Applying Bayesian ideas to the development of medical guidelines

Abstract

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 development of medical guidelines based on expert opinion, using ordinal categorical rater data. We develop guidelines for the use of coronary angiography following an acute myocardial infarction (AMI) for 890 clinical indications using statistical models fit to appropriateness ratings obtained from a nine-member expert panel. The main foci of our analyses were on the estimation of an appropriateness score for each of the clinical indications, an associated measure of precision, and functions of the underlying score. We considered 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 distribution, while permitting rater effects and indication heterogeneity in both. We estimated models using Markov chain Monte Carlo methods and constructed indices quantifying appropriateness based on posterior probabilities of selected model parameters. We compared our model-based approach to the standard approach currently employed in medical guideline development and found that the standard approach correctly identified 99 per cent of the appropriate indications while overestimating appropriateness 18 per cent of the time compared to our model-based approach.

Rating the Appropriateness of Coronary Angiography — Do Practicing Physicians Agree with an Expert Panel and with Each Other?

“…To compare more reliable estimates of the appropriateness of coronary angiography as rated by the surveyed physicians and the expert panelists and to determine the precision associated with these estimates, hierarchical regression models 21 were fitted separately to the observed ratings from each group. This method accounted for three important sources of variability: the underlying appropriateness of each indication, each physician's propensity to rate angiography as more or less appropriate in general, and measurement error.…”

Common sense and figures: the rhetoric of validity in medicine (Bradford Hill Memorial Lecture 1999)

“…We are slowly beginning to see this take place. For example, Mary Beth Landrum and Sharon-Lise Normand recently reported their use of Bayesian methods to develop guidelines for coronary angiography based on expert opinion [59].…”

Bayesian statistics in medicine: a 25 year review

“…More complex designs or issues were increasingly being studied mainly by full Bayesian approaches based on Markov chain Monte Carlo techniques including population pharmacokinetic modelling [370], population approaches to dose selection [371], prevalence estimates for depression in adolescents from two-stage sampling [372], bivariate survival models to jointly model hospital stay and community stay in assessing the effect of insurance policies on mental health care [373], hierarchical models to examine predictors of results in oral practice examinations in anaesthesiology [374], prevalence surveys in HIV accounting for imprecision in sensitivity and specificity [375], developing medical guidelines for coronary angiography following acute MI [376], back-calculating the time of transmission of HIV from mother to child [377], back-calculating age-specific incidence of HiB [378], classifying individuals based on predictors of random effects in HIV/AIDS [379] and risk of HIV infection as a function of duration of intravenous drug use [380]. Longitudinal models were developed further to cope with unequally spaced observations [381], and shrinkage estimates of immunological progression rates in HIV [382].…”