Medication Exposure Definitions in Perinatal Pharmacoepidemiology: Implications for Pregnancy Studies

BACKGROUND: Exposure definitions in real-world pregnancy studies vary across studies, data sources, treatments and outcomes, yet their impact on prevalence estimates remains unexamined. OBJECTIVES: To identify the most used exposure definitions during the 1st trimester and evaluate their impact on prevalence estimates: METHODS: We reviewed 400 randomly selected articles on medication use during pregnancy to identify exposure definitions. Using Danish health registers (1997-2024) and the Québec Pregnancy Cohort (1998-2015), we analysed chronic, as-needed, and acute medication use. Prevalence (per 1000) was calculated per definition, including: (1) ≥ 1 or (2) ≥ 2 filled prescriptions during 1st trimester; filled ≥ 1 prescription anytime between the 1st trimester or (3) 30 days, (4) 60 days, (5) 90 days, (6) 6 months before conception; filled a prescription during 1st trimester or before conception, (7) assuming 1 daily defined dose per day or (8) prescribed days overlapping 1st trimester. Trends across cohorts and medications were compared. RESULTS: We analysed 1,524,075 pregnancies in Denmark and 229,000 in Québec, Canada. The most common exposure definitions identified from the literature were Definition 1 (83%), followed by Definition 3 (13%). Prevalence varied across definitions, with differential impacts depending on medication and type of use. Definition 6 (1st trimester or within the 6 months before conception) yielded the highest prevalence; Definition 2 (≥ 2 fills in 1st trimester) the lowest. Broader time-windows of exposure increased prevalence, with variation by medication. The prevalence of medication use, whether chronic or acute, showed less variation across definitions than that of as-needed medication. CONCLUSIONS: Medication exposure prevalence varied by definition and medication. Using ≥ 1 filled prescriptions in the 1st trimester yielded prevalence estimates comparable to those from more complex definitions, supporting its use as a pragmatic standard. Sensitivity analyses and transparent reporting of exposure definition are essential to ensure replicability and critical appraisal.