Annotated bibliography

Delays in Global Disease Outbreak Responses: Lessons from H1N1, Ebola, and Zika

Inglobaldiseaseoutbreaks,there

are significant time delays be-

tween the source of an outbreak

and collective action. Some delay

is necessary, but recent delays

have been extended by insuffi-

cient surveillance capacity and

time-consuming efforts to mo-

bilize action.

Three public health emergen-

cies of international concern

(PHEICs)—H1N1,Ebola,andZika—

allow us to identify and com-

pare sources of delays and con-

sider seven hypotheses about

what influences the length of

delays. These hypotheses can

then motivate further research

that empirically tests them. The

three PHEICs suggest that de-

ferred global mobilization is a

greater source of delay than is

poor surveillance capacity. These

case study outbreaks support hy-

potheses that we see quicker re-

sponses for novel diseases when

outbreaks do not coincide with

holidays and when US citizens are

infected. They do not support

hypotheses that we see quicker

responses for more severe out-

breaks or those that threaten

larger numbers of people.

Better understanding the rea-

son for delays can help target

policy interventions and identify

the kind of global institutional

changes needed to reduce the

spread and severity of future

PHEICs. (Am J Public Health. 2018;

108:329–333. doi:10.2105/AJPH.

2017.304245)

Steven J. Hoffman, JD, PhD, LLD, and Sarah L. Silverberg, BArtsSc

Increasing global trade, travel,and climate change have accelerated the spread of disease outbreaks beyond national bor- ders to more quickly develop into international concerns. After the severe acute respiratory syndrome (SARS) epidemic, there were major efforts to build surveillance and response systems to identify outbreaks early, respond globally, and contain spread at the source.1

Yet despite these efforts, we still see prolonged time delays in severe outbreaks between their emergence and global collective action. To simplify, there are two main processes that can result in delays. First, there will be a delay between the emergence of an outbreak’s index case and the detection of the outbreak by health care providers, laborato- ries, and public health authorities. One goal of disease surveillance is to minimize this delay and maximize available information for guiding the public health response through ongoing data collection, analysis, and man- agement. Second, there will be a delay between an outbreak’s detection and widespread rec- ognition of the outbreak as an international concern. Should outbreaks involve international spread and require a coordinated international response, such rec- ognition is best evidenced by a declaration from the World Health Organization (WHO) that it constitutes a public health emergency of international concern (PHEIC). In making this declaration, WHO’s director general considers the prevention,

protection, and response needs of the situation and the advice of an emergency committee before potentially mobilizing efforts to address these needs.2

When the systems for recog- nizing and responding to dis- ease outbreaks act too slowly, the result is unnecessary delay, greater disease spread, additional people affected, and more lives lost.3–6 We used the last three PHEICs—H1N1, Ebola, and Zika outbreaks—as case studies to compare sources of delays and to screen seven potential hypotheses of what influences the length of delays. Our goal was to identify and consider possible reasons for the delays to motivate future hypothesis-testing research that empirically assesses them and ul- timately informs the design of evidence-based interventions that hasten future outbreak responses.

TECHNICAL VS POLITICAL DELAYS

If global surveillance was the main source of delay, there would be a significant lag between an outbreak’s index case and local health officials’ action that leads to laboratory diagnosis. In none of

the three PHEICs can surveillance capacity challenges be blamed for the majority of delay.

Surveillance capacity caused no meaningful delay in the 2009 H1N1 influenza outbreak. Mexico identified the unusual respiratory illness within three days of its emergence.7,8

Mexico’s Ministry of Health swiftly notified the Pan- American Health Organization (PAHO) and implemented in- fection prevention and control measures throughout the country as it tracked the virus’s spread.8

Although such quick de- tection was not possible in Guinea and other West African countries when Ebola emerged on December 26, 2013, much of the delay derived from the in- correct thought that deaths were from cholera or Lassa fever.9

Diagnostic accuracy must be improved, as insufficient sur- veillance and laboratory capacity delayed the Ebola response by 2.5 months—the time from the outbreak’s detection to when French scientists at the Institut Pasteur de Lyon confirmed it was Ebola on March 22, 2014.9

Although this was a delay, it accounts for a small percentage of the total time to full mobilization.

