Am Meteorol Soc. 2002;83:37–51.

60. Sinks T. Hazards of working in cold weather include frostbite, hypothermia. Occup Health Safety. 1988;57:20–25.

61. Lipman GS, Castellani JW. Nonfreezing cold-induced injuries. In: Auerbach PS, ed. Wilderness Medicine. 5th ed. Philadelphia: Mosby-Elsevier; 2007:188–195.

62. McCauley RL, Killyon GW, Smith DJ, et al. Frostbite. In: Auer- bach PS, ed. Wilderness Medicine. 5th ed. Philadelphia: Mosby- Elsevier; 2007:195–210.

63. Madden JF, O’Conner RE, Evans J. The range of medication storage temperatures in aeromedical emergency medical ser- vices. Prehosp Emerg Care. 1999;3:27–30.

64. DuBois WC. Drug storage temperatures in rescue vehicles. J Emerg Med. 2000;18:345–348.

65. Helm M, Castner T, Lampl L. Environmental temperature stress on drugs in prehospital emergency medical service. Acta Anaes- thesiol Scand. 2003;47:425–429.

66. Neuffer MC, McDivitt J, Rose D, et al. Hemostatic dressings for the first responder: a review. Mil Med. 2004;169:716–720.

67. Pusateri AE, Holcomb JB, Kheirabadi BS, et al. Making sense of the preclinical literature on advanced hemostatic products. J Trauma. 2006;60(3):674–682.

68. Jurkovich GJ. Environmental cold-induced injury. Surg Clin North Am. 2007;87:247–267.

69. Simon TD, Soep JB, Hollister JR. Pernio in pediatrics. Pediatrics. 2005;116:472–475.

70. Rustin MH, Newton JA, Smith NP, et al. The treatment of chilblains with nifedipine: the results of a pilot study. Br J Der- matol. 1989;120:267–275.

71. White JC, Scoville WB. Trench foot and immersion foot. N Engl J Med. 1945;232:415–422.

72. Prehospital Trauma Life Support Committee, National Asso- ciation of Emergency Medical Technicians. Thermal trauma: injuries produced by heat and cold. In: McSwain NE, Salomone JP, Frame SB, eds. Basic and Advanced Prehospital Trauma Life Support. 4th ed. St Louis, MO: Mosby; 2007:246–263.

73. Biem J, Classen D, Koehncke N, Dosman J. Out of the cold: man- agement of hypothermia and frostbite. CMAJ. 2003;168:305– 311.

74. Urschel JD. Frostbite: predisposing factors and predictors of poor outcome. J Trauma. 1990;30:340–342.

75. Bracker MD. Environmental and thermal injury. Clin Sports Med. 1992;11:419–436.

76. Syme D, International Commission for Alpine Rescue Medical Commission. Position paper: on-site treatment of frostbite for mountaineers. High Alt Med Biol. 2002;3:297–298.

77. Giesbrecht GG. Prehospital treatment of hypothermia. Wilder- ness Environ Med. 2001;12:24–31.

78. American Heart Association. Part 10.4: hypothermia. Circula- tion. 2005;112(24 Suppl):IV-136–IV-138.

79. Ho JD, Heegaard WG, Brunette DD. Successful transcutaneous pacing in 2 severely hypothermic patients. Ann Emerg Med. 2007;49:678–681.

80. Lawless J. Aborted air medical missions: a 4-year quality review of a Canadian province-wide air medical program. Air Med J. 2005;24:79–82.

81. Stewart RL, Black GB. Snowmobile trauma: 10 years’ experience at Manitoba’s tertiary trauma centre. Can J Surg. 2004;47:90– 94.

82. Haynes CD, Webb WA, Fenno CR. Chain saw injuries: review of 330 cases. J Trauma. 1980;20:772–776.

83. Istre GR, Tinnell C, Ouimette D, et al. Surveillance for injuries: cluster of finger amputations from snowblowers. Public Health Rep. 1989;104:155–157.

84. Proano L, Partridge R. Descriptive epidemiology of a cluster of hand injuries from snowblowers. J Emerg Med. 2002;22:341–344.

85. Owens BD, Wenke JC. Early wound irrigation improves the ability to remove bacteria. J Bone Joint Surg Am. 2007;89:1723– 1726.

86. Angerås MH, Brandberg A, Falk A, Seeman T. Comparison between sterile saline and tap water for the cleaning of trau- matic wounds. Eur J Surg. 1992;158:347–350.

87. de Alwis W. Fingertip injuries. Emerg Med Aust. 2006;18: 229–237.

88. Broder KR, Cortese MM, Iskander JK, et al. Preventing tetanus, diphtheria, and pertussis among adolescents: use of tetanus tox- oid, reduced diphtheria toxoid and acellular pertussis vaccines recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR. 2006;55(RR-3):1–34.

89. Zehtabchi S. Evidence-based emergency medicine (critically appraised topic: the role of antibiotic prophylaxis for preven- tion of infection in patients with simple hand lacerations. Ann Emerg Med. 2007;49:682–689.

90. DeGroot DW, Castellani JW, Williams JO, Amoroso PJ. Epi- demiology of U.S. Army cold weather injuries,1980–1999. Aviat Space Environ Med. 2003;74:564–570.

91. Bruen KJ, Ballard JR, Morris SE, et al. Reduction of the incidence of amputation in frostbite injury with thrombolytic therapy. Arch Surg. 2007;142:546–51; discussion 551–553.

92. Barker JR, Haws MJ, Brown RE, et al. Magnetic resonance imag- ing of severe frostbite injuries. Ann Plast Surg. 1997;38:275–279.

93. Cauchy E, Chetaille E, Lefevre M, et al. The role of bone scanning in severe frostbite of the extremities: a retrospective study of 88 cases. Eur J Nucl Med. 2000;27:497–502.

94. Cauchy E, Marsigny B, Allamel G, et al. The value of Techne- tium 99 scintigraphy in the prognosis of amputation in severe frostbite injuries of the extremities: a retrospective study of 92 severe frostbite injuries. J Hand Surg Am. 2000;25:969–978.

95. Ulrich AS, Rathlev NK. Hypothermia and localized cold injuries. Emerg Med Clin North Am. 2004;22:281–298.

96. Delaney KA, Vassallo SU, Larkin GL, Goldfrank LR. Rewarming rates in urban patients with hypothermia: prediction of under- lying infection. Acad Emerg Med. 2006;13:913–921.

97. Luna GK, Maier RV, Pavlin EG, et al. Incidence and effect of hypothermia in seriously injured patients. J Trauma. 1987;27:1014–1018.

98. McInerney JJ. Accidental hypothermia and active rewarming: the metabolic and inflammatory changes observed above and below 32 degrees C. Emerg Med J. 2002;19:219–223.

99. Rohrer MJ, Natale AM. Effects of hypothermia on the coagula- tion cascade. Crit Care Med. 1992;20:1402–1405.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

WI N T E R STO R M S ■ 607

100. Giesbrecht GG. Emergency treatment of hypothermia. Emerg Med (Fremantle). 2001;13:9–16.

101. Walpoth BH, Walpoth-Aslan BN, Mattle HP, et al. Outcome of survivors of accidental deep hypothermia and circulatory arrest treated with extracorporeal blood warming. N Engl J Med. 1997;337:1500–1505.

102. Ruttmann E, Weissenbacher A, Ulmer H, et al. Prolonged extra- corporeal membrane oxygenation-assisted support provides improved survival in hypothermic patients with cardiocircu- latory arrest. J Thorac Cardiovasc Surg. 2007;134:594–600.

103. Wik L, Kiil S. Use of an automatic mechanical chest compression device (LUCAS) as a bridge to establishing cardiopulmonary bypass for a patient with hypothermic cardiac arrest. Resuscita- tion. 2005;66:391–394.

104. Hernandez E, Praga M, Alcazar JM, et al. Hemodialysis for treat- ment of accidental hypothermia. Nephron. 1993;63:214–216.

105. Kjaergaard B, Bach P. Warming of patients with accidental hypothermia using warm water pleural lavage. Resuscitation. 2006;68:203–207.

106. Plaisier BR. Thoracic lavage in accidental hypothermia with cardiac arrest – report of a case and review of the literature. Resuscitation. 2005;66:99–104.

107. Jessen K, Hagelsten JO. Peritoneal dialysis in the treatment of profound accidental hypothermia. Aviat Space Environ Med. 1978;49:426–429.

108. Tomaszewski C. Carbon monoxide. In: Ford MD, Delaney KA, Ling LJ, et al., eds. Clinical Toxicology. Philadelphia: Saunders; 2001:657–667.

109. Henry CR, Satran D, Lindgren B, et al. Myocardial injury and long-term mortality following moderate to severe carbon monoxide poisoning. JAMA. 2006;295:398–402.

110. Satran D, Henry CR, Adkinson C, et al. Cardiovascular man- ifestations of moderate to severe carbon monoxide poisoning. J Am Coll Cardiol. 2005;45:1513–1516.

111. Raub JA, Benignus VA. Carbon monoxide and the nervous sys- tem. Neurosci Biobehav Rev. 2002;26:925–940.

112. Nelson LS, Hoffman RS. Inhaled toxins. In: Marx JA, Hockberger RS, Walls RM, et al., eds. Rosen’s Emergency Medicine: Concepts and Clinical Practice. 6th ed. vol 3. Philadelphia: Mosby-Elsevier; 2006:2432–2441.

113. Lueken RJ, Heffner AC, Parks PD. Treatment of severe carbon monoxide poisoning using a portable hyperbaric oxygen cham- ber. Ann Emerg Med. 2006;48:319–322.

114. Weaver LK, Valentine KJ, Hopkins RO. Carbon monoxide poisoning: risk factors for cognitive sequelae and the role of hyperbaric oxygen. Am J Respir Crit Care Med. 2007;176:491– 497.

115. Weaver LK, Hopkins RO, Chan KJ, et al. Hyperbaric oxygen for acute carbon monoxide poisoning. N Engl J Med. 2002;347: 1057–1067.

116. Thom SR. Hyperbaric-oxygen therapy for acute carbon monox- ide poisoning. N Engl J Med. 2002;347:1105–1106.

117. Rucker J, Tesler J, Fedorko L, et al. Normocapnia improves cerebral oxygen delivery during conventional oxygen therapy in carbon monoxide-exposed research subjects. Ann Emerg Med. 2002;40:611–618.

118. Takeuchi A, Vesely A, Rucker J, et al. A simple “new” method to accelerate clearance of carbon monoxide. Am J Respir Crit Care Med. 2000;161:1816–1819.

119. Schultz CH, Koenig KL, Noji EK. Disaster preparedness. In: Marx JA, Hockberger RS, Walls RM, et al., eds. Rosen’s Emer- gency Medicine: Concepts and Clinical Practice. 6th ed. vol 3. Philadelphia: Mosby-Elsevier; 2006:3010–3021.

120. Dealing with power failure: how Spokane hospitals survived the ice storm. Hosp Secur Safe Manage. 1997;17:3–4.

121. The ice storm of the century: how affected hospitals and com- munities dealt with the challenges of a unique, prolonged emer- gency. Hosp Secur Safe Manage. 1998;18:5–9.

122. The Oklahoma ice storm: a Y2K disaster that arrived one year later – how two rural hospitals coped and what they learned. Hosp Secur Safe Manage. 2001;22:5–8.

123. Hunter DG, MacDonald D, Peever L. Ice storm: a crisis man- agement diary. Hosp Q. 1998;1:69–73.

124. Department of Defense. A Mass Casualty Care Strategy for Bio- logical Terrorism Incidents: Neighborhood Emergency Help Cen- ter. Aberdeen Proving Ground, MD: Homeland Defense Office, United States Army Soldier and Biological Chemical Command; May 2001.

125. Department of Defense. A Mass Casualty Care Strategy for Bio- logical Terrorism Incidents: Acute Care Center. Aberdeen Proving Ground, MD: Homeland Defense Office, United States Army Soldier and Biological Chemical Command; December 2001.

126. Malilay J. Public health assessments in disaster settings: recom- mendations for a multidisciplinary approach. Prehosp Disaster Med. 2000;15:167–172.

127. Malilay J, Flanders WD, Brogan D. A modified cluster-sampling method for post-disaster rapid assessment of needs. Bull World Health Organ. 1996;74:399–405.

128. Centers for Disease Control and Prevention. Rapid Com- munity Needs Assessment Using Modified Cluster Sampling Methods. Available at: http://www.bt.cdc.gov/masscasualties/ research/community.asp. Accessed December 3, 2008.

129. Hlady WG, Quenemoen LE, Armenia-Cope RR, et al. Use of a modified cluster sampling method to perform rapid needs assessment after Hurricane Andrew. Ann Emerg Med. 1994; 23:719–725.

130. Centers for Disease Control and Prevention. Community needs assessment and morbidity surveillance following an ice storm – Maine, January 1998. MMWR. 1998;47(17):351–354.

131. Federal Emergency Management Agency. 9523.1 Snow Assis- tance Policy. Available at: http://www.fema.gov/government/ grant/pa/9523 1.shtm. Accessed December 3, 2008.

132. Centers for Disease Control and Prevention. Public Health Information Network. Messaging Services. Available at: http://www.cdc.gov/phin/library/documents/pdf/111759 PHINmessaging.pdf. Accessed December 3, 2008.

133. Centers for Disease Control and Prevention. IT Functions and Specifications (also known as the Public Health Informa- tion Network Functions and Specifications) Version 1.2. Avail- able at: http://www.bt.cdc.gov/planning/continuationguidance/ docs/appendix-4.doc. Accessed December 3, 2008.

134. National Disaster Education Coalition. Winter Storm. Avail- able at: http://www.redcross.org/images/pdfs/code/winter storms.pdf . Accessed December 3, 2008.

135. Collins KJ, Dore C, Exton-Smith AN, et al. Accidental hypother- mia and impaired temperature homoeostasis in the elderly. BMJ. 1977;1(6057):353–356.

136. Department of Health. Keep Warm – Keep Well. Available at: http://www.dh.gov.uk/keepwarmkeepwell. Accessed Decem- ber 3, 2008.

137. Wrenn K, Conners GP. Carbon monoxide poisoning during ice storms: a tale of two cities. J Emerg Med. 1997;15:465–467.

138. Sternbach G. Winter storms and great imitators. J Emerg Med. 1997;15:531–532. (Editorial)

139. Lin G, Conners GP. Does public education reduce ice storm- related carbon monoxide exposure?J Emerg Med. 2005;29:417– 420.

140. Krenzelok EP, Roth R, Full R. Carbon monoxide . . . the silent killer with an audible solution. Am J Emerg Med. 1996;14:484– 486.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

608 ■ JO H N M. WI G H T M A N, JA M E S A. FE N N O, A N D WI L L I A M H. DI C E

141. Yoon SS, Macdonald SC, Parrish RG. Deaths from unintentional carbon monoxide poisoning and potential for prevention with carbon monoxide detectors. JAMA. 1998;279:685–687.

142. Centers for Disease Control and Prevention. Use of carbon monoxide alarms to prevent poisonings during a power outage – North Carolina, December 2002. MMWR. 2004;53:189–192.

143. National Oceanic and Atmospheric Administration. Winter Storms . . . the Deceptive Killers: A Guide to Survival. Avail- able at: http://www.weather.gov/om/brochures/wntrstm.htm. Accessed December 3, 2008.

144. Centers for Disease Control and Prevention. Extreme Cold: A Prevention Guide to Promote Your Personal Health and Safety – Prepare Your Car for Winter. Available at: http://www.bt.cdc .gov/disasters/winter/guide.asp#car. Accessed December 3, 2008.

145. Centers for Disease Control and Prevention. Public health impact of a snow disaster. MMWR. 1982;31:695–696.

146. Lalikos JF, Hayden DB, Rothkopf DM. Untitled letter. J Trauma. 1997;42:348.

147. American Geriatrics Society, British Geriatrics Society, and American Academy of Orthopaedic Surgeons Panel on Falls Prevention. Guideline for the Prevention of Falls in Older Per- sons. J Am Geriatr Soc. 2001;49:664–672.

148. Kannus P, Khan KM. Prevention of falls and subsequent injuries in elderly people: a long way to go in both research and practice. CMAJ. 2001;165:587–588.

149. Betz M, Li G. Injury prevention and control. Emerg Med Clin North Am. 2007;25:901–914.

150. National Highway Traffic Safety Administration. National EMS Research Agenda. Washington, DC: Department of Transporta- tion and Department of Health and Human Services; 2001.

151. Quick G. A paradigm for multidisciplinary disaster research: the Oklahoma City experience. J Emerg Med. 1998;16:621– 630.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Extreme Heat Events

Carl Adrianopoli, Paul H. Brietzke, Irving Jacoby,

and Jerome H. Libby

INTRODUCTION

This chapter addresses the medical and public health implica- tions of extreme heat events (EHEs) and the associated mortality and morbidity. EHE conditions can be defined by summertime weather that is substantially hotter and/or more humid than average for a location during a comparable time period. History is filled with the failures of great civilizations caused by signif- icant climate changes reacting with human adaptations. Exam- ples include the collapse of the north African “Bread basket” for ancient Rome, the wind-swept droughts in Oklahoma’s “Dust bowl” during the 1930s, the vast European droughts in the Mid- dle Ages, and the severe 1921 drought in extensive areas of the former Soviet Union that resulted in millions of deaths.2 Even the genocide in the Darfur region of west Sudan has a weather- related component. The ongoing drought has pit herders against farmers, with the added elements of race and religion exacerbat- ing the situation.

There have been more than 20 serious EHEs across the world since 1901, including the deadly 2003 EHE in Europe that killed more than 35,000 people, with 15,000 dead in France alone.3,3a

In the United States, up to 800 died of EHEs in Chicago and Milwaukee in 1995. Additional thousands have died in Philadel- phia, St. Louis, Kansas City, and other major U.S. cities since the early 1990s. EHEs in U.S. mid-Atlantic and midwest cities can be accompanied by glaring sun with no cloud cover, tempera- tures in the 35◦C–40◦C range, and heat indexes (temperature and humidity) from 43◦C to 51◦C or more. This results in crowded hospital Emergency Departments (EDs) on diversion and media stories of the elderly found dead in tightly shut, overheated urban apartments.

EHEs are not determined by the absolute temperature alone, but are dependent on other conditions specific to each location. An EHE in sub-Saharan Africa in June, for example, will be much hotter in absolute temperature than an EHE in Minneapolis-St. Paul in a comparable period. In addition to overall temperatures, other environmental factors such as humidity, air circulation, building types, and nighttime temperatures can intensify the health effects of EHEs.

The first section of this chapter, “An Overview of Extreme Heat Events,” provides a background to understanding EHEs as disasters. The second section of this chapter, “The State of the Art: Extreme Heat Event Risk Factors and Medical Response,” is divided into five subheadings: 1) “Health Risk Factors from Extreme Heat Events,” 2) “Urban Heat Islands as Risk Factors,” 3) “Planning for Extreme Heat Events,” 4) “The Chicago Heat Wave of 1995,” and 5) “Extreme Heat Events in the Developing World.” The third section of this chapter, “Recommendations to Prevent or Mitigate the Health Effects of Extreme Heat Events,” includes recommendations for both policy change and future areas of research.

OVER V IEW OF EXTREME HEAT EVENTS

Health personnel are concerned with EHEs because rising tem- peratures result in rising heat-related mortality and morbidity. Global warming and climate change will only intensify this prob- lem. Thermal trends between 1880 and 2005 show a clear upward movement in both land and ocean temperatures (Figure 38.1). Unless scientists determine a way to alter the weather, the best approach to protection of people around the globe will remain for the nations of the world to develop plans for improving weather prediction and for strengthening public health and medical sys- tem responses to EHEs. Most serious health effects of EHEs can be effectively prevented or addressed by providing timely warning to vulnerable populations, access to air conditioning, adequate potable water (in the developing world), and shelter from the sun. As will be discussed, these simple solutions can be hindered by complex physiological, environmental, medical system, and even political and sociocultural restrictions.

Increasing global warming, urbanization, and population numbers require improvement in effective EHE planning and response activities. This is true not only across the developed world, but more significantly in the sprawling cites and barren rural areas of the developing world. The fact that there are more than 1 billion people living without access to potable water com- plicates other EHE-related problems.4 There is little likelihood

609 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

610 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

Jan. - Dec. Global Surface Mean Temp. Anomalies

Figure 38.1. Annual average global surface temperature anomalies 1880–2005. Source: Environmental Protection Agency, http://www .epa.gov/climatechange/science/recenttc.html. Modified by Tom Javorcic, 2007.

of technological and engineering solutions to global warming in the near future. The political will to address global warming by strong and effective restrictions and planning programs (e.g., fuel-efficient automobiles, less burning of high-carbon fuels such as coal and wood, carbon taxes/exchange programs, and strong landuse regulations) has been variable. Until better miti- gation strategies are in place to prevent EHEs, local and national

governments need to address growing heat-related morbidity and mortality with effective planning and responses.

In most years, more people in the United States die of EHEs than of hurricanes, lightning, tornadoes, floods and earthquakes combined. During the period from 1979 to 1999, there were 8,015 heat-related deaths.5 Despite this, EHEs have been among the most underrated of all deadly weather phenomena.6 Figure 38.2 compares fatalities from EHEs documented by death certificates to fatalities from other weather-related disasters over a 25-year period. EHE deaths often exceed those from all other weather- related sources.7

Mortality documented by death certificates will often be dwarfed by the number of actual heat-related deaths occurring during an EHE. Estimating the number of individuals who have died of temperature-related causes during an EHE is difficult. A traditional method has been to estimate “excess mortality,” defined as the difference between the number of deaths observed and the number expected, based on the crude death rates for the same geographical area, during the same period when no heat wave or other unusual circumstances were present.8 Substan- tial inconsistencies often exist between the “excess deaths” that are calculated for the period of an EHE and the exact number of deaths that have been certified as heat-related by a medical examiner or a coroner. For example, the average annual rate of heat-related deaths increased during EHEs in each age group, except for children aged 14 years and younger. This was particu- larly true for persons 55 years and older. Because other causes of death (e.g., cardiovascular and respiratory diseases) also increase during heat waves, heat-related deaths due to weather conditions represent only a portion of heat-related mortality.10–12

In 1980, when U.S. summer temperatures reached all-time high levels, there were 5,000 deaths above the expected number, with more than the 1,700 cases documented as having been caused by heat.8 In contrast to violent weather events, an EHE is a “silent killer” that is dramatically less apparent than other hazards, especially at the outset.7 There are credible estimates as high as 160,000 deaths annually across the world from EHEs and

Figure 38.2. Weather Fatalities. Source: NOAA, summer 2006.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 611

other weather disasters, with most of these deaths occurring in developing nations.12,13

Even advanced nations are not immune to what might oth- erwise appear to be a problem of teeming urban areas in poorer countries that lack adequate supplies of potable water, decent shelter for their populations, and a clear recognition of the dan- gers posed by EHEs. Kalkstein, using data from his study of 44 large American cities, estimates that 1,840 excess deaths occur annually due to the presence of high-risk air masses during a “present-day typical summer.”14 This estimate is consistent with studies demonstrating that only a portion of the increase in mortality during EHEs is documented on death certificates.8,9,15

Previous studies have estimated the combined EHE-attributable summertime mortality (excess deaths) for several vulnerable U.S. metropolitan areas is well above 1,000 deaths per year.1,16,17

The U.S. Centers for Disease Control and Prevention (CDC) has found that the diagnoses of heat-related deaths have been underestimated by 22%–100%.8,10,18 Although similar research on EHE-attributable mortality in rural areas has not been com- pleted, some studies found evidence of such an impact.19

Public awareness of potentially deadly EHEs has generally lagged behind the reality. In Europe, for example, despite cat- aclysmic, heat-related death tolls in recent years, the Euro- peans have had a difficult time in changing their basically benign, “friendly to people” view of the summer’s heat.20 Global warming/climate change data may change these perceptions.

