Validation of Brighton pediatric early warning score for predicting clinical deterioration in the emergency department

Abstract

Objectives: The objectives of the study were to validate the pediatric early warning score (PEWS) for predicting clinical deterioration in children attending the emergency department (ED) and its inter-rater agreement between nurses and pediatric trainees. Methods: This prospective observational study was done on patients aged 1 month–12 years attending the ED in tertiary care institute of South India, from June 2017 to June 2018. They were evaluated using PEWS at time of admission (S0) and after 1 h (S1) by both nurses and pediatric trainees. The scores were correlated with outcome. A receiver operating curve (ROC) was plotted to determine the cutoff value for PEWS in predicting deterioration. Inter-rater reliability between nurses and pediatric trainee was evaluated. Results: A total of 738 patients were included in the study. Patients admitted in pediatric intensive care unit (S0 2.623±2.075 and S1 2.024±1.592) had higher mean PEWS both at the time of admission and at 1 h, compared to patients admitted to the ward (S0 – 0.7551±1.129 and S1 – 0.5165±) (p<0.001). The optimal cutoff scores on the ROC for predicting clinical deterioration are S0 = 2 and S1 = 2 with areas under the curve of 0.76 and 0.78, respectively. The sensitivity, specificity, and likelihood ratios of these scores are as follows: S0 – sensitivity – 67.46%, specificity – 76.34%, positive likelihood ratio – 2.85, and negative likelihood ratio – 0.43; and S1 – sensitivity – 61.9%, specificity – 88.89%, positive likelihood ratio – 5.57, and negative likelihood ratio – 0.43. Inter-rater reliability between nurses and pediatric trainees was excellent (intraclass coefficient 0.99). Conclusions: Pediatric early warning score can be a useful clinical tool for identifying children at risk of deterioration. PEWS can be effectively used by nursing personnel in the triage room to identify a sick child.