Leveraging a National Pediatric Health Learning System for Phenotyping Pediatric Chronic Primary Pain to Advance Timely Patient-Centered Care

Chronic primary pain is common, impactful, and challenging to treat in children and adolescents. Unfortunately, many children do not get referred to, nor receive timely and effective pain treatment before persistent and disabling pain ensues. While the barriers to pediatric pain care are multifactorial, a major limitation in understanding these issues is the absence of reliable and valid approaches to identify children with chronic primary pain across the continuum from primary care services to tertiary care. With the advancements of machine learning and rule-based algorithms, we can leverage the electronic health record (EHR) to phenotype children and identify those at greatest risk for poor outcomes. The objective is to develop a precision health model for pediatric chronic primary pain that promotes timely and patient-centered care. To achieve this, we will leverage the infrastructure of the PEDSnet national pediatric learning health system. We will utilize retrospective, multi-cohort "big data" extracted from the EHR to develop a computable phenotype of chronic primary pain for tracking pain treatment and outcomes. Additionally, at four time points over a 12-month period, we will prospectively collect high-resolution individual and socio-environmental data from 600 children with chronic primary pain (English and Spanishspeaking) at three PEDSnet institutions to identify phenotypic characteristics that differentiate pain outcomes across the continuum of pediatric care. In parallel, we will follow the PROMIS methodology to refine and evaluate a brief, patient-reported outcome measure to monitor patients’ overall perception of changes with pain treatments using rigorous qualitative methods, psychometric evaluation, and establishing meaningful individual changes for core outcome measures in pediatric chronic pain. This will inform future implementation of screening and stratification tools to deliver personalized pain care. A Research Partners Group with lived experience will inform our research across all stages. In response to RFA-AR-24-007, we propose an interdisciplinary HEAL KIDS Chronic Pain Collaborating Research Team (CPT) with clinicians and researchers with expertise in pediatric learning health systems, behavioral science, applied clinical informatics, pain medicine, and innovative measurement methods, to manage our proposed studies and harmonize shared data elements, research data, methodologies, and outputs with the CPT Consortium and Resource Data Center. Our CPT has additional special strengths to contribute including expertise in ecological momentary assessment (EMA) based data collection and analysis, the use of actigraphy for assessing sleep patterns, and artificial intelligence and machine learning. Our CPT maintains active, ongoing research connections with well-characterized cohorts of children with chronic primary and secondary pain and has experience studying the transition from acute to chronic pain and long-term outcomes in young adulthood.