From Bayesian Diagnostics to Bayesian Health Economics

Abstract

This paper presents a Bayesian approach to clinical decision-making that incorporates utility and cost in the prediagnostic stage, which is the prevalence rate where a clinician is indifferent between treatment and no treatment given no further diagnostic information. The expected utility and cost of diagnostic information are then compared against this background. Our framework generalises the threshold concept in clinical decision analysis and adds to it the marginal cost-benefit criterion that determines the choice between several diagnostic strategies. The paper provides and discusses an application of this framework in cardiography.