Prescribed medicines among pregnant women attending health facilities in Eritrea: patterns, safety, and determinants.

Background: Prescribing unsafe medications during pregnancy poses significant risks for adverse maternal and fetal outcomes. This nationwide study assessed medication utilization patterns, safety risk classifications, and associated determinants among pregnant women in Eritrea. Methods: A cross-sectional study was conducted in 32 health facilities across Eritrea from October to December 2024. Data were collected through patient interviews, prescription reviews, and medical record abstractions. Medication safety was evaluated by translating Pregnancy and Lactation Labeling Rule (PLLR) narrative summaries into three clinical tiers (safe, cautious, and contraindicated) and comparing them with the legacy United States Food and Drug Administration (US-FDA) pregnancy risk categories. Descriptive statistics and logistic regression analyses were performed using IBM SPSS version 26.0 to identify factors associated with exposure to potentially unsafe medications. Results: A total of 740 pregnant women were included, with a mean (SD) age of 26.95 (5.75) years. A total of 1,031 medicines were prescribed, with a mean of 1.39 (SD = 0.65) medications per prescription. According to the legacy US-FDA system, the prevalence of prescriptions containing potentially risky medicines (Categories C and D) was 4.2% (95% CI: 3.0, 5.9). Under the PLLR framework, 96.7% of the medications were classified as "safe," 2.4% as "cautious," and 0.9% as "contraindicated." The most common "Cautious" medications were metronidazole (n = 10) and nitrofurantoin (n = 3), while "Contraindicated" agents primarily included diclofenac (n = 5) and ciprofloxacin (n = 2). While the agreement between the two systems was high (96.7%), the Cohen's kappa (kappa = 0.486) indicated moderate concordance. Type of healthcare facility (AOR = 2.38; 95% CI: 1.11, 5.12), presence of chronic illness (AOR = 12.56; 95% CI: 3.52, 44.88), and first-trimester gestational age (AOR = 3.39; 95% CI: 1.28, 8.99) were significantly associated with exposure to potentially unsafe medicines under the legacy system, while urban residence (AOR = 2.82; 95% CI: 1.23, 6.49) was a significant determinant under the PLLR framework. Conclusion: Prescribing practices during pregnancy were largely characterized by low polypharmacy and a predominant reliance on medications classified as lower-risk under both frameworks. However, the transition to the PLLR framework provides a more precise risk assessment, addressing the ambiguity of legacy categories. Targeted interventions are needed for high-risk groups, particularly women with chronic illnesses, those in early pregnancy, and urban residents.