Taking Baby Steps to Precision: Safety and Trainee Reflections from the Implementation of Kaiser Permanente Early Onset Neonatal Sepsis Risk Calculator (Kp-Src)

Abstract

Objectives The Kaiser Permanente Early Onset Neonatal Sepsis Risk Calculator (KP-SRC) is a recognised alternative framework in the UK to identify babies at risk of early onset neonatal sepsis (EONS) using a multivariate risk model.1 2Whilst it has demonstrated safe reduction in empirical antibiotic usage,2 there have been concerns regarding its sensitivity.3 Our aim was to develop a local framework to implement the KP-SRC in accordance with regional guidelines, whilst ensuring patient safety and efficient workflow for trainees. Methods We developed a local pathway for the use of KPSRC, including electronic 'smart-phrases', and teaching packages for trainees to ensure standardised local usage. We used the 'Plan, Do, Study, and Act' (PDSA) cycle for quality improvement, and undertook a retrospective audit of its use between February 2022 and January 2023. Anonymized patient data was collected and analyzed, and further qualitative feedback was obtained from health professionals on its use. Results A total of 63 babies were identified as appropriately scored on KP-SRC pathway. 42.9% (27/63) were stratified as low risk, hence avoided initial antibiotics and were placed on observations accordingly. In this subgroup, 37% (10/27) were subsequently screened based on clinical reasons during the observation period, including hypoglycaemia, hypothermia and signs of respiratory distress. There were no culture positive babies and no readmission with sepsis within the first week of life. Feedback from trainees were mixed. Some reported that it 'works well screening less babies which seems positive', whilst others felt 'uncomfortable that we are missing high risk babies'. In addition, due to trainees' concerns regarding lengthy documentation, we streamlined the documentation workflow for trainees and simplified 'smart-phrases'. Some patients that were not 'eligible' to be scored on KPSRC, e.g., those with only 1 maternal risk factor. Trainees felt it would still be useful to use KP-SRC for these patients to aid clinical decision making. Conclusion Our local experience demonstrates the successful use of this predictive analysis tool and the streamlined workflow integration to enable safe reduction in antibiotic usage. More broadly, this qualitative improvement project has shown the transformative potential of precision medicine in Neonatology. An ongoing continual prospective audit aims to evaluate patient safety, efficacy and sustainability.