Background Current guidelines encourage the measurement of rheumatoid arthritis (RA) disease activity and directing therapy to achieve a low disease state or remission (treat-to-target). Many RA patients with documented moderate to severe disease activity remain on their current therapy without change.

Objectives This study investigated Veterans Affairs (VA) clinical data to identify patient factors associated with a major change in RA therapy and to determine the relative importance of these different factors.

Methods US Veterans enrolled in the VARA registry were included if they had: 1) high/moderate disease activity (DAS28≥3.2) at index date, 2) 18 months of VA data prior to the index date and 3) two or more other DAS28 measured during the preceding 18 months separated by at least three months. A major change was report if within 7 days before to 90 days after the index date there was either: 1) initiation or escalation of DMARDs, 2) initiation or increase dose of prednisone and/or 3) ≥2 joint injections.

Baseline DAS28 was estimated during the observation period prior to the index date using an area under the curve calculation and compared to the DAS28 at the index date. Patients were categorized as have a worsening or improvement of disease if the DAS28 at index date was 0.6 higher or lower than the average DAS28 during the observation period respectively. Other patients were categorized as no change in DAS28.

Analyses of clinical variables including components of the DAS28 and patient and physician reported measures were compared in patients with and patients without a major change in therapy.

Results Of 941 patients who met study criteria, only 388/941 (41.2%) had a major change of therapy. Patients with worsening DAS28 were more likely to have a major change 183/369 (49.5%) than no DAS28 change 170/454 (37.4%) and improved DAS28 35/118 (29.6%) (P<0.001). Clinical variables were strongly associated with changes in therapy among patients with worsening disease activity and not as strongly associated with change in therapy in those with no change disease activity. Clinical variables were not significantly associated with major change in patients with disease improvement, though that group had the smallest sample size. Ten representative clinical variables with the highest association with major change are included below.

• Download figure
• Open in new tab
• Download powerpoint

Conclusions More than half of the patients with moderate disease activity did not have a major change in therapy. The likelihood of a major change in therapy increased with worsening disease activity. The clinical variables assessed were more strongly associated with change in therapy in patients with worsening of disease. Clinical disease activity measures are highly associated with the decision to initiate major changes. Future work will investigate the potential added value of administrative variables. This work emphasizes the need for methods to systematically collect and utilize clinical disease activity measurements, particularly longitudinally, to improve the treat-to-target strategy.

Acknowledgements Work Sponsored by VA Specialty Care Centers of Innovation, VA Health Service Research and Development, and Amgen.

Disclosure of Interest G. Cannon Grant/research support from: Amgen, C.-C. Teng Grant/research support from: Amgen, N. Accortt Shareholder of: Amgen, Employee of: Amgen, D. Collier Shareholder of: Amgen, Employee of: Amgen, S. Mehrotra Grant/research support from: Amgen, B. Sauer Grant/research support from: Amgen