Background/Purpose:

Current guidelines recommend that rheumatoid arthritis (RA) patients with poor response to their current regimen of disease modifying anti-rheumatic drugs (DMARDs) have therapy adjusted until reaching low disease activity or remission. We examined factors associated with timing of decisions to adjust DMARD therapy for RA patients with active disease and how the timing of decisions impacts resolution of active disease.

Methods:

Data came from the University of Pittsburgh Rheumatoid Arthritis Comparative Effectiveness Research (RACER) registry, which captures subjects’ disease activity status (DAS28-CRP) and medications at clinic visits.

We conducted survival analyses on time to DMARD therapy adjustment and time to resolution of active disease for RACER subjects with active disease. A Cox proportional hazards model was used to assess the impact of covariates on the survival times. For both analyses, followup begins when the subject is first known to have active disease (DAS28-CRP>3.2) and ends with the event of interest (DMARD therapy adjustment or resolution of active disease). For the analysis of time to therapy adjustment, we excluded patients who adjusted therapy at t=0 and those who exited active disease before adjusting therapy, and the model included covariates for age, gender, African-American race, comorbidities, duration of RA, and current use of a biologic therapy. For the analysis of time to resolution of active disease, the model included the same covariates plus an indicator for time to therapy adjustment.

Results:

We identified 562 subjects with active disease at a first timepoint and a followup DAS28-CRP measurement.

The analysis for time to therapy adjustment included 177 subjects (117 therapy augmentations observed). 364 subjects were excluded because they adjusted DMARD therapy at t=0 (n=196) or because they exited active disease status before adjusting therapy (n=162). The median time to therapy adjustment was 203 days. Age over 75 and longer duration of RA were significantly associated with longer times to therapy adjustment.

The analysis for time to resolution of active disease included 530 subjects (383 achieved DAS28-CRP ≤ 3.2). The median time to resolution of active disease was 257 days. African-American race and duration of RA >1 year were associated with longer times to resolution of active disease. (See Table 1)

Conclusion:

Among those with persistent active disease (n=394), 60% adjusted DMARD therapy within 90 days; however 50% of subjects took longer than 257 days to achieve DAS28-CRP ≤ 3.2. We found that age ≥ 75 and longer duration of RA were associated with longer times to DMARD therapy adjustment, while African-American race and duration of RA > 1 year were associated with poorer disease outcomes.  Future studies should further examine how these factors affect treatment choices as well as long term health outcomes.