Case–control study

A case–control study is an observational design that starts from the outcome rather than the exposure. Investigators recruit people who already have a particular disease or condition (the cases) and a comparable group who do not (the controls), then look back at each group's history to compare how often they were exposed to a possible cause, such as a food, drug, job or habit. If an exposure is much more common among cases than controls, it is associated with the disease.

Because researchers decide how many cases and controls to include, the design cannot measure incidence or prevalence. Instead, the association is expressed as an odds ratio, which approximates the relative risk when the disease is uncommon. Case–control studies are quick and inexpensive, suit rare diseases and those with long delays between exposure and illness, and can examine many exposures at once. Their weak points are choosing a fair control group and recall bias, since people who are ill often remember past exposures more thoroughly than healthy controls.

Outbreak investigators use this design constantly. Comparing what sick and healthy people ate, for example, can trace a hepatitis A outbreak to a single ingredient.