Abstract

Poor quality medical care is sometimes attributed to physicians’ unwillingness to act on evidence about what works best. Evidence-based performance standards (EBPSs) are one response to this problem, and they are increasingly employed by health care regulators and payers. Evidence in this instance is judged according to the precepts of evidence-based medicine (EBM); it is probabilistic, and the randomized controlled trial (RCT) is the gold standard. This means that EBPSs suffer all the infirmities of EBM generally—well rehearsed problems with the external validity of research findings as well as the inferential leap from study results in the aggregate to individual patient care. These theoretical weaknesses promise to have a practical impact on the care of patients. To avoid this, EBPSs should be understood as guidelines indicative of average effectiveness rather than standards to be applied in every case.