Abstract

OBJECTIVE--To review current knowledge of the effectiveness of medical audit programmes as a whole and of specific interventions within these programmes, as a means of changing clinical behaviour. CRITERIA FOR INCLUSION AND EXCLUSION OF PUBLISHED REPORTS--Articles listed on Medline from 1985-92 with key words "quality assurance" or "medical audit", and "evaluation" and relevant references from these articles, and from recently published reviews and reports on medical audit, were included. Excluded were simple descriptions of audit activity, replications of previous work, and publication in a language other than English. RESULTS--Evaluation of entire programmes of medical audit is unusual. Most reports concern specific interventions and focus particularly on the scientific and technical aspects of quality. These interventions may be classified by the means through which they attempt to achieve desired changes: patient characteristics; physician characteristics; administrative and organisational structures; and financial incentives. CONCLUSIONS--Knowledge about effective methods of bringing about specific changes in clinical behaviour is rudimentary. Impact is highly dependent on local factors, so generalisation of results to other settings is difficult. More qualitative research is needed to define the local factors which influence results.