Key points are not available for this paper at this time.
BACKGROUND: Understanding the aetiological organisms causing maternal infections is crucial to inform antibiotic treatment guidelines, but such data are scarce from Sub-Saharan Africa (SSA). We performed this systematic review and meta-analysis to address this gap. METHODS: Microbiologically confirmed maternal infection data were collected from PubMed, Embase, and African Journals online databases. The search strategy combined terms related to bacterial infection, pregnancy, postnatal period, observational studies, SSA. Exclusion criteria included colonization, asymptomatic infection, and screening studies. Pooled proportions for bacterial isolates and antimicrobial resistance (AMR) were calculated. Quality and completeness of reporting were assessed using the Newcastle-Ottawa and STROBE checklists. FINDINGS: We included 14 papers comprising data from 2,575 women from four sources (blood, urine, surgical wound and endocervical). Mixed-growth was commonly reported at 17% (95% CI: 12%-23%), E. coli from 11%(CI:10%-12%), S. aureus from 5%(CI: 5%-6%), Klebsiella spp. at 5%(CI: 4%- 5%) and Streptococcus spp. at 2%(CI: 1%-2%). We observed intra-sample and inter-sample heterogeneity between 88-92% in all meta-analyses. AMR rates were between 19% -77%, the highest with first-line beta-lactam antibiotics. Convenience sampling, and limited reporting of laboratory techniques were areas of concern. INTERPRETATION: We provide a comprehensive summary of microbial aetiology of maternal infections in SSA and demonstrate the paucity of data available for this region. We flag the need to review the current local and international empirical treatment guidelines for maternal bacterial infections in SSA because there is high prevalence of AMR among common causative bacteria. FUNDING: This research was supported by the NIHR-Professorship/NIHR300808 and the Wellcome-Strategic-award /206545/Z/17/Z. TRIAL REGISTRATION: Prospero ID CRD42021238515.
Chapuma et al. (Sat,) studied this question.