2.3 Selection and sources of cases and controls

noang.org/
In all case-control studies, as in all epidemiological studies, we must begin with a clearly defined question or hypothesis. We also need to establish a clear case definition at the very outset of the study. This lets us implement a case selection process using standardised diagnostic criteria. Selecting good controls in a case-control study can be problematic. We have to do it in a way that minimises bias but is also representative of the population from which cases are found.