An example of medical audit under political pressure.

Management of medical services necessitates decisions regarding the possibly ineffective treatment programmes, hospitals and (even) individual practitioners. Although rigorous evaluations of effectiveness can only be achieved by randomized trials there has been, in recent years, much increase in audit (or peer review) activity. Ideally medical audit is undertaken as a collaborative exercise, and follows preplanned methods of data collection and evaluation of outcome. However, real management situations frequently do not conform to the ideal, as when the medico-political finger of accusation points at one particular hospital and asks for an immediate enquiry. This audit analyses the mortality and morbidity after coronary artery bypass graft surgery in the 'suspect' hospital and in control hospitals. The paper describes the difficulties of obtaining comparable records post hoc and the interpretations of small differences, mostly non-significant statistically yet possibly 'real', both in the characteristics of patients operated on and in the outcomes of operations.