How to plan a good case-control study?

Case-control studies are a very frequently used restudy design in clinical and epidemiological research. They are popular because they are easy, quick and inexpensive to plan and conduct. Despite the risk of several types of biases (selection bias, recall and measurement bias, and confounding), their results are often fairly robust. This article explains the basic steps in the design and conduct of case-control studies, and also explains their usual pitfalls, from the perspective of a clinical researcher.