Violence against women is a major problem in Brazil, but data on its prevalence are scarce. We aimed to estimate the prevalence of physical violence against women in Brazil. We conducted a systematic review and meta-analysis of the prevalence of physical violence against women. Population-based researches that assessed physical violence in Brazilian women were searched on MEDLINE, Embase, Scopus, and VHL/BIREME. The last search update was carried out in March 2020. Two researchers selected the studies, extracted the data, and assessed the quality of the eligible studies. Summary of prevalence and 95% confidence interval (CI) was calculated using Freeman-Tukey double arccosine transformation, weighted by the official local population size. Heterogeneity was estimated by I 2 and investigated by meta-regression analyses. Of 3,408 reports, 13 studies carried out from 1999 to 2016 (n = 25,781 women) were included. Most studies had limitations on sample size (5/13) and response rate (7/13). The prevalence of physical violence was 22.4% in lifetime (95% CI [21.6, 23.2%]; I 2 = 99.0%), and 11.5% in previous year (95% CI [11.1, 11.9%]; I 2 = 99.5%). Assuring privacy during interview significantly increased the prevalence (p = .028; residual I 2 = 80.0%). Higher prevalence was also observed in studies with adequate sample source, validated questionnaire, and privacy (in both recall periods), potentially due to lower risk of nonresponse bias. Over two in 10 Brazilian women suffered physical violence during their lives, and over one tenth, in the previous year. Measurement of outcome affected the prevalence; privacy should be assured for the interviewee for future reliable estimates in the country.
Keywords: Brazil; domestic violence; meta-analysis; prevalence; systematic review; violence exposure; woman.