Risk Stratification to Support Antibiotic Stewardship and Breastfeeding Exclusivity in a Military Treatment Facility

To safely reduce unnecessary antibiotic exposure in neonates exposed to chorioamnionitis and inadequately treated Group B Streptococcus (GBS) using the early-onset sepsis (EOS) calculator for risk stratification and a 36-hour antibiotic duration. Evidence-based quality improvement initiative. Obstetric service at a midsized military treatment facility with approximately 2,000 births annually and no standard process for neonatal EOS risk assessment. Clinical nurse specialist, physicians, nursing leadership, unit-level nursing champions, and nurses assigned to the mother–baby and labor and delivery units. An interdisciplinary working group created a protocol to institute an EOS risk assessment calculator, a note for the electronic heath record, and interdisciplinary education for all staff providing care to neonates in our facility. Before implementation of the EOS calculator, 97.6% of neonates exposed to chorioamnionitis or inadequate maternal GBS treatment received antibiotics; after implementation, the mean rate dropped to 32%. Exclusive breastfeeding rates before discharge in neonates exposed to chorioamnionitis or inadequate maternal GBS treatment also increased during this time, from 40% to a mean of 89%. After implementation, there were no readmissions to our institution for culture-proven sepsis within 14 days of discharge. Multidisciplinary team-led implementation of the EOS calculator and of shortened antibiotic duration were associated with safely reduced antibiotic exposure in well-appearing neonates exposed to chorioamnionitis and GBS. In addition, dramatically improved rates of exclusive breastfeeding at discharge were observed in this population. Risk stratification using an early-onset sepsis calculator was associated with safely reduced antibiotic exposure in well-appearing neonates exposed to chorioamnionitis and Group B Streptococcus; improved exclusive breastfeeding rates were also observed.