Global warming/climate change is likely to result in pro- gressively more serious and frequent EHEs across the developed and developing world.21 Urban populations in nonindustrial- ized countries continue to be particularly vulnerable to the direct effects of climate change.22 The world political community has generally accepted the human involvement paradigm that the burning of carbon-based fossil fuels to a great degree causes the verified fact of continued global warming.23 At the 2005 United Nations (UN) Summit on Global Warming, Janez Dromvsek called for integrated worldwide planning, a search for solutions, and the raising of politicians’ and the public’s consciousness.24

Worldwide efforts to mitigate the effects of EHEs in developing nations generally have been neither extensive nor successful.

THE STATE OF THE AR T: EXTREME HEAT EVENT RISK FACTORS AND MEDICAL RESPONSE

Health Risk Factors from Extreme Heat Events

Physics, Physiological, and Meteorological Effects of Heat Exposure

Increasing heat and humidity affect the body’s ability to maintain its homeostatic balance, but similar heat indexes (tem- peratures and humidity) will affect individuals differently based on personal, geographical, sheltering, and other aspects of the miniclimates in which they live. The temperature of the air, its humidity and motion, and the amount of radiant heat energy to which an individual is exposed are the most important fac- tors in human heat stress. Of these, air temperature can have the greatest impact.25 Although there may be intense temperature fluctuations on the outer surfaces and extremities of the human body, thermal homeostatic mechanisms attempt to maintain a relatively stable core temperature. There are four aspects to this homeostatic process: 1) metabolic heat gain, 2) heat loss from perspiration/evaporation, 3) conductive and invective heat loss or gain, and 4) the effects of radiant energy.8 When air tem-

perature is low, heat generated metabolically is more easily lost from the body to the air. As air temperature increases, convec- tive heat loss is no longer possible, and heat can be gained from the air. High humidity limits the cooling effects of perspiration evaporation.8

The interpretation of any heat index value will be affected by differences in an individual’s age, medications, clothing, and body habitus. In addition, these numbers will fluctuate signifi- cantly when compared with other values obtained if one could measure the various microclimates to which individuals are exposed.8 For example, those older than age 52 tend to produce significantly less perspiration than those who are younger.26 Dif- ferences in hydration patterns can complicate the application of general heat indices to individuals or groups. Elderly populations with little shelter from the sun’s direct rays, or those shuttered tightly within steaming, unventilated brick buildings in inner- cities will experience drastically different reactions to the heat than middle class, middle-aged suburbanites who can escape to air-conditioned homes.

Increasing temperatures, humidity, and direct exposure to the sun can increase the heat stress that individuals will experi- ence during EHEs. Heat index tables often assume that temper- atures are taken in a shaded area, with little wind. In addition, most heat index tables note that direct sunlight can increase heat index figures by up to approximately 8◦C and that exposure to dry winds can further increase health risks by promoting rapid dehydration (Figure 38.3).27 Ultimately, any meteorological con- ditions that increase heat indexes will increase heat stress and health risk. All else being equal, the shock effect of the increased heat is greater the earlier in the summer the EHE occurs.28,28a In a similar fashion, health risks increase with the duration of the EHE, the amount of time spent above minimum temperature thresholds, and the rapidity of the rise in the heat index.1,29–33

Residents become increasingly acclimated to the heat as the sea- son progresses. It is not absolute temperature, but rather the extent of upward deviation from usual local summer tempera- tures that seems to be the key variable affecting mortality.32 As Table 38.1 demonstrates, there is striking similarity across the continents (with the exception of Antarctica) in the array of the highest ever recorded temperatures.

Diagnosis and Treatment of Heat Stress Illnesses There are several clinical syndromes that comprise heat stress

illnesses or conditions: 1) heat cramps, 2) heat edema, 3) heat syncope, 4) heat exhaustion, and 5) heat stroke. They represent variations and overlap in a continuum of heat illness, from minor complaints to overwhelming heat stress that can lead to death.33

Heat cramps present as pain and spasm in heavily exercised muscles. Their presentation can vary, from the parade marcher who complains of abdominal pain, to the athlete with cramping in the calves. The mechanism is thought to result from an imbal- ance in water and sodium intake, leading to hyponatremia, either measurably or locally at the cellular level. Clinically the body temperature is normal, with little evidence of frank dehydration. Measurement of electrolytes may reveal hypokalemia, hypona- tremia, respiratory alkalosis, hypomagnesemia, and hypophos- phatemia. On the scale of severity, this is usually a nonthreatening condition, treatable with rest, removal from the heat source, and oral or parenteral fluid and electrolyte replacement. Pitfalls that may occur include the attribution of abdominal cramping to heat cramps, when in fact the patient has an alternate diagnosis, such as infectious peritonitis or internal bleeding from a ruptured

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

612 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

Figure 38.3. Source: National Oceanic & Atmospheric Administration, http://www.srh .noaa.gov/ssd/html/heatwv.htm. Modified by Tom Javorcic, 2007.

viscus or hemorrhagic ovarian cyst. Additionally, rhabdomyoly- sis can occur in the setting of severe exertion or repetitive mus- cular contraction, leading to myoglobinuric renal failure and life-threatening hyperkalemia with its attendant effects on car- diac conduction. Thus, a reasonable evaluation in the setting of significant muscle pain will involve laboratory studies that can detect these conditions.

Heat edema is a mild condition resulting in swelling of the hands or feet, related to prolonged peripheral vasodilation fol- lowed by orthostatic pooling of blood in the extremities. It is usually responsive to elevation of the legs. It must be differenti- ated from renal failure or congestive heart failure in susceptible populations.

Heat syncope is characterized by sudden loss of conscious- ness and is also related to peripheral venous blood pooling with

subsequent orthostatic hypotension. It occurs with prolonged standing, or rising quickly from a sitting position. Such persons should not be “held up” or supported, but rather gradually low- ered to the ground. First aid treatment consists of laying the patient on the ground and lifting the legs up slightly to restore blood flow to the head. Maintaining the patient in an upright position may prolong the period of poor cerebral perfusion. Rehydration should ensue, and a cardiac rhythm strip should be obtained to ensure no heart blocks or other cardiac conduction abnormalities are occurring.

Heat exhaustion occurs in the setting of excess diaphore- sis in a hot, humid environment, leading to volume depletion. Core body temperature will be elevated above normal, but will remain less than 40.5◦C, which usually defines the level for heat stroke. Symptoms are profuse sweating, malaise, fatigue,

Table 38.1: Highest Temperature Extremes

Highest) Elevation Locator # Continent Temp (◦C) Place (m) Date

1 Africa 58 El Azizia, Libya 112 13 Sep 1922

2 N. America 57 Death Valley, CA (Greenland Ranch) –54 10 Jul 1913

3 Asia 54 Tirat Tsvi, Israel –220 22 Jun 1942

4 Australia 53∗ Cloncurry, Queensland 190 16 Jan 1889

5 Europe 50 Seville, Spain 8 4 Aug 1881

6 S. America 49 Rivadavia, Argentina 206 11 Dec 1905

7 Oceania 42 Tuguegarao, Philippines 22 29 Apr 1912

8 Antarctica 15 Vanda Station, Scott Coast 15 5 Jan 1974

∗ Note: This temperature was measured using the techniques available at the time of recording, which are different from the standard techniques currently used in Australia. The most likely Australian high-temperature record using standard equipment is an observation of 50.7◦C recorded at Oodnadatta in 1960.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 613

headache, dizziness, nausea, and vomiting. If left untreated, the condition will likely progress to classic heat stroke. Tachycardia and hypotension may be present, but major neurological dys- function does not occur. Treatment consists of cooling, oral rehy- dration, or intravenous rehydration in someone who is hypoten- sive or fails to respond to oral fluid replacement within a few hours.

Heat stroke is the most life-threatening condition related to heat stress. It is defined as an elevated core body temperature usu- ally equal to or greater than 40.5◦C in association with significant acute mental status or behavioral changes. Mental status changes can consist of confusion, bizarre behavior, hallucinations, delir- ium, unresponsiveness, seizures, posturing, or coma. The obser- vation that sweating ceases is found in only 50% of cases, and those patients with exertion-related hyperthermia are more likely to be sweating. Loss of sweating as a mechanism for body cooling is a late finding. This true medical emergency occurs when heat production exceeds physiological cooling capacity such that heat dissipation is no longer occurring. Hyperthermia is character- istically differentiated from fever, in that fever occurs due to an upward adjustment of the temperature set point in the hypotha- lamus. In hyperthermia, the hypothalamus set point is normal, but the body is unable to eliminate acquired heat, leading to excessive body temperature.

With the onset of heat stroke, widespread cellular and organ damage ensues, characterized initially by tachycardia, increased cardiac index, and central venous dilation. Critical deterioration continues, progressing to hypotension and cardiovascular col- lapse. Manifestations include coagulopathy in association with hepatic and cardiac failure. Hemostatic disturbances are marked by drops in platelet count and fibrinogen and consumption of clotting factors.

Classic heat stroke is described during EHEs, particularly in the elderly. Exertional heat stroke may also occur in young, fit populations such as athletes and military recruits undergo- ing training. Several high-profile deaths in professional athletes were reported by the media in recent years, demonstrating the importance of prevention when weather conditions increase the risk of such heat stress. In that setting, marked rhabdomyol- ysis and myoglobinuric renal failure are often observed along with acute hepatic failure and disseminated intravascular coagu- lation. Mortality ranges from 10% to 70% in series of heatstroke patients, with higher mortality rates found when treatment is delayed more than 2 hours.34 Predictors of multiorgan dysfunc- tion include respiratory failure, metabolic acidosis, elevated cre- atinine phosphokinase, and liver function elevations greater than twice normal.35

Treatment requires rapid cooling to avoid further cellular and organ damage by extreme hyperpyrexia. In the field or upon arrival in the ED, ice or cold liquids should be placed in contact with the patient especially in the axillae and groin, and additional cooling measures instituted as available, while airway, breathing, and circulation are assessed and managed. Airway management may include endotracheal intubation for control of the airway. If severe hyperpyrexia is suspected, use of a rapid sequence par- alytic agent that does not induce hyperkalemia is indicated. Due to altered mental status, an immediate glucose determination is needed to allow rapid diagnosis and treatment of hypoglycemia. Breathing assessment should include oxygen saturation monitor- ing. Circulatory support includes cardiac monitoring and fluid resuscitation when indicated to support blood pressure, perfu- sion, and urine output. Caution is necessary when dealing with

the geriatric population that may have antecedent cardiac, pul- monary, and/or renal disease. In addition to ice packing, when fans are available, cooling can be accomplished by spraying the undressed patient with tepid (not cold) water while blowing air from large fans across the body surface to maximize evapora- tive heat loss. More invasive methods such as ice water lavage of the stomach through a nasogastric tube or the peritoneum through a peritoneal lavage catheter, and even cardiopulmonary bypass have had reported anecdotal success. These aggressive treatments have limited evidence of benefit, however, and may actually be harmful and are therefore not recommended by some experts.36,37 A newer modality of using intravenous catheters containing cooling coils offers an additional treatment option. Frequent monitoring of the core body temperature is essential to avoid overshooting and creation of further problems due to hypothermia. It is not necessary to reduce body temperature to normal, but only to the level thought not to produce cellular injury; a goal of 38.9◦C is reasonable. Risks and benefits of each of these techniques have been reviewed.38

In addition, the number of fatal cocaine overdoses appears to correlate with higher ambient temperatures. In a New York City study, the mean daily number of cocaine overdose deaths increased by 33% on days with a maximum temperature of 31.1◦C or higher, in comparison to days when the mean tem- perature was below this point.39

The Emergency Department and Extreme Heat Events In nations with well-developed emergency services, the ED

plays a significant role during EHEs. EDs are the major point of entry for most victims suffering heat-related illnesses and there- fore have a large role in preventing escalating morbidity and mor- tality. EDs should certainly be prepared to manage surges of EHE patients with adequate cooling equipment. Thus, ED managers must ensure sufficient supplies of water spray bottles, cooling packs, fans, cooling catheters, and ice during EHEs. Informa- tional handouts can be printed as part of discharge instructions in advance of such emergencies as well as during the events. This should become part of the general community educational pro- gram, much as is done for other types of safety issues. Increased visits to the EDs in affected areas should be expected.

Using a wide variety of measures to predict EHEs is a com- plex process that can involve the National Weather Service, local health departments and emergency management agencies, first response emergency medical services agencies, hospitals, medical examiners/coroners’ offices, and many other local agencies and community organizations. No matter how effective the ED may be as a heat mortality sentinel, it cannot provide sufficient warn- ing to reduce the impact of these emergencies. Systems designed to predict EHEs in a timely manner should augment and in no way substitute for the longer-range prediction times that air mass monitoring systems allow. The EHE has already started by the time patients begin presenting to EDs.

EHEs tend to generate so many patients so quickly that, in systems where it is permitted, many EDs implement diver- sion status. As three recent Institute of Medicine (IOM) reports delineate, the U.S. emergency medical care system “is woefully inadequate and unprepared for a pandemic, bioterrorist attack, natural disaster or other national crisis.”40 The IOM found the U.S. emergency care system to be underfunded, too fragmented to communicate and cooperate effectively across levels and geo- graphical areas, and possessing little surge capacity to manage a disaster. The IOM also found that emergency care staff members

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

614 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

are often inadequately trained to respond to large-scale disasters or to care for pediatric patients.

During Chicago’s July 1995 EHE, there were 1,072 more hos- pital admissions than average for comparative weeks, with 838 (35%) more patients aged 65 years and older being admitted than expected. There was also strong anecdotal evidence of increased ED visits. An analysis of excess hospital admissions during the heat wave defines who was admitted and why. The primary rea- sons for a hospital visit were dehydration, heat stroke, or heat exhaustion. The susceptible population at risk for admission had comorbid cardiovascular illnesses, endocrine disorders, liver and kidney diseases, or nervous system disorders. Within this popu- lation, the elderly were disproportionately represented, in large part due to their altered thirst perception and related condi- tions.41 On the second day of the EHE, only a few Chicago EDs were on diversion and directing ambulances to other hospitals. By the fourth day, however, 18 city EDs were diverting patients to other facilities.7 A study of the 1993 heat wave in Philadelphia found a 26% increase in total mortality and a 98% increase in cardiovascular mortality associated with the EHE. In adjacent counties, the risk for dying of cardiovascular disease rose sig- nificantly for people older than 65 years, for both sexes and all races.42

During the European Heat Wave of 2003, heat-related deaths in a Parisian hospital occurred mostly in elderly patients (mean age 84), and 69% were women. Patients who died differed from those who survived.43 The former were characterized by greater levels of dependency and by a more abnormal initial clinical presentation (such as elevated temperature, lower blood pres- sure, and altered mental states). They were also more likely to have existing ischemic cardiomyopathies and to be taking psy- chotropic medications.43 In London, during the same period, 2,091 deaths occurred (17% more than for the same period in ear- lier years). Twenty-three percent of the deaths were among those 75 years of age or older.44 Similar findings come from Australia, where high environmental temperatures are common, although it is rare that these exceptional conditions produce elevated levels of heat-related morbidity and mortality. In four major teaching hospitals in Adelaide, most patients presenting with heat-related conditions (85%) were 60 years or older, with 20% coming from institutional care, and 30% with poor mobility. Peak presenta- tion followed high daily temperatures for four consecutive days. Severity was related to existing cognitive impairment, diuretic use, presenting temperature, heart rate, blood pressure, plasma sodium, and plasma creatinine. The mortality rate was 12%. Seventeen percent required a more dependent level of residential care upon discharge.45 Similar findings came from a 1999 study in Wisconsin where heat-related illnesses led to 21 deaths. Death rates were highest among the elderly; particularly those aged 65– 84 years (2.2/100,000). Heat was the underlying cause of death for 12 of the 21 victims. Cardiovascular conditions resulted in another eight deaths and were a contributing cause for an addi- tional seven.46

Hospital/Emergency Department Surge Capacity and Extreme Heat Events

For years, medical disaster planners have considered “surge capacity” in their catastrophic planning (see Chapter 3). Although an EHE would severely test the in-house surge capacity of existing hospitals and EDs, it is unlikely that most EHEs would last long enough and generate enough patients to necessitate the use of external surge facilities. A quick review of surge capacity,

however, is helpful to the extent that the same forces that will expand external surge capacity can be used to support and extend services internally.

U.S. hurricanes in 2004 and 2005 provide an example of the devastating health and medical effects of weather disasters. Hur- ricanes Katrina and Rita and subsequent flooding caused the same types of damage to many local care facilities as they did to other types of building. Many hospitals and federal and state medical support agencies were forced to establish operations in temporary locations such as shuttered retail stores and veteri- nary hospitals.47 Freestanding or support/augmentation facilities were also constructed at airports, sports complexes, and adja- cent to existing institutions. These “surge” hospitals addressed the increase in demand for medical care and contained triage, treatment, and sometimes surgical capacities.47 Advice, guid- ance, and funding to state health departments and hospital sys- tems for surge planning is available from the U. S. Department of Health and Humans Services’ Hospital Preparedness Program, as well as from the integrated, all-hazard medical system planning within the Federal Emergency Management Agency’s Metropoli- tan Medical Response System program.

Using the U.S. emergency management organization as an example, federal disaster support to hospitals is only a small part of the national emergency management system. This system is a complex network of public, private, and nonprofit organizations (ranging from the American Red Cross to professional organiza- tions such as the American Hospital Association, local hospital councils, and even local community groups in the vicinity of hos- pitals) and individual benefactors. It also includes federal, state, and local government agencies, special districts and quasigovern- mental bodies, nonprofit service and charitable organizations, ad hoc volunteer groups and individuals, and private sector firms that provide governmental services by contract.48 Dealing with this huge array of emergency preparedness agencies and entities in a coordinated and effective manner requires hospital prepared- ness staff whose perspectives are broader than just the individual hospital’s concerns. They must also be knowledgeable about the complex and rapidly growing scientific evidence base related to disasters.

When U.S. hospitals implement surge planning, they enter an integrated, but complicated web that includes Joint Commission requirements for: 1) establishing hospital incident command systems as the chain of command for disasters; 2) all-hazard emergency response plans; 3) mutual aid agreements and pro- cesses with other hospitals, systems, and local, state, and federal agencies; 4) coordination with local emergency management agencies; and 5) the requirement to maintain comprehensive documentation on decision making, victim destinations, patient tracking, and reimbursement.49 Existing rules and waivers under the federal Medicare program can also have significant effects on a hospital’s ability to address surge conditions (Table 38.2). Other recommendations that address surge planning appear in the pub- lication Medical Surge Capacity and Capability: A Management System for Integrating Medical and Health Resources during Large- Scale Emergencies.50 Those that have the most relevance to EHEs include

■ Redundancy. Developing redundancy in hospital operations systems to ensure backup capability during an emergency. Backup systems should be evaluated for their vulnerability to hazards, particularly those most likely to affect primary systems

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 615

■ Testing of backup and support systems. Establishing pro- grams for testing, inspection, and preventive maintenance of backup systems and facility safety features

The National Foundation for Trauma Care has made a num- ber of key recommendations for improving the capacity of trauma centers to provide care to victims of a terrorist event.51

Those relevant to EHE surge planning include

■ Fund disaster medical care at cost and develop sustainable funding because existing federal programs (predisaster and postdisaster) do not provide enough fiscal support

■ Sustainment (e.g., staff and supplies) for more than 3 days is required

■ Fund statewide (and multistate) resource monitoring sys- tems

■ Provide adequate funding to train staff, based on the prox- imity and the threat of the hazard

■ Provide aftercare for the chronically ill and displaced persons ■ Mutual aid agreements and memoranda of understanding

must be developed

Auf der Heide published response strategies for hospitals during disasters. Those that are directly applicable to EHEs are

■ Establish EMS/hospital radio networks to rapidly collect hos- pital status information and direct the flow of those casualties who are transported by ambulance. Since a truly interoper- able system that is effective, affordable, and easy to use does not currently exist, overall communications redundancy is desirable

■ Ensure that hospitals/EMS radio systems are established to facilitate early warning to hospitals from responders in the field. A number of sophisticated systems exist that report hospital bed status availability, hospitals on diversion, and other key information

The Joint Commission found that legal and reimbursement issues are among the most critical, nonpatient care–related prob- lems that hospitals face when developing surge capacity.47 Dur- ing surge conditions, state and federal waivers can protect emer- gency medical workers. During Hurricane Katrina, for example, the Governor of Louisiana waived state licensure restrictions for those licensed out of state. The Department of Health and Human Services afforded liability protection to healthcare work- ers who volunteered, and waived the Emergency Medical Treat- ment and Active Labor Act. A number of recent U.S. laws and agreements (the Emergency System for Advanced Registration of

Table 38.2: External Support for Hospital Emergency Departments before/during EHEs

■ Longer shifts, intra- and interhospital agreements ■ Mutual aid (provided by state/local agencies, groups, or through

FEMA’s Emergency Medical Assistance Compact – EMAC) ■ State/Local Departments of Public Health Assets, e.g., The Illinois

Mobile Emergency Response Team (IMERT), or the Special Opera- tions Response Team (SORT) in North Carolina

■ The National Disaster Medical System (described earlier) ■ The Medical Reserve Corps (situated in the Office of the U.S. Surgeon

General)

Volunteer Health Professionals) as well as a proposed hospital- based credentialing system have made it potentially easier to use volunteers who respond to a disaster.

Other U.S. federal and state programs that can provide healthcare staff resources during surge conditions include the National Disaster Medical System, the U.S. Public Health Ser- vice Commissioned Corps, and the Medical Reserve Corps. The Emergency Management Assistance Compact, administered by the National Emergency Management Association, can also pro- vide volunteers. Individuals recruited by this program (31,000 for Hurricanes Katrina and Rita alone) played important roles in the responses to the four hurricanes in 2004 and to Katrina and Rita the next year. It must be stressed that all of these are “borrowed” resources to the extent that the nation does not have a standing pool of available healthcare personnel. Each individ- ual has an existing job, so the problem of mobilizing personnel with existing commitments is one that occurs in every major disaster.47,52

Mortality and Morbidity from Heat Exposure There is a relatively consistent correlation between mortality

and increases in heat measured by temperatures, heat index (a measure of temperature and humidity), or by air-mass condi- tions.12,14,53–55 Using sophisticated air-mass models, researchers have demonstrated a clear relationship between heat-related mortality and EHEs in 44 major U.S. metropolitan areas.14 In a study of 28 metropolitan areas within the United States, heat- related deaths during EHEs significantly exceeded the expected totals for time of year and were in substantial agreement with the previous findings.17 Overall, however, heat-related mortal- ity trends have declined between 1964 and 1998, as tempera- tures and heat stress conditions have risen. This suggests that a relative “desensitization” of the U.S. metropolitan population to weather-related heat stress has occurred. Such desensitization can be attributed to a variety of factors, including improved med- ical care, increased utilization of air conditioning, better public awareness programs, and both human physiological and urban medical/emergency response systems adaptations. The irony is clear: traditionally hotter metropolitan areas have lower heat- related mortality.

