Measuring preference weights for american college of rheumatology response criteria for patients with rheumatoid arthritis

Objective. To estimate weights for health states comprising American College of Rheumatology (ACR) response and different levels of adverse events associated with rheumatoid arthritis (RA) treatments. Methods. A survey was mailed to 748 patients with RA from southern California. In addition to several questionnaires commonly used for patients with RA, patients were instructed to evaluate 10 hypothetical health states, in which they could have an ACR response and/or adverse events due to new treatments, with a visual analog scale (VAS). Patients also evaluated their current health with a VAS question and a time tradeoff (TTO) question. Linear extrapolation was used to derive 6 more health states. The Pearson correlation coefficient was used to validate VAS and TTO results. Results. A total of 487 (65%) patients returned the survey. Among the 10 health states evaluated with VAS directly, the health state in which a patient has ACR70 with no adverse events had the highest VAS weight (0.84), followed by the one having an ACR50 response with no adverse events (0.80). Correlation coefficients ranged from 0.63 for the correlation between VAS and physical component summary to -0.18 between TTO and pain and tender joint count; the correlation coefficients were all statistically significant, indicating there was convergent validity of the VAS and that VAS functioned differently from TTO in how it measured weights. Conclusion. VAS weights for 16 ACR response health states of patients with RA were derived. These weights could be used for cost-utility analyses of interventions for patients with RA.