Treat‐to‐Target Approach in Rheumatoid Arthritis: A Quality Improvement Trial. Issue 2 (1st February 2021)

• Record Type: Journal Article
• Authors: Desai, Sonali P.; Leatherwood, Cianna; Forman, Malka; Ko, Eunji; Stevens, Emma; Iversen, Maura; Xu, Chang; Lu, Bing; Solomon, Daniel H.
• Is Part Of: Arthritis care & research. Volume 73:Issue 2(2021)
• Page Start: 207
• Page End: 214
• Publication Date: 2021-02-01
• Subjects: Arthritis -- Periodicals; Rheumatism -- Periodicals; 616.72
• DOI: 10.1002/acr.24114
• Languages: English
• ISSNs: 2151-464X

Abstract

Abstract : Objective: Using a quality improvement approach, our objective was to integrate a treat‐to‐target approach for rheumatoid arthritis (RA) through routine electronic collection of patient‐reported disease activity scores and a multidisciplinary learning collaborative for rheumatologists. Methods: RA patients completed a patient‐reported outcome measure, the Routine Assessment of Patient Index Data 3 (RAPID3), at check‐in. Nine rheumatologists and their patients were allocated to a learning collaborative intervention group focused on a treat‐to‐target approach and 13 were allocated to a control group. The primary outcome was documentation of a treat‐to‐target implementation score: disease activity score, disease activity score used in the medication change decision, the presence of a treatment target, and an indication of shared decision‐making. A primary analysis of patient visits with medication changes was conducted using an interrupted time‐series analysis. Results: We studied 554 individual rheumatology patients with 709 patient visits. Treat‐to‐target implementation scores among intervention rheumatologists (mean ± SD 44.6% ± 1.63%) were 12.4% higher than in the control group (mean ± SD 32.2% ± 1.50%; P < 0.0001). We observed differences in treat‐to‐target implementation score components, comparing intervention group to control group rheumatologists: disease activity score present, 77.2% versus 68.0% ( P = 0.02); disease activity score used in the medication change decision, 45.2% versus 30.0% ( P < 0.01); treatment target, 9.0% versus 0.4% ( P < 0.01); and shared decision‐making, 46.9% versus 30.0% ( P < 0.01). Secondary analysis of patient visits with high RAPID3 scores found that medication changes were 54% less likely in the intervention versus control group (odds ratio 0.46 [95% confidence interval 0.27–0.79], P = 0.005). Conclusion: This nonrandomized, interrupted time‐series trial demonstrated a modest but significant impact of a learning collaborative intervention on rheumatologist documentation of a treat‐to‐target approach in RA.