Individual case studies in clinical research

Case studies have acquired an unmerited reputation as being anecdotal, unscientific and intrinsically inferior to group studies. The subsequent disregarding of individual patients as the focus of investigation has led to the neglect of an extremely useful clinical research method, and has probably impaired the pace of therapeutic innovation. The purpose of this paper is to clarify the scope and nature of case studies and promote their rehabilitation. Case studies can, in principle, be used to test any theory that has implications for individual patients. There are two crucial methodological stages. The first is to identify scientifically plausible general theories and derive from them specific hypotheses or models of sufficient precision to have implications for individual cases. The second is to test these hypothetical models against 'pure' cases, selected so as to exclude interfering variables. There are two main types of case study--those made by serendipity (unplanned case observations which challenge an implicit theoretical framework); and formal case studies (designed prospectively to collect pure cases to test a prior hypothesis). The difference between serendipity and planned case studies roughly corresponds to the difference between surveillance and screening. A worked-example of a formal case study is described here in order to illustrate the method. Individual case studies deserve fresh consideration by researchers, since they are a clinician-friendly method with a unique potential for incorporation into routine practice.