RESUMEN
OBJECTIVE: To compare the incidences of early and late-onset neonatal sepsis, including group B streptococcus (GBS) and Escherichia coli (E. coli) before and after implementation of universal screening and intrapartum antibiotics prophylaxis (IAP). DESIGN: Retrospective cohort study. SETTING: Eight public hospitals and 31 Maternal and Child Health Centres (in Hong Kong. POPULATION: 460 552 women attending routine antenatal service from 2009 to 2020. METHODS: Universal culture-based GBS screening has been offered to eligible women since 2012. Total births, GBS screening tests, maternal GBS colonisation and neonatal sepsis with positive blood or cerebrospinal fluid were retrieved from clinical and laboratory database. MAIN OUTCOME MEASURES: Maternal GBS colonisation rate, early- and late-onset neonatal sepsis (including GBS and E. coli). RESULTS: Of 318 740 women with universal culture-based screening, 63 767 women (20.0%) screened positive. After implementation of GBS screening and IAP, the incidence of early-onset neonatal sepsis decreased (3.25 versus 2.26 per 1000 live births, p < 0.05), including those caused by GBS (1.03 versus 0.26 per 1000 live births, p < 0.05). Segmented regression showed that change in early-onse GBS sepsis incidence after screening was the only significant variable in the outcome trend. There was no significant evidence of increase in incidence of late-onset neonatal sepsis including those caused by GBS. CONCLUSIONS: Universal culture-based GBS screening and IAP were associated with reduction in early-onset neonatal sepsis including GBS disease. Although an increase in incidence of late-onset neonatal sepsis including those caused by GBS cannot be totally ruled out, we did not identify significant evidence that this occurred.