ABOUT THE AUTHORS Steven J. Hoffman is with the Global Strategy Lab, Dahdaleh Institute for Global Health Research, Faculty of Health and Osgoode Hall Law School, York University, Toronto, ON, Canada, and is also with the Department of Global Health and Population, Harvard T. H. Chan School of Public Health, Harvard University, Boston, MA. Sarah L. Silverberg is with the Faculty of Medicine, University of Toronto, Toronto.

Correspondence should be sent to Steven J. Hoffman, Global Strategy Lab, York University, 4700 Keele Street, Toronto, ON, M3J 1P3, Canada ([email protected]). Reprints can be ordered at http://www.ajph.org by clicking the “Reprints” link.

This article was accepted November 15, 2017. doi: 10.2105/AJPH.2017.304245

March 2018, Vol 108, No. 3 AJPH Hoffman and Silverberg Peer Reviewed Commentary 329

AJPH PERSPECTIVES

Surveillance was not the main source of delay in the response to Zika in Latin America. Brazilian authorities swiftly noted the rise in microcephaly cases, even though the root cause of the microcephaly was not immedi- ately identified.10 After initial data of a microcephaly cluster became available on October 17, 2015,11 local health officials im- mediately began investigating the cause, gathering further in- formation in the face of un- certainty about the cluster, and Brazil’s Ministry of Health established emergency opera- tions centers in affected areas only a month later.12

If surveillance is not a main source of delay, slow political mobilization following outbreak detection may be a culprit. In fact, overall, on the basis of these three PHEICs, poor mobilization is responsible for 1.9 times more delay than is insufficient surveil- lance capacity (i.e., 241 days vs 127 days cumulatively; Figure 1).

One cause of delayed mobi- lization is the contextual and scientific uncertainty in eachcase.

As more data become available, interpretation of data, charac- terization of risk, and epidemio- logic analysis become clearer and inform decisions to mobilize. Yet, there were delays after in- creased scientific certainty in the three PHEICs, indicating that there are additional causes of delayed political mobilization beyond surveillance and un- certain scientific evidence.

After laboratories isolated H1N1 on March 18, 2009, just three days after H1N1’s emer- gence, it took WHO one month to declare a PHEIC. The Mex- ican government started much of the effort needed to track and control the spread in the interim. By April 17, 2009, all hospitals were asked to report any case of severe respiratory illness and collect specimens, and by April 23, 2009, a case definition had been developed.8 In this case, there was relatively quick global mobilization once the nature of the pandemic was understood.

In stark contrast to the speed of global action in 2009 was the delay in responding to Ebola in

2014. After Ebola was confirmed on March 22, 2014,9 some action did happen: WHO mobilized its Global Outbreak Alert and Re- sponse Network, and Médecins Sans Frontières established a field presence.9,13 As the virus spread to Sierra Leone, WHO created an emergency response team.8

But despite growing evidence of the outbreak’s uncontrolled spread, a PHEIC was not de- clared until August 8, 2014—138 days after Ebola’s detection.14

Most recently, there was delayed global political mobili- zation for the Zika outbreak, despite Brazil’s declaration of a national emergency on No- vember 11, 2015.12 Although PAHO formally warned of the link between Zika, neurologic syndromes, and microcephaly on December 1, 2015,15 it took another two months before WHO declared a PHEIC on February 1, 2016.16 There was little international news media coverage of Zika from early December 2015 to mid-January 2016 despite steadily increas- ing infection rates and an

understanding of the link be- tween the virus and neurologic abnormalities.

SCREENING HYPOTHESES ON FASTER RESPONSES

Why is global political mo- bilization much faster for some outbreaks than others? With so many unique factors playing a role in each outbreak, it is difficult to pinpoint exact de- terminants of delay. But these three PHEICs provide an op- portunity to identify some likely hypotheses and consider whether they are consistent with what was actually seen during these outbreaks (seven hypotheses summarized in Table 1). Our exploratory analysis provides a basis for future empirical re- search to diagnose causes of delay and inform efforts to respond quickly to future outbreaks.

