Impact of New Evidence-Based Sepsis Guidelines on Breastfeeding Rates

Purpose: Traditional management of asymptomatic full term neonates with risk factors for early-onset sepsis has potentially negatively affected maternal-infant bonding and breastfeeding success. Beginning in January 2016, the Lenox Hill Hospital NICU adopted new evidence-based early-onset sepsis guidelines for the evaluation of asymptomatic full term neonates. While the length of stay in the NICU was decreased to just hours in some cases, overall antibiotic utilization did not decrease. In June 2017, the guidelines were modified using the Kaiser Permanente Neonatal Early-Onset Sepsis Calculator to further stratify risk of sepsis. Infants with a low risk of sepsis were transferred to the mother’s room after a short period of observation in NICU. The goal of the new guidelines was to decrease antibiotic utilization, and to decrease length of stay in the NICU of asymptomatic term newborns with sepsis risk factors. Our hypothesis is that with less separation of the newborn from the mother, in-hospital breastfeeding success will improve. Methods: All asymptomatic full term neonates admitted to the NICU for sepsis evaluation in 2 periods (January 29, 2016 - June 4, 2017 and June 5, 2017 - February 11, 2018) were identified. Demographic data and additional variables were collected. The Mann-Whitney test or chi-square test were used to assess the comparability of the two cohorts. Univariable linear regression was utilized to compare patient variables to mean total breastmilk percent (primary outcome). Once confounders were identified, the outcome was analyzed using multivariable linear regression with backward elimination. Results were considered statistically significant when p<0.05. Results: Period 1 included 268 babies, and Period 2 had 138 babies. Demographic characteristics between the two groups were similar (Table 1). There was a significant decrease in NICU length of stay and antibiotic exposure, with a significant increase in exclusive breastfeeding at discharge (Table 2) in the second period. Variables that were significant in the univariable analysis but were removed after backward elimination included hours in hospital, hours on antibiotics, race and mode of delivery. Maternal age (p < 0.002), hours in NICU (p < 0.0001), and first feed breastmilk (p < 0.0001) remained in the multivariable model. For each hour increase in NICU length of stay, there is a statistically significant decrease in predicted total breastfeeding percent value when other variables are held fixed. Conclusion: By implementing a new evidence-based, early-onset sepsis protocol, we have effectively and safely decreased neonatal antibiotic exposure and NICU length of stay. We suggest that the decrease in separation time of the mother from baby has resulted in increased total amount of breastmilk consumed by the baby and an increase in exclusive breastfeeding at discharge. By optimizing breastfeeding success in the hospital environment, we aim to improve breastfeeding success post discharge.