Diagnostic accuracy of presepsin and C-reactive protein in the early stages of neonatal sepsis: A prospective study.

Introduction. Neonatal sepsis is a common complication in the neonatal intensive care unit (NICU). Biomarkers have certain limitations, and recent studies suggest that presepsin (PSEP) could be used for early diagnosis. Objective. To determine the early diagnostic accuracy of PSEP in newborns with suspected sepsis and to compare it with C-reactive protein (CRP). Population and methods. A prospective, cross-sectional study of newborns with suspected early- or late-onset sepsis. We defined confirmed sepsis as a positive blood culture or cerebrospinal fluid culture and excluded patients with congenital anomalies, those currently receiving antibiotic treatment, and those for whom parental informed consent was not obtained. ROC curves were plotted for PSEP and CRP. Results. A total of 150 suspected sepsis events in 122 newborns were included. The median gestational age was 32 weeks (IQR 29-37) and the median birth weight was 1530 grams (IQR 990-2705). The incidence of confirmed sepsis was 12% (95% CI, 7.3-18.3), and the median PSEP was 511 pg/mL (IQR, 392-785), with no statistically significant differences among confirmed sepsis, clinical sepsis, and no sepsis. The ROC curve for PSEP in confirmed sepsis was 0.62 (95% CI 0.48-0.76) and 0.77 (95%CI 0.68-0.86) for CRP. The optimal cutoff value for PSEP was 447 pg/mL and 14.8 mg/dL for CRP. Conclusion. The diagnostic accuracy of PSEP is low in the early stages of sepsis.