Epi 505 – ActivEpi Lesson 2 Outline

How to Obtain Data

· Interview (Face to face, telephone, mail, e-mail, …)

· Review existing records (medical records, work records, …)

· Laboratory

· Examination (physical, CAT scan, ultrasonography …)

Select Appropriate Study Design (Lesson 3)

· Clinical Trial

· Cohort

· Case-Control

· Cross-Sectional

· Other

In class example: relationship between

Diet -> Breast cancer

Age

Smoking

Obesity

Physical activity

Measuring the Variables - Decisions needed on exact definitions. For example

· Illness in Sydney beach study – self report vs. physician dx?

· Define “swimming”

· Define “water quality”

Measures of Disease Frequency (Lesson 4)

Risk of disease in exposed: R(E)

Risk of disease in nonexposed: R(not E)

Risk Ratio

Consider potential biases/flaws (Lessons 7-11):

· Study design

· Selection of participants (selection bias)

· Information collected from/on participants (information bias)

· Confounding

Methods of data analysis

Biases and flaws may lead to incorrect conclusions

Yes-P = swam in polluted water

Yes-NP = swan in non-polluted water

No = did not swim

Test for trend (Lesson 14)

Examples of Epidemiologic Studies

Alcohol consumption and breast cancer in the Nurses Health Study

Example of cohort study

Identify group of disease-free subjects

Collect exposure information

Follow over time

Results: alcohol associated with breast cancer with dose response effect

Bogalusa Outbreak

Legionnaires disease in Bogalusa

Example of case-control study

Result: associated with misting machine for vegetables at a grocery store

The Rotterdam Study of Alzheimer’s Disease

Cohort study of 8000 elderly

Assessed relation of smoking to Alzheimer’s accounting for age, gender, education, and alcohol consumption.

Result: smoking associated with increased risk of Alzheimer’s