Abstract/summary Leveraging the electronic health record to characterize and optimize care delivery for children with cerebral palsy: Cerebral palsy (CP) is the most common physical disability of childhood, but it is highly heterogeneous with respect to its severity, response to therapy, care needs, and impact on wellness for the child and family. To optimize health and wellness throughout life and enable new research avenues to be effectively tested, it is critical to develop a comprehensive clinical care and biopsychosocial data model. Development of a comprehensive model would both accelerate and improve understanding, care, and further research, including the identification of novel targets for interventions. The overriding objective of this proposal is to develop a precision health model for CP-related phenotypes, health status, care activities, and psychosocial well-being that will individualize care. To accomplish this objective, we will automate the collection, cleaning, and integration of multi-dimensional, multi-domain and multi-cohort based "big data" extracted from the electronic health record (EHR), and combine this EHR data with prospectively collected, high-resolution clinical, functional, environmental and psychosocial data. The focus of this proposal will be on children between ages 6 and 12 years. Preliminary work indicates that the medical center provides care for approximately 1,800 patients with CP who are between the ages of 6 and 12 years and have at least three years of EHR data. From this EHR cohort, we will prospectively recruit 200 children and their families for detailed phenotyping. Recruitment will be stratified by Gross Motor Functional Classification System (GMFCS) Levels: 60% GMFCS I, II, or III (able to walk) and 40% GMFCS IV or V (use a wheelchair). Using this multi-cohort design will allow for robust characterization of multi-dimensional factors that impact care receipt, functional outcomes, quality of life and participation. Aim 1 will focus on creating a diverse and comprehensive data repository using both retrospective and prospective data to characterize actual versus optimal care (defined by current evidence-based literature). Aim 2 will lead to development of a receipt of care coefficient score and characterize how the degree of optimal care relates to function, quality of life and participation, controlling for functional status and age. [Models developed in Aim 2 will be translated into a clinical decision tool prototype. Aim 3 will demonstrate proof of concept for scalability of machine learning algorithms with the PEDSnet Learning Health System. This project innovatively combines retrospective EHR data with prospective clinical data to elucidate individual, treatment, family, and environmental factors associated with greater receipt of evidence-based care and/or better outcomes. This project will move the field toward precision medicine for CP and create a foundation for develo...