Investigators conducted a prospective comparative cohort study among 580 pregnant women delivering at five health facilities in Mwanza City, Tanzania, between August 2023 and February 2025. The study team collected a maternal lower vaginal swab before delivery for culture and antimicrobial susceptibility testing. Neonates were observed for seven days after birth using standardized clinical and laboratory criteria to identify early-onset neonatal sepsis (EONS); culture confirmation was performed where applicable. The authors reported ethical approvals from institutional review bodies and made study data available in the article supplement. The source did not provide additional detail about swab timing relative to membrane rupture or intrapartum antibiotic administration beyond the methods summary.
The cohort yielded 89 neonates diagnosed with EONS, producing a cumulative incidence of 15.3% and an incidence rate of 2.42 per 100 person-days (95% CI: 1.97–2.98). Of those 89 cases, 24 (27.0%) were culture-confirmed; the remaining 65 (73.0%) were clinically diagnosed based on the study's standardized criteria. Kaplan–Meier survival analysis and Cox proportional hazards regression were used to estimate incidence over the seven-day risk window and to identify independent predictors. The authors specified these statistical approaches but did not supply the full model covariate list in the abstract; those details are present in the full preprint text.
Data indicated a marked difference in sepsis incidence between neonates born to colonized versus non-colonized mothers. Neonates of colonized mothers experienced an incidence rate of 3.41 per 100 person-days, compared with 0.34 per 100 person-days among neonates of non-colonized mothers. After adjustment, maternal vaginal pathogenic bacterial colonization emerged as the strongest predictor of neonatal sepsis (adjusted hazard ratio [aHR] 9.52, 95% CI: 3.47–26.09; p<0.001). The study authors reported that colonization status was determined by lower vaginal culture, but the abstract did not enumerate the overall prevalence of maternal colonization in the enrolled population; readers should consult the full text for prevalence tables and organism-specific maternal carriage rates.
The regression analysis identified multiple independent maternal predictors. Maternal HIV infection independently increased the hazard of neonatal sepsis by 76% (aHR 1.76, 95% CI: 1.14–2.73; p=0.011). Maternal malaria was associated with a more than twofold increase in hazard (aHR 2.74, 95% CI: 1.43–5.22; p=0.002). Even after accounting for these conditions, maternal vaginal pathogenic colonization remained the dominant risk factor. The source did not report additional adjusted effect estimates for obstetric interventions, duration of labor, or intrapartum antibiotics in the abstract; those covariates may be addressed in the full manuscript.
Among the 24 culture-confirmed neonatal infections, Escherichia coli accounted for 58.3% of isolates. Klebsiella pneumoniae and Staphylococcus aureus each comprised 20.8% of culture-confirmed pathogens. Antimicrobial susceptibility testing was performed on cultured isolates; however, the abstract did not provide resistance patterns or susceptibility percentages. The authors therefore noted pathogen distribution but did not report detailed antibiograms in the abstract text.
The study authors recommended strengthening maternal infection prevention measures, improving neonatal surveillance, and expanding microbiological diagnostic capacity in this setting. They stated that prospective intervention studies are needed to determine whether targeted screening and prevention of maternal bacterial colonization can reduce neonatal sepsis. The publication is a preprint and has not undergone peer review; that status limits immediate applicability to clinical practice. The abstract did not provide full methodological detail on swab timing, intrapartum antibiotic use, or comprehensive covariate adjustments in the regression models—gaps that warrant attention when interpreting causality. Future work should address whether screening and decolonization strategies reduce EONS incidence and should report organism-specific maternal carriage prevalence and antimicrobial resistance patterns in full, peer-reviewed analyses.