Patterns of Empiric Antibiotic Administration for Presumed Early-Onset Neonatal Sepsis in Neonatal Intensive Care Units in the United States.

Objective To evaluate current patterns in empiric antibiotic use for early-onset neonatal sepsis (EONS). Study Design Retrospective population-based cohort study of newborns admitted on postnatal day 0 to 1 and discharged from NICUs participating in the Pediatric Health Information System (PHIS 2006-2013). Analyses included frequency of antibiotic initiation within 3 days of birth, duration of first course, and variation among hospitals. Results Of 158,907 newborns, 118,624 (74.7%) received antibiotics on or before postnatal day 3. Within 3 days of treatment, 49.4% (n = 58,610) were discharged home or remained hospitalized without antibiotics. There was marked interhospital variation in the proportion of infants receiving antibiotics (range: 52.3-90.9%, mean 77.9%, SD 11.0%) and in treatment days (range: 3.2-8.6, mean 5.3, SD 1.4). Facilities with higher number of newborns started on antibiotics had longer courses (r = 0.643, p < 0.001). The cost of admissions for infants born at ≥35 weeks started on antibiotics and discharged home after no more than 3 days of antibiotics was $76,692,713. Conclusion Site variation in antibiotic utilization suggests antibiotic overtreatment of infants with culture unconfirmed EONS is common and costly. Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.