Evaluating Chronic Disease for Heterogeneous Populations

Abstract

Cost-effectiveness evaluation for health care programs often involves the use of quality-adjusted life-year (QALY) estimates to measure morbidity losses from health conditions. Current techniques for measuring morbidity losses are often subjective, inflexible, impractical, and subject to bias. We sought to examine the impact of population heterogeneity on QALY values for arthritis sufferers by estimating an alternative health-adjusted life-year (HALY) measure based on self-assessed health status. We present a feasible approach for the assessment of improved QALY estimates for chronic conditions affecting heterogeneous populations. An ordered probit model, using data from the National Health Interview Survey (NHIS), is used to calculate expected HALY losses from arthritis for distinct population subgroups. These measures are used to scale existing QALY measures that have been calculated for distinct homogeneous populations. : We find that QALY losses from chronic arthritis vary by age, time since onset, and type of arthritis. When we apply these results to prevention programs aimed at reducing the incidence of Salmonella enteritidis infections (and the resulting reactive arthritis sequelae), we find that age-invariant QALYs underestimate the true discounted lifetime QALY losses from arthritis by 15%. Our results indicate that a failure to account for population heterogeneity can lead to biased health loss estimates. The modified HALY measure presented here can be used to help inform policymakers faced with heterogeneous populations.