Outcome measures in economic evaluations of rheumatoid arthritis

Objectives: The primary objectives of this study were to: 1) compare the properties of commonly utilized indirect utility assessment instruments (the Health Utilities Index Mark 2 and 3 [HUI2 and HUI3], the EuroQol [EQ-5D], and the Short Form 6-D [SF-6D] in terms o f feasibility, reliability, construct validity and longitudinal construct validity (responsiveness) in rheumatoid arthritis (RA); and 2) determine if, when utilized to act as quality weights in the estimation of quality adjusted life years (QALYs) in an economic evaluation, the application of scores from the different instruments would result in different incremental cost per QALY ratios. The primary hypotheses of this study were that there would be differences between these instruments in terms of their properties and that using their scores to estimate QALYs in an economic evaluation of an intervention for RA would result in significantly different estimates. Methods: Three hundred and twenty patients between 19 and 90 years of age diagnosed with RA residing in the Greater Vancouver Regional District or rural Okanagan region of British Columbia were recruited. Patients were administered a questionnaire containing the HUI2, HUI3, EQ-5D, SF-6D, a disease-specific instrument (the Rheumatoid Arthritis Quality of Life [RAQoL] questionnaire, and a disability index (the Health Assessment Questionnaire [HAQ]). In addition, questions were asked regarding RA management (including drug use and toxicity), RA severity (including swollen and tender joints, pain visual analogue scale, RA duration, patient global assessment of disease activity VAS , and self-perceived RA severity and control), socio-economic status, and RA health utilization. Questionnaires were administered at baseline, three and six months thereafter. In a subset of patients, an additional questionnaire was administered within five weeks of the three month questionnaire to determine reliability. Results: Scores obtained with the HUI2, HUI3 , EQ-5D, and the SF-6D were significantly different, had low agreement, and appeared to be measuring mostly physical function and pain. All the instruments displayed cross-sectional construct validity and were able to discriminate between different levels of severity of RA. However, when their scores were used to estimate QALYs in an economic evaluation of RA , there was a two fold difference between the lowest (using the HUI3) and highest (using the SF-6D) incremental cost per QALY ratios. Further examination revealed that the scores achieved with the indirect utility assessment instruments were influenced by annual household income despite adjustment for RA severity and other chronic diseases. Finally, in longitudinal analyses, the disease-specific RAQoL displayed the highest reliability and sensitivity to change with the HUI3 and SF-6D scores being the most responsive of the indirect utility assessment instruments in measuring positive change. Conclusions: Although all indirect utility assessment measures appear to be able to assess generic HRQL in RA , when used as quality weights to estimate QALYs in an economic evaluation, they yielded vastly different estimates of the incremental cost-effectiveness ratio that could result in different policy recommendations. The scores of these instruments could also be influenced by income leading to possible bias in cost-effectiveness analyses. The HUI3 and SF-6D were responsive to positive changes in RA. The RAQoL displayed excellent properties and is a suitable disease-specific HRQL instrument for RA.Medicine, Faculty ofPopulation and Public Health (SPPH), School ofGraduat