Physical Therapy Database Development at Children's Hospital of Philadelphia: An Administrative Case

A.M. Glanzman, PT, DPT 1, A.T. Harrington, PT, PhD 1,2, H.L. Atkinson, PT, DPT 1, P. Camacho, MS 3, J. Aberdeen, PT, DPT 1, M. Sullivan, MSW, MBA 1, EHR revision committee, L.A. Prosser, PT, PhD 4,5 Author Affiliations 1 Department of Physical Therapy, Children's Hospital of Philadelphia, 3401 Civic Center Boulevard, Philadelphia, PA 19104, USA; GLANZMANA@chop.edu (corresponding author) 2 Department of Physical Therapy, Arcadia University, Glenside, PA, USA 3 Department of Biomedical and Health Informatics, Children's Hospital of Philadelphia, Philadelphia, PA 4 Division of Rehabilitation Medicine, The Children's Hospital of Philadelphia, Philadelphia, PA, USA 5 Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

Poster # 87

Our goal was to create a pediatric rehabilitation research database at our multi-site tertiary care medical center for the purpose of tracking functional outcomes in children with injury and disability. The database includes all clinical physical and occupational therapy and rehabilitation medicine data to facilitate a Learning Health System (LHS) that can quantify the natural histories of function and impairment and help us better understand the link between rehabilitation practice and functional outcomes. Our institution employs >100 physical therapists across inpatient care, neonatal/infant intensive care, inpatient rehabilitation, 9 outpatient locations, 4 sports medicine locations, and 29 multidisciplinary clinics. Service line representatives collaborated with electronic health record (EHR) analysts and research bioinformatics to build a research database with automated EHR data extraction. Data capture reliability was optimized using discrete documentation fields. The patient cohort was validated against hospital business data and manual record review. The database is housed in our institution's big data research repository host environment (ARCUS). The database captured examinations by 231 physical therapists for a total of 959,802 physical therapy visits from 89,966 unique patients from 2009-2021. Inpatient and outpatient physical therapy programs from 2012-2021 provided 385,835 and 506,181 visits to an average of 44,600 and 12,900 patients a year respectively. Between 2009 and 2022 visit volume has tripled and clinic volume has grown 110% now representing data from over 30 different specialty clinics. Our database facilitated access to an abundant and heterogeneous trove of clinical data which can serve as a foundation for practice-based evidence. By leveraging big data with ongoing reflection and analysis of practice, a Learning Health System has the potential to improve PT practice and ultimately patient outcomes.