First, one might expect the speed of political mobilization to increase with disease severity. Yet, the H1N1 outbreak caused the least severe disease of the three PHEICs and attracted the fastest global mobilization. It therefore seems unlikely that severity, especially as demon- strated by the early outbreak, dictates speed of mobilization.

A second hypothesis is that when more countries are af- fected, global response is faster. This makes sense because PHEICs,bydefinition,areinvoked to mitigate diseases’ international spread; evidence of such spread should trigger discussion and even follow-up action. Yet the outbreak of microcephaly related to Zika virus contradicts this hypothesis, as the virus affected 21 countries before a PHEIC was finally declared. Only three countries had H1N1 infections

3

87

37

38

138

Days from index case to detection

Days from detection to PHEIC declaration

65

0 50 100 150 200 250

H1N1

Ebola

Zika

Number of Days

102 days

225 days

41 days

1 : 1.8

1 : 1.6

1: 12.7

Note. PHEIC = public health emergencies of international concern. The H1N1 outbreak began March 15, 2009, was detected March 18, 2009, and was declared a PHEIC April 25, 2009. The West African Ebola outbreak began December 26, 2013, was detected March 22, 2014, and was declared a PHEIC August 8, 2014. For the Zika outbreak, the main concern was microcephaly rather than Zika virus infections; thus we consider this outbreak to have begun October 22, 2015, when the rise in microcephaly cases was first identified, and we consider the outbreak detected November 28, 2015, once there was strong evidence of a link between the virus and the microcephaly cases. The Zika outbreak was declared a PHEIC February 1, 2016.

FIGURE 1—Delays From Index Case to Outbreak Detection to PHEIC Declaration

AJPH PERSPECTIVES

330 Commentary Peer Reviewed Hoffman and Silverberg AJPH March 2018, Vol 108, No. 3

when that outbreak was declared a PHEIC.

A third hypothesis is that the number of people at risk predicts mobilization, whereby diseases that easily spread to whole pop- ulations might be considered more concerning than those dangerous to smaller groups. H1N1, an airborne disease, could easily spread across whole pop- ulations. Although Zika infection is widespread, microcephaly is concentrated within a small subset of the population: fetuses within pregnant women. Ebola falls between the two; in West Africa, it primarily infected those caring for the sick and often spread to family and health care

providers, including overseas volunteers. This hypothesis is therefore not supported by the three cases.

A fourth hypothesis is that disease novelty breeds fears of doomsday scenarios that moti- vate quicker precautionary action and that familiarity correspond- ingly causes global quiescence that lulls people into a false sense of security. This possibility can- not be ruled out on the basis of the last three PHEICs: H1N1 was a novel strain of influenza with quick mobilization, whereas the well-known Ebola virus elicited the slowest response. Although there was significant global experience with influenza

outbreaks, Ebola had never af- fected large enough populations to cause an international emer- gency. The novelty of the West African setting and transmission within urban areas might even have precipitated delays in mo- bilization rather than action. Additionally, greater knowledge about familiar entities, such as influenza, might hasten mobili- zation attributable to a more complete understanding of the risk.

A fifth hypothesis is that the mode of transmission and its perception by the public affects mobilization. Airborne diseases, like H1N1, can spread more broadly, put more people at risk,

be more difficult to contain, in- still greater fear, and require more stringent precautions than do diseases like Ebola that spread only through contact and drop- lets. Vector-borne diseases like Zika are more predictable than are airborne diseases—probably eliciting less fear—but are harder for individuals to protect them- selves against than are contact- borne infections. Vector-borne diseases require different infec- tion prevention and control mechanisms, as quarantine and hygienic medical practice will not prevent their spread.

