A Proposed Method for Monitoring U.S. Population Health: Linking Symptoms, Impairments, and Health Ratings

We propose a method of quantifying non-fatal health on a 0-1 QALY scale that details the impact of specific symptoms and impairments and is not based on ratings of counterfactual scenarios. Measures of general health status are regressed on health impairments and symptoms in different domains, using ordered probit and ordinary least squares regression. This yields estimates of their effects analogous to disutility weights, and accounts for complex non-additive relationships. Health measures used include self-rated health status on a 5-point scale, EuroQol 5D (EQ-5D) scores, and ratings of current health using a 0-100 rating scale and a time-tradeoff. Data are from the nationally representative Medical Expenditure Panel Survey (MEPS) year 2002 (N=34,615), with validation in an independent sample from MEPS 2000 (N=21,067) and among 1420 adults age 45-89 in the Beaver Dam Health Outcomes Study. Decrement weights for symptoms and impairments are used to derive estimates of overall health-related quality of life, laying the groundwork for a detailed national summary measure of health.

To purchase a copy of the earlier version of this paper, please contact the Working Papers department directly at (617) 588 1405.