Case-control study

A case-control study works in reverse compared with a cohort study. Researchers assemble a group of people who already have the outcome (cases) and a comparison group without it (controls), then compare how often each group was exposed to a suspected risk factor in the past. This design is efficient for rare diseases or outcomes that take a long time to develop, because you do not have to follow a huge population and wait. It typically yields an odds ratio. When reading one, focus on how controls were selected (they should come from the same population that produced the cases) and on recall bias, since cases may remember past exposures differently from controls. Selection of controls is where these studies most often go wrong. A well-known example linked diethylstilbestrol taken during pregnancy to a rare vaginal cancer in daughters, detected by comparing affected young women with unaffected ones. The question it lets you ask: were cases more likely than controls to have had the exposure?