Age-specific high-sensitivity C-reactive protein is superior to age-specific procalcitonin for diagnosing early- onset bacterial infection in term neonates

Abstract

Abstract Purpose: To investigate the predictive usefulness of age-adjusted cutoff values of high-sensitivity C-reactive protein (hs-CRP) and procalcitonin (PCT) for early-onset bacterial infection (EOBI) in term neonates. Methods: Serumhs-CRP and PCT levels in term neonates were compared among the no-symptom reference (REF), non-infectious respiratory failure (NIRF), and suspected bacterial infection (SBI) groups at birth and 1-18 h, 19-36 h, and 37-72 h postnatally. The predictive cutoff values of serum hs-CRP and PCT for EOBI were analyzed using a receiver operating characteristic (ROC) curve for each time frame. Results: Serum levels of hs-CRP and PCT in the NIRF group were significantly higher than those of the REF group at 1-72 h postnatally. Serum hs-CRP levels were significantly higher in the SBI group than in the NIRF group at all time points, but serum PCT levels had significance only at 1-18 h postnatally. ROC analysis for predicting EOBI showed that hs-CRP had a better area-under-the-curve, specificity, positive predictive value, and negative predictive value than did PCT at all time points. The best predictive cutoff value of serum hs-CRP for EOBI was 18.4 mg/L at 19-36 h postnatally, which was higher and occurred earlier than the most used serum hs-CRP cutoff value of 10 mg/L in the first several days of life in neonates. Conclusions: Age-adjusted hs-CRP cutoff values had better predictive power than age-adjusted PCT cutoff values for EOBI in term neonates during the first 72 h postnatally. Using age-specific hs-CRP may reduce prophylactic antibiotic administration for suspected EOBI in term neonates.