RESUMEN

INTRODUCTION: Pediatric Early Warning Score (PEWS), also known as Brighton PEWS or Monaghan PEWS, was developed to identify children at risk for clinical deterioration in hospitals. We designed this study to describe the epidemiology of unintentional injuries in children admitted to the critical care unit in a tertiary healthcare setting, and to determine the predictive properties of PEWS in these injuries. METHODS: This is a cross-sectional study. Injury-related data were based on Haddon's matrix of agent, host, and environment factors. Each child was evaluated using PEWS on admission. We noted the following outcomes: duration of stay in the intensive unit; major intervention required; and death. RESULTS: We analyzed data from 157 children. Most of the children were in the age group of one to five years (57.7%), followed by more than five to 12 years old (37.6%). The most common injuries were bites (35.7%), falls (24.2%), and poisoning (21.7%). These injuries occurred at home (52.7%) and in the presence of a caretaker (40.0%). On admission, 11% of children were classified as green, 40% as yellow, 36% as orange, and 13% as red by PEWS. Classification of red versus the rest (orange/yellow/green) had a high sensitivity (100%), specificity (88.3%), and negative predictive value (100%) for "death" as an outcome. CONCLUSION: Most of the injuries occurred at home and at a time when the caretaker was around. Thus, it will be useful to develop "safe home interventions" and train parents in first aid to take care of these injuries on-site. Baseline PEWS was a good predictor of "poor" as well as "positive" outcomes. It may be worthwhile to implement this score regularly in the management of childhood injuries in hospitals.