EHEs also increase morbidity, although the data are less compelling. The majority of studies emphasize the most seri- ous incidents that result in ED or hospital admissions. Nonethe- less, the general trend of research examining all conditions is clear.56 Semenza studied 1,072 hospital admissions during the 1995 Chicago EHE and found the majority of excess admissions were due to dehydration, heat stroke, and heat exhaustion among people with existing, underlying conditions.41 Rydman also stud- ied intense heat-related morbidity from the same event through an analysis of ED visits and found that heat-related morbidity was an antecedent of mortality.57 Kilbourne searched for the most effective responses to heat-related illnesses and determined that universal access to air conditioning may be the most effec- tive intervention, even with the relatively steep costs of providing this service to the poor.32 More recent studies found that basic behavioral changes and adaptations (e.g., use of air conditioning, adequate hydration, heat emergency plans, warning systems, and illness management plans) could significantly affect heat-related outcomes.58

The most precise definition of mortality related to heat waves is that given by a medical examiner or a coroner in the formal determination of death. After serious EHEs in Chicago in 1995,

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

616 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

Figure 38.4. The demographic and individual risk factors of heat-related mortality.

and in Philadelphia in 1993, the National Association of Medical Examiners recommended the following definition of heat-related death: “a death in which exposure to high ambient temperature either caused the death or significantly contributed to it.” The committee also recommended that the diagnosis of heat-related death be based on a history of exposure to high ambient temper- ature and the reasonable exclusion of other causes of hyperther- mia. The diagnosis may be established from the circumstances surrounding the death, investigative reports concerning envi- ronmental temperature, or measured antemortem body temper- atures at the time of collapse being at least 40.6◦C. Under those conditions, the cause of death should be certified as heatstroke or hyperthermia. In cases in which the antemortem body temper- ature cannot be established, but the environmental temperature at the time of collapse was high, appropriate heat-related diag- nosis should be listed as the cause of death or as a significant contributing condition.59

Demographic and Individual Factors in Heat Exposure Those individuals at the highest risk of becoming ill or

dying during EHEs are the very young, the elderly (socially isolated, without access to air conditioning, bedridden, and with ischemic heart disease or other chronic conditions), the poor, minorities, and those taking certain medications such as neuroleptics or antiparkinson agents.8,26,45,55,57,60–65 Behaviors that can result in heat stroke from dehydration and impaired judgment include strenuous exercise in hot or humid weather (even by the young and physically fit), alcohol consumption, and the use of some nonprescription drugs (e.g., antihistamines and sleeping pills).8,66 Cocaine overdose, which is associated with hypertension, tachycardia, coronary vasospasm, arrhyth- mias, and increased core temperature, was linked with a signif- icant increase in mortality in Marzuk’s 1998 study of New York City.39,64

Use of cooling centers by individuals was not shown to be significantly protective, probably because so few visited them.60

Walking down flights of stairs, with the mobility limitations often accompanying older age, and crossing potentially unsafe streets

to attend cooling centers are unlikely options for many of the at-risk elderly.60 European mortality patterns in the August 2003 heat wave mirrored those seen in the United States, with 70% of those dying from heat-related causes being 75 years or older.67

Accurate demographic estimates and projections of those sick- ened or dying from heat-related causes are complicated because susceptible groups often remain in the city, creating a bias in predicted excess deaths.68 Most often, there are more females affected by adverse heat-related conditions. When stratifying by age group, there is probably some residual confounding related to sex, because the increase in mortality seems to be greater among women. This probably reflects the higher proportion of women in the elderly population, their possible higher susceptibility, and their higher rates of living alone (Figure 38.4).68

During Chicago’s heat wave of 1995, hospital admissions were up 11% for the week of the heat event, with a 35% increase for patients 65 years of age and older. The majority of the excess admissions (59%) represented patients requiring treatments for dehydration, heat stroke, and heat exhaustion. With the excep- tion of acute renal failure, no other primary discharge diagnoses were significantly elevated. In contrast, analysis of comorbid conditions revealed 23% excess admissions of underlying car- diovascular diseases, 30% for diabetes, 52% with renal diseases, and 20% excess admissions related to nervous system disorders. Patient admissions for emphysema and epilepsy were also signif- icantly elevated during the heat event.41

Geographical Factors in Heat Exposure Average summer temperatures appear to have no effect on

heat deaths. Kilbourne observes that it is not the absolute tem- perature value, but rather the extent of upward deviation from the usual summer temperature that seems to be the key variable affecting mortality.32 For example, southwestern cities such as Phoenix have fewer heat-related deaths but have higher average temperatures than midwestern cities such as St. Louis or Chicago, which have experienced high mortality from heat waves. Neither excess mortality nor prominent heat-related health effects were noted in Phoenix in July 1980 despite temperatures that averaged

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 617

Table 38.3: Threshold Heat Temperatures that Result in an Increased Local Mortality Rate

Threshold Location Temperature (◦C)

Atlanta, Georgia 34

Chicago, Illinois 33

Cincinnati, Ohio 33

Dallas, Texas 39

Denver, Colorado 32

Detroit, Michigan 32

Kansas City, Kansas 37

Los Angeles, California 27

Memphis, Tennessee 37

Miami, Florida 32

Minneapolis, Minnesota 34

New York City, New York 33

Philadelphia, Pennsylvania 33

St. Louis, Missouri 36

Salt Lake City, Utah 35

San Francisco, California 29

Seattle, Washington 31

2.4◦C above normal and a highest monthly temperature of 46◦C. As expected, based on reports in the international literature, cities that normally have a cool climate (those located in the north) reported the highest excess mortality.69 Reasons for these differ- ences have not been studied extensively. Possible explanations include differences in population age/acclimatization, architec- tural style/building materials, and air conditioning use. Research done at the University of Delaware resulted in a list of the tem- perature levels that affect mortality and morbidity in select, large American cities whose populations have been subjected to vari- ous degrees of heat stress (Table 38.3).

There is some evidence of a geographical and physiological basis for the lower death rates in urban areas that have higher average temperatures. Kilbourne makes the same observation that heat seems to cause fewer health problems in characteristi- cally warm areas than in those that are more variable in climate; the temperature level required to increase mortality is actually higher in hotter climates.14,32,71 DiMaio has observed that when individuals live in a temperate zone, many of the sweat glands with which they were born become permanently inactive dur- ing childhood.72 If, however, the individual lives in the tropics, the glands remain functional throughout life. Other adaptations include reduction in sodium loss from sweat to 3–5 g/day, after 4–6 weeks of acclimatization. A person who sweats profusely may lose as much as 15–30 g/day of sodium chloride until becoming acclimated.

Chestnut demonstrated geographical patterns in heat-related mortality.73 The highest hot weather–related mortality rates are in northern metropolitan areas of the U.S., even though aver- age summer temperatures are higher in southern metropolitan

areas. This suggests that biological/behavioral adaptations occur in areas that are consistently hot, but not where minimum daily temperature variability is great. The availability of air condi- tioning, standards of living, and housing quality contribute to differences in mortality, but these explain a much smaller share of the fatalities than does variability in minimum daily tempera- tures.73 Kalkstein found that consecutive days of hot, oppressive weather caused a continued rise in mortality.74 Chestnut also reported that areas with higher average temperatures and more frequent hot weather episodes do not necessarily experience more hot weather–related mortality during summer months.73 Sev- eral southern metropolitan areas with very hot, humid summer weather experience much lower or no statistically significant hot weather–related-mortality. It remains unclear how many of these data apply to the degree of mortality or morbidity from increases in the temperature, duration, and frequency of hot weather episodes in the tropics.2

Death rates related to EHEs do not occur on the days with the highest average temperatures. In a study of heat-related mortality from the September 1970 heat wave in New York and other eastern seaboard cities, the highest mortality levels occurred on the third day, when the temperatures were less than those on the first 2 days (Figure 38.5).8,75,75a In Chicago’s 1995 EHE, a slightly different pattern manifested itself, with the number of deaths peaking 2 days after the maximum heat-index recording.76 This is consistent with previous studies demonstrating that during EHEs, the maximum number of heat deaths tends to lag behind the days with the highest temperatures.77,78 It has also been suggested that EHE-related mortality may reflect the tendency of heat stress to precipitate death in persons who are already ill from a wide variety of chronic diseases and would die in the near future anyway.8 Evidence of this potential effect has been sought but not found. For example, no decrement in the number of deaths was reported following the EHE in New York in 1972.8,75,75a

Figure 38.5. Heat-related deaths – Chicago, July, 1995. This graph tracks maximum temperature (Tmax), heat index (HI), and heat- related deaths in Chicago each day from July 11–23, 1995. The light gray line shows maximum daily temperature, the dark gray line shows the heat index, and the bars indicate the number of deaths for the day. Source: By the National Synthesis Team. U.S. Global Change Research Program, published in 2000.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

618 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

Figure 38.6. Urban heat island profile. Source: Environmental Protection Agency, http://www.epa.gov/heatisland/about/. Modified by Tom Javorcic, 2007.

Urban Heat Islands as Risk Factors

The Urban Heat Island and Bioclimates The long-established concept of the “Urban Heat Island” is

pervasive in the American and the European literature on EHEs, and applies, to a lesser extent, to urban areas in developing coun- tries.79,79a This phenomenon describes urban and suburban tem- peratures that are 1◦C–6◦C hotter than nearby rural areas. Ele- vated temperatures can impact communities by increasing peak energy demand, air conditioning costs, air pollution levels, and heat-related illness and mortality.80 A typical urban heat island profile is shown in Figure 38.6. Important aspects of the urban heat island literature include themes regarding poverty, social isolation, social class, race/minority status, crime, poor housing, inadequate healthcare, and mobility problems. Ultimately, most of these factors are related to poverty and unemployment. Stud- ies of heat islands do not generally indicate bioclimate aspects (effects of weather and climate on life forms, including humans) and are therefore of limited use to urban planning.81

Urban areas tend to be warmer because of their “masses of stone, brick, concrete, asphalt, and cement.”25 Their darker surfaces, which absorb more solar heat during the day, radi- ate this energy into the environment at night.70 All of these factors including diminished wind and cooling air currents cre- ate the heat storing and absorbing “Urban Heat Island.” The Atlanta Metropolitan Area Heat Profile appears in Figure 38.7 with dark and light gray denoting hotter and colder areas. Heat wave response planning is of interest in intensely developed

Figure 38.7. Source: Environmental Protection Agency, http://www.epa.gov/heatisland/about/measurement.html. Modified by Tom Javorcic, 2007.

urban areas that are large enough, dense enough, and often old enough to have potentially damaging urban heat islands. U.S. studies analyzing data from the last 100 years confirm this insight and recognize clear associations among 1) large masses of cityscape (e.g., cement, asphalt, and high percentages of mul- tiple unit dwellings), 2) a relative lack of trees and other veg- etation, and 3) poor wind and air circulation patterns. These areas have experienced relatively high EHE-related morbidity and mortality.8,31,70,73 Klinenberg has called urban heat islands “Urban Loneliness Islands,” referring to the physical and social isolation that affects many elderly heat victims.82

After the deadly Chicago heat wave of 1995, the National Oceanic and Atmospheric Administration observed7

There is sufficient circumstantial evidence to conclude that the urban heat island was at least partially responsible for . . . conditions in Chicago’s south side . . . the elderly and infirm . . . in urban areas are in the greatest danger during heat waves.

During EHEs, these effects can magnify health risks by increasing both the potential maximum temperature to which residents are exposed and the duration of time in which this exposure occurs.3

Urban populations in nonindustrialized countries are likely to be particularly vulnerable to such effects when compounded by climate change.83 Worsening economic conditions will only exacerbate this impact. A recent statement by Lord May, Pres- ident of the (British) Royal Society, makes clear that climate change and its related and unexpected costs could ruin efforts to elevate Africa out of poverty.4 Vulnerable populations can, however, be spared many of the effects of EHEs through rela- tively uncomplicated human interventions, including the use of advanced early warning systems and providing access to water, shelter, appropriate emergency medical care, and air condition- ing – if available.1

Concentrations of asphalt, concrete, stone, and brick are relatively new, and make up relatively smaller parts of urban areas in the developing nations. Poor immigrants to the city, and even the emerging middle class, usually live in a differ- ent environment. These areas are characterized by quasitradi- tional, owner-constructed, single- or extended-family housing, frequently densely packed, and usually with dirt rather than asphalt or cement roads. These typically large barrios, favellas,

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 619

kampungs, or townships are not really heat islands, due to their lack of a genuine urban infrastructure, and they vary widely as to their degree of criminality. Yet these communities are where almost all deaths and illnesses from EHE events will occur.

Heat in the Indoor Environment Kovats and Jendritzky have found that there are three main

factors associated with indoor heat exposure.84

1) Thermal capacity of the building – A heavy building will warm up more slowly in a heat wave, particularly if it is well insu- lated.

2) Position of apartment – The upper floors of a building will generally be hotter than the lower floors because the roof provides inadequate insulation and warmer air tends to rise.

3) Behavior and ventilation – Occupants will adapt as best they can to hot environments using fans (which can actually cause harm at higher temperatures) and opening windows to let in cool evening air. The fact that many people tend to wear the same clothing regardless of season or temperature demon- strates the complexity of decision processes and determinants of behavior.

Givoni has categorized the variables that influence the inner bioclimate.81,85,86

■ The geometrical configuration of the building ■ The orientation of the building ■ The size and location of the windows ■ The properties of the building materials ■ The colors of the external surfaces

Givoni has also categorized the aspects of external design that affect urban climate.

■ Size and density of the built up area: – The microclimate in the immediate vicinity of green

spaces differs from that prevalent in unplanted areas – Vegetation has lower heat capacity than building mate-

rials – Solar radiation is absorbed so that reflected radiation is

very small (low albedo) – Green spaces have higher evapotranspiration rates than

unplanted areas – Plant leaves can filter dust out of the air

■ Layout and width of streets and their orientation to prevailing winds

■ The height, shape, and relative location of buildings ■ Shading conditions along streets and parking areas ■ Ensuring short distances for walking

There is a need for further health research on the relationship between housing types and heat effects. In addition, there is little information about how people behave in their homes. When do they use their air conditioners? What is the effect of electrical costs on the poor? When do they open their windows for cooling or when do they close them to keep out pollution or noise?

Air Conditioning Air conditioning is a “special case” in Europe, which relies

much less on air conditioning than the United States, even in many healthcare institutions. The European research on air con-

ditioning is far from definitive, but supports the use of unit-wide, central air conditioning as opposed to single room air condition- ing. The latter has been found to be minimally, if at all protective, unless the housing unit has enough window air conditioners to approximate the coverage of central air conditioning.87 Although the Europeans recognize the protective nature of air condition- ing in EHEs, they stress that air conditioning requires sealed buildings that create stagnant, polluted air issues, and that air conditioning itself uses energy that contributes to global warm- ing. When power grids fail, people are often left in relatively airtight buildings.81 In Europe, air conditioning is advised only in cases in which ill health is present.84

Planning for Extreme Heat Events

The Extreme Heat Event Planning Process Despite the many examples of deadly EHEs, a recent review of

plans from 18 U.S. cities at risk for heat-related mortality found that many had inadequate plans or no plans at all.88 Another study of the nation’s 120 largest cities found that only 29 had developed single-purpose response plans of varying quality and scope.89,90 Although there is no mechanism for coordinated U.S. EHE response planning, this decade has seen significant fed- eral activity. These efforts include the development and mass circulation of a comprehensive, interagency extreme heat event guide (2006), the federally sponsored Heat Wave Workshop in 1996, the National Weather Service’s growing use of sophisti- cated air mass prediction systems, the CDC’s coverage of EHEs in its Morbidity and Mortality Weekly Reports, and the studies found on multiple web pages by the CDC, Environmental Protec- tion Agency, and National Oceanic and Atmospheric Adminis- tration.

Urban EHE response planning has developed into a unique policy area with its own literature that is scattered among larger disciplines. It is supported by a growing public awareness con- stantly reinforced by heat wave alerts. During recent decades, a modestly growing number of urban governments in the United States, Canada, and Europe have reacted to EHEs by developing EHE response plans. These include alert/watch/warning systems based on an analysis of deadly air masses that provide more “early warning time” than previous systems. In addition to such strategies, existing urban EHE response plans have included the following: 1) public utility bill forgiveness or modified bill pay- ment, 2) free air conditioner distributions, 3) public cooling cen- ters, 4) hotlines, 5) public education and information systems, 6) registries for the elderly and other at-risk populations, and 7) aggressive outreach programs. Such policies could be adopted in developing countries, although international donors may need to supply the necessary resources. The most plausible cooling sites are usually government offices and elite hotels, places often reluctant to admit poor people.

Khogali and Rosenfeld break down response strategies into primary, secondary, and tertiary levels. Their recommendations are specific and proactive in reducing heat-related disease in the workplace and sports venues.91

Primary prevention addresses adequate and effective building design to maximize comfortable cooling, good ventilation, and reduction of radiant and convective heat with mechanical aids.

Secondary prevention includes a wide variety of workplace and sports-related preventative measures that are placed in two groups. The first is referred to as selection and acclimation. These interventions include pre-event employment and placement

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

620 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

medical examination in occupational settings and pre-event par- ticipation medical examination for sports activities. The second measure is appropriate administrative personnel behavior. This emphasizes modification of the work-rest cycle or the exercise– rest cycle and provision of cool rest areas with fluids.

Tertiary prevention strategies aim at diagnosing heat illness syndromes as early as possible. Khogali and Rosenfeld stress workplace and organized athletic participation.

Certain characteristics are associated with individual munici- palities that have developed an EHE response plan. A logistic regression analysis of the 29 cities with such documents showed that cities with larger populations, higher percentages of multiple unit dwellings, fewer residents age 25 or older with bachelor degrees, higher violent crime rates, and a member of a heat wave advocacy coalition working or living within the jurisdiction were more likely to have developed a plan. Violent crime was a significant predictor. If such correlations are not spurious, they offer further evidence of widespread perceptions that heat waves are law-and-order problems. Cities with a political rather than a professional administration are more likely to have a plan, suggesting “political” solutions as more plausible than purely “technocratic” ones.

Well-designed EHE plans can be cost effective. Semenza, for example, concluded that those at greatest risk of dying during Chicago’s heat wave of 1995 were people with medical illnesses, socially isolated, and without air conditioning. Such groups could clearly benefit from simple interventions.55

In both Chicago and Milwaukee during the 1995 events, the National Weather Service issued warnings of the developing heat wave several days in advance, which were quickly broad- cast by the local media. Given this advance warning, many of the heat-related deaths associated with this event were preventable.93

Despite these timely warnings and effective media coverage, this information either failed to reach or was not used effectively by the people who could have prevented heat-related deaths. This included the victims themselves as well as members of the healthcare community who did not always understand the scope of the impending calamity.7 Much more than effective media announcements are needed to avoid serious heat wave mortal- ity and morbidity.55,57,94–97 The development of pre-event plans that integrate public and private assets are necessary to coordi- nate and effectively deploy life-saving resources.

It is difficult to use scientific evidence to judge the post facto efficiency of EHE response plans because no two heat waves are the same, and because of the large number of interrelated variables. Researchers have analyzed the 1995 EHEs in St. Louis and Chicago and compared them to the 1999 EHEs in those same cities. Although noting differences, they saw more effective and faster responses in 1999 that appeared to save lives that would have been lost in 1995. These cities learned from prior experiences and crafted effective EHE response plans.

The only intervention that was not widely reflected in the content of existing plans was equipment grant subsidies (e.g., for air conditioners) for poor residents. Not surprisingly, bill forgive- ness, bill postponement, and related programs involve real losses for utility companies. Such programs, as they exist, reflect a con- fluence of creativity, resources, cooperation, and public spirit. In developing countries, air conditioners would rank low among priorities for the poor as compared with receiving advanced notice of the impending increase in temperature, better educa- tion, potable water, decent shelters, and healthcare. In addition, a large number of air conditioners would likely cause local circuits

to fail or create a demand that electricity-generating facilities could not fulfill.

Community Prevention and Mitigation Plans for Extreme Heat Events

The U.S. National Center for Environmental Health within the CDC conducted a nationwide survey of local heat wave prepared- ness plans in an effort to develop guidelines for cities.6,98 CDC’s National Center for Environmental Health reviewed EHE pre- paredness plans from 12 U.S. cities at risk for heat wave–related morbidity. Cities were selected based on location and population. Examination of their respective plans focused on assessing key facets. These included 1) the inclusion of community organiza- tions, 2) plans for early information dissemination, 3) targeting of high-risk populations, 4) heat monitoring methods and plan activation, 5) interventions, and 6) evaluation. The important elements of these plans are discussed.

Community Participation

Effective heat wave response plans require collaboration between a wide variety of government agencies including departments of public health, emergency management agencies, urban and regional planning departments, agencies for the elderly, and community and volunteer organizations. Community organi- zations can lend and donate equipment or supplies, provide air-conditioned cooling sites, and assist with membership lists to conduct outreach for high-risk populations. Although the departments of health and community organizations were gen- erally included in plans, incorporation of the police, media, hos- pitals, utility companies, and local businesses varied. Communi- cations systems between organizations appeared in all plans, but only six provided complete and updated contact information.

Early Information Dissemination

Heat-related illnesses can be prevented by the dissemination of early public information, and such educational endeavors are essential to successful prevention efforts. These should include recognition of health threats posed by high heat and humidity, appropriate precautionary and response measures, and encour- agement to check on family members and neighbors in high-risk groups. Eight of 12 plans included early educational messages to city residents.

High-risk Populations

Although eight of the 12 surveyed plans targeted educational messages and interventions toward the elderly populations, only four developed specific methods to reach socially isolated individuals. Furthermore, only one plan addressed individuals with chronic illnesses, and two addressed those on medications affecting thermoregulation. Three focused on the disabled and two on the homeless. Clinics, pharmacies, physicians, visiting nurses, and home health workers can help with public educa- tion, but were not included in many plans.

Heat Monitoring Methods and Plan Activation

Each plan involved methods for monitoring heat. Typically, mon- itoring was done by the local office of the National Weather

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 621

Table 38.4: Elements of Effective EHE Plans

■ Ensuring prediction of EHE conditions 1–5 days in advance ■ Quantitative links between temperature and health effects ■ EHE notification and effective governmental responses: hotlines,

public information, at-risk registry, direct action ■ Active mitigation by urban builders and urban governmental regu-

lators

Service. Six plans included the medical examiner’s office and EDs for evaluating heat-related mortality and morbidity. Ten plans used a three-level public warning system for advisory, watch, and warning levels. Only three described criteria for deactivation.

