Validity of International Classification of Diseases Codes for Identifying Alcohol Use and Repeated or Harm-associated Alcohol Use in Pregnancy.

Abstract

Objectives: To describe the validity of alcohol use disorder (AUD), the International Classification of Diseases (ICD) codes for alcohol use and repeated or harm-associated use in pregnancy.

Methods: We conducted a retrospective study of pregnancies in individuals who sought care at a medical center from May 2014 to August 2023. We selected all pregnancies with an ICD 9th (ICD-9) or 10th (ICD-10) revision AUD code (ICD-9: 303.9x and 305.x; ICD-10: F10.x and O99.31x) and calculated their positive predictive value (PPV) for capturing alcohol use and repeated or harm-associated use in pregnancy using review of health record as the reference standard. We defined alcohol use as any alcohol consumption during pregnancy and repeated or harm-associated alcohol use as a repeated pattern associated with adverse consequences.

Results: AUD ICD codes were associated with 305 unique pregnancies. The most common AUD ICD code group was 305.x (n=177, 56.4%), followed by F10.x (n=105, 33.4%), O99.31x (n=25, 8.0%), and 303.9x (n=7, 2.2%). The PPV of AUD codes for capturing alcohol use in pregnancy ranged from 10.0% (95% confidence interval [CI], 8.9%-11.2%) for 305.x to 100% (95% CI, 80.9%-99.5%) for O99.31x. The PPV of AUD codes for capturing repeated or harm-associated use in pregnancy ranged from 1.7% (95% CI, 1.6%-1.8%) for 305.x to 28.0% (95% CI, 21.3%-35.9%) for O99.31x.

Conclusions: The PPV of AUD codes for identifying alcohol use and repeated or harm-associated use in pregnancy varies widely across ICD-9 and ICD-10 code groups. These results should be considered when estimating alcohol exposure in pregnancy from administrative data.