INTRODUCTION: Although routine episiotomy is not recommended, it remains widely performed during childbirth, with rates varying substantially across countries. This systematic review synthesizes evidence on the prevalence and associated factors of episiotomy in Brazil. METHODS: This is a systematic review of prevalence, conducted in accordance with the recommendations of the Joanna Briggs Institute (JBI). The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the identification number CRD42024621526. Seven electronic databases were searched with no restrictions regarding publication period or language: MEDLINE/PubMed (via the National Library of Medicine), Web of Science, SCOPUS, CINAHL (EBSCO), Cochrane, EMBASE (Elsevier), and the Latin American and Caribbean Health Sciences Literature (LILACS). The JBI Critical Appraisal Checklist for Studies Reporting Prevalence Data was used by two authors to assess the quality of the selected studies. RESULTS: A total of 65 studies were included. The prevalence of episiotomy practice among women varied from 0.5% to 94%. Only 16 studies (24.6%) reported episiotomy rates close to the recommended benchmark of 10%. Thirty-three studies (50.8%) analyzed factors associated with the practice, with the most frequently reported factors being: primiparity, primigravidity, nulliparity, adolescence, youth, higher education, higher income, delivery in private healthcare settings, physician-assisted births, use of oxytocin, instrumental delivery, absence of previous vaginal birth, prematurity, newborn weight over 4000 g, and horizontal delivery position. CONCLUSIONS: A limited number of studies have met recommended episiotomy rates, underscoring the challenge of translating evidence into practice. Lower rates were seen in births attended by nurse midwives. Continuous professional training focused on evidence-based care is essential to reduce unnecessary interventions and align rates with recommended targets.
Rodrigues et al. (Mon,) studied this question.