Interventions

All 12 cities surveyed had plans to provide air-conditioned shel- ters. Ten municipalities used hotlines, and six provided trans- portation to the shelters. Two provided air conditioners and “check-in” services for the elderly, and two could procure emer- gency funds for cooling packs, beverages, and transportation services. Four provided water, suspension of utility disconnec- tion, and extension of community swimming pool hours. Two cities provided translation services and four provided telecom- munication devices for the deaf.

Evaluation

Only one proposal outlined a method of evaluating the plan following its implementation. After the EHE and the response conclude, the lead organization should conduct an evaluation of the city’s efforts and document the results in an After Action Report. This evaluation should include data/input from partici- pating organizations and from the public. This will help improve future responses to heat wave emergencies.

Elements of Effective Extreme Heat Event Plans

Effective EHE preparedness and response plans vary, based on different local circumstances. The EHE experts who developed the Excessive Heat Events Guidebook reviewed multiple informa- tion sources, including the preparedness and response literature, the views of other experts in the field, and the best examples of existing plans.1 They selected a number of elements key to the development of effective EHE preparedness and response plans (Table 38.4).

Extreme Heat Event Prediction: Ensure Prediction of Extreme Heat Event Conditions 1–5 Days in Advance

Adequate advanced warning is probably the most impor- tant factor in preventing EHE related morbidity and mortality. EDs need time to augment staffing and hospitals must adjust to summer vacations, staff rotations, and the suspension of elective procedures. Ambulance services require similar preparation and coordination. If importing medical support from other jurisdic- tions is necessary, pre-event planning is critical. Since the pool of paramedics and emergency medical technicians is a relatively inelastic group, the practice of assigning individuals to serve in a variety of capacities will likely be ineffective and only highlight existing shortages and limited staff availability.

Forecasting the development and characteristics of an EHE is critical to both EHE risk assessment and implementation of notification and response systems. The National Weather Ser- vice provides this information across the U.S. and is moving toward the use of more sophisticated air mass models that pro- vide improved notification. Toronto and a growing number of U.S. cities use this sophisticated air mass–based prediction sys- tem that incorporates local factors. The system was originally developed by Kalkstein at the Center for Climatic Research at the University of Delaware. Air mass occurrences can be predicted up to 48 hours in advance with the use of model output statistics guidance forecast systems.99

Extreme Heat Event Risk Assessment of Potential Health Impacts from Rising Temperatures

Whichever agency coordinates the response to EHEs, it must develop quantitative estimates of potential health impacts from the rising temperatures. These impacts will range from effects on EDs and hospitals to populations whose special needs and cir- cumstances make them more vulnerable. Emergency managers can access records of facilities and locations with significant con- centrations of high-risk individuals to use as the basis for health needs projections. Estimating resource needs based on the risk assessment will result in outreach activities to those most likely to be affected.

Extreme Heat Event Notification and Effective Response Actions

Coordinating public broadcasts of information about the anticipated timing, severity, and direction of EHE conditions as well as information regarding implementation of various pro- tective measures is important. Emergency facilities, healthcare providers, and hospitals require notification regarding the tim- ing of EHEs so that emergency staffing plans can be invoked. Emergency phone lines must be created as soon as early weather warnings are issued.

Mutual aid plans are necessary so that staff and facility short- ages can be managed effectively. In addition to mutual aid assis- tance, in the U.S., a variety of state and federal programs may provide staffing during EHEs, including the National Disaster Medical System, the Emergency Management Assistance Com- pact, and the Medical Reserve Corps. At the time of this writing, some of these federal response assets are coordinated through the staff of the Assistant Secretary for Preparedness and Response. Local hospital councils and hospital networks may also provide mutual support.

Outreach activities for high-risk groups including the elderly and homeless must be undertaken as soon as possible after early weather warnings are issued. Utility shutoffs should be suspended during the period of the EHE and the recovery phase. Cooling shelters must be staffed and operational, with transportation and security measures taken. Rescheduling of some public events may be required.

Extreme Heat Event Mitigation Strategies currently exist that can reduce the effects of urban

heat islands and general EHE conditions. City managers can develop and publicize urban garden and vegetation programs. Contractors can construct buildings using lighter colors (partic- ularly on roofs) and use the reflective quality of certain building materials, particularly in public structures, to reduce EHE effects. A summary of basic strategies that various levels of government

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

622 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

Table 38.5: Effective Urban Strategies to Reduce Heat Island Effects

■ Cities foster urban gardens/roof gardens ■ Builders use light colors (especially on roofs or any reflective surfaces) ■ Cities support tree and shrub intensity ■ Cooling airways planned

can follow to reduce the risks of heat-related mortality from EHEs are shown in Table 38.5.1

The Philadelphia Extreme Heat Event Response Plan

In response to a deadly EHE in 1993, a multiagency task force developed an integrated EHE preparedness and response plan. As the Excessive Heat Events Guidebook observes, the Philadelphia plan is often described as the “Benchmark” Plan. The Philadel- phia Plan includes1

■ Public announcements associated with intense education and preparation of the media

■ Buddy system advocacy for checking on local high-risk resi- dents throughout the event (includes block captains)

■ Heatline activation ■ Home visits by health department staff ■ Halt to service shutoff during high heat warning periods ■ Increased emergency medical service staffing ■ Increased outreach for the homeless ■ Cooling shelter/senior refuge ■ Outreach by public posting of key numbers on local land-

mark buildings

Basic strategies for reducing heat-related mortality based on a review of the EHE response literature and several model city plans are presented in Table 38.5.

The Chicago Heat Wave of 1995

The Chicago Heat Wave of 1995 was a relatively short, but intense event that resulted in deaths within vulnerable populations and damaged the reputation of the political system that was slow in recognizing and coping with its consequences. The effects of heat- related emergency extend beyond morbidity and mortality and include intensely emotional and political components. During the Chicago heat wave, a short but intense confrontation between the medical examiner and the city’s mayor ensued regarding the definition of heat-related deaths. By varying accounts, Chicago experienced approximately 500–700 excess deaths, most of which were classified as heat-related during the roughly 5-day heat wave period.55 The African-American community had the highest rate of heat-related mortality of all minority groups.100 In Chicago’s case, there was an existing heat wave emergency plan, but it was not as detailed or as extensive as the one that was developed after the event. The plan in effect during the heat wave was very brief and not comprehensive. It was the focus of an intense media debate regarding its value, the lack of urgency of the city’s initial response, and the definition and numbers of “excess” heat-related deaths. In the midst of a growing controversy, two reporters for the Chicago Sun Times wrote

Chicago’s 11/2 page heat plan is thin on details about getting relief to the people who need it most. And during last week’s heat emergency, they failed to follow that plan until the death toll started rising.101

During the first days of the heat wave, the mayor was under attack for his response to the deadly event, and, like most Chicagoans, he did not anticipate the problem. The Cook County medical examiner reported more than 370 heat-related deaths, and accurately predicted that the final count would be more than 400. The newspapers reported that when confronted with such numbers, the mayor responded to the medical examiner by saying, “Every day people die of natural causes . . . You can’t put everything as heat-related.” The nightly news aired images of the crowded funeral homes and the refrigerated trucks outside the county morgue.102 The medical examiner’s figures and methods were supported by later research.63 Once the magnitude of the crisis became clear, an improved plan was quickly developed. In the interim, community groups called for the resignation of top cabinet members involved in the heat crisis.103

Extreme Heat Events in the Developing World

A scientific consensus has developed regarding the belief that global temperatures are warming and that there is an association with humankind’s growing use of carbon-based fuels. In 1988, under the guidance and auspices of the United Nations (UN), scientists and government officials from around the world inau- gurated the Intergovernmental Panel on Climate Change. This body has produced massive world assessment reports in 1990, 1995, 2001, and most recently, in 2007, further refining and cor- roborating the global warming paradigm as both anecdotal and scientifically validated evidence continues to grow.104 A meta- analysis was reported in the journal Science in 2004, stating that of 928 papers presented, “none of the papers” disagreed with the thesis that global warming exists and that human activity is a cause. A significant amount of planning and analysis has emerged from the United States and Europe in response to deadly heat waves. Given that 2007 had the highest January temperatures on record to that date, planning for EHEs across the world will likely intensify (Figure 38.8).1,8,105,105a,105b Planning related to global warming has its limitations. Robert J. Samuelson has said, “we don’t know enough to relieve global warming, and – barring major technological breakthroughs – we can’t do much about it.” He cites projections from the International Energy Agency demonstrating that “unless we condemn the world’s poor to their present poverty – and freeze everyone else’s living stan- dards, greenhouse gases will double by 2050. No government will adopt the draconian restrictions on economic growth and personal freedom that might curb global warming.” World pop- ulation growth and development will likely proceed as expected exemplified by China’s construction of one new coal-fired power plant each week.106 The urban megacities of the developing world are subject to the interlocking problems of shortages of potable water, weak government services, lack of access to air- conditioned structures, and crowded shelters (Table 38.6).

Despite the history of EHEs, there is consensus that many adverse outcomes are preventable.1,107 Reducing future unde- sirable outcomes requires improving the awareness of public health officials and the general public regarding the associated health risks while continuing to develop and implement effective

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 623

Figure 38.8. Temperature anomalies january 2007 (with respect to a 1961–1990 base period). Source: National Oceanic & Atmospheric Association, http://www.noaa.gov. Modified by Tom Javorcic, 2007.

EHE notification and response programs.106 EHEs are public health threats because they often increase the number of daily deaths and other nonfatal adverse health outcomes in affected populations. The most vulnerable groups can be found living in urban heat islands and EHEs can increase mortality and morbid- ity in these vulnerable populations.92,106,108 Future unfavorable health impacts could be reduced through early forecasting of EHEs and subsequent implementation of low-cost and effective responses.

The effects of global warming are intensified by two cru- cial issues: population growth and the shift to living in cities. For those residing in developing countries, the cities often pro- vide only rudimentary infrastructure and poverty significantly impacts quality of life.109 At the end of 2008, the world’s popula- tion reached more than 6.87 billion and it is expected to reach 7.5 billion by 2015. Two-thirds of these individuals will live in cities by 2020 with estimates of up to 1 billion now living as squatters in informal housing, mainly across the developing world.110,111

In 2015, there are predicted to be 23 “megacities” (>10 mil- lion inhabitants each), with 19 of these in the developing world. This does not imply, however, that urban density is necessarily negative. London’s stylish Kensington and Chelsea have densities three times those of poorer boroughs.108

The available support from international organizations is insufficient to ameliorate EHE disasters in meaningful ways, so a search for solutions to these growing public health problems is a

Table 38.6: Interlocking EHE Problems in Urban Megacities of the Developing World

■ Shortages of potable water ■ Weak government services ■ Lack of access to air conditioned structures ■ Crowded shelters ■ Unplanned, tightly packed neighborhoods with casual structures

pressing global need. Industrialized nations can play a significant role as informed leaders of public opinion, selecting the best domestic and international policies. The situation in several cities will be highlighted to illustrate the problems.

The following discussion will focus on two cities from South America (Caracas, Venezuela and Rio de Janeiro, Brazil), Africa (Cairo, Egypt and Nairobi, Kenya); and southeast Asia (Jakarta, Indonesia and Kuala Lumpur, Malaysia). There are many inter- esting differences among these cities, but the cities’ surprising similarities will be discussed. Because political and economic constraints on adequate preparations for EHEs are significant, these will be addressed for developing nations.

Each of the cities mentioned previously has the following characteristics: 1) a gleaming high-rise commercial center rep- resenting a heat island that is smaller in relation to the city’s total surface area than in the west, 2) spacious elite residential areas occupied by foreigners and ethnic minorities whose success can stir resentment among the poor, 3) huge residential areas occupied by lower socioeconomic status individuals in search of employment (often recent migrants from rural areas), and 4) a better-educated, emerging middle class that cannot afford elite residences. The areas where the poor reside are called bar- rios, favellas, alleys, townships, or kampungs. There, the physical infrastructure, roads, electricity, sewage facilities, potable water, communication networks, schools, clinics, and hospitals range from grossly overstretched to virtually nonexistent. For exam- ple, in some regions, extreme, monsoon-like rains cause sewage- laden flooding that does not drain because the open ditches built by local residents do not connect to each other.112 Also lacking are social safety nets and bureaucratic competence. Individuals have few net benefits from citizenship. This exacerbates prob- lems with any response to EHEs. While military and security institutions do exist, these powerful resources are not generally available to address EHEs.

Each vulnerability exacerbates the others and compounds the impact of disasters, ensuring persistent malnutrition, insecure

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

624 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

residential status, and deteriorating public safety.113 Traditional coping techniques are ineffective in these urban areas. Personal savings do not exist so people cannot afford to replace their homes when destroyed by disasters.108 The poor are essentially segregated from mainstream markets and the economic oppor- tunities these markets offer, as well as from basic education and healthcare. Even under conservative assumptions about popula- tion growth, landuse, and industrial production in developing countries, computer models predict serious health consequences from a warmer world.115

High levels of air and water pollution in these cities will exac- erbate the deleterious effects of EHEs for the majority who cannot afford any mitigating interventions, even bottled drinking water. Breathing the air of Jakarta (population 14 million) has the esti- mated health effect of smoking two packs of cigarettes a day. This impact is magnified because many Jakartans do smoke two or more packs of cigarettes a day as well. On a hot day, the collective struggle for breath is palpable. Carbon dioxide concentrations and diesel exhausts contribute to high levels of asthma and other respiratory problems. The absence of any interconnected sewer drains means a greatly increased risk of hepatitis, gastroenteri- tis, cholera, meningitis, typhoid, and even polio – especially for children.116

It is extremely difficult for the poor to help themselves, or compel the authorities to provide assistance. There are few if any realistically available “cooling centers” in these six cities. The main venues – government offices and elite hotels – will not welcome temporary occupation by the poor. It might be thought that tax incentives would prompt corporations to occasionally open their cooled offices to those living in poverty, but some corporations have already made arrangements to avoid paying taxes, and others do not provide such shelter for practical or security-related reasons. Some of the poor can attempt to seek shelter in the badly polluted underground car parks or similar structures, but will likely be deterred by inhospitable police. The prevalence of rice paddies in the suburbs of Asian cities offers some cooling, as does the prevalence of Australian eucalyptus trees, but their deep roots lower the water table.

Elitism, or rule by a small and powerful politically insu- lated group, is more frequent, even in ostensibly democratic countries in the developing world. The six cities and countries discussed have varying degrees and types of democracy. Elitism is far from absent in western politics, but there are counter- vailing powers and pressures on elitism in the west, so it does not always determine political outcomes. Analogous powers and pressures have yet to emerge in most developing countries. In the cities/countries under discussion, elitism simply correlates with increasing inequality and a resistance to reform.117 In these cities, one of the only recourses to for the poor is to engage in political demonstrations or civil unrest. These tactics usually do not improve their situation to any great degree. This is a major reason why elites, and the police and military forces they support, treat EHEs primarily as law-and-order problems.

The state of administration in most developing countries is also problematic. In the west, disaster planning and implemen- tation is the province of trained and motivated professionals. In contrast, there is little bureaucratic altruism in most develop- ing countries and little evidence of professionally trained man- agement. Elite interests are typically far removed from EHEs. Attempts to plan and implement EHE policies are thwarted by low levels of administrative competence, especially at the munic-

ipal level, and public mistrust, because few benefits are conferred by this bureaucracy. It is impractical for a bureaucrat to specialize in healthcare planning, because these administrators are regu- larly transferred for reasons unrelated to their expertise or the public’s needs. Low bureaucratic salaries and sometimes a cul- ture of corruption often mean that actual management decisions run contrary to any reasoned elaborations of the public inter- est. Persons imbued with an ethos of professionalism will be less likely to exhibit such behavior.118

All problems cannot be attributed to political and admin- istrative underdevelopment. Even the best of intentions are stymied by an acute shortage of resources. The western mix of clinics, hospitals, and specialty service providers is in short sup- ply. The few services that exist are mostly dedicated to elite use. Emergency responses are usually directed toward regime stabil- ity, and often involve the military and police rather than public health. Because overall elite fortunes are also tied to issues such as severe acute respiratory syndrome and avian influenza, there can be public health responses in these areas. Ideally, UN agen- cies, western governments, and nongovernmental organizations (NGOs) such as Oxfam could, and no doubt will, coordinate efforts in future events.

Egypt (Cairo), Kenya (Nairobi), and Indonesia (Jakarta) are acutely dependent on aid from multilateral donors. Organiza- tions such as the International Monetary Fund, the World Trade Organization, the World Bank, and the U.S. Agency for Interna- tional Development provide EHE resources and expertise. They could demand implementation of effective policies as conditions for providing assistance; however, this is unlikely to occur. Many of the best intentioned NGOs and international aid organizations are not sensitive enough to EHE-related needs at their highest administrative levels. At present, donors are unlikely to consider extreme weather before other types of development projects, although growing worldwide awareness of global warming may change this.112

Could international organizations improve on extreme- weather preparedness for developing countries? Will they de- velop a regional or global warning system for EHEs and a strat- egy for more effective city planning before an extreme-weather equivalent of the 2004 tsunami occurs? This is what Anna Tibai- juka, the head of UN Habitat, calls “the biggest problem con- fronting humanity in the 21st Century.”110 The background paper for the 3rd World Urban Forum recommends that cities form new partnership networks with multilateral institutions and donors, national and provincial governments, the private sector, and the urban poor to create resource commitments for planning, implementation, and upgrading of slums.119 As the Hyogo Framework for Action, 2005–2015 notes120

Disaster loss is on the rise with grave consequences for the survival, dignity and livelihood of individuals, particularly the poor . . . [since this loss interacts with] changing demographic and socio-economic conditions, unplanned urbanization . . . environmental degradation, climate variability and change [including El Niño/ La Niña], . . . competition for scarce resources, and the impact of epidemics such as HIV/AIDS.

Events of hydrometeorological origin [i.e., floods, droughts, landslides, tropical cyclones, hurricanes and typhoons] constitute the large majority of disasters.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 625

The Yokohama Strategy . . . [of 1994, addressed] disas- ter risks in the context of sustainable development [and identified gaps and challenges that still remain unfilled and unmet] in five main areas. (a) Governance: orga- nizational, legal and policy frameworks; (b) Risk iden- tification, assessment, monitoring and early warning; (c) knowledge management and education; (d) Reduc- ing underlying risk factors; (e) Preparedness for effective response and recovery.

The World Health Organization seems ideally placed to undertake these actions, especially because it already collaborates effectively with similar organizations such as the UN Habitat and the UN Environmental Program on other issues. Additional plausible coordinators are the World Bank, which has recently augmented its mission to take account of global warming, or the International Monetary Fund. Regardless of what organization takes the lead, the same stumbling blocks are likely to remain: a scarcity of resources and political/ bureaucratic entangle- ments.

International leadership is just one issue. The question also arises as to which level of government is most appropriate for managing EHEs in developing countries? In Europe, citizens look toward national ministries or even the national chief executive. Responsibility in the U.S. falls almost exclusively on the munici- pal level by default. In developing countries, mayors or regional governors are frequently unskilled and lack leadership training. Instead, they are usually people who have performed well in elite patron–client networks and are given their positions as a reward by prominent national politicians. It is these latter politicians that the public will hold responsible. Therefore, one solution is to assign formal authority for EHE response planning to these individuals.

Ideally, such a plan would be technocratic in its creation and supported by sufficient political power and legitimacy to enhance implementation. Matters are frequently complicated, however, as the concept of elitism and the lack of a technocracy in devel- oping countries illustrate. One potential innovation that can partially circumvent uncooperative administrations and supply needed technocracy is use of an Advocacy Coalition. These enti- ties are groups of respected politically independent professionals who work together to identify a problem and to develop solu- tions. Often these groups are composed of weather professionals, government officials, public health personnel, and medical and emergency management experts who advocate for EHE plan- ning.121 Such groups exist in the United States and in Europe. With a legitimacy based on their nonprofit volunteerism, such advocates could collaborate directly with interested municipali- ties and national NGOs.

Factors that correlate highly with injury and death from EHEs in the U.S. are also associated with death from these events in the developing world. These factors include poverty and social isolation, race, age older than 75 years, the percentage of peo- ple living in multiunit dwellings, population density, and high rates of violent crime. Many of these are often combined into the single variable identified as the urban heat island, described previously. Other than poverty, however, many of these factors are absent from the cities under discussion. For example, levels of violent crime are low in Cairo, Jakarta, and Kuala Lumpur. The additional factors relevant to developing countries remain unclear. Ideal research projects designed to answer such ques-

tions will encompass large populations. The relative scarcity of resources and the wider scope of competing needs in developing countries suggest that smaller, less sophisticated evaluations are more relevant.

In contrast, Europeans demonstrate a broader, more inte- grated approach that better links building codes, wetlands/ shoreline protections, landuse regulations, controls over rates of urbanization and deforestation, and control over industry locations. Some extreme-weather events are so new to Europe that the tailoring of responses to a specific event remains insuf- ficient, as illustrated by the lack of sufficient numbers of air- conditioned hospitals and cooling centers. Developing countries are can select between adaptable U.S. and European approaches to create a strategy that meets local needs.

The generally modest external support available to devel- oping nations for solving EHE-related health problems is less problematic than it appears. Developing nations, generally with sophisticated external support, have implemented cost-effective interventions and accomplished major public health successes even under conditions of extreme poverty, weak or nonexistent healthcare infrastructure, and civil war or unrest. For example, as recently as 1988, 125 countries were endemic for polio; however, by the end of 2003, just six countries were reporting polio cases. Efforts were launched in 1974 to kill the black fly that spreads dis- ease and blindness. By 2002, these interventions prevented up to 600,000 cases of blindness.122 Nevertheless, the challenges asso- ciated with EHEs requires a much more rapid response (a matter of days) than that for traditional health problems resulting from infectious diseases.

RECOMMENDATIONS TO PREVENT OR MITIGATE THE HEALTH EFFECTS OF EXCESSIVE HEAT EVENTS

Hospitals/Emergency Departments in the U.S.

EDs Can and Should Be Used to Predict the Healthcare Burden and Epidemiological Consequences of Environmental Disasters

Computerized networks of ED databases should be devel- oped and incorporated into existing reporting mechanisms.57

These networks would report the prevalence of heat-related con- ditions treated in the ED in real time. Although ED data create an accurate picture of the EHEs current impact, they will not pro- vide the 1–5-day lead time for area-wide EHE preparations and response. Air mass monitoring systems are necessary to supply this advanced warning.123

Individual Hospitals and Medical Systems Should Arrange for Extra Medical Staffing in Support of Emergency Department Services

The rising demand for medical care associated with EHEs will place additional burdens on individual hospitals and medical sys- tems to increase emergency services capacity.1 Increased staffing of EDs, individual hospitals, and hospital systems in response to a forecasted EHE can prevent the emergency medical system from becoming overwhelmed. This enhances the opportunity to avert negative outcomes or at least address them at an earlier and less severe stage.1 EDs must be aware that many of the addi- tional support staff needed in emergency situations have multiple commitments. They are often members of the local, state, and

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

626 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

federal emergency teams, as well as the National Guard and other organizations.

