PRECIPITATING HEALTH BEHAVIOR CHANGE: THE ROLE OF TECHNOLOGY AND THE SOCIAL ENVIRONMENT

Abstract

Health behaviors represent the largest determinant of a person’s health and impact healthcare practice and delivery. Responding to the need to uncover mechanisms that instigate health behavior change, in this dissertation I investigate the effects of two important drivers of variation: health information technology and social health influence. I conduct empirical analyses using a unique medical and administrative dataset on 820,000 U.S. Army soldiers over four years. In the first study, I examine the effects of a patient portal implemented by the Army in 2011. Patient portals are used to facilitate greater patient engagement with health and increase patient activation, defined as the ability and desire to improve one’s health. A critical, overlooked factor is the reciprocal of patient activation: “provider activation,” the provider's knowledge and desire to get patients more activated. I examine the discrete and complementary effects of patients’ healthcare needs and provider activation and demonstrate each significantly impact patient activation. Using a novel matching method to minimize selection bias, I investigate the impacts of a patient portal on healthcare utilization and outcomes. Patient portal usage is shown to complement healthcare utilization, improve access to services, and increase medication adherence. In study two, I investigate social health influence within Army units. Large variations in health behaviors have been observed across different locations. I assert this variation is a result of distinct “health cultures” or norms of health behaviors, influencing individual soldier behaviors. To examine the effect of these health cultures on soldiers’ health behaviors, while minimizing selection bias, I exploit a unique feature of the Army: the exogenous assignment of soldiers to units. The hierarchical structure inherent to the Army provides an opportunity to examine leader and subordinate effects, in addition to peer effects, which I demonstrate have significant, differential impacts on the spread of obesity, tobacco use, and alcohol abuse. This dissertation contributes to the field’s understanding of drivers of health behavior change, through the examination of patient and provider activation and the role of social influence in health behaviors. It offers important recommendations for policy makers seeking to improve the effectiveness and efficiency of healthcare systems.