Presepsin to safely reduce antibiotics in preterm infants (PRESAFE): a randomized controlled trial with a concurrent observational study.

Background: Accurate and rapid diagnosis of early-onset neonatal sepsis (EOS) in preterm infants remains problematic due to lack of specific symptoms and diagnostic tools. Following the Dutch EOS guidelines, over 80% of infants born <32 weeks of gestation are empirically started on antibiotics directly after birth, while the actual incidence of EOS varies between 1-2%. Unnecessary antibiotic exposure leads to severe short and long term complications. The biomarker presepsin, also known as soluble CD14 subtype, may be used to reduce antibiotic prescription in preterm infants after birth. The objective of this study is to investigate whether adding a presepsin-guided decision to the guideline safely reduces antibiotic exposure directly after birth in preterm infants. Secondly, the diagnostic accuracy of presepsin for EOS will be evaluated. Methods: The PRESAFE trial is a multicentre, randomized controlled trial (RCT), including a concurrent observational study. Presepsin levels are determined from umbilical cord blood or during first regular blood draw in all infants born <32 weeks gestation. Infants who qualify for empirical antibiotics according to the Dutch EOS guideline are categorized as moderate or high risk for EOS based on prespecified high risk criteria. Infants not qualifying for empirical antibiotics are categorized as low risk. Preterm infants with a moderate risk for EOS are randomized 1:1 into an intervention arm and a comparator arm. In infants allocated to the intervention arm, empirical antibiotics will only be started above a prespecified presepsin level of ≥645pg/ml. In the comparator arm infants will be treated according to standard of care following the Dutch EOS guideline, equivalent to starting empirical antibiotics. The co-primary outcomes of the RCT are the incidence of culture-proven EOS (non-inferiority) and unnecessary antibiotics administration (superiority). The required sample size for the RCT is 900 patients. Infants with a high- or low-risk of EOS are excluded for randomization but included in a concurrent observational study, and treated according to the Dutch EOS guideline. The primary outcome of this part is diagnostic accuracy of presepsin. Discussion: The findings of the RCT will provide evidence for safe and effective reduction of administration of antibiotics for suspected EOS in infants born <32 weeks of gestation. The observational study will provide more insight in diagnostic accuracy of all infants born <32 weeks of gestation. Trial registration: ClinicalTrials.gov NCT06100614. First registered on October 25, 2023.