Affiliation:
1. Department of Pediatrics, Zeynep Kamil Maternity and Children's Training and Research Hospital, University of Health Sciences, Istanbul, Türkiye
2. Division of Neonatology, Department of Pediatrics, Zeynep Kamil Maternity and Children's Training and Research Hospital, University of Health Sciences, Istanbul, Türkiye
Abstract
Abstract
Objective Late-onset sepsis (LOS) is a major public health burden globally. Newborns with very low birth weight (VLBW) are at high risk of mortality related to LOS. The protective effect of breast milk feeding against many infections has been studied but data on the effect of breast milk feeding on LOS mortality are limited. In this study, we aimed to evaluate the risk factors for LOS-related mortality.
Methods This single-center, case–control study was conducted retrospectively from August 2013 to July 2018. VLBW newborns with ≤32 weeks of gestational age who had culture-proven LOS were included in the study. Demographics, clinical and laboratory data, and parenteral and enteral feeding details within 72 hours before LOS episodes were extracted from electronic and paper medical records.
Results A total of 190 LOS episodes were identified in 168 VLBW newborns with a median birth weight of 888 (695–1,143) g. Lower birth weight (adjusted odds ratio [aOR] 0.81, 95% confidence interval [CI] [0.69–0.96], p = 0.01), C-section delivery (aOR 0.38, 95% CI [0.17–0.84], p = 0.02), gram-negative (aOR 4.97, 95% CI [2.01–12.28], p = 0.001) and polymicrobial sepsis (aOR 6.29, 95% CI [1.34–29.47], p = 0.03), and lower breast milk feeding 72 hours before LOS episodes (aOR 0.89, 95% CI [0.80–0.99], p = 0.03) were independently associated with higher odds of LOS related death.
Conclusion Gram-negative sepsis was associated with higher odds of LOS mortality and C-section delivery was associated with lower odds of LOS mortality. Additionally, every 10 mL/kg/day increase in breast milk feeding was associated with 11% lower odds of LOS mortality although this finding should be interpreted cautiously as there may be unadjusted confounders due to the study design.