Furthermore, Zika was found to be sexually transmissible weeks before the PHEIC was

TABLE1—Hypotheseson WhatElicits Faster Outbreak Responses andEvidencefor Each HypothesisFrom the H1N1,Ebola, andZika Outbreaks

Hypothesis H1N1 (Fastest) Zika (Intermediate) Ebola (Slowest) Supported by 3 Cases? Explanation

More severe

disease

Least severe Intermediate severity Most severe Not supported The more severe diseases did not elicit

faster responses than did the less

severe diseases

More countries

affecteda 3 countries 21 countries 5 countries Not supported H1N1 elicited fastest response after

only 3countries,whereasZika’s slower

response came after 21 countries

were affected

More people

at risk

Everyone at risk Mostly future babies at risk Close contacts and caregivers

at highest risk

Not supported Ebola, which risked an intermediate-

sized population, had the slowest

response

Disease novelty No experience Some experience Considerable experience Supported Novel H1N1 virus elicited a fast

response, whereas the known viruses

elicited slower responses

Greater ease of

transmission

Airborne Vector, mosquito, sexual Contact, sexual Supported H1N1, an airborne infection, led to fast

response; Ebola, with more easily

preventable transmission, had delayed

response

Not spread

during holiday

seasons

No holiday Christmas Summer Supported Zika and Ebola, which worsened when

many public health professionals took

vacations, elicited slower responses

Impact on

US citizens

First US citizen infected

10 d before PHEIC

First transmission in the US

confirmed 7 d before PHEIC

First infected person in the US

6 d before PHEIC

Supported PHEICs consistently declared within

10 d of having a direct impact on

US citizens

Note. PHEIC = public health emergencies of international concern.This table provides a summary of the evidence that either supports or does not support the 7 identified hypotheses. We considered hypotheses supported if their categorical ordering across the 3 PHEICs matches the order of how quickly the world responded to those PHEICs. aNumber of countries with at least 1 infected person at the time of PHEIC declaration. For Zika, this is the number of countries with confirmed cases of Zika infection, not clusters of microcephaly.

AJPH PERSPECTIVES

March 2018, Vol 108, No. 3 AJPH Hoffman and Silverberg Peer Reviewed Commentary 331

declared; although the evidence was sparse, this could have con- tributed to mobilization because of additional public fear of a sec- ond mode of transmission that may be harder to detect and the virus’s ability to spread to areas without the main Aedes aegypti mosquito vector. If airborne diseases are most difficult to contain, they would more easily trigger an emergency. This is supported by the speed of PHEIC declaration during the H1N1 outbreak and delays for the other two. Airborne transmission may also more easily spread across in- ternational borders, which is the concern that would trigger a PHEIC. Although Ebola was found in 2015 to also be sexually transmitted, this was long after local and international commu- nities had taken action.

A sixth hypothesis is that faster responses can be expected when diseases worsen outside holiday seasons, when public health au- thorities (e.g., WHOstaff) are not on vacation. Perhaps it is just a wicked coincidence, but the disease with the slowest mobili- zation (Ebola) worsened over Geneva’s long summer holidays, and the disease with the second slowest mobilization (Zika) spread over the shorter Christmas holidays. The fastest mobilization (H1N1) was mounted when public health professionals, both locally and internationally, were at work—consistent with this hypothesis. Although data on staff vacation periods are not publicly available, this co- incidence leads one to wonder whether staff availability played any role in delayed mobilization.

A final hypothesis is that outbreaks affecting US citizens, whose country is the world’s most powerful, are responded to most quickly; or, moreconcerning,that direct impact on the United States is a necessary condition for a

PHEIC’s declaration. Evidence from all three case studies supports this hypothesis. For H1N1, a PHEIC was declared 10 days after the first US citizen had a documented infection, on April 15, 2009.17 Ebola was declared a PHEIC only six days after in- fected health care providers ar- rived for the first time in the United States on August 2, 2014, for treatment, although it co- incided with the introduction of Ebola into Nigeria via a traveler from Liberia. Zika was declared a PHEIC two weeks after a Zika-positive microcephalic baby was born in Hawaii on January 15, 2016, and one week after the United States’ first laboratory-confirmed case of Zika on January 25, 2016, in the US Virgin Islands. Although this may simply reflect the na- ture of a PHEIC that there be demonstrable potential for in- ternational spread and the po- tential populations at risk, other examples of international spread in these outbreaks do not appear to be as closely tied to initiation of the global political response.

CONCLUSIONS Understanding why collective

action is delayed in global disease outbreaks provides guidance for hastening future responses. Using these three PHEICs as case studies, we have seen that greater delay seems to be attributable to poor political mobilization rather than technical surveillance ca- pacity. This is in sharp contrast to the post-Ebola calls for increased surveillance as the key to ap- propriate outbreak response.18–20

More of the solution might actually reside in improving global institutional structures that currently slow political mobilization.

