Measuring Health Outcomes for Cost-Effectiveness Studies: Are all Quality Adjusted Life Years Created Equal?

Abstract

A number of different approaches are available for assessing the effectiveness of medications and other medical interventions. Desirable qualities for a measure of effectiveness include: 1. usefulness for evaluating outcomes across different diseases and interventions, 2. a combination of the impact of treatment and disease on health-related quality of life and survival, and 3. good psychometric characteristics. Preference/utility measures meet most of these criteria, however, they are not without problems. The most important issues are that different preference scaling methods yield different scores; different scaling methods vary in cognitive complexity and subject understanding; and variance in preferences are introduced due to population and contextual factors. Preference/utility measures are less responsive than health status measures to small, clinically meaningful changes. Therefore, the quality-adjusted life years generated in health economic evaluations may not be comparable between studies. Comparability in quality-adjusted life years between studies depends on whether the same scaling methods are used and similar populations produce the preferences/utilities.