Non-clinical sciences/Stats & evidence · Study design

Case-control study

[object Object]

What it means

Starts from outcome: cases (with disease) vs controls (without) compared for past exposure → odds ratio. Efficient for rare diseases and multiple exposures; cheap/fast; prone to recall and selection bias.

Worked example

Comparing thalidomide exposure in mothers of babies with vs without phocomelia is a case-control study.

In the exam

To study a rare malformation, investigators identify affected and unaffected children and look back at their mothers' drug exposures.

What settles it

Case-control = backward from outcome (OR, rare disease, recall bias); cohort = forward from exposure (RR).

Classically confused with

Cohort study

Source: StatPearls