Evidence of Construct Validity for Stoke and Arthritis in a Population Health Survey

Background The Health Utilities Index Mark 3 (HUI3) is a comprehensive, compact health status classification and health state preference system. The HUI3 system has been implemented in 4 Canadian population health surveys. Objectives To evaluate the construct validity of the HUI3 for the measurement of health related quality of life (HRQL) and attribute-specific morbidity of respondents to the 1990 Ontario Health Survey reported to have arthritis or stroke. We tested (1) whether those with stroke, arthritis, and both conditions had lower HRQL than those with neither condition; (2) whether HUI3 detects morbidity in specific health attributes affected by arthritis and stroke. We expected stroke to affect primarily speech and cognition, arthritis to affect primarily pain and neither condition to affect vision or hearing. Research Design Linear regression models of global HRQL, 8 measures of attribute-specific utilities and logit regression models of 8 measures of attribute-specific functional disability were estimated as a function of 3 illness indicators (stroke only, arthritis only, both) and a set of variables included to reduce confounding. Results Subjects with stroke, arthritis and both conditions had substantially lower HRQL than those with neither condition. Stroke subjects had greater morbidity in speech and cognition than arthritics; somewhat surprisingly stroke subjects endured burdens of pain similar to arthritics; neither condition affected vision or hearing. The tests were robust to several different model specifications. Conclusions The HUI3 system appears to be valid for the measurement of health status and HRQL for stroke and arthritis.