RESUMO
OBJECTIVE: The aim of this study was to describe the inflammatory markers studied in umbilical cord blood and to analyze the performance of the three markers most frequently studied for the prediction of early-onset neonatal sepsis. DATA SOURCES: An integrative review from 1995 to 2021 was performed, with a search in the MEDLINE, Embase, Cochrane Library, SciELO, and gray literature databases, using the terms "neonates," "newborns," "neonatal sepsis," "early-onset neonatal sepsis," "neonatal infection," "inflammatory markers," "biomarkers," "cord blood," "fetal blood." STUDY SELECTION AND DATA EXTRACTIONS: Study evaluation was limited to primary studies, prospective, observational or intervention, descriptive or analytical, that assessed the diagnosis of early-onset neonatal sepsis using inflammatory markers in umbilical cord blood, in Portuguese, English, or Spanish. Qualitative studies, reports, review studies, and case series were excluded. Only studies with a punctuation ≥ 6 in the Newcastle-Ottawa scale were included. RELEVANCE TO PATIENT CARE AND CLINICAL PRACTICE: Sixteen studies were included in the qualitative synthesis. Procalcitonin, C-reactive protein, and interleukin-6 were the most frequently studied markers. The best performance for C-reactive protein was observed at a 0.2 mg/L cutoff, with a sensitivity of 82% and a negative predictive value of 99%. Procalcitonin presented the best performance at a 0.5 ng/mL cutoff with 87.5% sensitivity and 98.7% negative predictive value. Interleukin-6 presented the best performance at a 108.5 ng/mL cutoff, with 95% sensitivity and 97.4% negative predictive value. CONCLUSION: The evaluation of markers in the umbilical cord for the diagnosis of early-onset neonatal sepsis, could contribute to a more assertive therapy for the neonate and anticipate sepsis screening. Since the cost is less and technically easier, C-reactive protein is recommended for routine use.