Abstract

Background: Functional capacity is an important outcome in rheumatoid arthritis and is generally measured using the Health Assessment Questionnaire disability index (HAQ). Functional limitation incorporates both activity and damage. Because irreversible damage increases over time, the HAQ may be less likely to show improvement in late than in early rheumatoid arthritis.

Objective: To determine the relation between sensitivity to change of the HAQ and duration of rheumatoid arthritis in reports of clinical trials.

Methods: Data were pooled from clinical trials that measured responses of HAQ scores at three or six months. The effect size of the HAQ was calculated and linear regression used to predict the effect size by duration of rheumatoid arthritis at group level. Treatment effect was adjusted for by including the effect sizes of pain scores and of tender joint counts as additional independent variables in separate models. Subgroup analysis employed contemporary regimens (methotrexate, leflunomide, combination therapies, and TNF inhibitors) only.

Results: 36 studies with 64 active treatment arms and 7628 patients (disease duration 2.5 months to 12.2 years) were included. The effect sizes of the HAQ decreased by 0.02 for each additional year of mean disease duration using all trials, and by 0.04/year in the subgroup analysis (p⩽0.01 for both analyses, except for pain adjusted models at three months).

Conclusions: In individual trials, less improvement in the HAQ might be expected in late than in early rheumatoid arthritis. Comparison of changes in HAQ among rheumatoid arthritis trials should take into consideration the disease stage of the treated groups.