Application of Unified Medical Language System (UMLS) to Standardize Pediatric Drug Data.

Pediatric Electronic Health Records (EHRs) contain drug/medication data. Despite the importance of standardizing drug data to identify drug class information and enable interoperability between computer systems, sometimes no biomedical vocabulary is used, and therefore the data is not standardized. This paper employed UMLS vocabularies to standardize Cincinnati Children’s Hospital Medical Center (CCHMC) EHR drug data to use it build models for pediatric mental health trajectories. We present an approach that identifies a UMLS Concept Unique Identifier (CUI) for a drug for the scenarios when the RxNorm code is available, and also for when it is not available. We employed an offline UMLS database to identify UMLS CUI for a drug when the RxNorm code was available. We used UMLS API and a fuzzy string matching approach to determine UMLS CUI for a drug when the RxNorm code was not available. Our early results demonstrate that drug data can be standardized using the UMLS vocabulary and this approach is helpful for anomaly detection.