G287(P) Can ‘pews’ scores be helpfully employed in paediatric major trauma?

Abstract

Aims Evaluate the performance of established Paediatric Early Warning Scores (‘PEWS’) in paediatric major trauma. Methods Retrospective Emergency Department (ED) data was obtained from all paediatric trauma calls in a major trauma centre (1334 patients), over a 4 year period (2011–2015), to study five widely used ‘PEWS’ systems: Brighton, Cardiff Vale, COAST, ManChews and PAT-POPS. Individual ‘PEWS’ scores were calculated using initial physiological parameters, staff concerns and comorbidities. Each patient’s TARN (Trauma Audit and Research Network) Injury Severity Score (ISS) was also obtained, for comparison. ED disposition (Home, Ward, PICU or Theatre) was correlated to the score’s sensitivity/specificity and areas under the receiver operating curves (AUC), with respect to ED disposition, were calculated. To date, 141 patient’s data has been studied (findings presented below). Results Sample size: 141 patients. Median age: 13 years. No recorded deaths in ED. AUCs for the different PEWS systems against outcomes are displayed in table 1. Findings Accepting relatively small numbers were evaluated, COAST and PAT-POPS equally performed best for ward admission (AUC’s 0.68); Brighton and Cardiff-Vale best predicted PICU admission (AUCs 0.76 and 0.73 respectively); theatre admission and discharge home did not demonstrate statistically reliable predictability using any of the PEWS systems studied except for PAT-POPS for discharge home (AUC 0.67). Conclusion COAST and PAT-POPS best predict the need for ward admission, with Brighton and Cardiff-Vale best predicting PICU admissions, (although the AUCs for ward admission are considered poorand those for PICU admission only fair). When compared with all of the remaining scores, TARN’s ISS (calculated post-event) had better predictive value for ward admission.