Utility of Pediatric Early Warning Sign Score in Predicting Outcome of PICU Admissions at a Suburban Tertiary Care Hospital

Abstract

Assessment of the severity of illness is very important in intensive care unit care for quality assessment, assessing prognosis, and proper counseling. The goal of the study was to see how well the Pediatric Early Warning Sign (PEWS) score predicted the outcome of pediatric intensive care unit patients. This prospective cross-sectional study included children younger than 18 years. PEWS was calculated at presentation. The outcomes analyzed were mortality (primary outcome), need for mechanical ventilation, inotropic support, and length of stay (LOS). A median score was calculated and compared across the outcome groups. The performance of the PEWS was assessed for calibration and discrimination, and the best cutoff was determined. This study included 237 patients with a median score of 6 (range 4–9). Twenty-two (9.3%) patients required ventilator support and 66 (26.6%) inotropic support. The overall mortality rate was 5.1%, and 16.4% had prolonged LOS (>4 days). The median score of patients was significantly higher among those who died (8.5 vs. 6; p = 0.001), required ventilator support (8 vs. 6; p = 0.001), inotropic support (7 vs. 6; p = 0.030), and prolonged LOS (7 vs. 6; p = 0.001). On calibration, PEWS was found to have a good fit to predict mortality, the need for ventilator support, inotropic support, and prolonged LOS. Receiver operating characteristic curves for the PEWS model yield an area under the curve of 0.966 for mortality, 0.951 for ventilator support, 0.626 for inotropic support, and 0.760 for prolonged LOS. A cutoff value of > 7 was found to be the best to predict the outcome. PEWS is a robust tool to easily prognosticate the patient on the basis of clinical parameters.