PS2-19: Developing an Automated Surveillance for Fetal Alcohol Syndrome Using Electronic Medical Records

Abstract

Background/AimsPrenatal alcohol exposure is the leading preventable cause of birth defects and developmental disabilities. The full diagnosis of Fetal alcohol syndrome (FAS) requires assessment of:Using integrated information available at one source, we assessed the feasibility of developing an FAS surveillance system using the Kaiser Permanente Georgia (KPGA) Health Plan EMR.MethodsUsing EMRs, we extracted relevant information on the three main areas of FAS case definition (facial features, restricted growth, and CNS abnormalities). This was done among children up to the age of 10 years at the time of the diagnosis (including birth). Due to lack of ICD-9 codes for many of the facial features, we scanned the physician progress notes for relevant terms. These data were synthesized and applied to the FAS case definition algorithm.Results(at the time of the submission ? these results may change with further analyses):Overall, preliminary analyses showed that 23,678 children met either the criteria for growth restriction (n=5,052) or CNS abnormalities (17,296). Of these children, 1,330 met the criteria for both growth restriction and CNS abnormalities. An algorithm is to conduct text string searches for facial dysmorphic features within physician progress notes among these 1,330 children is under development and results will be presented at the HMORN conference. Results on linking these children to mothers, along with alcohol use during pregnancy will also be presented.DiscussionDevelopment of a surveillance system for FAS using EMRs is challenging. Information on two (CNS and growth) of the three main criteria can be extracted to create a pool of potential FAS cases, while information on dysmorphic facial features is more difficult to obtain. Despite these challenges, further development of this algorithm is required and the benefits of applying a well-developed algorithm across the HMORN sites will far outweigh many of the challenges in the long term. NOTE: more detailed methods and results will be available at the time of the conference.