Building a Learning Health System in Pediatric Rheumatology

Esi M. Morgan ^1,2,3* Constance Mara ^4,5 Sheetal S Vora ^6,7 Beth S Gottlieb ^8,9

• ¹ Division of Pediatric Rheumatology, Seattle Children's Hospital, Seattle, United States
• ² Department of Pediatrics, School of Medicine, University of Washington, Seattle, Washington, United States
• ³ Seattle Children's Research Institute, Seattle, Washington, United States
• ⁴ Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, United States
• ⁵ Department of Pediatrics, College of Medicine, University of Cincinnati, Cincinnati, Ohio, United States
• ⁶ Department of Pediatrics, School of Medicine, Wake Forest University, Winston-Salem, North Carolina, United States
• ⁷ Atrium Health Levine Children's Hospital, Charlotte, United States
• ⁸ Cohen Children's Medical Center, New Hyde Park, New York, United States
• ⁹ Department of Pediatrics, Northwell Health, New Hyde Park, United States

To explore the impact of LHS in this field, we put out a call for a special Research Topic on "Building a Learning Health System in Pediatric Rheumatology". Our goal was to highlight real-world clinical practice and network interventions to improve the quality of healthcare delivery and outcomes in pediatric specialty care. The editorial team is pleased to highlight a broad overview of the 12 articles selected by peer review for publication, representing 23 distinct health systems across North America.Key themes reflecting drivers of improving chronic illness care emerged across these contributions. A study by Vora,et al. (3) addressed delays in referral and low access to care. The authors describe a single center intervention to increase timely referrals to rheumatology care from safety net primary care clinics through education, raising awareness of presenting symptom bundles of rheumatic diseases, streamlining referral processes, and establishing a triage system with expedited scheduling for urgent cases. This work exemplifies the capability of a single clinic to influence access and timeliness of care within the confines of a larger health system.Another major theme was addressing variation in treatment practices. Balay-Dustrude, et al. (4) describes unwarranted variation in care in a then nascent learning health network (Pediatric Rheumatology Care and Outcomes Improvement Netowrk, PR-COIN), through evaluation of use of intraarticular corticosteroid injections (IACI) following release of American College of Rheumatology JIA treatment guidelines (in 2011, 2013). Data from 2011-2015 was analyzed on whether IACI were used as primary treatment for oligoarticular JIA. Although there was no network-level intervention, centers worked locally to try and achieve a network stated goal of IACI within 2 weeks of diagnosis. Despite local efforts, results revealed lower-than-expected IACI rates and regional practice variation, highlighting the need for standardized care approaches within a LHS.Multiple articles touched on the importance of data capture to assess patient outcomes to evaluate for potential gaps in care or disparities, and of the reliable collection of disease specific activity measures across patient population. Goh, et al (5), and Pan, et al (6), conducted surveys of PR-COIN LHS members to understand current state of data collection across network centers, and understand barriers and facilitators to reliable collection. Goh focuses on the role of 'critical data elements' and offers unique insights on the challenge of collection during tele-rheumatology visits. Patient centered care requires assessment of outcomes directly from the patients (PROs), Pan shares best practices for systematic PRO collection based on network experience. Although health equity is a key principle of quality of care, a fundamental prerequisite is data completeness which confers the ability to identify disparities. Banschbach, et al, (7) through an evaluation of PR-COIN learning network registry data identified missing demographic data in 1/3 of patients. With interventions, this was remediated in about 94%. The article noted that patients missing race data are likely to be missing other critical data elements, and challenges us to equitably measure disease activity and ensure data capture for vulnerable groups.Two studies detailed single-center interventions leveraging electronic health records (EHR) for datadriven improvement. Timmerman, et al (8), and Barbar-Smiley, et al (9), each provide single center interventions to standardize disease activity data collection in the electronic health record (EHR). Timmerman describes partnership with local IT resources to establish a center dashboard to capture data on all clinic patients to support QI initiatives. Barbar-Smiley's group leveraged features within the EHR collection system to support and sustain a clinic effort to capture systemic lupus (SLE) disease activity, a process that also benefitted their research registry participation.Huang, et al (10) describe use of technology to leverage clinical registry data of an LHS to refine algorithms supporting selection of personalized treatment based on real world experience of other similar patients, which can be used at point of care in shared decision making with patients.Mental health and self-management support were also focal points. Harper,et al. (11) present an intervention to systematically collect mental health data in childhood SLE patients, and an intervention to respond with mental health supports. Argraves, et al, (12) address the perpetual challenge of assessing and supporting transition readiness in children reaching age of majority. They leverage process automation to achieve reliability in assessment (90%). The intervention also requires provider review and self-management support education. That multiple QI interventions were published from one center suggests a strong culture of QI, and effective use of QI as a strategy for operational improvement.Two articles highlighted outcomes improvement and parent engagement within PR-COIN, a pediatric rheumatology LHS with over 7,200 active patients across 23 centers. Harris, et al (13) and Ferraro, et al. (14), discuss network-wide interventions and outcome improvements demonstrating the effectiveness of the LHS model in driving sustainable improvements. Of vital importance to the network's success is the engagement of parents and patients in design and conduct of interventions contributing the patient perspective, and keeping focus on patient outcomes.What can we take away from this issue? The inspiration of pediatric rheumatology clinical teams working in coordination together with parents to move the needle on healthcare improvement for a vulnerable patient population. A culture of QI, rigorous data collection, and the use of technology are central to this work. By focusing on equitable access, timely diagnosis and treatment, mental health integration, proactive care teams, engaged families, and health system support for QI, a LHS organizational framework has been shown to improve processes and patient outcomes in pediatric rheumatology. This special issue serves as a testament to the power of adopting a LHS model in transforming care in pediatric chronic illness care.