Local Emergency Medical Systems Need Fiscal and Institutional Support at the Regional, State, and Federal Levels

It is likely that the onset of EHE conditions will result in significant increases in demand for emergency care in the form of emergency medical services calls and visits to EDs. Exist- ing local and state resources supported by mutual aid agree- ments, state emergency medical assistance compacts, as well as the resources of the federal government may provide the neces- sary additional medical support that individual hospitals or hos- pital systems require. Evaluation of the applicability and avail- ability of these resources must occur before the onset of EHE conditions.1

Studies published by the U.S. Institute of Medicine empha- size that although demands on emergency and trauma care have grown dramatically, the capacity of the system has not kept pace.124 Meeting the need for medical care in the face of increas- ing patient volume and limited resources remains challenging. Balancing existing system utilization with the huge demands that EHEs could place on EDs requires extensive area-wide planning coupled with an effective integration of existing response assets and systems.

Hospitals Should Employ a Full- or Part-time/Shared Medical Disaster Manager

The increasing probability of EHEs as well as other disaster events makes such positions a wise investment. In some cases, federal and state funds may be available to support hiring emer- gency managers. For smaller facilities, a single individual may be employed by a number of institutions or by a hospital trade group. These disaster experts can conduct hazard vulnerabil- ity analyses for the areas surrounding local facilities as well as perform evidence-based disaster research and analysis.

Local and Regional Governments

Urban Areas at Substantial Risk for Extreme Heat Events Should Develop Listings of Vulnerable Individuals, Especially Those Older than 60 Years of Age

Detailed registries are difficult to develop and can be costly. In their absence, however, the municipal government’s ability to conduct effective outreach to those vulnerable to EHEs is extremely limited. Information for compiling such lists can be obtained from religious and social organizations. Another option is requiring owners of dwellings in which the elderly are resi- dents to report essential data to local health and welfare agencies. City departments of urban and regional planning and Adminis- tration on Aging offices can also be very helpful in supporting the development of registries.

Research Exploring the Effectiveness of Extreme Heat Event Preparedness and Response Strategies as well as the Association Between Extreme Heat Events and Morbidity Is Necessary to Aid in the Development of Evidenced-based Plans

McGeehin and Mirabelli have discussed the need for broad- based research into EHEs, including further investigations of critical weather parameters as they relate to EHEs and urban design.58

International Endeavors

An International Institute Is Needed to Coordinate the Various Aspects of Extreme Heat Wave Preparedness and Response, Focusing on Emergency Management and Planning Efforts Related to Weather, Public Health, Medical, Structural, Environmental, and Urban and Regional Planning Issues

A number of federal, state, and municipal organizations as well as universities have served as focal points for research and development activities for EHEs. To continue and intensify the growing interest in EHE response planning and the related areas of medical, weather, and emergency management research, a single-purpose institute dedicated to the study of EHEs should be developed.

Proposed Solutions to the Problems of Extreme Heat Events in the Cities of the Developing World

What should response organizations, administrators, and politicians do about EHEs in developing countries? The U.S. and, to some extent, European experiences yield a long list of EHE preparedness and response items for developing countries to consider. In practical terms, these are more likely to be addressed by international NGOs and be coordinated by entities such as the World Health Organization.

Weather Prediction Based on Air Mass Modeling from Entities Such as the National Oceanic and Atmospheric Administration and the European Union Should be Used as the Basis for Alert/Watch/Warning Systems

These air mass monitoring systems can provide an additional 1–5 days of early warning regarding EHEs. They are already in place in a growing number of urban areas across the world and are being instituted in the U.S. by the National Oceanic and Atmospheric Administration’s National Weather Service.

Education and Communication with the Residents of Urban Areas Is Necessary. These Efforts Must Anticipate Cultural and Religious Diversity and be Externally Funded

EHE preparedness and response efforts must be extremely sensitive to local religious and cultural practices. As in most pro- posals discussed in this section, external funding from NGOs and international aid coordinated by organizations such as the World Health Organization are required. These efforts may be compli- cated by local practices and the general lack of governmental and political support.

Targeted Funding for Specific Public Health Measures Is a Practical Necessity

The underfunded status of emergency medical and public health systems in most developing nations is likely to continue, as is the lack of governmental resources and political determination to address EHEs. Therefore, financial support for specific public health measures is required. Otherwise, assistance will continue to come from international NGOs and other external sources of international aid.

Utilization of Police and Military to Improve Disaster Response

This is a resource and political choice issue that can best be addressed by the developing nations’ themselves. External

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 627

pressures to appropriately utilize these resources can be help- ful. However, addressing EHE-related issues without confronting larger medical and public health systems shortfalls may not be realistic.

Implement More Effective Urban and Regional Planning in Cities that Feature Extensive Areas of Slum Housing

Unless these EHE infrastructure issues can be addressed along with other medical and public health systems needs, there is little chance of addressing existing resource shortfalls.

Distribution of Bottled Water and Portable Water Systems Can be Effective Short-term Solutions to Extreme Heat Events

Although initially effective, these interventions do not address the long-term needs of developing nations regarding adequate supplies of potable water. It highlights the traditional approach of meeting short-term disaster needs as opposed to addressing the much larger, long-term resource and political/ governmental problems that are the root cause of most medical and public health problems.

FUTURE RESEARCH DIRECTIONS FOR EXTREME HEAT EVENT PREPAREDNESS AND RESPONSE

The following examples illustrate the types of research endeav- ors that could improve the effectiveness of EHE preparedness and response plans. These areas of research have been generally underinvestigated by public health and medical planners. The research selections that appear reflect significant aspects of the urban reality from which EHEs have developed. Unlike most public health and medical research, it is not primarily empiri- cally derived. Rather, it is process driven, and recognizes that good plans will not automatically result in good outcomes unless time and thought are spent on the implementation aspects. In fact, a whole research literature in political science and public adminis- tration has been dedicated to how laws, plans, and concepts are implemented.125

Experts in the field of urban and regional planning have spent decades studying how plans are implemented into effective methods of solving specific problems. Implicit in this research is how the public, the urban governance leadership (public and private), and other groups are brought together to turn plans into reality. The types of responses that various U.S. or European cities crafted cannot be directly transferred to cities in the developing world without some modification. There is, however, much to be learned from most aspects of these strategies. Although it will not be possible to implement some systems that depend on effective and relatively well-equipped bureaucracies, potentially useful technologies can be applied successfully if the particular political, ethnic, sociocultural, and other urban-based realities are understood. The projects can be grouped in eight categories.

1) Remote sensing.126 Such techniques could have an impor- tant role in refining and quantifying the widely accepted effects of urban heat islands. These procedures are not referenced in any of the plans reviewed in this chapter, although the weather prediction and heat island literatures are becoming more incorporated into related areas of investi- gation. For example, the U.S. National Aeronautics and Space Administration Landsat space imaging shows a clear corre-

lation between demographic changes and physical landscape changes. Future international coordination efforts, taken in conjunction with such prediction innovations as air mass modeling, can be expected to predict localized risks for heat- related morbidity and mortality.

2) The effects of information on participants in the problem- solving process.127 Information has a transformative func- tion that can alter perceptions of those participating in urban issues. Consensus building requires broad informa- tion access. Urban and regional planning research constantly recognizes the value of information and that it is part of the problem-solving process. Simply stated, consensus building with accurate and pertinent information among agencies or constituents cannot be undervalued. A good example is the widespread and effective public information campaign that educated citizens about the buddy system in Philadelphia. This plan had a positive and life-saving impact as people evaluated each other’s state of health during EHE events.

3) Rational arguments and irrational audiences.128 When institu- tions are involved in technological issues such as power plant locations, they often confront hostile and apparently irra- tional audiences. Sometimes frustrated technocrats will turn to behavioral science for solutions. Whether or not the audi- ence is really “irrational” depends on understanding their perceptions of the problem. In the developing world, belief systems and social and class orientations may dissuade peo- ple from what appears to be the most prudent technological solutions. An example would be resistance to the brief use of large, air-conditioned spaces in tourist hotels to prevent heat-related deaths.

4) Consensus building: mistaken interests in the discourse model of planning.129 Researchers have demonstrated that mistaken judgments about individual and group self-interest could be addressed by structured rational discourse; however, more work is needed. In the complicated ethnic and political milieus of modern urban areas, mistaken motives still lead to failure. For example, many urban EHE plans include cooling centers that require individuals to use public transporta- tion, travel through dangerous areas, and overcome personal mobility issues. In this case, the planners either need to mod- ify their consensus to ensure the public can easily access the sites or abandon the concept.

5) Planning through consensus building.130 In an investigation of eight case studies, consensus building with stakeholders was an effective planning strategy. This research proved that those whose interests are directly impacted will actively par- ticipate if they are approached. Borrowed from the urban development experience, the author followed eight projects and found that arriving at a consensus involving the major stakeholders is normally required if a plan is to be effectively implemented. In U.S. cities, this means that EHE interven- tions such as cooling centers must be tested among urban elderly poor, as well as with the utility companies and major government agencies. In developing nations, it means arriv- ing at consensus will require participation by NGOs, local and/or world advocacy groups, impoverished citizens, and the local elite, whose acquiescence is required.

6) Discussion and rationality.131 Researchers found some indica- tion that discussion enhanced rationality among the partici- pants; however, increased rationality among the participants may not be sufficient to foster the outcomes that are desired by planners.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

628 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

7) Confronting a fractured public interest.132 The researcher reviewed models addressing urban ethnic conflict in Johan- nesburg, Belfast, and Jerusalem. In Johannesburg, govern- ment planners confronted racism, poverty, and vast cultural gaps as they improved services to the squatter-shack areas in Alexandria Township. They concentrated their efforts on addressing the poor quality of water, sanitation, hous- ing, and public health services. The planners balanced the need to reduce apartheid patterns and provide access to the more affluent white areas for blacks while serving as medi- ators for the interest of all groups. Addressing urban and regional planning problems required that political, social, and racial problems were also resolved. This type of organi- zational research on the political and ethnic environments in the developing world also applies to the increasingly diverse urban and rural areas in the U.S.

8) Planning and chaos theory.133 Some of the ideas about ran- domness or chaos emerging in various fields of the nat- ural, social, and applied sciences have major implications for EHE planning. Research suggests that the world may be both easier and more difficult to understand than previously believed and that untidy cities may not be as dysfunctional as is assumed. This is another way of saying that just because an order is not perceived, does not mean there is none. This knowledge is very important in the cultures of urban areas in developing nations as emergency management professionals work with supportive local networks to advocate for imple- mentation of EHE plans.

In summary, the technologic aspects of EHE planning are far less complicated than are the influences of political, sociocultural, and ethnic issues. Only through the application of coordinated transdisciplinary approaches to research will the morbidity and mortality of EHEs be reduced.

REFERENCES

1. United States Environmental Protection Agency. Excessive Heat Events Guidebook. EPA contract 430-B-06–005, June 2006, Washington, DC: Environmental Protection Agency.

2. Menne B, with contributions from Wolf T. In: Kirch W, Menne B, Bertollini R, eds. Extreme Weather Events and Public Health Responses. Berlin/Heidelberg: Springer-Verlag; 2005:xxviii.

3. Kobasa P, ed. Heat Waves: Library of Natural Disasters. (2008). Chicago: World Books, Inc; 2008:10.

3a. The National Weather Service. Available at: http://www.wrh .noaa.gov/sto/heatwave.php. Accessed December 11, 2008.

4. Kinver M. Water Policy “Fails world’s poor,” BBC One Minute World News. Updated March 9, 2006, Based on the UN World Water Development Report, Outlined at the World Water Forum in Mexico, 2006, retrieved on March 24, 2007 from http://news.bbc.co.uk/2/science/nature/4787758.stm.

5. U.S. Centers for Disease Control and Prevention. Heat Ill- nesses and Death, CDC Media Relations: MMWR News Syn- opsis for July 4, 2003. Available at: www.cdc.gov/od/oc/media/ mmwrnews/n030704.htm. Accessed December 11, 2008.

6. U.S. Centers for Disease Control and Prevention. Heat Wave Emergency Response: A Review by the National Center for Envi- ronmental Health. Atlanta: The National Center for Environ- mental Health, Division of Environmental Hazards and Health Effects; 1999.

7. National Oceanic and Atmospheric Administration. July 1995 Heat Wave: National Disaster Survey Report. Washington, DC: U.S. Department of Commerce, National Weather Service; 1995;viii.:17–52.

8. Kilbourne E.M. Heat waves and hot environments. In: Noji EJ, ed. The Public Health Consequences of Disasters. New York: Oxford University Press; 1997:249.

9. Ellis FP. Mortality from heat illness and heat-aggregated illness in the United States. Environ Res. 1972;5:1–58.

10. Ellis FP, Prince HP, Lovan G, Whitington RM. Mortality and morbidity in Birmingham during the 1976 heatwave. Q J Med. 1980;49:1–8.

11. Heat-related mortality – United States, 1997. MMWR. 1998; 47:3–5.

12. Bhattacharya S. “European heatwave caused 35,000 deaths.” Available at: http://www.newscientist.com/article/dn4259- european-heatwave-caused-35000-deaths.html. Accessed De- cember 11, 2008.

13. Doyle A. 160,000 said dying yearly from global warm- ing. Planet Ark. Available at: http://www.planetark.org/ dailynewsstory.cfm/newsid/22420/story.htm. Accessed Decem- ber 11, 2008.

14. Kalkstein LS, Greene JS. An evaluation of climate/mortality relationships in large U.S. cities and the possible impacts of climate change. Environ Health Perspect. 1997;105(1):84– 93.

15. National Center for Health Statistics. Mortality Public Use Com- puter Data Tapes for the Years 1979–1991. Hyattsville, MD: National Center of Health Statistics, 1994.

16. Kalkstein LS. Climate and human mortality: relationship and mitigating measures. Adv Bioclimatol. 1997;5:161–177.

17. Davis RE, Knappenberger PC, Michaels PJ, Novicoff WM. Changing heat-related mortality in the United States. Environ Health Perspect. 2003;11:12–18.

18. Henschel A, Burton LL, Margolies L, Smith JE. An analysis of the heat deaths in St. Louis during July 1966. Am J Public Health. 1969;59:2232–2242.

19. Sheridan SC, Dolney TJ. Heat, mortality, and level of urban- ization: measuring vulnerability across Ohio, USA. Clim Res. 2003;24:255–266.

20. Cohen JC, Veysseire JM, Bessemoulin P. Bio-climatological aspects of summer over France. In: Kirch W, Menne B, Bertollini R, eds. Extreme Weather Events and Public Health Responses. Berlin/Heidelberg: Springer-Verlag; 2005:34.

21. Meehl GA, Tebaldi C. “More intense, more frequent, and longer lasting heat waves in the 21st Century.” Science. 2004;305:994– 997.

22. Kovats S, Haines A. The potential health impacts of climate change: an overview. Med War. 1995;11:168–178.

23. Oreskes N. Beyond the ivory tower: the scientific consensus on climate change. Science. 2004;306:1686.

24. Outcomes of the 2005 World Summit, TDN International Edition. Daily Travel & Tourism Newsletter for the Interna- tional Traveler since 1999. Retrieved on August 17, 2008, from http://traveldailynews.com/stiles print.asp?central id (943 See also UN News Centre, retrieved on September 17, 2005 from http://www.UN.org/News.

25. Kilbourne EM. Heat waves. In: Gregg MB, ed. The Public Health Consequences of Disasters. Washington, DC: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention; 1989:51

26. Kilbourne EM, Choi K, Jones TS, Thacker SB. Risk factors for heat stroke. A case-control study. JAMA. 1982;247:3332– 3336.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 629

27. Excessive Heat Events Guidebook, 2006. Washington, DC: Envi- ronmental Protection Agency. Available at: http://www.epa.gov/ heatisland/about/pdf/EHEguide final.pdf. Accessed December 11, 2008.

28. Kalkstein LS, Davis RE. Weather and human mortality: an evo- lution of demographic and interregional responses in the United States. Ann Assoc Am Geograph. 1989;79:44–64.

28a. Sheridan SC, Kalkstein LS. Progress in heat watch-warning sys- tem technology. Bulle Am Meteorolog Soc. 1989;85:1931–1941.

29. Greene JS, Kalkstein LS. Quantitative analysis of summer air masses in the eastern United States and an application to human mortality. Clim Res. 1996;VOL:43–53.

30. Stein, Kaplan. Chicago Tribune. 1995;July 26:1, 13, Sec. 1. 31. Kunkel KE, Chagnon SA, Reinke BC, Arritt RW. The July 1995

heat wave in the midwest: a climatic perspective and critical weather factors. Bull Am Meteorolog Soc. 1996;7:1507–1517.

32. Kilbourne EM. Heat-related illness: current status of prevention efforts. Am J Prevent Med. 2002;22:328–329.

33. Vlum LN, Bresolin LB, Williams MA for the council on Scientific Affairs: Heat-related illness during extreme weather emergen- cies. JAMA. 1998;279:1514.

34. Hoppe J, Sinert R, Kunihiro A, Foster J. Heat exhaustion and heat stroke. The continually updated medical site. Available at: http://www.emedicine.com/emerg/topic236.htm. Accessed December 11, 2008.

35. Varghese GM, John G, Thomas K, Abraham OC, Mathai D. Predictors of organ dysfunction in heatstroke. Emerg Med J. 2005;22:185–187.

36. http://emedicine.medscape.com/article/770413-treatment. Accessed December 20, 2008.

37. http://usariem.army.mil/heatill/histroke.htm. Accessed Decem- ber 20, 2008.

38. Bouchama A, Knochel JP. Heat stroke. N Engl J Med. 2002; 346:1978–1988.

39. Marzuk P, Tardiff K, Leon A, Hirsch C, et al. Ambient tem- perature and mortality from unintentional cocaine Overdoses. JAMA. 1998;279:1795–1800.

40. Institute of Medicine. Emergency health system unprepared for disasters, June 20 New Release by the Center for Infectious Dis- ease Research and Policy, Academic Health Center – University of Minnesota. CIDRAP News. p. 1.

40a. IOM’s “Future of Emergency Care” Available at: http:// www.iom.edu/CMS/3809/16107/35007.aspx.

41. Semenza JC, McCullough JE, Flanders WD, McGeehin MA, Lumpkin JR. Excess hospital admissions during the July 1995 heat wave in Chicago. Am J Prevent Med. 1999;16:269–277.

42. Wainwright SH, Buchanan SD, Mainzer HM, Parrish RG, Sinks TH. Cardiovascular mortality – the hidden peril of heat waves. Prehosp Disaster Med. 1999;14:222–231.

43. Patzak, Davido A, Dart T, et al. Risk factors for heat-related death during the August 2003 heat wave in Paris, France, in patients evaluated at the emergency department of the Hos- pital European ‘Georges Pompidou.” EMJ 23 (2006): 515– 518.

44. Johnson H, Kovats R, McGregor G, et al. The impact of the 2003 heat wave mortality and hospital admissions in England. Health Stat Q. 2005;25:6.

45. Faunt JD, Wilkinson TJ, Henschke P, Webber M, Penhall RK. The effete in the heat: heat-related hospital presentations during a ten day heat wave. Aust NZ J Med. 1995;25:117.

46. Rajpal RC, Weisskopf MG, Rumm PD, et al. Wisconsin, July 1999 heat wave: an epidemiologic assessment. WJM. 2000;99: 41–44.

47. The Joint Commission on Accreditation of Healthcare Orga- nizations and Joint Commission Resources. Surge Hospitals: Providing Safe Care in Emergencies, (2006), iv.

48. Waugh WL. Living with Hazards, Dealing with Disasters: An Introduction to Emergency Management. Armonk, NY: M. E. Sharpe; 2000:3, 4.

49. Landesman LY. Public Health Management of Disasters: The Prac- tical Guide. Washington, DC: American Public Health Associa- tion; 2001:35, 6.

50. CNA Medical Surge Capacity and Capability: A Management System for Integrating Medical and Health Resources During Large-Scale Emergencies. This document was prepared under Contract Number 233–03-0028 for the Department of Health and Human Services. CAN Corp.; 2004:2–8.

51. The National Foundation for Trauma Care. United States Trauma Care Proposals for a Terrorist Attack in the Commu- nity: The Study of the Impact of a Terrorist Attack on Individ- ual Trauma Centers. Atlanta: Centers for Disease Control and Prevention; 2006:3–5.

52. Delinger RF, Gonzenbach K. The two-hat syndrome: determin- ing response capabilities and mutual aid limitations. In: Kayyem JN, Pangi RL, eds. First to Arrive: State and Local Responses to Terrorism. Cambridge, MA: The MIT Press; 2003:193–205.

53. Patz J, McGeehin M, Bernard S, et al. The potential health impacts of climate variability and change for the United States: executive summary of the report of the health sector of the U.S. National Assessment. Environ Health Perspect. 2000;108:367– 376.

54. Patz J, Engleberg D, Last JJ. The effects of changing weather on public health. Ann Rev Public Health. 2000;21:271.

55. Semenza JC, Rubin K, Selanikio J, et al. Heat-related deaths during the July 1995 heat wave in Chicago. N Engl J Med. 1996;335:84–90.

56. United States Environmental Protection Agency. Excessive Heat Events Guidebook, 2006. Citing the American Medical Associ- ation Council on Scientific Affairs, 1997. Heat-Related Illness During Extreme Weather Emergencies. Report 10 of the Coun- cil on scientific Affairs (A-97).

57. Rydman RJ, Rumoro DP, Silva JC, Hogan TM, Kampe LM. The rate and risk of heat-related illness in hospital emergency departments during the 1995 Chicago heat disaster. J Med Syst. 1999;23: 41,53–54.

58. McGeehin M, Mirabelli M. The potential impacts of cli- mate variability and change on temperature-related morbid- ity and mortality in the United States. Environ Health Perspect. 2001;109(Suppl 2):185–189.

59. Donoghue ER, Graham MA, Jentzen JM, Lifschultz BD, Luke JL, Mirchandani HG. Criteria for the diagnosis of heat-related deaths. National Association of Medical Examiners. Position paper. National Association of Medical Examiners Ad Hoc Com- mittee on the Definition of Heat-related Fatalities. Am J Forensic Med Pathol. 1997;18:11–14.

60. Naughton MP, Henderson A, Mirabelli MC, et al. Heat-related mortality during 1999 heat wave in Chicago. Am J Prevent Med. 2002;22:221–227.

61. Adams BE, Manoguerra AS, Lilja GP. Long RS, Ruiz RW. Heatstroke associated with medications having anticholinergic effects. Minnesota Med. 1977;60:103–106.

62. Heat-wave-related mortality – Milwaukee, Wisconsin, July 1995 JAMA.1996;276:275.

63. Whitman S, Good G, Donoghue ER, Benbows N, Shou W, Mou S. Mortality in Chicago attributed to the July 1995 heat wave. Am J Public Health, Public Health Briefs. 1997;9:87.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

630 ■ CA R L AD R I A N O P O L I, PAU L H. BR I E T Z K E, IRV I N G JACO B Y, A N D JE RO M E H. LI B B Y

64. Heat-related illnesses, deaths, and risk factors – Cincinnati and Dayton, Ohio, 1999, and the United States 1979–1999. MMWR.2000;VOL:473.

