Session 2 Lecture 2: Introduction to Epidemiological Study Designs

Question 1

How do you conduct an Ecological study?

Answer 1

-Identify groups to study

-Define the characteristics to be studied
Exposure and outcome

-Decide how analysis is conducted :
Counting categorical data (nominal or ordinal scale) or measuring continuous data (interval or ratio scale)

• Gather data on the group level characteristics
• Unit of analysis is groups

Question 2

Name 4 specific issues with Ecological studies

Answer 2

• definition of characteristics
• measurement variation
• confounding factor such as ecological fallacy
• chance (random error)

Question 3

What is ecological fallacy ?

Answer 3

Falsely inferring individual level association from group level association

Question 4

What are the features of a cross-sectional study?

Answer 4

• unit of analysis is individuals
• snapshot of data at 1 point in time
• routinely collected

Question 5

Name 3 specific issues with cross-sectional studies

Answer 5

• sampling bias
• responder/participant bias
• chance (random error)

Question 6

How do you Conduct a case-control study? (always retrospective: looking at old data)

Answer 6

• identify a group of cases on outcome status
• identify a suitable group of non cases (controls) on outcome status
• discover previous exposure status of everyone
• compare level of exposure in case and controls
• odd ratio used as an analysis

Question 7

Name 4 specific issues with case-control studies

Answer 7

• selection bias: controls should reflect study pop and should be comparable to cases
• information bias
• confounding factors
• chance (random error)

Question 8

State the two type of cohort studies

Answer 8

• historical (data collection in past but still follow up overtime)
• prospective ( allows for temporal sequence)

BUT cohort studies are all prospective

Question 9

How do you conduct a cohort study?

Answer 9

• identify outcome free individuals
• group individuals according to level of exposure
• follow up over time until you discover outcome for everyone
• analysis can be incidence rates or odds ratio
• compare rates for each exposure groups

Question 10

Name 4 specific issues with Cohort studies

Answer 10

1. Loss to follow up (people dropping out)
   - differential bias
   - survivor bias
2. Information bias, especially for outcome
   - differential misclassification
   - non-differential misclassification
3. Confounding
4. Chance (random error)

Question 11

What method should be used when describing or finding a study?

Answer 11

PICO

P - What is the Population to be studied?
I - What is the Intervention / Exposure of interest?
C - What is the Comparison / Control of interest?
O - What is the Outcome of interest?