Electronic health record and patterns of care for children with cerebral palsy.

AIM To characterize the patterns of care of children with cerebral palsy (CP) in a tertiary healthcare system. METHOD Electronic health record data from 2009 to 2019 were extracted for children with CP. Machine learning hierarchical clustering was used to identify clusters of care. The ratio of in-person to care coordination visits was calculated for each specialty. RESULTS The sample included 6369 children with CP (55.7% males, 44.3% females, 76.2% white, 94.7% non-Hispanic; with a mean age of 8y 2mo [SD 5y 10mo; range 0-21y; median 7y 1mo]) at the time of diagnosis. A total of 3.7 million in-person visits and care coordination notes were identified across 34 specialties. The duration of care averaged 5 years 5 months with five specialty interactions and 21.8 in-person visits per year per child. Seven clusters of care were identified, including: musculoskeletal and function; neurological; high-frequency/urgent care services; procedures; comorbid diagnoses; development and behavioral; and primary care. Network analysis showed shared membership among several clusters. INTERPRETATION Coordination of care is a central element for children with CP. Medical informatics, machine learning, and big data approaches provide unique insights into care delivery to inform approaches to improve outcomes for children with CP.