Epidemiology Master Level Quiz

Descriptive Study Designs: Case-Series, Cross- Sectional, and Ecologic Studies

David Celentano, ScD, MHS Johns Hopkins University

JHU Vision of Epidemiology

2

Main Types of Epidemiologic Study Designs

3

Study type Characteristics

Experimental • Studies prevention and treatment of disease • Investigator actively manipulates which groups receive the study agent

Observational • Studies causes, prevention and treatment for diseases • Investigator watches as natures takes its course

Cohort • Examines multiple health effects of an exposure • Subjects defined by exposure levels and follow for disease occurrence

Case-control • Typically examines multiple exposures in relation to a disease • Subjects are defined as cases and controls and exposure histories compared

Cross-sectional • Examine relationship between exposure and disease prevalence in a defined population at one point in time

Ecological • Examines relationship between exposure and disease with population-level data rather than individual data

The material in this video is subject to the copyright of the owners of the material and is being provided for educational purposes under rules of fair use for registered students in this course only. No additional copies of the copyrighted work may be made or distributed.

Descriptive Epidemiology

!  Definition

Descriptive Epidemiology

5

!  Distribution of disease and its determinants is the domain of descriptive epidemiology !  Analysis of disease patterns according to the characteristics of the person, place

and time

!  Who is getting the disease?

!  Where is it occurring?

!  How is it changing over time?

Descriptive Epidemiology

6

The Epidemiologic Triad

7

Descriptive Epidemiology: Person

!  Age

!  Gender

!  Race and ethnicity

!  Socioeconomic status

!  Occupation

!  Religion

!  Marital status

8

!  Defined in geopolitical units or natural geographic features

!  Encompasses aspects of the environment: !  Physical environment (climate, water, air) !  Biological environment (flora and fauna) !  Social environment (cultural traditions)

Descriptive Epidemiology: Place

9

Malaria Risk Map in Sub-Saharan Africa: 10 Sept 2013

10

Descriptive Epidemiology: Time

11

!  Case reports and case series

!  Cross-sectional surveys

!  Exploratory ecological designs

Descriptive Study Designs

12

Case reports and case series

13

!  A 74-year-old woman was experiencing airway obstruction when food was lodged in her trachea

!  The Heimlich maneuver was successful in dislodging the food

!  In the ER, she complained of abdominal pain and distention

!  A 2 cm rupture of the lesser curvature of stomach was detected and corrected by surgery

Case Report: Profile of a Single Individual

14

Case Series: a Small Group of Patients with Similar Diagnoses

15

MERS-CoV Outbreak: Biological Association? Epidemiology?

16

MERS coronavirus

Egyptian tomb bat

Camels

June 5, 2014

Case Series of MERS-CoV in 2013

Source: http://www.nature.com/polopoly_fs/7.12529.1379429726!/image/MERS.jpg_gen/derivatives/fullsize/MERS.jpg 17

MERS-CoV Outbreak KSA 2012–2013

18

Influenza Season, USA, 2013–14

19

Influenza-Like Illnesses USA: Recent Years

20

The material in this video is subject to the copyright of the owners of the material and is being provided for educational purposes under rules of fair use for registered students in this course only. No additional copies of the copyrighted work may be made or distributed.

Cross-Sectional Studies

2

Cross-Sectional Studies: A Snapshot in Time

3

Cross-Sectional Study—1

4

Cross-Sectional Study—2

5

Cross-Sectional Study Comparisons

► First, identify a population of n persons to study

► Determine exposure and disease for each study participant ► a persons who were exposed and have disease ► b persons who were exposed but without disease ► c persons who have the disease but not exposed ► d persons who have not been exposed nor have the disease

6

OR

Cross-Sectional Study—3

► Prevalence of disease in exposed compared to non- exposed

► 𝑎𝑎

𝑎𝑎+𝑏𝑏 𝑣𝑣𝑣𝑣. 𝑐𝑐

𝑐𝑐+𝑑𝑑

► Prevalence of exposure in diseased and non- diseased

► 𝑎𝑎

𝑎𝑎+𝑐𝑐 𝑣𝑣𝑣𝑣. 𝑏𝑏

𝑏𝑏+𝑑𝑑

7

Key Features of Cross-Sectional Studies

► Examine exposure prevalence and disease prevalence simultaneously

► Cannot infer temporal sequence between exposure and outcome (especially if exposure can change)

► Preponderance of prevalent cases of long duration

► Healthier participants/volunteers a concern

8

How Do We Interpret an Association?

