Background/Purpose: DMARDs are the standard of care for rheumatoid arthritis (RA), however multiple studies find that not all patients use these agents.  We examined predictors of DMARD stopping and DMARD starting among a  large cohort of patients with RA.

Methods: Study participants were drawn from an open longitudinal cohort of 1,346 participants with RA recruited from rheumatologists’ practices in Northern California.  We examined patterns and predictors of DMARD stopping and starting, including non-biologic and biologic DMARDs, based on annual questionnaires.  Stopping was defined as stopping ALL DMARDs and starting was defined as transitioning from NO DMARDs to any DMARD across consecutive years. Predictors were categorized as related to RA (disease duration, HAQ score, tender and swollen joints, and use of oral steroids), sociodemographics (age, gender, race/ethnicity, education, income), or comorbidities (index).  Calendar year was also included.  Generalized linear mixed regression models for binary outcomes were constructed that accounted for the non-independence of multiple pairs of years from individual participants, and model fit was assessed using the c-statistic.

Results: The analysis of determinants of starting DMARDs included 471 subjects with 1,974 pairs of years with no DMARD use in the first of two consecutive years from which 313 (16.3%) started DMARD use before year two. The analysis of determinants of stopping DMARDs included 1,026 subjects with 7,595 pairs of years with DMARD use in the first of two consecutive years from which 423 (5.6%) stopped DMARD use before year two. Over the 23 years of follow-up (1987 – 2009), the percent starting DMARDs between two consecutive years was stable at approximately 10%, but the percent stopping DMARDs from one year to the next decreased from 9% to 3%.  In fully adjusted models, significant predictors of starting DMARDs included younger age (OR 0.85, 95% CI 0.75 – 0.95, per 5-year decrease), Hispanic ethnicity (OR 1.88, 95% CI 1.06 – 3.33), shorter disease duration (OR 0.90, 95% CI 0.80 – 1.00, per 5-year decrease), and the use of oral steroids (1.90, 95% CI 1.36 – 2.66).  In separate fully adjusted models, predictors of stopping DMARDs included older age (OR 1.05, 95% CI 1.00 – 1.10, per 5-year increase), Hispanic ethnicity (OR 1.54, 95% CI 1.02 – 2.30), lowest annual income quartile (OR 1.83, 95% CI 1.13 – 2.96, compared with highest), and more tender joints (OR 1.03 , 95% CI 1.00 – 1.07 per joint increase).  The c-statistics for RA-related factors were 0.60 (stopping a DMARD) and 0.62 (starting a DMARD), suggesting that they were relatively weak predictors of stopping or starting a DMARD.  Including sociodemographic factors and comorbidities in the fully adjusted models improved the model fit for both sets of models – c-statistics 0.68 (stopping a DMARD) and 0.69 (starting a DMARD).

Conclusion: Predictors of stopping and starting DMARDs include non-RA related factors as well as RA-related factors.   More frequent starting and stopping of DMARDs in Hispanic subjects may reflect barriers to continued use.  The significance of non-RA related factors such as race/ethnicity and income suggest that there are disparities in DMARD use despite clear clinical guidelines about their use.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-starting-and-stopping-disease-modifying-anti-rheumatic-drugs-for-rheumatoid-arthritis-a-23-year-longitudinal-cohort/