Health Utilities Index mark 3 demonstrated construct validity in a population-based sample with type 2 diabetes

Abstract

Objective

To assess the cross-sectional construct validity of the Health Utilities Index mark 3 (HUI3) in type 2 diabetes using population health survey data.

Study Design and Setting

Data used were from 5,134 adult respondents of Cycle 1.1 (2000–2001) of the Canadian Community Health Survey (CCHS) with type 2 diabetes. Analyses of covariance models were used to assess differences in overall and single-attribute HUI3 scores between groups hypothesized a priori to differ in HRQL. The association between health-care resource use (i.e., hospitalizations and physician and emergency room visits) and overall HUI3 scores was assessed using logistic regression models.

Results

For overall HUI3 scores, clinically important and statistically significant differences were observed between all groups expected to differ in HRQL. Depression was the comorbidity associated with the largest deficit (−0.17; 95% confidence interval CI = −0.22, −0.12), followed by stroke (−0.15; 95% CI = −0.21, −0.10) and heart disease (−0.08; 95% CI = −0.11, −0.05). Insulin use and comorbidities were associated with clinically important deficits in pain. Overall HUI3 scores were significantly predictive of all three categories of health-care resource use.

Conclusion

Observed differences between groups contribute further evidence of the construct validity of the HUI3 in type 2 diabetes.

Introduction

Research on health-related quality of life (HRQL) in diabetes has focused mainly on the development and application of diabetes-specific instruments [1], [2]. Specific HRQL measures focus on the impact of a condition or treatment on health and functioning and are intended to provide detail concerning outcomes associated with a condition [3]. In contrast, generic HRQL measures (such as multiattribute utility measures) provide information on general function and well-being. The more general content of generic measures may make them better suited for particular applications. In diabetes, for example, disease-specific measures fail to capture the additional HRQL deficits associated with comorbidities that contribute to disease burden [4], [5], [6].

The Health Utilities Index (HUI) group is a family of multiattribute utility measures. Evidence of construct validity of the Health Utilities Index mark 3 (HUI3) in type 2 diabetes has been generated in clinical samples [7], [8], [9]. Further evidence of construct validity of the HUI3 comes from stroke and arthritis patients [10] and from high-risk primary care patients [11]. When we previously assessed the construct validity of the HUI3 in type 2 diabetes, data limitations kept us from assessing assess the ability of HUI3 to detect differences according to specific comorbidities [7], [8], [9]. Our objective for the present study, therefore, was to provide further evidence of the cross-sectional construct validity of HUI3 in type 2 diabetes by assessing (a) the ability of the overall HUI3 and single attributes to detect clinically important differences according to treatment intensity, duration of diabetes, and comorbidities and (b) the association between health-care resource utilization and overall HUI3 scores.