Predicting Early Onset Neonatal Sepsis in asymptomatic infants ? 34 weeks gestation: to kaiser or not to kaiser?

Abstract

Objectives In 2021, National Institute for Health and Care Excellence (NICE) NG195 recommended Kaiser Permanente Sepsis Risk Calculator (KP-SRC) as an alternative framework to predict the risk of early onset neonatal sepsis (EONS) in babies born at -34 weeks.1 There are local and regional differences in its use across the UK. The East of England guideline uses risk factors stipulated by NG195 for initial screening, and babies with one red flag or -2 amber risk factors who would have received antibiotics based on NG195 are eligible for KP-SRC assessment.1 2 We aimed to evaluate KPSRC usage, safety and reflect on our local experience following its implementation. Methods Retrospective analysis of babies assessed with the KPSRC between February 2022 and January 2023 was undertaken. We also retrospectively applied KP-SRC on all babies born at - 34 weeks with culture positive EONS who were not initially assessed. Results Eighty-five (85) babies had KP-SRC assessment at birth. 74% (n=63) were appropriately scored, and 25.8% (n= 22) were scored despite not meeting local criteria. Median (IQR) gestation was 39.5 (38.3 - 40.4) weeks. Amongst those appropriately scored (n=63), 42.9% (27/63) were recommended for 'no culture, no antibiotics', 42.9% (27/63) for ' blood culture only' and 14.3% (9/63) were recommended for 'blood culture and empirical antibiotics'. Amongst babies recommended for 'no culture, no antibiotics', 37% (10/27) were started on antibiotics for clinical concerns whilst seventeen babies underwent observations without antibiotics, representing 26.9% (17/63) reduction in antibiotics use. There were no missed cases of sepsis and no readmissions for sepsis within first 7 days of life. We identified six culture-confirmed EONS that were not assessed by KP-SRC. Four were symptomatic soon after birth, and antibiotics were commenced following clinical assessment. Two babies were asymptomatic, one had an amber risk factor (maternal pyrexia) and subsequently developed a fatal E. coli sepsis with meningitis. Interestingly, retrospective application of KP-SRC would have prompted a septic screen at birth for this baby. The second infant would not have been recommended for a septic screen based on KP-SRC or NICE NG195. Conclusion Implementation of KP-SRC has safely reduced antibiotics use in our centre. Restricting its use to a subset of infants already identified by another screening tool may limit its negative predictive value and overall performance. We suggest that all infants born at -34 weeks with risk factors for EONS should be assessed on KP-SRC.