► If there seems to be an association between exposure and outcome (e.g., increased cholesterol and CHD), we have to confront several issues… 1. Identify prevalent cases (not incident cases), which may not be representative of the

population of cases ● Including only prevalent cases excludes deaths ● The association may reflect survival after developing the disease (CHD) and not the

risk of developing the disease

9

Another Issue in Interpreting an Association

► If there seems to be an association between exposure and outcome (e.g., increased cholesterol and CHD), we have to confront several issues… 2. Cannot establish temporality between exposure and outcome in cross-sectional

studies ● If exposure does not precede a disease, association cannot reflect a causal relation

10

Features of Cross-Sectional Studies

► Interest is on identifying possible relationship between an exposure and an outcome ► For example, high-risk HPV and cervical dysplasia among women 15 to 24 years

► Cases are prevalent cases of the disease—we do not know when the disease began

► Hence, also known as a prevalence study

11

Sources of Bias in Cross-Sectional Studies

► Response bias—those who participate are systematically different from those who do not respond or who refuse

► Acquiescent response style—respondents try to “look good” for the interviewer

► Interviewer biases (leading the witness)

► However, these biases are often seen in other epidemiologic study designs as well!

12

Conclusions: Cross-Sectional Studies

► Cross-sectional surveys suggest possible risks for an outcome if an association is found

► If based on probability sample of a known population, highly generalizable

► Can be conducted “quickly” and may be comparatively less expensive

► Generates hypotheses for other epidemiologic studies ► Need to rely upon other study designs to establish etiologic relationships

13

HIV among MSM in India (N=12,021)

Source: Solomon, S. S., et al. (2015). High HIV prevalence and incidence among MSM across 12 cities in India. AIDS, 29(6), 723-31. http://dx.doi.org/10.1097/QAD.0000000000000602

14

National Center for Health Statistics Surveys

15

Age-Sex- Adjusted Percentage of Adults Aged 18 Years and Over Who Engaged in Regular Leisure- Time Physical Activity, by Race and Ethnicity, US, 2008

The material in this video is subject to the copyright of the owners of the material and is being provided for educational purposes under rules of fair use for registered students in this course only. No additional copies of the copyrighted work may be made or distributed.

Ecological Studies

Section C

!  Examination of rates of disease in relation to a factor described on a population level

!  “Units of analysis are populations or groups of people rather than individuals”

!  Aggregate measures that summarize: !  Individuals in a population (e.g., individuals > 65 years) !  An environmental measure (e.g., air pollution) !  A geographic location (e.g., county) or !  A global measure that has no individual analogy (e.g., population density)

Ecological Studies

2

Per Capita Egg Consumption and Colon Cancer Mortality among Men (n=31 Countries)

3

!  Key features that differentiate ecological studies: !  Population = unit of analysis (not individuals) !  Exposure status = property of the population

!  Often the first step in determining whether an association exists

!  Problem: we do not know if individual risk equates to group risk

Ecological Studies

4

!  The hypothesis is new

!  Adequate measurement of individual-level variables is not possible

!  Perhaps not ethical to conduct an individual-level study

!  We may be interested in the effect from ecological variables, for which there is no correlate at the individual level

!  Limited time or funds to do the study

Why Conduct an Ecological Study?

5

!  Biological: causation only through biological pathway

!  Ecological: causation only through group characteristic

!  Contextual: causation through biological and group characteristics

Causal Inference in Ecological Studies

6

!  Biological !  Among unvaccinated persons, low-income people have increased risk for Hepatitis B !  Having low income increases stress, which increases HBV risk

!  Ecological !  Among unvaccinated persons, people living in a low-income community have

increased risk for HBV because low-income communities have high exposure to HBV

!  Contextual !  Among unvaccinated persons, both high-income and low-income people who live in

a low-income community have increased risk for HBV because of increased stress and increased exposure

Causal Inference in Ecological Studies

7

Dietary Fat Intake and Breast Cancer by Country

8

A Recent Example to Outline Interpretation

9

!  Aggregate variables (measure exposure and outcomes) !  A summary or composite measure derived from values collected from individuals

(e.g., BP, DM2) !  Aggregate variables can measure exposures (BP) or outcomes (DM2) !  Limitation = variation within the population—not have the average blood pressure

!  Environmental variables (measure exposures) !  A measure of the physical characteristics of the environment in which people

reside, work, recreate or attend school (measure exposure); same limitations

!  Global variables (measure exposures) !  A measure of the attributes of groups, organizations, or places for which there is no

analogue at the individual level (e.g., population density)

Types of Ecological Variables

10

!  “An association observed between variables on an aggregate level does not necessarily represent the association that exists at the individual level”

!  Cannot necessarily infer the same relationship from the group level (egg consumption at the county or country and colon cancer among men) to the individual level (residents of a location)

!  Cannot fill in the 2x2 table from the data available in a traditional ecologic study

Ecological Fallacy or Ecological Bias

11

!  Presented information on case studies and case series as informative for possible associations

!  Described the design of cross-sectional studies—and their value and limitations

!  Demonstrated how ecologic studies fit in with other epidemiologic study designs

Summary

12

Your feedback is very important and will be used for future revisions. The Evaluation link is available on the lecture page.

Lecture Evaluation

13