6893 Assessing the utility and potential harms of NICE NG51 screening for sepsis guideline in the paediatric emergency department

Objectives The difficulty in distinguishing between a febrile child and one with serious bacterial infection keeps many paediatricians awake at night. Whilst NICE Fever in under 5s (NG143)¹ provides a practical evidence-based framework to structure assessment and management, the 2016 'Sepsis: recognition, diagnosis and early management' (NG51)² can be interpreted as recommending strict physiological criteria to trigger rapid investigation and treatment for sepsis. In practice these thresholds feel intuitively wrong, but with the retrospect-o-scope firmly levelled on any cases of missed sepsis it is a difficult balance applying NICE guidance even when studies have shown it to overtrigger.³ ⁴ Our aim was to quantify how many children presenting to an children's emergency department (ED) would be over treated/harmed by rigid application of the NG51 guideline. Methods For a feasibility study we extracted the observations of all children seen in our ED during Oct 2022 from eVitals, linking them to attendance information and lab results extracted from our EPR(Careflow). We defined 'concerns of sepsis' as measurement of CRP and blood culture, followed by inpatient admission for >48hrs. We applied age based NG51 screening thresholds to initial, temperature adjusted and (as a proxy for observation) lowest recorded heart rate during their emergency department &/or observation ward stay. Sensitivity and specificity were calculated from the proportion of patients admitted with concerns over sepsis against those triggering NICE heart rate thresholds. All analyses were done using R version 4.3.1. Results 2909/4289 children seen during this period had electronic observations, 75 met the definition of 'concerns of sepsis'. NG51 thresholds labelled 610 children as high risk (figure 1) of whom 587 were discharged. 51 children were admitted with concerns of sepsis but did not trigger NG51 heart rate criteria. NG51 heart rate screening thresholds had a positive predictive value of 4% (95CI 3–6%) and a negative predictive value of 98% (95%CI 97–98%). These were not significantly affected by adjusting heart rate for fever; obtaining a lower heart rate through observation worsened both PPV and NPV. Conclusion Identification of sepsis from heart rate thresholds alone is of limited use as a high proportion of ill rather than septic children will trigger it. That a child was tachycardic during their attendance is a very poor marker that they were septic. We are now expanding this analysis to an entire calendar year and evaluating the other NG51 criteria as well as alternative PEWS screening tools. References National Institute for Health and Care Excellence. Fever in under 5s: assessment and initial management [Internet]. 2021. Available from: https://www.nice.org.uk/guidance/ng143/. National Institute for Health and Care Excellence. Sepsis: recognition, diagnosis and early management [Internet]. 2017. Available from: https://www.nice.org.uk/guidance/NG51. Carter MJ, Stilwell PA, Nijman RG, Eisen S. Identification and treatment of paediatric sepsis: getting the balance right. Arch Dis Child. 2018 Dec;103(12):1185–6. Powell R, Jeavons K. Identifying paediatric sepsis: the difficulties in following recommended practice and the creation of our own pathway. Arch Dis Child. 2018 Jan;103(1):114–114.