Aims and Objectives Identifying unwell children in the high pressure setting of the paediatric emergency department (ED) is difficult. Multiple Paediatric Early Warning Score (PEWS) systems have been used to track inpatient deterioration and assess need for escalation but have not been widely used/validated in the ED. The English national PEWS (nPEWS) was recently announced, a standardised system combining physiological parameters to summarise a child’s condition. Whilst not developed for ED, it would be logical to use consistent methods to assess patients through their journey. Here we evaluate the utility of triage nPEWS as a predictor of admission or deterioration in the ED for the first time.

Method and Design Administrative data were retrospectively collected from attendances to the Bristol Royal Hospital for Children (BRHC) ED in 2023. We created an anonymised dataset with all observations from attendances linked to outcomes. To determine prognostic ability of nPEWS, scores were retrospectively calculated for triage observations; risk of general and PICU admission were calculated for each score. Area under the receiver operating characteristic curves (AUCs) were plotted. Statistical analyses were performed using R version 2023.12.1.

Results and Conclusion 29,170 patients were included. Whilst higher nPEWS scores were indicative of an increased likelihood of admission, absolute risks were very low, implying poor clinical utility. Children with low scores were still admitted to PICU and children with high scores were frequently discharged. However, a score < 3 correlated with a decreased admission risk, and a high likelihood of admission was observed from scores of 10 onwards. The first nPEWS score predicted admission with an AUC of 0.67 (95% CI, 0.66 – 0.68).

Triage nPEWS is a poor predictor of admission and PICU attendance in an ED-setting. We are now evaluating the utility of serial nPEWS assessment and plan to replicate in other centres.