Observational study design

Cohort, case control and related metrics

Prospective studies: cohort. Population at risk and incidence are directly observed over time

Retrospective studies: CC.

In cohort studies and RCT, participants are selected based on exposure status; in CC studies, they are selected based on outcome status.

Relative risk (risk ratio) RR, odds ratio OR

Case control studies

Odds ratio is usually used over risk ratio for the following reasons

• CC studies aim to compare the exposure history (risk factor) between cases and controls. As subjects are selected based on outcome, you can’t compute risk (which measures the population)
• incidence or risk calculation: it is not directly measureable you do not have total number of people developing the disease. In CC typically matched as a ratio
• odds is computable
• in rare outcomes, OR approximates RR
• logistic regression estimates OR

Interview questions

• Why use OR instead of RR