Relationship between rheumatoid arthritis disease severity, health-related utility, and resource use in Australian patients: A cross-sectional, multicenter study.

OBJECTIVES: This research explores the relationship between rheumatoid arthritis (RA) severity measures, clinical characteristics, and patient preference-based health-related quality of life (ie, utility) in Australian RA patients. A secondary objective was to investigate the relationship between a range of disease severity measures, clinical characteristics, and the cost of RA-related resource use.
METHODS: This was a cross-sectional, multicenter study of consecutive patients with RA aged > or =18 years attending routine clinical appointments. Patients completed a questionnaire comprising general demographic, resource use, and disease-specific questions; the RA-specific Health Assessment Questionnaire (HAQ); and 2 multiattribute preference-based quality-of-life (utility) instruments, the Health Utilities Index Mark 3 (HUI3) and the EuroQol 5 Dimensions (EQ-5D). A second questionnaire was completed by the patient's rheumatologist, with questions on key clinical data pertinent to RA, including the number and location of tender joints and swollen joints, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) concentration, current and previous treatments, and details of important comorbid conditions. Data on RA-specific resource use were also collected, including health-professional visits (eg, general practitioner, specialist, nurse, occupational therapist, paid caregivers), hospital admissions, surgery, and RA-associated home modifications. Spearman nonparametric correlation, simple linear, multiple, and stepwise regression analyses were used to explore the relationship between utility scores (HUI3 and EQ-5D) and variables including HAQ, RA-related resource use, and key demographic, physical, and biochemical measures.
RESULTS: A total of 170 patients were recruited from 4 centers in Australia. Consistent with the epidemiology of RA, the ratio of women to men was approximately 3:1 (126 vs 44, respectively). Male and female patients were of similar age (mean [SD], 59.2 [12.9] and 58.9 [12.2] years, respectively). Time since diagnosis of RA was significantly shorter for men than for women (mean difference, -4.52 years; 95% CI, -8.65 to -0.38; P = 0.03). Of all the disease severity measures and clinical characteristics investigated, patients' HAQ scores predicted their utility most closely (for HUI3, R(2) = 0.626, P < 0.001; for EQ-5D, R(2) = 0.403, P < 0.001). The Disease Activity Score 28 severity index provided the next best relationship with a patient's utility; however, its explanatory power was poor (for HUI3, R(2) = 0.085, P < 0.001; for EQ-5D, R(2) = 0.042, P = 0.008). ESR, CRP, and, RA-affected joint counts had negligible explanatory power for patient utility. In analyses of the relationship between a range of key variables and the direct costs associated with RA, the HAQ score explained 22% of the variability in log costs (P < 0.001).
CONCLUSION: This study found that of all the disease severity measures and clinical characteristics investigated, patients' HAQ scores predicted their utility most closely. 2010 Excerpta Medica Inc. All rights reserved.