65. Lee DH. Seventy-five years of searching for a heat index. Environ Res. 1980;VOL:331–356.

66. Poulton T J, Walker RA. Helicopter cooling of heatstroke victims. Aviation Space Environ Med. 1987;VOL: 358.

67. Vendentorren S, Empereur-Bissonnet P. Health impacts of the 2003 heat-wave in France. In: Kirch W, Menne B, Bertollini R, eds. Extreme Weather Events and Public Health Responses. Berlin/Heidelberg: Springer-Verlag; 2005:82.

68. Michelozzi P, de’Donato F, Bisanti L, et al. Heat waves in Italy: cause specific mortality and the role of educational level and socio-economic conditions. In: Kirch W, Menne B, Bertollini R, eds. Extreme Weather Events and Public Health Responses. Berlin/Heidelberg: Springer-Verlag; 2005:126.

69. Conti S, Meli P, Minelli G, et al. Epidemiological study of mor- tality during summer 2003 in Italian regional capitals: results of a rapid survey In: Kirch W, Menne B, Bertollini R, eds. Extreme Weather Events and Public Health Responses. Berlin/Heidelberg: Springer-Verlag; 2005:119.

70. Department of Emergency Management City of New York. Heat Emergency Preparedness Guide. 2002.

71. Keating WR, Donaldson CG, Cordioli E, et al. Heat related mortality in warm and cold regions of Europe: observational study. BMJ. 2000;321:670–673.

72. Di Maio DJ, Di Maio VJM. Forensic Pathology. London: CRC Press;1992:379.

73. Chestnut LG, Breffle WS, Smith JB, Kalkstein LS. Analysis of dif- ferences in hot-weather-related mortality across U.S. metropoli- tan areas. Environ Sci Pol. 1998;1:59.

74. Kalkstein LS. A new approach to evaluate the impact of climate on human mortality. Environ Health Perspect. 1991;96:145.

75. Ellis FP. Mortality from heat illness and heat-aggravated illness in the United States. Environ Res. 1978;5:1–58.

75a. Ellis EP, Nelson F, Pincus L. Mortality in the elderly in the heat wave in New York City July 1972 and August and September, 1973. Environ Res. 1975;10:1–13.

76. Centers for Disease Control. Heat-related mortality – Chicago, July 1995. MMWR. 1995;44:577–579.

77. Lyster WR. Deaths in summer. Lancet. 1976;2:469. (Letter) 78. Oeschli FW, Buechley WB. Excess mortality associated with

three Los Angeles September hot spells. Environ Res. 1970;3:277– 284.

79. Okle TR. City size and the urban heat island. Atmosphere Envi- ron. 1972;7:769–779.

79a. Akbari H, Rosenfeld A. Cooling our Communities: A Guidebook on Tree Planting and Light-Colored Surfaces, Washington, DC: US Environmental Protection Agency; 2003.

80. U.S. Environmental Protection Agency. Climate Change – Health and Environmental Effects. Available at: http://www.epa .gov/climatechange/effects/health.html. Accessed December 12, 2008.

81. Koppe C, Kovats S, Jendritzky G Menne B. Heat-Waves: Risks and Responses. Geneva: World Health Organization; 2004.

82. Klinenberg E. Heat Wave: A Social Autopsy of Disaster in Chicago. Chicago: University of Chicago Press; 2002.

83. Kovats S, Haines A. The potential health impacts of climate change: An overview.” Med War, 11;1995:168–178.

84. Kovats RS, Jendritzky G. Heat-waves and human health. In: Menne B, Ebi KL, eds. Climate Change and Adaptation Strategies in Human Health. Darmstadt: Steinkopf Verlag; 2004:79–82.

85. Givoni B. Design for climate in hot, dry, cities. In: Oke TR, ed. Urban Climatology and its Applications with Special Regard to

Tropical Areas. Proceedings of the Technical Conference, Mex- ico D.E. 26–30 November 1984. Geneva: World Meteorological Organization (WMO No. 652); 1986:87–513.

86. Givoni B, et al. Outdoor comfort research issues. Energy Build- ings. 35;2003:77–86.

87. Rogot E, Sorlie P, Backlund E. Air-conditioning and mortality in hot weather. Am J Epidemiol. 136;1992:106.

88. Bernard SMJ, McGeehin MA. Municipal heat wave response plans in practice. Am J Public Health. 94;2004:1520–1522.

89. McGowan KJ, ed. Terrorism and Disaster Management: Prepar- ing Healthcare Leaders for the New Reality. ACHE Manage- ment Series. Chicago: Health Administration Press; 2004:129– 130.

90. Adrianopoli C, Culhane P. Heat Waves and Heat Response Plan- ning in American Cities. Presented at the Annual Meeting of the Midwest Political Science Association, Chicago, The Palmer House, April 7, 2005.

91. Khogali M. Heat illness alert program. Practical implications for management and prevention. Ann NY Acad Sci. 1997;813:530– 532.

92. Palecki MA, Chagnon SA, Kunkel KE. The nature and impacts of the July 1999 heat wave in the Midwestern United States. Bull Am Meteorolog Soc. 2001;82:1353–1367.

93. National Oceanic and Atmospheric Administration 96–21. Many of the 1995 Heat Wave Deaths Were Preventable Accord- ing to NOAA Report. Available at: http://www.publicaffairs .noaa.gov/pr96/apr96/noaa96-21.html. December 13, 2008.

94. CDC. Heat-related illnesses and deaths – United States, 1994– 1995. MMWR. 1995;VOL:465–468.

95. CDC. Heat-wave related mortality – Milwaukee, Wisconsin, July 1995. MMWR. 1996;45:505–507.

96. Kellerman AL, Todd KH. Killing heat. N Engl J Med. 1996;335: 126–127. (Letter)

97. Dematte JE, O’Mara K, Buescher RW, et al. Morbidity and mor- tality associated with the July 1980 heat wave in St. Louis and Kansas City, MO. JAMA. 1982;247:327–331.

98. The CDC Extreme Heat Bibliography retrieved from the Extreme Heat Site, on March 15, 2008 from http://emergency .cdc.gov/disasters/extremeheat/bibliography.asps.

99. Kalkstein L, Jamason PF, Greene JS, Libby J, Robinson L. The Philadelphia hot weather-health watch. Warning system: devel- opment and application. Bull Am Meteorolog Soc. 1996;7:1519– 1528.

100. Kaiser R, Le Terte A, Schwartz J, CA, et al. The effect of the 1995 heat wave in Chicago on all-cause and cause-specific mortality. Am J Public Health. 2007;97:158–162.

101. Spielman F, Mitchell MA. City ignored emergency plan. Chicago Sun Times. 1995;July 18:1, 6.

102. Neal S. Daley’s leadership wilted in heat crises. Chicago Sun Times. 1995;July 25:25.

103. Mitchell MA. Daley details heat emergency plan: call for alerts, outreach to elderly. Chicago Sun Times. 1995;July 21:3.

104. Mooney C. Some like it hot. Mother Jones. 2005;May/June:42. 105. World Health Organization. Working Paper of the Fourth Min-

isterial Conference on Environment and Health, Budapest, Hun- gary, 23–35 June 2004. Geneva: World Health Organization Europe; 2004

105a. Kirch W, Menne B, Bertolini R, eds. Extreme Weather Events and Public Health Responses. Berlin/Heidelberg: Springer-Verlag; 2005.

105b. Menne B, Ebi KL, eds. Climate Change and Adaption Strategies for Human Health. Berlin/Heidelberg: Springer-Verlag; 2006.

106. Editorial comments. The Washington Post Weekly. 206; July 10– 16:26.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

EXT R E M E HE AT EV E N TS ■ 631

107. Centers for Disease Control and Prevention. Extreme Heat: A Prevention Guide to Promote Your Personal Health and Safety. Available at: http://www.bt.cdc.gov/disasters/extremeheat/heat guide.asp. Accessed December 13, 2008.

108. American Medical Association Council on Scientific Affairs. Heat-Related Illness During Extreme Weather Emergencies. Report 10 of the Council on Scientific Affairs (A-97). Presented at the 1997 AMA Annual Meeting.

109. Madon S, Sahay S. ICTs and cities in the developing world: A network of flows. Inform Technol People. 2001;14(3):273–286.

110. Ajayi, Jare. Development: From Slums to Sustainability. (June 16, 2003) From the IPS (Inter Press Service) for Journalism and communications for Global Change. Retrieved on January 13, 2007 from http://ipsnews.net/news.asp?idnews/33650.

111. Neuwirth R. Shadow Cities: A Billion Squatters. A New Urban World. New York/London: Routledge; 2006:9.

112. World Bank. Poverty and climate change: Reducing the Vulnerabil- ity of the Poor Through Adaption. Available at: http://siteresources .worldbank.org/INTCC/817372-1115381292846/20480614/ PovertyAndClimateChangePresentation2003.pdf. Accessed December 13, 2008.

113. World Urban Forum III. Our Future: Sustainable Cities. UN Habi- tat. (Background Paper);(2006).

114. Brietzke P. Urban development and human development. Indi- ana Law Rev. 1992;25:741–798.

115. Perkins S. Dead heat: the health consequences of global warm- ing could be many. Science News. 2004;166(1):10. Available at: http://www.phschool.com/science/science news/articles/ dead heat.html. Accessed December 13, 2008.

116. Valente M. 2006. Open Sewers a Health and Environmental Risk. From the IPS (Inter Press Service) for Journalism and commu- nications for Global Change. Retrieved on Jan. 13 January 2007 from http://ipsnews.net/news.asp?idnews/33012.

117. Brietzke P. The politics of legal reform. Washington Univ. Global Studies Law Rev. 2004;3(1):1–47, at 4–12. Id. at 4–5, 6–9, 11.

118. Brietzke P. Democratization and . . . administrative law. Okla- homa Law Rev.YEAR; 52(2023):1–47, at 43–47, passim.

119. A Brisbane – Building Research & Information, 2006 – informaworld.com . . . World Urban Forum ‘Our Future: Sus- tainable Cities – Turning Ideas into Action’ Vancouver, Canada 19–23 June 2006 wb: http://hq.unhabitat.org/wuf/2006

retrieved August 6 and 13, 2008 from www.unhabitat.org/ downloads/docs/GRHS.2007.7.pdf

120. The Hyogo Framework for Action 2005–2015: Building the Resilience of Nations and Communities to Disasters. Hyogo: International Strategy for Disaster Reduction Available at: www.unisdr.org/wcdr/intergover/official-doc/L-docs/Hyogo- framework-for-action-english.pdf. Accessed December 13, 2008.

121. Sabatier PA, Jenkins-Smith HC. The advocacy coalition frame- work: an assessment. In: Sabatier PA, ed. Theories of the Pol- icy Process. University of California, Davis: Westview Press; 1999:117–166.

122. Jamison D, Breman J, Measham A, et al. Priorities in Health: The Disease Control Priorities Project. Washington, DC: The World Bank; 2006: 27–31.

123. Sheridan SC, Kalkstein LS. Progress in heat watch-warning sys- tem technology. Bull Am Meteorolog Soc. 2004;85:1931–1941.

124. Committee on the Future of Emergency Care in the United States Health System, Institute of Medicine of the National Academies. Hospital-Based Emergency Care: At the Breaking Point. Washing- ton, DC: National Academies Press; 2006:1.

125. Hill M, Hupe P. Implementing Public Policy: Governance in The- ory and in Practice. London: Sage Publications; 2002:1–40.

126. Ryznar Wagner, 2001, “Using Remotely Senses Imagery to Detect Urban Change: Viewing Detroit from Space”, JAPA, vol. 76, (2001): 327.

127. Hanna KS. The paradox of participation and the hidden role of information. JAPA. 2000;66:398.

128. Kartz JD. Rational arguments and irrational audiences. JAPA. 1989;55: 445.

129. Taylor NT. Mistaken interests and the discourse model of plan- ning. JAPA. 1998;64:64–75.

130. Innes JE. Planning through consensus building: a new view of the comprehensive planning ideal. JAPA. 1996;62:460–472.

131. Wilson RW, Payne M, Smith E. Does discussion enhance ratio- nality? A report from transportation planning practice. JAPA. 2003;69:354–367.

132. Bollens SA. “Urban planning and intergroup conflict: con- fronting a fractured public interest. JAPA. 2002;68:22–42.

133. Cartwright TJ. Planning theory and chaos. JAPA. 1991;57:44– 56.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

Volcanoes

Peter J. Baxter

OVERVIEW

Introduction

An understanding of how people are killed and injured in vol- canic eruptions is essential for advancing disaster mitigation measures as well as for managing casualties in actual events (Table 39.1).1 The main aim of disaster medicine is the preven- tion of direct human deaths and injuries and reducing economic losses that can indirectly affect health through increasing poverty and inequality. A surprisingly large number of lethal phenom- ena are associated with eruptions, and health hazards can arise even when volcanoes are in a state of apparent repose. Com- pared with other types of disasters such as floods, wind storms, and earthquakes, volcanic eruptions occur much less frequently. In recent decades there were on average only two–four events worldwide per year that involved fatalities and people living near an active volcano have only a low statistical risk of a serious event happening in their lifetimes.

The reactivation of a volcano in a densely populated area should trigger a state of emergency, like a warning for an approaching hurricane or storm surge. Forecasting an eruption in a state of unrest requires a team of experienced volcanologists equipped with the latest monitoring technology who can make a rapid appraisal based on what is known about the particular volcano and by drawing analogies from similar volcanoes and how they behave. Although most of these potential crises do not lead to a major eruption and states of unrest can remain in place for years, the uncertainty of forecasting eruptive behavior and the possibility that a serious eruption can happen with little to no warning means that evacuation decisions may have to be made very quickly.2 As the development of emergency planning at most volcanoes is still in its infancy, a state of unrest at a vol- cano usually arises in a community with a poor state of disaster preparedness. This chapter describes the key elements of volcanic disaster planning for health sector workers.

Recent Historical Record of Volcanic Eruptions

The most comprehensive database for global volcanic disasters and incidents in the 20th century (Tables 39.2 and 39.3) shows

how the numbers of dead, injured, homeless and evacuated, or otherwise affected persons are attributable to a relatively small number of large events.3 A typical crisis at the end of the cen- tury developed at Guagua Pichincha volcano, which lies close to Quito, the capital city of Ecuador, and the densely populated Inter-Andean valley. Its resumption of activity in 1998 raised widespread concerns about a repetition of its last major eruption in 1660. Heavy ash falls affected Quito’s city center, and lahars (mud flows), triggered by heavy rain on the ash deposits, dev- astated the slopes of the volcano complex, which today are an integral part of the city. In the event, this “worst-case” eruption failed to materialize, and evacuation or other mitigation mea- sures were not needed; however, the uncertainty affected the city population of over 1 million (Table 39.2) and its economy for several years.

The historical record is not a good guide to present-day volcanic risk. By the end of the 20th century approximately 10% of the world’s population was living in areas of active volcanism, with a significant proportion in megacities whose inhabitants were largely oblivious to the hazard, as shown in the example of Quito and the Inter-Andean valley, which is flanked on both sides by active volcanoes.

Vulnerability of Cities to Volcanic Disasters

What could happen to a city or any other inhabited area in the worst-case volcanic crisis was brought to world attention by the cataclysmic explosive eruption of Mount St. Helens in a wilderness area of the Cascade Range of the western United States on May 18, 1980. Despite scientists close monitoring of the volcano for 2 months after a state of unrest began, the devastating blast occurred without any immediate warning. The forest was devastated by a directed lateral blast and pyroclastic surge that destroyed trees as far away as 28 km and killed 58 of the more than 100 people who were in the area at the time. What if, in the place of the 10 million downed trees, there had been buildings and people?4 The world was shocked by the horrific scenes at Nevado del Ruiz volcano, Colombia, in 1984 when 23,000 people died buried in massive volcanic mud flows, known as lahars, which had been triggered by a sudden summit eruption

632 Koenig and Schultz's Disaster Medicine : Comprehensive Principles and Practices, edited by Kristi L. Koenig, and Carl H. Schultz, Cambridge University Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/apus/detail.action?docID=564432. Created from apus on 2018-03-09 07:38:45.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

VO LC A N O E S ■ 633

Table 39.1: Main Injury Agents in Volcanic Eruptions

Hazard Injury Agent Impact

Flow Processes

Pyroclastic flows and surges Heat Skin and inhalation burns

Dynamic pressure Body displacement

Hot particles Asphyxia

Missiles Multiple trauma

Lahars Lateral loading and crushing Multiple trauma

Traumatic asphyxia

Slurry Asphyxia, infected wounds

Lava Radiant heat Skin burns

Fall Processes

Tephra Collapsed roofs Multiple trauma

Roofs with ash infill Asphyxia

Respirable particles Asthma

Crystalline silica Silicosis (chronic effect)

Ballistic projectiles Multiple trauma

Pumice/lava fragments Multiple trauma

Other

Gases Irritant acid gases and aerosols Asthma, bronchitis

Earthquakes Collapsing houses Multiple trauma

that partially melted the glacier and sent torrents of melt water and pyroclastic debris toward the town of Armero. A warning system was in place, but it was inadequate, and there was no evacuation plan.5 The blindness of the responsible authorities in preventing this entirely foreseeable and largely avoidable disaster was an important stimulus for the United Nations declaration of the 1990s as the International Decade for Natural Disaster Reduction, the main goal of which was to transfer technology and know-how to developing countries to enable them to improve their preparedness and mitigation measures for reducing disaster risks.

Mitigation measures need to be directed toward reducing this global failure of sustainability, which is beginning to have huge societal impacts in most countries with active volcanoes. One famous example is Vesuvius, Italy, which encapsulates the sociopolitical challenge of evacuating large populations at risk when the size, timing, and duration of a foreseeable devastating eruption cannot be predicted with any certainty, while the con- sequences of inaction could lead to the loss of thousands of lives.

Table 39.2: Best Estimates of the Human Impacts of 20th Century Volcanic Events3,46

Number of Number of Human Consequence Events People

Killed 260 91,724

Injured 133 16,013

Homeless 81 291,457

Evacuated/affected 248 5,281,906

Even if only a small (or no eruption) occurred, the disruption caused by the threat of a major eruption and the evacuation of over 600,000 people for possibly months would have unprece- dented politicoeconomic implications.

There have been other notable volcanic crises that have left their marks, either because of devastating eruptions or because the threat failed to materialize (Table 39.1). These have included El Chichon, Mexico, in 1984 when more than 2,000 people died in pyroclastic flows, and Mt. Pinatubo, Philippines, in 1991, in which more than 300 people died in the ash fallout that collapsed roofs;6 however, more than 50,000 people were evacuated in time from the paths of lethal pyroclastic flows and surges.2 At some other volcanoes, such as Galeras, Colombia and the city of Pasto (200,000 people) or Tungarahua, Ecuador and the small city of Banõs (25,000 people), crises have continued for some years and up to the time of this writing, with thousands of people enduring elevated levels of risk on a daily basis. Thus, although many eruptions can be small and short-lived or occur in remote places some volcanic crises and eruptions can continue for years and lead to severe socioeconomic consequences. This distinguishes most volcanic eruptions from sudden-onset events such as floods and earthquakes with which they are often misleadingly linked by disaster planners.

CURRENT STATE OF THE AR T

Types of Volcanoes and their Eruptions

A simplified way to approach volcanoes and their hazards is to learn whether they are mainly explosive or effusive (a nonexplo- sive outpouring of fluid lava) in behavior.7 An essential guide to

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

634 ■ PE T E R J. BAXT E R

Table 39.3: Top 10 Volcanic Events in the 20th Century by Impact3,47

Killed Injured Homeless Evacuated/Affected

Rank Event No. People Event No. People Event No. People Event No. People

1 Pelée, 1902 29,000 Nevado del Ruiz, 1985

4,470 Pinatubo, 1991 53,000 Guagua Pichincha, 1999

1,200,400

2 Nevado del Ruiz, 1985

23,080 Awu, 1966 2,000 Kelut, 1919 45,000 Pinatubo, 1991 967,443

3 Santa Maria, 1902 8,750 Ambrym, 1979 1,000 Galunggung, 1982 22,000 Pinatubo, 1992 787,042

4 Kelut, 1919 5,110 Dieng, 1979 1,000 Pinatubo, 1992 15,700 Agung, 1963 332,234

5 Santa Maria, 1929 5,000 Lake Nyos, 1986 845 Tokachi, 1926 15,000 Vesuvius, 1906 100,000

6 Lamington, 1951 2,942 Taal, 1965 785 El Chichón, 1982 15,000 Popocatépetl, 1994 75,000

7 El Chichón, 1982 2,000 El Chichón, 1982 500 Merapi, 1930 13,000 Soufrière Guadeloupe, 1976

73,500

8 Lake Nyos, 1986 1,746 Merapi, 1994 500 Merapi, 1961 8,000 Mayon, 1984 73,000

9 Soufrière St. Vincent, 1902

1,565 Merapi, 1998 314 Soufrière Hills, 1995 7,500 Arenal, 1976 70,000

10 Merapi, 1930 1,369 Vesuvius, 1906 300 Colo (Una Una), 1983 7,101 Galunggung, 1982 62,755

Total 80,562 11,714 201,301 3,741,374

this is the stratigraphic (geological) record, which provides the evidence from past eruptions. Explosive eruptions involve rapid energy changes whereby magma, or molten rock, is so violently expelled from a deeper environment that it turns into pyroclas- tic fragments. The silica-rich composition of the magma lends it a high viscosity and these volcanoes are found in subduction zones, an example of which is the Ring of Fire around the Pacific Ocean. Mount St. Helens belongs to this category. The most hazardous products of explosive eruptions are pyroclastic flows and surges, and tephra (ash and glass) fallout, including ash and pumice blocks, as well as expelled ejecta. Fragmentation may also be produced by steam explosions and the thermal shock when magma comes into contact with water below the earth’s surface or when it erupts into surface water, ice, or wet sediments. Explosive eruptions are often accompanied by small, viscous lava flows.

At the other end of the spectrum, volcanoes with silica-poor rocks mainly occur at midocean ridges, and more commonly have fluid lava eruptions. The most well known volcanoes with effusive eruptions are on the Big Island of Hawaii. Volcanoes of intermediate magmas can have features of both types. Mount Etna, Italy, is an example of a mainly effusive volcano whose hazard is mostly lava flows, but spectacular eruptions in 2001 and 2002 reminded people that it also had explosive potential.

Explosiveness is also governed by the percentage of volatile components, or dissolved gases, in the magma. Viscous magmas are more resistant to deformation from the action of forces and mechanical stress, and so the gases do not so readily escape, and this leads to increased gas pressure. Typically, only low levels of pressure develop in eruptions involving more fluid magmas. The main volatiles are water, carbon dioxide, and sulphur diox- ide. Only a small percentage of emissions consist of other gases and these include hydrogen chloride (hydrochloric acid) and hydrogen sulphide, and trace gases including hydrogen, carbon monoxide, and methane.7 Volcanic gas emissions and gas bursts are human hazards in populated areas.

