THE EVALUATION OF HEALTHCARE INTERVENTIONS

Abstract

This dissertation studies statistical inference methods and optimization models to address, methodologically and practically, the emerging challenges in three increasingly recognized healthcare interventions. First, how to decide among multiple medical treatment interventions in the presence of treatment heterogeneity? Second, what is the consequences of potential interventions for a prevalent occupational symptom at different scales of decision-makings? Third, how to design dynamic interventions in hospital operations to achieve some specific operational criteria? These problems are important in health delivery yet lack satisfactory answers thus far, because of either the difficulty entailed in available methods and data, and/or the insufficient attention given to the problem. In this dissertation, we explore either methodological or practical solutions to these real-world problems.