Transferability of Medical Decision Support Systems Based on Bayesian Classification

Abstract

This study tested the hypothesis that probabilities derived from a large, geograph ically distant data base of stroke patients could form the basis of an accurate Baye sian decision support system for locally predicting the etiology of strokes. Perform ance of this "extrainstitutional" system on 100 cases was assessed retrospectively, both by error rate and using a new linear accuracy coefficient. This approach to patient classification was found to be surprisingly accurate when compared to classi fication by physicians and to Bayesian classification based on "low cost" local and subjective probabilities. We conclude that for some medical problems Bayesian clas sification systems may be significantly more transferable to new sites than is gener ally believed. Furthermore, this study provides strong support for the utility of clin ical databases in building, transferring, and testing Bayesian classification systems in general. (Med Decis Making 3:501-509, 1983)