This article is freely available to all

Article Abstract
It is increasingly well-recognized that the results of highly controlled randomized clinical trials (RCTs) provide an imperfect foundation for the practice of evidence-based medicine. Such trials typically exclude too many people to be readily generalizable to the larger population of affected individuals. Moreover, the very controls that ensure the internal validity of an RCT, including the use of a placebo comparison group and "blinded" assessment of outcome according to reliable, standardized measures, have little ecological validity for practicing clinicians and their patients. The results of RCTs thus must be complemented by other types of data, including illustrative case series of more representative groups of patients.