C-reactive protein and the neonatal early-onset sepsis calculator for the diagnosis of neonatal sepsis.

ABSTRACT

Our aim was to evaluate the utility of the neonatal early-onset sepsis risk calculator (NEOSC) to the utility of C-reactive protein (CRP) for diagnosing neonatal EOS. This retrospective study reviewed the records of neonates who underwent sepsis workups due to equivocal symptoms and compared their CRP values to the calculator's recommendations and their cultures. A total of 382 newborns who underwent sepsis work-up due to equivocal symptoms were included in our study. The calculator's recommendations would have reduced the number of newborns who underwent sepsis workups by 82.5% and antibiotic treatment by 83.4% (n = 315). Considering that 373 of 382 (97.6%) ultimately had no sepsis, the calculator's specificity was higher than that of CRP (83.9% versus 76.1%). When comparing the maximal CRP value with the risk according to the neonatal sepsis calculator, a significant correlation was found between them (P < 0.01), but the relationship was not strong (Pearson's correlation = 0.27). We found a significant correlation between the risk of sepsis according to the NEOSC and the CRP values, although the correlation was not strong. The calculator's high specificity enables safe avoidance of multiple blood tests and antibiotic treatments for suspected neonates who are not infected. CRP tests can reduce the number of infected newborns the calculator may miss, at the cost of unnecessary blood tests and antibiotic therapy to many newborns.