Relative Importance of Disease-Modifying Antirheumatic Drug Attributes for Patients with Rheumatoid Arthritis

The significance of fatigue reduction in rheumatoid arthritis (RA) is well recognized. However, prior patient preference studies on disease-modifying antirheumatic drugs (DMARDs) have not adequately explored the relative importance of fatigue reduction compared to other DMARD attributes or accounted for preference heterogeneity. The objective of this study is to determine the relative importance of DMARD attributes, including fatigue reduction, from the patient perspective. We conducted a web-based discrete choice experiment (DCE) survey among RA patients in the US. Six DMARD attributes were included: chance of pain reduced by 50% or more, chance of physical function improved by 50% or more, chance of fatigue reduced by 10 points or more, chance of serious side effects, the route and frequency of administration, and out-of-pocket cost per month. Choice sets were constructed using a Bayesian efficient design. Mixed logit (ML) and latent class (LC) models were developed to determine preference weights and to calculate the conditional relative importance of each attribute. Of 228 patients, the ML model showed that the chance of pain reduction had the highest conditional relative importance (2.4), followed by out-of-pocket cost (2.1), the chance of physical function improvement (1.6), the chance of fatigue reduction (1.5), the chance of experiencing serious adverse events (0.6), and the route and frequency of administration (0.09). Preference heterogeneity was observed. The LC model identified two patient classes. The chance of fatigue reduction and the out-of-pocket cost were the most important attributes for class 1 and class 2, respectively. Pain reduction, physical function improvement, fatigue reduction, and out-of-pocket cost were relatively important DMARD attributes for RA patients. However, the presence of preference heterogeneity underscores the need for individualized treatment. Future studies should explore fatigue preferences and generalizability in the broader RA population.