Summary
The episode works through four questions that students commonly find difficult in case-control design. It uses a hypothetical potluck outbreak to show why an odds ratio of six can correspond to a risk ratio of three when the outcome is common, and it explains when the odds ratio estimates a risk ratio or a rate ratio. It examines control selection through the 1950 Doll and Hill study of smoking and lung cancer and the 1981 study of coffee and pancreatic cancer, and it works a hypothetical example in which differential recall moves an odds ratio from two to about 2.4. The episode then discusses the case-crossover design using a Toronto study of cellphone use and collisions, and the hosts close by debating whether case-control studies deserve their place below cohort studies in evidence hierarchies.