Validity of the Pediatric Early Warning Score and the Bedside Pediatric Early Warning Score in Classifying Patients Who Require the Resources of a Higher Level Pediatric Hospital

Introduction: The pediatric early warning score (PEWS) and the bedside pediatric early warning score (BPEWS) are validated tools that help determine the need for critical care in children with acute medical conditions. These tools could be used by EMS and have not been evaluated outside of the hospital. This study retrospectively tested the validity of these tools in the prehospital setting to identify children who needed a hospital with higher level pediatric resources. Methods: This was a multi-center retrospective validation of screening tools using prehospital and in-hospital data obtained from 3 EMS agencies. EMS patient records from April 1, 2013 to April 30, 2015 were used to identify subjects for this analysis. Pediatric patients were retrospectively classified using the PEWS based on the clinical information documented in the EMS medical record. Those with PEWS scores greater than 4 were matched to a subject with scores less than 4 based on age, gender, and paramedic primary impression. Hospital medical record review was then used to determine whether the patient required a hospital with higher level pediatric resources. These classifications were used to calculate sensitivity, specificity, and resultant 95% confidence intervals. The analysis was repeated for included subjects who had sufficient data to calculate BPEWS. Results: There were 386 patients enrolled. A PEWS ≥ 4 demonstrated a sensitivity of 62.8 (95% CI 53.6-71.4) and a specificity of 55.9 (95% CI 49.6-61.9) in identifying a patient who required a hospital with higher level pediatric resources. There were 44 pairs of patients that had sufficient EMS data documented to calculate a BPEWS. A BPEWS ≥ 7 demonstrated a sensitivity of 46.4 (95% CI 27.5-66.1) and a specificity of 76.7 (95% CI 64.0-86.6) to correctly classify a patient who required a hospital with higher level pediatric resources. Conclusion: In the prehospital setting neither PEWS nor BPEWS exhibited sufficient sensitivity for clinical use to accurately identify children who need a hospital with higher level pediatric resources. Further research should be conducted to identify variables that are captured by prehospital care providers and are associated with children who need a hospital with higher level pediatric resources.