Location of Volcanoes

Volcanoes are not found everywhere on earth. Volcanoes and their gas emissions were responsible for the formation of the earth’s atmosphere and today plate tectonics and the volcan- ism arising from it are essential for maintaining the atmosphere in a state that enables life to continue on earth. Eruptions and earthquakes are the deleterious events that maintain this plane- tary homeostasis. Eighty percent of volcanoes are in subduction zones at convergent plate boundaries, which include the Ring of Fire around the Pacific Ocean (including the Cascades and Mount St. Helens in the United States), and island arcs, such as Indonesia and the islands of the Caribbean; these are explosive. In contrast, approximately 10% of volcanoes occur at midocean ridges and are effusive.7 There are approximately 700 subaerial volcanoes in the world, of which 170 are currently active.

Scale of Eruptions

The most devastating eruptions are explosive in nature; large eruptions are uncommon in comparison to small events. Vol- canoes that erupt the most frequently are likely to be the best studied, for obvious reasons; scientists regularly monitor a pro- portion of these, especially if they are in densely inhabited areas. All of the most frequently active volcanoes in Italy and Japan are closely monitored, whereas a lack of resources makes it possible to focus attention on only a handful of the dozens of volcanoes in Indonesia. When a volcano that has been dormant for many years shows signs of activity, scientists will be needed to install monitoring devices or intensify current monitoring efforts to keep the volcano under close surveillance. If experts determine that a volcano is moving toward an eruption, they will advise authorities to declare a state of unrest.

The size of an eruption relates to the magnitude of the erup- tion (mass of material discharged, or tephra volume) and the intensity or rate of discharge (mass flow rate). A volcanic explosive

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

VO LC A N O E S ■ 635

index is a measure of both of these parameters.7 The largest erup- tions are in the supervolcano league, such as Yellowstone, whose eruptions are at average intervals of 100,000 years; an event at Yellowstone would impact most of the U.S. and produce a global catastrophe through persistent atmospheric effects lasting sev- eral years after the event. None of these tremendous eruptions have been witnessed in recent history. Eruptions on the scale of Tambora, Indonesia, in 1815, which temporarily lowered global temperatures and triggered crop failures, occur once every 1,000 years on average. A Krakatau-type (Java, 1883) occurs approxi- mately twice every 100 years, and one the size of Mount St. Helens (1980) once every 10 years somewhere in the world. Types of explosive eruptions include the largest and most energetic and destructive – the plinian and subplinian, which have major ash falls and pyroclastic flow activity – to the smaller strombolian and vulcanian, which do not. Plinian eruptions produce high ash and gas-laden atmospheric columns that penetrate into the stratosphere and can interfere with climate and weather on a global scale.8

The main eruptive hazards can be divided into fall and flow processes. Air falls of tephra – airborne fragments of rock and lava of any size or shape expelled during explosive eruptions – are the most common and most hazardous fall processes.

Pyroclastic flows and surges, lahars and debris flows, and lava flows can, for emergency planning purposes, be envisaged by “thinking visually” or intuitively, but their behavior belongs to the world of flow, or fluid dynamics, on which the under- standing of the blood circulation, meteorology, and aeronautics is based. The concepts that describe the motions of fluids – vis- cosity, vorticity, waves, instability, and turbulence – apply to the complex behavior of pyroclastic flows and surges, which move along the ground and over obstacles with features of both gases and liquids. Fluid dynamics underlies the computer modeling of such phenomena, as will be described later.

Human Impacts of Pyroclastic Flows and Surges

Pyroclastic flows and surges are the most hazardous eruptive phenomena. They are hot mixtures of particles and gases that flow as density currents and are the most devastating phenom- ena, associated mainly with explosive volcanic eruptions. Their hazard is related to their unsurvivable high temperatures and high overpressures (higher-than-hydrostatic pressures), or a lat- eral loading effect akin to blast waves from nuclear or conven- tional weapons, and the high speeds with which they move and overcome topographical obstacles in their paths.9 They comprise fine magma fragments (“ash”); pulverized rock and air become entrained and heated as they move.

Pyroclastic surges form more dilute and turbulent suspen- sion clouds than pyroclastic flows, which settle when they stop moving into a thin layer of well-sorted and fine-grained deposits. Pyroclastic flows are very dense with high particle concentrations and therefore have much more potential for destruction; they are more topographically controlled in that they readily follow valley bottoms, and settle as a massive, poorly sorted deposit.7,8

These so-called pyroclastic density currents can be formed in several different ways. Among the most dangerous are lateral blasts, as occurred at Mount St. Helens, when the side of the volcano suddenly gave way after 2 months of volcanic unrest and the surge obliterated a 180◦ sector as far as 28 km from the crater.4 The Soufrière Hills volcano on Montserrat has been erupting since 1995 with its most significant events being caused

by the growth and collapse of large domes of heaped-up, viscous lava, extruded like toothpaste from its tube, with the material in each major collapse forming a flow or surge into low-lying areas. Large explosive eruptions often begin with vertical erup- tion columns, which can last several hours or even 1 or 2 days, but at some stage parts of the column may lose their buoy- ancy and descend down the flank of the volcano, as occurred in the famous eruption of Vesuvius in ad 79 that destroyed Pompeii. Toward the end of such eruptions dense flows can be formed by a process known as “boiling over” the crater rim. In all instances the hazards in the areas affected by the surges or flows are similar. Whether they form surges or flows can be unpredictable, but an important feature of pyroclastic flows is that surges can detach from them and extend much further than the pyroclastic flow would normally travel. All the phenomena are driven by gravity and some will have the extra impetus from explosions.

For emergency planning purposes, the event that typifies the risk scenario in terms of likelihood and severity is the violent surge of a small volume (covering areas <10 km3).9 This type of event has been modeled for the next eruption of Vesuvius and considered to be the most likely.9 A plausible emergency scenario is one in which people have returned to an exclusion zone around an explosive volcano against scientific advice, or evacuation has been incomplete prior to the eruption, such as when many people refuse to move despite warnings from the authorities. A surge should be regarded as unsurvivable in the open, with the effects of high temperature (>200◦C), dense and irrespirable concentrations of ash, and lateral loading (dynamic pressure) forming a lethal combined impact. At ordinary temper- atures, the minimum concentration of inhalable dust (<100 µ) capable of causing asphyxia is reported to be 0.1 kg/m3, but this figure is not well founded.10 Autopsies of victims at Mount St. Helens in 1980 showed ash occluding the tracheas of victims found in locations close to the volcano where this concentration was undoubtedly greatly exceeded.11

Exposure to dry, motionless air at 200◦C –230◦C can report- edly be survived by a lightly clad person for 2–5 minutes, but in the open and in the absence of protective clothing the convective heat transfer from a fast moving flow at this temperature quickly causes severe thermal injury and rapid death. Inhaling dry air free from hot particles at this temperature is also tolerable for only a few minutes, but the presence of steam, or water vapor, or inhalable amounts of hot, fine ash will reduce the tempera- ture that can be tolerated to below 100◦C, because of the risk of thermal injury to the airways in the absence of respiratory protection.10 In a dome collapse – generated flow on Montserrat in 1997, 19 people died in the open when they were caught in the surge that detached from the flow running along a valley. They were killed instantly by the surge temperature, which was as high as 400◦C and ignited their clothing.12

At a similar eruptive event at Mount Unzen in Japan in 1991, 41 people, including two volcanologists, were killed, and seven survivors in the open developed severe laryngeal edema within 25 minutes of exposure, but they were rescued in time to undergo life-saving tracheostomy at the hospital. Laryngeal obstruction is therefore likely to be an important cause of death in victims whose rescue is delayed. The seven survivors and 10 other victims subsequently died from acute respiratory distress syndrome as a result of their inhalation injuries and extensive skin burns, as did two badly burned loggers at the Mount St. Helens eruption in 1980.13

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

636 ■ PE T E R J. BAXT E R

Some surges can also be saturated with water vapor, which will increase the amount of heat delivered to the respiratory tract and skin. Humans cannot survive long breathing saturated air over approximately 60◦C. At 50◦C the heat transfer to the deep lung could be comparable with that of dry air at 200◦C, while at 80◦C the oxygen concentration of saturated air would be 10%– 11%, a critically low level not even allowing for the presence of ash particles.10 Thus, inhalation of saturated air that also contains an abundance of inhalable ash particles at temperatures between 50◦C and 100◦C could be very hazardous, giving rise to acute bronchoconstriction, pulmonary injury, and hypoxia and could explain the sudden deaths recorded in people sheltering in houses who otherwise appear externally unharmed. A base surge from a steam explosion due to magma–water interaction would also present this hazard. Irritant volcanic gases such as sulphur dioxide might also be present in the surge cloud mixed with the entrained air.

Direct contact of the skin with the hot ash in the surge will cause partial and full thickness burns. In the outdoors, clothing may offer little protection and become combusted in the heat of the surge, or be torn off the body by the violence of the surge impact. Indoors or outside under some protection against the hot current in the periphery of a surge, the heat transfer would be much briefer so that even light clothing could limit the total body surface area (TBSA) affected. Thus a person wearing a suit would receive 20% TBSA burns if the hot ash was in contact with the unprotected exposed skin, whereas for someone with a t-shirt and shorts the figure would be approximately 40% TBSA; these patterns have been observed, for example, in the fatal Unzen eruption in 1991.

A provisional burns scale that predicts mortality risk in vol- canic eruptions has been devised by the author for mass casualties in a pyroclastic surge striking an urban area. Many victims would be killed outright, including most of those caught outdoors, as already explained. In locations sheltered by topography or other buildings, burns to the skin and lung would be the main cause of injury

Level 1. Minor effects of heat, such as singed hair, superficial burns to uncovered skin. Low mortality.

Level 2. Burns to 20%–40% TBSA; moderate inhalational lung injury. Mortality dependent on treatment resources, patient age, and existing illnesses. Without early treatment in a large- scale disaster most would die.

Level 3. Burns greater than 40% TBSA; severe inhalation injury. Mortality greater than 90% in a disaster.

Level 4. Rapid death from heat or asphyxia. Mortality 100%.

Other injuries can be expected, such as lacerations from broken glass when windows are imploded by the pressure impact of the surge, and smoke inhalation from fires ignited by the hot ash in contact with combustible materials in streets and inside dwellings.

No published data on survival in a large series of volcanic eruption victims exist, but some (unpublished) evidence was obtained after an eruption of Merapi, Indonesia, in 1994, when approximately 30 people were killed outright and 86 survivors (52 males and 29 females) were admitted to four hospitals in Yogyakarta after exposure to a pyroclastic surge. They had been attending a wedding at a spot with a view of the volcano’s summit. This event was totally unexpected and occurred when the lava dome collapsed, as it did at times, sending a pyroclastic flow

along a different valley from its usual course. The mean TBSA burned of those hospitalized was 44% (as recorded in the hospital records and probably overestimated) and the overall mortality was 78%. Most of those who died were very badly burned – all had TBSA burns of over 50% – and must be assumed to have had inhalation injury as well.

The Baux index is a long-recognized guide for triaging burn casualties in an event on a mass scale, as would be envisaged here. The index is calculated by adding the age of an adult (>17 years) to the TBSA burned. This sum is the probability of mortality. This index does not account for inhalation injury from hot ash particles, which is likely to be severe in individuals with greater than 40% TBSA burns and would greatly increase their mortality risk.

More than a small number of casualties with 20%–40% TBSA burns would exceed the national burn unit capacity of most countries, yet this group could comprise a large num- ber of the volcano victims and their survival would be largely dependent on the speed of search and rescue operations and time taken until they received definitive treatment. National and international burn centers would need to be available to receive patients in transfer when local facilities were overwhelmed (see Chapter 3).

At the time of this writing, no events of mass burn casualties on this scale have previously occurred, with the exceptions of the Los Alfaques camping ground explosion in Spain in 197814

and the Ufa gas explosion in the Soviet Union in 1989.15 The scenario in a volcanic eruption would be altogether different. Rescue attempts would be hampered by the continuing eruption, with previous ash falls and resuspended ash preventing access. A thick deposit of hot ash on the ground would prevent immediate entry to the disaster area by road (tires can catch fire and shoes ignite on contact with thick pyroclastic deposits before they have cooled). Concern over an additional pyroclastic surge would keep rescuers away for hours until it was believed that the activity had subsided. Helicopters would be needed for rapid movement of rescuers and patients, but these might not fly if too much ash were in the air because this could interfere with the engines.

Explosion Hazards to Scientists and Tourists

Although certain eruptions and volcanoes are referred to as explosive, the release of energy is much slower and in a different form than with nuclear weapons or conventional explosives. In a “blast wave” in a typical pyroclastic surge the dynamic pressure is solely from the lateral loading of the gravity current and there is no peak overpressure as produced in a supersonic wave. Minor explosions from volcanoes can be steam driven, or phreatic, with rocks being hurled without warning from fumaroles or craters, killing those within range. Six scientists and three tourists were killed in an explosion in the crater at Galeras, Columbia, in 1994 due to multiple trauma resulting from flying rocks blasted from the lava dome.16

In most explosive eruptions shock or blast waves do not present hazards far from the eruptive vent and instead it is the dynamic pressure and heat of a pyroclastic surge or flow, together with the kinetic energy of entrained loose materials and projec- tiles that cause the injurious impacts. Velocity measurements of projectiles have been estimated for explosions at the craters in the 1968 eruption at Arenal Volcano (300–400 m/s) and at Ngau- ruhoe in 1975 (220–260 m/s).17 Although loud sound waves can rattle or even break windows, supersonic blast waves are

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

VO LC A N O E S ■ 637

uncommon and their direct effects will mostly be limited to the vicinity of the crater.

The Impacts of Lava Flows

Lava flows from most volcanoes are slow moving and so, in con- trast to pyroclastic flows and surges, usually cause few deaths and injuries because people are able to move out of the lava path in time. The main hazards are from their high temperature (800◦C –900◦C), which can cause combustion of buildings by their direct contact or just by radiant heat. Lava flows are especially mechan- ically destructive and can cause collapse of buildings in their paths. It is for their destructiveness that they are most feared, although some success has been achieved with diverting lavas, the best example being the intervention of the Italian Civil Pro- tection in the flow from the eruption of Mount Etna in 2003. Fifty-one lateral vent eruptions have occurred at Mount Etna in the last 330 years, but these have been of minimal threat to human life. Volcanogenic earthquakes arise when fissures open and these can weaken and destroy buildings and occasionally kill people from their cumulative shaking effect.

Secondary hazards from lava flows can arise in several dif- ferent ways. Lava flows entering the sea will generate clouds of steam and hydrogen chloride gas. Occasionally people close to lava flows can be asphyxiated in heavy rain because of trapped rainwater suddenly flashing to form a jet of steam. Lava flowing over densely vegetated areas can trap pockets of methane and organics, which can then explode and scatter lethal projectiles of lava.18 More rarely, a lava flow moving down a slope can disintegrate on meeting a sudden incline and form a small but potentially lethal pyroclastic flow.

A few volcanoes in the world can erupt very fluid and fast moving lava and their lava flows can be highly hazardous. An example of this type of volcano is Nyiragongo, located near the city of Goma in the Eastern Democratic Republic of Congo and in the midst of an humanitarian crises. More than 4 million lives have been lost as a result of conflict and its consequences since 1998.19 In 1977, 500 people living high up on the flanks of the vol- cano died in a sudden rifting eruption, which drained the crater lava lake. Then, in January 18 and 19, 2002, the rifting suddenly extended further down the flank toward Goma, trapping 170 villagers in flows moving as fast as 60 km/hour. Two main flows reached Goma, by which time they were moving much more slowly, the largest arriving at approximately 1800 hours on the first day. More than 300,000 people escaped from the advancing lava as it drove its way through the center of the city, destroying the main commercial areas and more than 120,000 homes before the end of the day.20,21

The chaotic self-evacuation of the population and then their rapid return within 2 days before the main lava flow had even ceased flowing raised many concerns about the health hazards they would face. The most important was the risk of outbreaks of endemic cholera and dysentery from the consumption of unchlo- rinated water from Lake Kivu. International aid agencies flew workers to the disaster zone and helped to erect emergency chlo- rination stations and water tankers along the side of the lake. The loss of food supplies and charcoal for cooking would have led to hunger very quickly in a population with an already high preva- lence of acute malnutrition, but aid workers averted a food crisis. Crowding increased the risk of infectious disease outbreaks and responders rapidly initiated an immunization campaign. Most people had to escape from the lava flows with relatively little

warning and left all their possessions behind. The psychological stress of losing all their property and livelihoods to the lava flows was one of the most significant factors for the population. Only approximately 13,000 people had to be accommodated in relief camps, the remainder being taken in by relatives or others from the same ethnic group.20

The absence of governmental organizations due to the exist- ing complex health emergency meant that the population was entirely dependent on a swift and effective response by the inter- national agencies and without this, the loss of life from epidemics and other secondary consequences of the eruption would have been much greater.

The Impacts of Lahars

Lahars, or volcanic mud flows, are slurries of water and sediment (60% or more by volume), which can flow at speeds of a few tens of kilometers per hour to more than 100 km/hour on the steep slopes of a volcano. They can flow and set like concrete. In the disaster caused by the Nevado del Ruiz volcano in which 23,000 people died, as much as 85% of the town of Armero was left covered in 3–4 m of hardened mud. In the rescue operation over the following 5 days 1,244 survivors, mainly from Armero, were admitted to hospitals and 138 of these subsequently died. The lahar struck the town at approximately 1130 hours and was preceded by a river of water that flowed along the streets that was fast and deep enough in places to overturn cars and sweep away people.5 When the lahar arrived it was traveling at an estimated velocity of 12 m/second and flowed for 10–20 minutes, during which time most of the town was devastated as buildings collapsed under the load of the moving flow. Survivors clung to moving pieces of debris or were swept along on top of the mud. Overall, the several inundations of mud lasted approximately 2 hours, with two slower moving major pulses being accompanied by several smaller pulses over this time period.

The head of the lahar would have been turbulent and con- tained all manner of debris, including large boulders. The slurry mass would have likely buried victims or hurled them against sta- tionary objects, or they would have been contorted and crushed by debris such as trees and collapsed parts of buildings. Stones and other sharp objects would have caused deep lacerations. The slurry would have forced its way into the mouth, and into the eyes, ears and open wounds. The pressure against the chest would have inhibited breathing in those buried up to the neck as they were swept along in the flow resulting in death by traumatic asphyxia.

The main injuries sustained by the hospitalized patients were severe crush injury with open fractures, hemorrhagic, hypovolemic or traumatic shock, chest trauma (flail chest, pneu- mothorax, mud aspiration, and wound sepsis). More than two- thirds of the in-hospital deaths were ascribed to overwhelm- ing infection, such as gas gangrene, ischemic gangrene, tetanus, and generalized sepsis or septic shock. Some patients had limbs amputated for the treatment of wound infections.22

An important cause of wound sepsis was primary suturing of infected or deep or extensive wounds after inadequate debride- ment, instead of leaving them open for 5 days before closing them (delayed primary closure).

A small number of victims who were rescued after being immersed in mud for at least 3 days developed life-threatening necrotizing fasciitis caused by anaerobic and aerobic organ- isms in synergistic combinations. Zygomycetic organisms were

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

638 ■ PE T E R J. BAXT E R

identified in patients with more severe and rapidly progressing lesions. This dreaded complication was due to normally non- pathogenic soil organisms replicating in wounds in the absence of oxygen and is notoriously resistant to medical treatment.23

The reason for so many people surviving when trapped in the slurry was due to its granular or sandy consistency. Lahars diluted by water in this way are called noncohesive, because they contain relatively little clay, in contrast to the much denser cohesive types, which contain much clay derived from chemically altered rocks. The sandy consistency made access to the injured difficult and many had to be air lifted from the slurry by helicopters. The lahar was originally hot and scalded some victims, and many parts must have retained heat over several days otherwise the trapped victims would have died from hypothermia before they were rescued.

Mount Rainier in the U.S. state of Washington is the most hazardous volcano in the Cascade range because of its potential for forming massive lahars, which can travel long distances into areas that have become heavily populated.24 Mount Rainier – at 4,393 m – is the highest peak in the Cascade range and its load of glacier ice exceeds that of any mountain in the conterminous United States.

Emergency planning and land utilization in the volcano’s footprint are based on the lahar hazard, which according to vol- canologists dominates the risk scenario above all other eruptive phenomena as based on its history and potential hazards. More fluid, noncohesive lahars can form in future volcanic activity if huge quantities of melt water are produced from glacier melt water (as in the Nevado del Ruiz tragedy), but cohesive lahars can also form with little, if any warning, or in the early stages of activ- ity when magma is forcing its way to the surface. This can occur because the flanks are formed by substantial volumes of unstable, water-rich hydrothermally altered rock that can collapse in the form of a dry debris flow or a dense lahar.24

In Ecuador, a vast, fast-moving lahar developed at Cotopaxi (5,897 m high) from the melting of the summit ice in its last eruption in 1877. At the time of this writing, the worst expected volcanic disaster in the country is a new eruption of this volcano that would lead to a repeat of this event in the Chillos valley, which has become much more populated since the 19th century, with the town of Lattacunga (population 80,000) only 50–70 minutes travel time for the lahar along its path. The damage to property and infrastructure, as well as the potential for extreme loss of life, would dwarf the losses suffered in the Nevado del Ruiz disaster given the greater economic development that has occurred in this high-risk zone.

Even small, noncohesive lahars can be very dangerous, and they arise commonly when thick ash deposits produced by erup- tive activity become mobilized by heavy rains almost anywhere on a volcano. A serious complication in an eruption is the trig- gering of lahar formation when heavy rainfall is produced as large quantities of fine ash particles mix with rain clouds. Lahars can present lethal hazards for several years after large eruptions, which leave thick ash deposits on flanks, as was seen follow- ing the Mount Pinatubo eruption in 1991, with resultant deaths and other very disruptive consequences for thousands of people living along the far end of the slopes of the volcano.

Tephra Fall Impacts

Tephra is the ash and glass material emitted in the atmosphere during a volcanic eruption. The temperature of the erupted mate-

rial and the mass eruption rate determine the height of the erup- tion column, which along with wind strength and direction, are the principal controls on the long-distance transport and fallout of the tephra. Ash typically is finer grained and forms thinner deposits with increasing distance downwind from the eruptive vent.7,8

Ash falls can affect very wide areas downwind of an erupt- ing volcano as far as hundreds of kilometers away and can have many different impacts on health and human activity over these long distances. The mixture of ash, gases, and aerosols that are injected into the atmosphere can circle the earth and have tem- porary effects on global weather and climate.7,8 The most impor- tant aspects for disaster medicine are the immediate health and safety hazards that arise in a typical heavy ash fall, including those created by damage to infrastructure. A large conurbation would be most vulnerable to many of these, such as disrup- tion to transport, breakdown of utilities, and the injuries and loss of life from weaker structures collapsing under the accumu- lated weight of ash. The experience at the eruption of Mount St. Helens in 1980 is a good model for disaster planning for ash falls at most explosive volcanoes that have moved into a state of unrest.25

The first experience downwind of an eruption of this size is the cloud passing overhead and darkness growing as ash starts to fall. In some instances, pumice and other clast material in the fallout can be large enough to smash windshields and cause head injuries, so people should rapidly seek shelter. When fine ash settles on windshields it soon leaves smears if the windshield wipers are turned on; when the spare wiper fluid is depleted, the driver can no longer see out of the windshield. If the ash fall is heavy, visibility can become zero. When ash has settled motor vehicles resuspend it in the air and visibility becomes severely limited. Rain accompanying an ash fall can make the roads very slippery, resulting in collisions. Even without rain, steep inclines can become hazardous as car wheels lose their grip on the road. Ash soon clogs engine air filters and can bring vehicles to a halt. All transport, including planes and trains, will grind to a halt if visibility becomes bad, a situation that lasted for 5 days in central Washington State after the eruption of Mount St. Helens on May 18, 1980, until an unseasonable rainfall in a normally arid area cleared the air of ash that was being constantly resuspended by winds. Volcanic ash severely damages jet engines and airplanes’ superstructure, so airports will close after even a light deposit of ash on the runway.

