Dissertation Title:  "Methods for Modeling and Valuing Life Expectancy Gains"

Over the past decade, there has been increasing concern over the inability of society to afford the growing number of health-related interventions. Decision-makers have turned to the tools and techniques of economic analysis to illuminate the trade-offs associated with different policy decisions. This thesis presents three studies that contribute to the improvement of methods for the economic evaluation of interventions that extend human lives.

The first study focuses on methods used to predict life expectancy gains. We use a novel combination of observed epidemiological relationships and disease simulation to infer knowledge about the underlying effects of aspirin use on colorectal cancer. We find that aspirin likely acts at multiple points in the disease process by exerting a strong influence early in the disease and a relatively weaker influence later in the disease, a result with implications for the effectiveness (and cost-effectiveness) of aspirin as a chemopreventive agent for colorectal cancer. From a methodological perspective, this study illustrates the ability of simulation modeling to generate important new insights about the disease process.

The second paper focuses on estimating the monetary value of life expectancy gains by using stated willingness to pay for two life-extending interventions. Our analysis reveals a troubling disparity between estimates of the value of a statistical life year derived from the two goods. That the disparity cannot be explained suggests that the value of a statistical life year may not be a readily generalizable metric for monetizing life expectancy gains.

The third study focuses on examining alternative methods for communicating the benefits of interventions that extend human lives. We find that survey respondents valued benefits expressed as a life expectancy gain more highly than identical benefits expressed as a reduction in annual mortality risk. Using sensitivity to scope as a validity test, we find that the life expectancy format demonstrates greater validity, implying that the monetary value of longevity benefits may be more likely to reflect true preferences if elicited using a life expectancy format. In summary, the three studies presented herein contribute to the ongoing search for methods to credibly estimate the magnitude and value of life expectancy gains. Enhancing the validity of these methods will help us more efficiently allocate society’s scarce resources among the growing number of competing programs that extend human lives.