Some of these hypotheses, if true, could be seen as justifiable reasons for delayed political ac- tion. It would make sense that mobilization was slower for dis- eases that were less severe, af- fected fewer countries, or risked fewer people. Unfortunately, none of these hypotheses are consistent with the three recent PHEICs. Instead, half of the hypotheses that remain are ac- tually extremely troubling, in that they point to the hazards of holidays and US exceptionalism.

Properly testing the four remaining hypotheses—that dis- ease novelty, ease of transmission, transmission during nonholiday seasons, and outbreaks that af- fect US citizens hasten global mobilization—requires studying more cases. Process tracing from political science is one important tool for better understanding why different outbreaks receive different responses. Qualitative analysis of news media and po- litical speeches can also be help- ful.21 There are additional insights to be gained from probing why other recent out- breaks, such as the Middle East Respiratory Syndrome corona- virus, were not declared PHEICs. More research should also be conducted to better understand the role that scientific uncertainty may play in delaying political responses to outbreaks, as well as to better understand the addi- tional factors beyond lagging political mobilization that con- tribute to delayed action.

Ultimately, we hope the causes of political delays are sci- entifically and ethically justifi- able, even though we recognize that delays are probably multi- factorial and include both en- lightened and unsavory elements. Once the causes are known, global decision-makers should be able to better target policy in- terventions, open important

conversations about how we prioritize disease control efforts, and identify the kinds of global institutional changes needed to reduce the spread and severity of pandemics.22–26 Further research may necessitate changes to the way we evaluate and declare outbreaks as PHEICs if the causes of political delays are confirmed to be scientifically or ethically suspect. In the meantime, we should plan additional hypothesis-testing research and experiment with mechanisms to reduce political delays so that global disease outbreaks are addressed as quickly as possible.

CONTRIBUTORS S. J. Hoffman conceptualized the idea and designed the study. S. L. Silverberg led the data analysis. Both authors cowrote the article, approved the final article, and guarantee the work.

ACKNOWLEDGMENTS This work was completed as part of the International Collaboration for Capitaliz- ing on Cost-Effective and Life-Saving Commodities, which is funded by the Research Council of Norway’s Global Health & Vaccination Programme (GLOBVAC project 234608). S. J. H. is additionally funded by the Canadian Institutes of Health Research and the Ontario government’s Ministry of Re- search, Innovation and Science.

S. J. H. previously worked for the World Health Organization.

REFERENCES 1. Hoffman SJ. The evolution, etiology and eventualities of the global health se- curity regime. Health Policy Plan. 2010; 25(6):510–522.

2. Ottersen T, Hoffman SJ, Groux G. Ebola again shows the international health regulations are broken: what can be done differently to prepare for the next epi- demic. Am J Law Med. 2016;42(2–3): 356–392.

3. World Health Organization. Report of the Ebola Interim Assessment Panel—July 2015. Geneva; 2015.

4. Moon S, Sridhar D, Pate MA, et al. Will Ebola change the game? Ten essential reforms before the next pandemic. The report of the Harvard-LSHTM in- dependent panel on the global response to Ebola. Lancet. 2015;386(10009): 2204–2221.

5. Commission on a Global Health Risk Framework for the Future. The Neglected

AJPH PERSPECTIVES

332 Commentary Peer Reviewed Hoffman and Silverberg AJPH March 2018, Vol 108, No. 3

Dimension of Global Security: A Framework to Counter Infectious Disease Crises. Wash- ington, DC: National Academies Press; 2016.

6. United Nations. Protecting Humanity From Future Health Crises: Report of the High-Level Panel on the Global Response to Health Crises. New York; 2016.

7. Hsieh YH, Ma S, Hernandez JX, Lee VJ, Lim WY. Early outbreak of 2009 influenza A (H1N1) in Mexico prior to identification of pH1N1 virus. PLoS One. 2011;6(8):e23853.

8. Centers for Disease Control and Pre- vention. Outbreak of swine-origin influenza A (H1N1) virus infection— Mexico, March–April 2009. MMWR Morb Mortal Wkly Rep. 2009;58(17): 467–470.

