Developing Computable Phenotypes (CPs) for Identifying Children, Adolescents, and Young Adults with Diabetes Using Electronic Health Records in the DiCAYA Network

The DiCAYA Network, which is funded by the CDC and NIDDK, is working to create a nationwide electronic health record-based diabetes surveillance system. The feasibility and performance of existing diabetes computable phenotypes in this system were unknown, as they had only been tested in individual health systems. This study aimed to develop a DiCAYA-wide EHR-based computable phenotype (CP) to identify prevalent cases of diabetes by type. We conducted network-wide chart reviews for 1830 youth (age < 18y) and 1830 young adults (age 18 - <45y) among individuals with possible diabetes, identified as those with any potential indication of diabetes in the EHR (any diabetes-related diagnosis, medication, or abnormal lab test). We assessed the performance of the DiCAYA CP (two outpatient diagnoses or one inpatient diagnosis by type) for identifying diabetes and diabetes type in children and young adults in the reviewed sample, using chart review as the gold standard. (Table 1). We developed a highly accurate EHR-based CP for diabetes in children and young adults based on EHR data from the DiCAYA network. Our results support the feasibility of deploying a unified CP across a nationwide EHR network for diabetes surveillance, with consistently high accuracy across the participating sites. Disclosure H.Shao: Consultant; Lilly Diabetes. B.E.Dixon: Consultant; Merck & Co., Inc. E.Lustigova: None. M.Rosenman: None. V.W.Zhong: None. S.Conderino: None. J.Divers: None. L.Thorpe: None. A.D.Liese: None. C.Rudisill: Stock/Shareholder; Pfizer Inc., Viatris Inc., Teladoc Health, Catalent. J.S.Obeid: None. Y.Guo: None. M.E.Pavkov: None. G.Imperatore: None. J.A.Ewing: None. J.Bian: None. P.Li: None. D.Dabelea: None. R.B.Conway: None. T.L.Crume: None. B.S.Schwartz: None. A.G.Hirsch: None. Funding Centers for Disease Control and Prevention (U18DP006512)