Comparison of a Paper-Based Perinatal Infection Risk Score and the Neonatal Sepsis Calculator by Kaiser Permanente

Abstract

Objective

We aimed to compare a paper-based perinatal infection risk (PIR) score with a validated computer-based early onset sepsis calculator by Kaiser Permanente (KP-EOSC) to assess the their respective performance on neonatal sepsis detection. 

Methods

All newborn babies admitted at the department of neonatology with an increased infection risk between January 2019 and December 2021 were retrospectively included. The PIR score was designed to support the decision whether to perform additional monitoring, tests, and/or administer antibiotics. A PIR score and KP-EOSC score were calculated and compared. A PIR score ≥6 or a KP-EOSC score ≥1/1000 would trigger further investigations and observation.

Results

A total of 105 babies were included, corresponding to 3.4% of the local newborn population. Of all patients born at Jan Palfijn Hospital (Antwerp, Belgium), culture proven sepsis incidence was 0.09%. When comparing the PIR score and KP-EOSC score for well appearing children, a minimal correlation was seen (Cramér’s V 0.199, Cohen' Kappa -0.014, p=0.077) but scores tended to yield similar outcomes more frequently when clinical appereance shifted to equivoqual (Cramér’s V 0.180, Cohen’ Kappa 0.022, p=0.190).

Conclusion

When comparing the PIR tool and the KP-EOSC tool,  the PIR tool looks less specific but has an overall good sensitivity in identifying children with neonatal sepsis. In this pilot study, it appears that neonatal risk stratification can be done based on several scores, while the need for a computer and/or clinician might limit the KP-EOSC tool in certain circumstances. Additional studies are warranted to validate these findings.