9. World Health Organization. Origins of the 2014 Ebola epidemic: one year into the Ebola epidemic. 2015. Available at: http://www.who.int/csr/disease/ebola/ one-year-report/virus-origin/en. Accessed December 15, 2017.

10. Centers for Disease Control and Prevention. Increase in reported preva- lence of microcephaly in infants born to women living in areas with confirmed Zika virus transmission during the first trimester of pregnancy—Brazil, 2015. MMWR Morb Mortal Wkly Rep. 2016; 65(9):242–247.

11. Ministério da Saúde. Ministério da Saúde divulga boletim epidemiológico. 2015. Available at: http://portalsaude. saude.gov.br/index.php/cidadao/ principal/agencia-saude/20805- ministerio-da-saudedivulga-boletim- epidemiologico. Accessed December 12, 2017.

12. Ministério da Saúde. Portaria GM no 1.813, de 11 de Novembro de 2015. 2015. Available at: http://bvsms.saude.gov.br/ bvs/saudelegis/gm/2015/prt1813_11_ 11_2015.html. Accessed December 15, 2017.

13. Vetter P, Dayer JA, Schibler M, et al. The 2014–2015 Ebola outbreak in West Africa: hands on. Antimicrob Resist Infect Control. 2016;5(1):17.

14. World Health Organization. State- ment on the 1st meeting of the IHR Emergency Committee regarding the 2014 Ebola outbreak in West Africa. 2014. Available at: http://www.who.int/ mediacentre/news/statements/2014/ ebola-20140808/en. Accessed December 15, 2017.

15. Pan-American Health Organization. Neurological Syndrome, Congenital Malfor- mations, and Zika Virus Infection. Implication for Public Health in the Americas. Wash- ington, DC; 2015.

16. World Health Organization. WHO statement on the first meeting of the In- ternational Health Regulations (2005) Emergency Committee on Zika virus and

observed increase in neurological disor- ders and neonatal malformations. 2016. Available at: http://www.who.int/ mediacentre/news/statements/2016/1st- emergency-committee-zika/en. Accessed December 15, 2017.

17. Centers for Disease Control and Prevention. Swine influenza A (H1N1) infection in two children—Southern California, March–April 2009. MMWR Morb Mortal Wkly Rep. 2009;58(15): 400–402.

18. Gates B. The next epidemic—lessons from Ebola. N Engl J Med. 2015;372(15): 1381–1384.

19. Tomori O. Will Africa’s future epi- demic ride on forgotten lessons from the Ebola epidemic? BMC Med. 2015;13:116.

20. Kalra S, Kelkar D, Galwankar SC, et al. The emergence of Ebola as a global health security threat: from ‘lessons learned’ to coordinated multilateral containment ef- forts. J Glob Infect Dis. 2014;6(4):164–177.

21. Hoffman SJ, Justicz V. Automatically quantifying the scientific quality and sensationalism of news records mention- ing pandemics: validating a maximum entropy machine-learning model. J Clin Epidemiol. 2016;75:47–55.

22. Hoffman SJ, Caleo GM, Daulaire N, et al. Strategies for achieving global col- lective action on antimicrobial resistance. Bull World Health Organ. 2015;93(12): 867–876.

23. Hoffman SJ, Røttingen JA, Frenk J. Assessing proposals for new global health treaties: an analytic framework. Am J Public Health. 2015;105(8):1523–1530.

24. Gopinathan U, Watts N, Hougendobler D, et al. Conceptual and institutional gaps: understanding how the WHO can become a more effective cross-sectoral collaborator. Global Health 2015;11(46):1–13.

25. Hoffman SJ. Mitigating inequalities of influence among states in global decision making. Global Policy. 2012;3(4):421–432.

26. Hoffman SJ, Røttingen JA. Assessing implementation mechanisms for an in- ternational agreement on research and development for health products. Bull World Health Organ. 2012;90(12): 854–863.

AJPH PERSPECTIVES

March 2018, Vol 108, No. 3 AJPH Hoffman and Silverberg Peer Reviewed Commentary 333

Copyright of American Journal of Public Health is the property of American Public Health Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.