Ash settling on unprotected electrical insulators at substa- tions can lead to outages from short circuits as it is a good conductor of electricity when it is wet. These outages in turn can stop water supplies that depend on pumping. Drains and sewage systems can become blocked with the mass of ash when it is mobilized by rain. Electronic equipment and computers are readily penetrated by fine ash (laptops depend on a small inside fan for cooling) and can be rendered useless. Hospital systems and equipment can fail. Falling ash and lightening strikes (fre- quent in some eruptions) can lead to serious communications disruptions, including the lack of availability of wireless phones, television, radio, and any other technology requiring transmit- ters or repeaters.

Rates of accumulation of ash on horizontal surfaces can be as high as 10–20 cm/hour and weak roofs begin to collapse with an accumulated depth of wet ash of only 10 cm. The type of roof and its condition will determine the extent of damage. A collapsing roof can kill or injure the building’s occupants by striking them

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

VO LC A N O E S ■ 639

with roof members and by burial in the onrushing ash, causing asphyxia.26

Ash particles in explosive eruptions can be very fine and of respirable size. The high inhalable particulate matter concentra- tions can irritate the airways and trigger asthma attacks in people with asthma and asthmatic bronchitis. The ash from silica-rich lava dome eruptions can contain very elevated levels of crystalline silica (cristobalite, quartz), and long-term exposure to repeated ash falls over many years has the potential to cause silicosis. This risk is apparent on Montserrat and long-term health checks and chest x-ray surveys are needed to confirm the effectiveness of the measures for preventing the disease in the general and working populations.27

Ballistic projectiles do not usually travel farther than 5 km from the crater. They are an important reason for evacuating a population living close to a crater, even though other more important hazards may dominate the risk scenario. Hot ballistics even of small size can smash through corrugated metal roofs leaving small holes of entry but then explode their hot fragments over furniture inside and trigger fires. Obviously, direct impact can cause serious injury as well. Blocks a meter or more in size can leave large impact craters. Experiencing a shower of ballistics is very frightening and can leave an area with damaged or burnt- out houses, roads, and infrastructure such as, downed power lines.28

Hazards of Volcanic Gases

A few volcanoes emit substantial amounts of gases in the plumes from their craters for years without showing any other activity. The summits of most volcanoes are so elevated that the plume does not result in any hazardous concentrations in the ambient air. At a few low-lying volcanoes (e.g., Masaya, Nicaragua; Poas, Costa Rica) gas plume emissions can present health hazards for kilometers downwind of a volcano’s vent due to the effects of sulphur dioxide on the airways, while soluble acid gases damage vegetation by direct fumigation or by making rain water intensely acid. Volcanologists entering crater areas to take samples are at risk of being overcome by carbon dioxide and hydrogen sul- phide.29

Soil gases can permeate into the atmosphere from the slopes of volcanoes in repose and their concentrations can accumu- late in the indoor air of buildings with gas permeable founda- tions.30,31 Carbon dioxide is the main hazard, and the gas can also act as a carrier of the radioactive gas radon, which is known to be a causal factor for lung cancer and can accrue inside poorly ventilated houses.

A change in the composition of plume or soil gas and an increase in emission rate from a volcano can be the first sign of magma on the move, so these gases are routinely monitored for evidence of a renewal of volcanic activity.

A rare but important hazard in volcanic areas is an accumula- tion of carbon dioxide in solution in a pressurized hydrothermal system of a volcano or the build up of huge amounts of carbon dioxide dissolved under the hydrostatic pressure of deep lakes. An event that triggers the sudden release of the carbon dioxide can lead to the formation of a lethal, denser than air cloud of asphyx- iating gas. In the Dieng Plateau, Java, in 1979, carbon dioxide was suddenly released and flowed down a slope and killed 149 vil- lagers fleeing from an incipient minor eruption.32 Two thousand people died when a cloud of carbon dioxide was released late at night from Lake Nyos in 1986, with victims lying unconscious

Table 39.4: Volcanic Crises and Cities since 1980

Campi Flegrei, Italy 1982–1984 Naples

Galeras, Colombia 1989–ongoing Pasto

Popocatapetl, Mexico 1994 Mexico City, Puebla

Colima, Mexico 1994 Colima

Guagua Pichincha, Ecuador 1998–2001 Quito

Tungurahua, Ecuador 1999–ongoing Banõs

Reventador, Ecuador 2002 Quito

Nyiragongo, Dem. Rep. Congo 2002–ongoing Goma

on the ground for hours before dying or regaining consciousness when the gas dispersed in the early morning.33 Many deep lakes in volcanic areas around the world have been studied for carbon dioxide accumulation since that time, but only Lake Kivu in the Eastern Democratic Republic of Congo and Lake Albano near Rome have been identified so far as potential candidates for such releases. In a future eruption of Nyiragongo, the fissuring could extend into Goma and Lake Kivu, which is a highly stratified and deep lake (maximum depth 485 m) containing a huge quantity of carbon dioxide and methane dissolved in its deeper waters. These gases, derived from volcanic sources under the lake and from biodegradation of organic matter, have accumulated over the centuries and would be highly dangerous if suddenly released in a cloud that was blown over the populated shore. While the technology for degassing lakes before they become saturated with gas and highly hazardous (as occurred with Lake Nyos) is avail- able, the technology does not exist for identifying hydrothermal systems that are waiting to explode when disturbed by the onset of volcanic activity.

Emergency Planning for Cities and Islands

As with many other hazards, densely populated areas increase vulnerability to volcanic eruptions. The key infrastructure of cities is very vulnerable to volcanic activity. This, coupled with the large numbers of people, leads to a challenge for health offi- cials responsible for making timely evacuation decisions as the consequences of a judgment error could be large human or eco- nomic losses. At the time of this writing, there have been no recent major eruption impacts on a city or megacity, but there are cities in proximity to volcanoes where the potential for major disaster exists (Table 39.4). A glimpse of what could happen can be seen by analyzing the worst volcanic disaster of the 20th cen- tury that occurred in 1902 when the city of St. Pierre on the French Caribbean island of Martinique was struck by a pyro- clastic surge, killing 28,000 people within minutes and leaving only two survivors in the city itself.9 At least half a million in a megacity of 4 million people in the Bay of Naples area are at risk from pyroclastic surges and flows in a future eruption of Vesu- vius. Lava flows have not entered a modern city, but the invasion of Goma by two lava flows in 2002 described previously shows how the destruction can be severe and in a complex emergency threaten large loss of life from ethnic conflict and expose thou- sands to infectious disease epidemics and the risk of starvation. At Mount Rainier, the horrors of Armero in 1984 could be vis- ited on the large population in the Puget Sound lowland area,

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

640 ■ PE T E R J. BAXT E R

but monitoring and warning systems and emergency planning are in place to help prevent such a disaster.24 Although actual experience is limited in terms of the direct and indirect impacts of ash on the health and safety of an urban population (and what mitigation measures will be feasible to offset the immediate risks), a large ash fall with a depth of 0.5–1 m on a city would most likely lead to a complete paralysis of the city’s function- ing. Massive and costly clearance operations to remove the vast quantities of ash would be required to restore the city’s infra- structure.

Very similar considerations apply to islands, where lack of space means that people may live in proximity to an erupting volcano or leave the island to live elsewhere because of safety concerns or to find remunerative work. The island of Montserrat is only 17 × 10 km and half of the island around the volcano has become an exclusion zone.34 Of an original population of 12,000 people, only 4,500 remained in 2007 when the volcano was still erupting.

Until recently, volcanologists have been using a hazard-based approach to decision making in volcanic crises, which means advising authorities on the most likely maximum event (maxi- mum expected eruption),35 rather than using the evidence from past eruptions to determine all the main eruption possibilities and their probabilities. Thus, at Vesuvius the maximum expected event is on the scale of the largest eruption of the last 1,000 years; the 1631 eruption is the reference eruption, with the likelihood of pyroclastic surges and a heavy ash fall across the breadth of Italy, which dominate the risk scenario for emergency planning. The drawback to the hazard-based approach is that it is all or nothing, and the authorities will lose the confidence of large numbers of people having to be evacuated from their homes for more than 1 or 2 months if a large eruption fails to occur. Recognition that smaller-scale eruptive events are much more probable is needed in a long drawn out crisis, as has been demonstrated repeatedly in the eruption of the Soufrière Hills volcano, Montserrat, since it began erupting in 1995.34

This problem of scientific uncertainty and evacuation deci- sion making is embodied in the 1976 Guadeloupe crisis. More than 70,000 people were evacuated from the Basse Terre area in Guadeloupe for up to 9 months because the Soufrière vol- cano threatened a major eruption. The foreseen eruption did not occur, which led to intense recriminations among the sci- entists involved, as well as huge economic and political conse- quences.2 The Tungurahua volcano, Ecuador, renewed its activ- ity on September 14, 1999 and the President of Ecuador ordered a hurried evacuation of the town 3 days later. There were no preparations in place for an evacuation of 25,000 people and by the end of December the people had forced their way back into their homes despite the risk of death from pyroclastic flows that had been widely predicted by volcanologists. At the time of this writing, the disaster risk is high because the eruptive activity continues with approximately 17,000 people living in the city.

RECOMMENDATIONS FOR FUR THER RESEARCH

A more risk-based approach for volcanoes is under develop- ment that assigns probabilities to all the reasonably foreseeable eruptive events and their emergency scenarios. The full range of eruptions and their impacts can be codified in a decision tree whose branches can be populated with conditional probabili- ties through a process of elicitation and expert judgment. This

approach was first applied to a volcanic crisis in Montserrat in 199734 and is analogous to the evidence-based methods that are used in medicine in diagnosis or treatment in the face of clini- cians’ uncertainty.36 This methodology is beginning to be more widely used as a powerful way to quantify risk and to identify the most important eruptive scenarios for emergency planning purposes.

An evidence-based approach to volcanic crises is becoming possible through the development of numerical simulation mod- eling in recent years to predict probabilistically the impacts and consequences of pyroclastic flows and surges and ash falls in developed areas. The extent of lahar run-out can also be mod- eled. Models are not absolutely predictive but they are useful for beginning to quantify risk assessment, as long as their uncer- tainties and limitations are understood. They are also beginning to be used to support decision making in crises. The MESIMEX exercise in the fall of 2006 in Naples was the first “real-time” exercise to simulate a state of unrest lasting 5 days and escalat- ing toward an eruption. The exercise incorporated models for forecasting the direction of the wind and the building damage footprint under the plume, as well as the limits of a pyroclastic surge and the extent of disruption and damages caused by houses collapsing from volcanogenic earthquakes.

Computer simulations of volcanic hazards37 and their impacts, together with casualty types and estimated numbers, provide the opportunity for probabilistic risk modeling and re- search into mitigation measures, a quantum leap compared with past approaches. Thus, models to predict casualties from tephra fallout37–39 and pyroclastic surges40–42 based on the vulnerabil- ity of buildings and their occupants to impacts quantified by numerical simulation modeling are now being developed. For the first time, three-dimensional numerical simulation model- ing of pyroclastic surges by using supercomputer technology43

opens up the future possibility of quantifying risk for disaster planning in cities and a much greater interaction between volca- nologists and decision makers. At present, these tools are helping to define the limits of uncertainty in decision making, but one day they will become essential adjuncts for developing propor- tionate approaches in volcanic crises. This work needs to be based on an improved understanding of the causes of death and injury in eruptions as well as collecting data on the range of impacts to the built environment and infrastructure in actual events.

Probabilistic-based methods are also being applied to evacu- ation decision making in cities where the political and economic consequences of false alarms and relocating large numbers of people can be enormous, especially if an eruption fails to mate- rialize. On the other hand, an evacuation decision that is too late will expose the same population to large losses of life. This dilemma was illustrated in the delayed call by the Mayor of New Orleans in the face of Hurricane Katrina resulting in thousands of trapped people in the flooded city.

Conclusion

Volcanic eruptions are infrequent compared with other hazards. They do not attract the same high level of general attention as earthquakes, floods, and windstorms, at least until a dangerous volcano moves out of a state of repose and threatens to erupt and devastate a populated area. Despite this, a threatened or actual volcanic eruption is of much higher complexity compared with these typically short-lived events, and the emergency manage- ment of a volcanic crisis, which can last months or years, is a

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

VO LC A N O E S ■ 641

unique challenge to all those involved, including the population at risk.

Although mitigation of human casualties by timely evacu- ation is the main goal in emergency management of volcanic threats, disaster planners must also be familiar with rescue and emergency treatment measures and prepare for unique condi- tions associated with volcanic eruptions. Complete or partial evacuations may last months and measures to support displaced persons are essential, or people may soon prefer to accept a high risk and return to their homes against the advice of scientists and the authorities. In a prolonged crisis in which a forecasted major eruption fails to occur, public perceptions can quickly become highly distorted as a disillusioned population becomes habitu- ated to risk taking and starts to ignore official warnings. With the limited ability of scientists to accurately forecast the timing and size of most future eruptions,44 the potential for events with mass casualties is always present and remains throughout the duration of most volcanic crises, not just when an eruption begins.45

REFERENCES

1. Baxter PJ. Medical effects of volcanoes. 1. Main causes of death and injury. Bull Volcanol. 1990;52:532–544.

2. Newhall CG, Punongbayan RS. The narrow margin of successful volcanic risk mitigation. In: Scarpa R, Tilling RI, eds. Monitoring and Mitigation of Volcano Hazards. Berlin: Springer; 1996;809– 838.

3. Witham CS. Volcanic disasters and incidents: a new database. J Volcanol Geotherm Res. 2005;148:191–233.

4. Newhall CG. Mount St. Helens, master teacher. Science. 2000;288:1181–1183.

5. Voight B. The management of volcanic emergencies: Nevado del Ruiz. In: Scarpa R, Tilling RI, eds. Monitoring and Mitigation of Volcano Hazards. Berlin: Springer; 1996:719–769.

6. Spence RJS, Pomonis A, Baxter PJ, Coburn AW, White M, Dayrit M. Building damage caused by the Mount Pinatubo eruption of June 15, 1991. In: Newhall CG, Punongbayan RS, eds. Fire and Mud: Eruptions and Lahars at Mount Pinatubo, Philippines. Seattle: University of Washington Press; 1996:1055–1061.

7. Schmincke H-U. Volcanism. Berlin: Springer, 2004. 8. Oppenheimer C, Francis P. Volcanoes. 2nd ed. Cambridge:

Cambridge University Press, 2004. 9. Baxter PJ, Boyle R, Cole P, Neri A, Spence R, Zuccaro G. The

impacts of pyroclastic surges on buildings at the eruption of the Soufrière Hills volcano, Montserrat. Bull Volcanol. 2005;67:292– 313.

10. Baxter PJ, Neri A, Todesco M. Physical modeling and human survival in pyroclastic flows. Natural Haz. 1998;17:163–176.

11. Eisele JW, O’Halloran RL, Reay DT, Lindhock GR, Lewman LV, Brady WJ. Deaths during the May 18, 1980, eruption of Mount St Helens. N Engl J Med. 1981;305, 931–936.

12. Loughlin SC, Baxter PJ, Aspinall WP, Darroux B, Harford CL, Miller AD. Eyewitness accounts of the 25 June 1997 pyroclas- tic flows and surges at Soufrière Hills volcano, Montserrat, and implications for disaster mitigation. In: Druitt TH, Koke- laar BP, eds. The Eruption of Soufrière Hills Volcano, Montser- rat, from 1995 to 1999. London: Geological Society; 2002:211– 230.

13. Parshley PF, Kiessling PJ, Antonius JA, Connell RS, Miller SH, Green FHY. Pyroclastic flow injury. Mount St Helens . May 18, 1980. Am J Surg. 1982;143:565–568.

14. Arturson G. The Los Alfaques disaster: a boiling-liquid, expand- ing vapour explosion. Burns. 1981;7:233–251.

15. Kulyapin AV, Sakhautdinov VG, Temerbutalov VM, et al. Bashkiria train-gas pipeline disaster: a history of the joint ASSR/USA collaboration. Burns. 1990;16:339–342.

16. Baxter PJ, Gresham A. Deaths and injuries in the eruption of Galeras volcano, Colombia, 14 January 1993. J Volcanol Geotherm Res. 1997;77:325–338.

17. Morrisey MM, Mastin LG. Vulcanian eruptions. In: Sigurdsson H, Houghton BF, McNutt, SR, Rymer H, Stix J, eds. Encyclopedia of Volcanoes. San Diego: Academic Press; 2000:463–475.

18. Tilling RI, Peterson, DW. Field observation of active lava in Hawaii: some practical considerations. In: Kilburn CRJ, Luongo G, eds. Active Lavas: Monitoring and Modeling. London: UCL Press; 1993:147–174.

19. Salama P, Spiegel P, Talley L, Waldman R. Lessons learned from complex emergencies over past decade. Lancet. 2004;364:1801– 1813.

20. Baxter P, Allard P, Halbwachs M, et al. Human health and vul- nerability in the Nyiragongo volcano eruption and humanitar- ian crisis at Goma, Democratic Republic of Congo. Acta Vul- canolog. 2002–2003;14–15:109–114.

21. Komorowski J-C, Tedesco D, Kasareka M, et al. The January 2002 flank eruption of Nyiragongo volcano (Democratic Republic of Congo): chronology, evidence for a tectonic rift trigger, and impact of lava flows on the city of Goma. Acta Vulcanolog. 2002– 2003;14–15:27–62.

22. Organización Panamericana dela Salud. Cronicas de Desastres. Erupción Volcánica en Colombia, Noviembre 13, 1985.

23. Patiño JF, Castro D, Valencia A, Morales P. Necrotizing soft tissue lesions after a volcanic cataclysm. World J Surg. 1991;15: 240–247.

24. Hoblitt RP, Walder JS, Driedger CL, Scott KM, Pringle PT, Vallance JW. Volcano Hazards from Mount Rainier, Washing- ton. Open File Report 98–428. Denver: U.S. Geological Survey; 1998.

25. Buist AS, Bernstein RS, eds. Health effects of volcanoes: an approach to evaluating the health effects of an environmental hazard. Am J Public Health Med. 1986;76 (Suppl):1–90.

26. Pomonis A, Spence R, Baxter P. Risk assessment of residential buildings for an eruption of Furnas volcano, São Miguel, the Azores. J Volcanol Geotherm Res. 1999;92:107–131.

27. Horwell CJ, Baxter PJ. The respiratory health hazards of vol- canic ash: a review for volcanic risk mitigation. Bull Volcanol. 2006;69:1–24.

28. Blong RJ. Volcanic Hazards: a Sourcebook on the Effects of Erup- tions. Sydney: Academic Press; 1984.

29. Allen AG, Baxter PJ, Uttley CJ. Gas and particle emissions from Soufrière Hills Volcano, Montserrat. Bull Volcanol. 2000;62: 8–19.

30. Baxter PJ, Baubron J-C, Coutinho R. Health hazards and disaster potential of ground gas emissions at Furnas volcano, Sao Miguel, Azores. J Volcanol Geotherm Res. 1999;92:95–106.

31. Carapezza ML, Badalamenti B, Cavarra L, Scalzo A. Gas haz- ard assessment in a densely inhabited area of Colli Albani Vol- cano (Cava dei Selci, Roma). J Volcanol Geotherm. Res. 2003;123: 81–94.

32. Le Guern F, Tazieff H, Faivre Pierret RX. An example of health hazard: people killed by gas during a phreatic eruption Dieng Plateau (Java), Indonesia, February 20th 1979. Bull Volcanol. 1982;45(2):153–156.

33. Baxter PJ, Kapila M, Mfonfu D. Lake Nyos disaster, Cameroon, 1986: the medical effects of large scale emission of carbon dioxide? Br Med J. 1989;298:1437–1441.

34. Druitt TH, Kokelaar BP, eds. The Eruption of Soufrière Hills Vol- cano, Montserrat, from 1995 to 1999. London: Geological Society; 2002.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .

642 ■ PE T E R J. BAXT E R

35. Rosi M. Quantitative reconstruction of recent volcanic activity: a contribution to forecasting of future eruptions. In: Scarpa R, Tilling RI, eds. Monitoring and Mitigation of Volcano Hazards. Berlin: Springer; 1996:631–674.

36. Aspinall WP, Woo G, Voight B, Baxter PJ. Evidence-based vol- canology: an application to eruption crises. J Volcanol Geotherm Res. 2003;128:273–285.

37. Dobran F, Neri A, Todesco M. Assessing the pyroclastic flow hazard at Vesuvius. Nature. 1994;367:551–554.

38. Barberi F, Macedonio G, Pareschi MT, Santacroce R. Mapping the tephra fallout risk: an example from Vesuvius, Italy. Nature. 1990;344:142–144.

39. Spence RJS, Kelman I, Calogero E, Toyos G, Baxter PJ, Komorowski J-C. Modeling expected physical impacts and human casualties from explosive eruptions. Natural Haz Earth Syst Sci. 2005;5:1003–1015.

40. Spence RJS, Kelman I, Baxter PJ, Zuccaro G, Petrazzuoli S. Residential building and occupant vulnerability to tephra fall. Natural Haz Earth Syst Sci. 2005;5:1–18.

41. Spence RJ, Zuccaro G, Petrazzuoli S, Baxter PJ. Resistance of buildings to pyroclastic flows: analytical and experimental stud- ies and their application to Vesuvius. Natural Haz Rev. 2004; 5(1).

42. Spence R, Kelman I, Brown A, Toyos G, Purser D, Baxter P. Residential building and occupant vulnerability to pyroclastic density currents in explosive eruptions. Natural Haz Earth Syst Sci. 2007;7:219–230.

43. Neri A, Ongaro TE, Menconi G, et al. 4D simulation of explosive eruption dynamics at Vesuvius. Geophys Res Lett. 2007;34:L04309.

44. Sparks RSJ, Aspinall WP. Volcanic activity: frontiers and chal- lenges in forecasting, prediction and risk assessment. In: The State of the Planet: Frontiers and Challenges in Geophysics. Geophysi Monograph 150, IUGG 2004;19:359–373.

45. Simkim T, Siebert S, Blong R. Volcano fatalities – lessons from the historical record. Science. 2001;291:255.

46. Used with permission from Elsevier. 47. Used with permission from Elsevier.

C op

yr ig

ht ©

2 00

9. C

am br

id ge

U ni

ve rs

ity P

re ss

. A ll

rig ht

s re

se rv

ed .