Background/Purpose:   Polyarticular juvenile idiopathic arthritis (Poly-JIA) causes pain, functional disability, and joint damage.  Variation in Poly-JIA clinical assessment and treatment likely has a negative impact on outcomes.  Our goal was to improve Poly-JIA outcomes by standardizing disease activity scoring, ensuring disease activity review at the point of care, and implementing clinical decision support (CDS) to reduce treatment variation.  

Methods: In January 2016, we began documenting JIA disease activity at outpatient visits.  For Poly-JIA, we iteratively designed and tested CDS algorithms to standardize medication selection, dosing, and treatment duration.   In April 2016, we implemented CDS (Phase 1) for Poly-JIA reactivation.  In October-November 2016, we began disease activity target review for all JIA patients and CDS (Phase 2) for patients with new Poly-JIA and those in remission. Process measures included visit-level clinician target attestation (goal > 50%) and CDS use (goal > 15%).  The outcome measure was the three-variable clinical Juvenile Arthritis Disease Activity Score (cJADAS), which is the sum of the physician and patient/parent global assessment (both 0-10), and active joint count (maximum 10).  We calculated the monthly population cJADAS mean using the patientsÕ last recorded value in patients with two or more visits.  Our goal was to improve the Poly-JIA cJADAS by ≥ 10%.  Data were analyzed for special cause variation using standard statistical process control (SPC) and run chart methodology. 

Results: From February 2016-March 2017, we longitudinally assessed 112 patients with Poly-JIA and 358 with other JIA subtypes.  After October 2016, the mean monthly disease activity target attestation in all JIA patients was 74%. In Phases 1 and 2, we used CDS in 5-17% and 41-88% of Poly-JIA encounters, respectively. From July 2016-June 2017, Poly-JIA cJADAS scores decreased from a baseline of 5.5 to 3.5 (36%) (Figure). Special cause variation was identified in January 2017, when the cJADAS score decreased to 4.0, which was less than the lower confidence limit.  During this period, the proportion of patients in clinician-defined remission increased from 31% to 49%. Significantly fewer joint injections were performed from December 2016-June 2017. In other JIA subtypes observed over the same period in whom target attestation was used without CDS, cJADAS scores decreased from 3.9 to 3.4 (13%), which did not achieve special cause variation. 

Conclusion: After exceeding our target attestation and CDS use goals, Poly-JIA disease activity improved by 36%.  We estimate that approximately 22 more patients are currently in remission.  Therefore, using a treat to target approach paired with CDS can improve Poly-JIA disease activity substantially.  We plan to study changes medication use, assess pain and functional outcomes, and develop additional CDS algorithms.