A Bayesian approach to health project estimation

Abstract

This paper illustrates how a Bayesian statistical approach was used to estimate the outcomes of the National Tuberculosis Program in India. Such an estimate, it is argued, is necessary for a proper judgment about a project's social usefulness. The process of medical care delivery is reduced to a set of conditioned probabilities. The numbers are estimated using as source material medical records, the results of medical research, and the opinion of experts. Bayesian methods of estimation are used and their value is discussed. The final discussion contains a brief treatment of the role of project analysis in public decision making. The place of Bayesian methods in project analysis is briefly illustrated, demonstrating their operational value in the field of public health decision making.