750 - Prolonged Empiric Antibiotics Use at Birth in Infants Born Less than 25 weeks

Publication Number: 2731.750

Gesca H. Borchardt, Arnold Palmer Hospital for Children, Orlando, FL, United States; Thais O. Queliz, Arnold Palmer Hospital for Children, Winter Park, FL, United States; Mary Schreck, Orlando Health, Orlando, FL, United States

Background: Infants born at the threshold of viability face extremely high mortality and morbidity. At our center, sepsis is the leading cause of death among infants born at 22–24 weeks’ gestation, despite advances in perinatal and neonatal care. Empiric antibiotics are universally initiated after birth, yet the optimal duration of therapy remains uncertain. Whether extending empiric antibiotics beyond the 48-hour mark improves survival or alters morbidity in this fragile population remains unknown.

Objective: Primary outcomes were survival to discharge and 30-day mortality; secondary outcomes included late-onset sepsis (LOS), necrotizing enterocolitis (NEC), and spontaneous intestinal perforation (SIP).

Design/Methods: We reviewed 296 inborn infants at 22–24 weeks’ gestation (2018–2023). For the 48-hour analysis, those who died before 48 hours (n = 22) or had early-onset sepsis (n = 28) were excluded (n = 246). For the 72-hour comparison, infants who died before 72 hours were also excluded (n = 242). Groups were stratified by empiric antibiotic duration (≤48 vs >48 hours; ≤72 vs >72 hours). Statistical comparisons used Wilcoxon tests or Fisher’s exact tests.

Results: Among all 22–24 week infants, survival and 30-day mortality did not differ at either the 48- or 72-hour cutoffs. In 22–23 week infants, 30-day mortality decreased from 33.8% to 14.8% (p = 0.02) and overall mortality from 43.1% to 24.1% (p = 0.03) when treated >72 hours compared to ≤72 hours. There were no differences in LOS or NEC rates. LOS occurred later with prolonged antibiotics (median 17.5 vs 9 days, p < 0.0001). SIP was more frequent in the prolonged group (p ≈ 0.04), an association likely reflecting antibiotic extension in infants with evolving intestinal disease rather than a causal effect.

Conclusion(s): Among periviable infants, empiric antibiotic therapy >72 hours was associated with significantly lower 30-day (33.8% vs 14.8%) and overall mortality (43.1% vs 24.1%) in 22–23 week infants, but not in the overall 22–24 week cohort. Prolonged exposure did not increase LOS or NEC, though LOS presented later. The apparent rise in SIP likely reflects antibiotic extension in infants with evolving intestinal disease rather than a direct effect of therapy. These findings support individualized antibiotic stewardship and further prospective evaluation in this population.

30-day mortality and overall mortality of infants born at 22-23 weeks
IMG_0049.jpegThis table shows the difference in mortality at 30-day and overall for infants born at 22 and 23 weeks based on duration of empiric antibiotic use. The columns are divided into two different groups: Infants who received < 48 hours vs > 48 hours of empiric antibiotics, and infants who received < 72 hours vs > 72 hours of empiric antibiotic.

Survival curve for 22 and 23 weeks infants based on antibiotics duration
IMG_0050.jpegThe graph shows that the curves of infants who received <72 hours vs > 72 hours of empiric antibiotics are statistically different both at 30 day mortality and overall mortality.

Secondary outcomes: Late Onset Sepsis (LOS), Spontaneous Intestinal Perforation (SIP) and Necrotizing Enterocolitis (NEC) for infants at 22 to 24 weeks
IMG_0051.jpegRate of LOS, SIP and NEC for infants born 22-24 weeks when separated into two different comparison groups: < 48 hours vs > 48 hours of empiric antibiotics, and < 72 hours vs > 72 hours empiric antibiotics.