Applying Bayesian ideas to the development of medical guidelines

Measurements of the quality of health care, in particular the underuse and overuse of medical therapies and diagnostic tests, often involve employment of medical practice guidelines to assess the appropriateness of treatments . This paper presents a case study of a Bayesian analysis for the developme nt of medical guidelines based on expert opinion, using ordinal categorical rater data. We develop guidelines for the use of coronary angiography foll owing an acute myocardial infarction (AMI) for 890 clinical indications usi ng statistical models fit to appropriateness ratings obtained from a nine-m ember expert panel. The main foci of our analyses were on the estimation of an appropriateness score for each of the clinical indications, an associat ed measure of precision, and functions of the underlying score. We consider ed two classes of models that assume the ratings are either in the form of grouped normal data or are ungrouped variables arising from a normal distri bution, while permitting rater effects and indication heterogeneity in both . We estimated models using Markov chain Monte Carlo methods and constructe d indices quantifying appropriateness based on posterior probabilities of s elected model parameters. We compared our model-based approach to the stand ard approach currently employed in medical guideline development and found that the standard approach correctly identified 99 per cent of the appropri ate indications while overestimating appropriateness 18 per cent of the tim e compared to our model-based approach. Copyright (C) 1999 John Wiley & Son s, Ltd.