The case-control study is often a very simple, cheap and quick way to test the hypothesis that a particular factor causes a disease or that a particular treatment is helping patients who receive it. There also seems to be an inherent logic in comparing patients and well people to see how they differed in the past, or in comparing those who did and did not survive a certain illness to see how their treatment differed. Despite these appealing features, the case-control approach is subject to a number of pitfalls. These can usually be avoided by an appropriately designed study, and by careful collection and analysis of the data. Even so, the results of a case-control study are a lower order of evidence than those produced from a properly conducted randomised controlled trial. (Author/publisher)
Abstract