Background Guidelines for the prevention of perinatal group B streptococcal (GBS) disease were updated in 2010, including a revised algorithm for management of newborns with respect to risk for early-onset GBS disease (EOD-GBS).

Aim To know the impact of this new algorithm on EOD-GBS evaluations, hospital admissions, and detection of EOD-GBS cases in a newborn unit.

Methods Retrospective cohort study of infants of GBS-colonised mothers born at ≥36 weeks gestational age in two periods of time: from July to December 2010, and from July to December 2012. The following variables were analysed: gender, gestational age, chorioamnionitis, indication for and prescription of antibiotics to the mother, EOD-GBS evaluations, infant admission and outcome. Continuous data were compared by using t test; discrete data using chi square. Preventable fraction in the exposed (Pf_e) was used to quantify the impact or the new algorithm.

Results One hundred and fifty-two neonates were included in 2010 and 130 in 2012. No significant differences between both groups were found regarding gender, gestational age, chorioamnionitis, obstetric care and antibiotic prophylaxis received by mothers. There were no cases of GBS infection in both periods. The new algorithm avoided 88% evaluations in EOD-GBS screening from 2010 to 2012 (Pf_e = 0.88, 95% confidence interval [CI]: 0.39–0.96). The number of infants admitted for suspected EOD-GBS was reduced by 48.1% (Pf_e: 0.481, 95% CI: -0.648–0.864).

Conclusions Implementation of the 2010 algorithm resulted in a decrease of EOD-GBS evaluations and the number of newborn admissions for suspected sepsis.