Global prevalence of antibiotic consumption during pregnancy: A systematic review and meta-analysis

Antibiotic use during pregnancy is widespread with notable variations across regions.

This systematic review and meta-analysis (Prospero protocol CRD42023418979) examines the prevalence and variability of antibiotic use in pregnancy globally and regionally, considering different methodologies and maternal characteristics. We searched Embase, PubMed, and Web of Science for observational studies published in English from the year 2000 and onwards. Random-effect meta-analyses were used to pool the prevalence of antibiotic consumption during pregnancy, presented as percentages with 95% confidence intervals (CI). Joanna Briggs Institute Critical appraisal checklist for prevalence studies was used for bias assessment.

Overall, 116 studies (14 from Africa, 24 from the Americas, six from Eastern Mediterranean, 57 from Europe, four from South-East Asia and 11 from Western Pacific) were included (33,821,194 pregnancies). The majority of studies (84.5%) were appraised with a low risk of bias. The prevalence of antibiotic consumption during pregnancy ranged between 0.04 to 90%, with a pooled estimate of 23.6% (95% CI: 20.1-27.5, I² =100%). Low-income countries had the highest pooled prevalence (45.3%, 95% CI: 15.4-79.1, I² =99.6%). Regionally, the Western Pacific had the highest pooled prevalence (34.4%, 95% CI: 13.4-64.1, I² =100%). The prevalence of antibiotic consumption during pregnancy increased over time in the Americas and Western Pacific. The studies exhibited considerable heterogeneity (I² >95%), and the trim-and-fill method estimated a potential 10% underestimation of the overall pooled prevalence, suggesting publication bias.

This meta-analysis suggests that about 1/4 of women worldwide use antibiotics during pregnancy. This study suggests a high prevalence of antibiotic consumption during pregnancy with disparities according to region and level of country income, ethnicity and whether antibiotics were prescribed or self-medicated. There was a variability in reported findings across age categories, potential bias from small sample sizes, and language bias from including only studies published in English.