This retrospective study reviewed medical records for 305 pregnancies associated with alcohol-related diagnostic codes at a medical center between 2014 and 2023. It assessed how reliably ICD-9 and ICD-10 codes identified any alcohol use and repeated or harm-associated alcohol use during pregnancy.
Positive predictive values differed substantially across code groups. For any alcohol use, values ranged from 10.0% for code group 305.x to 100% for O99.31x. For repeated or harm-associated use, values ranged from 1.7% to 28.0%. The authors caution against treating administrative diagnostic codes as uniformly accurate measures of prenatal alcohol exposure.
Original summary for clinical review. Read the published study.
