• Jill S Sweney, W Bradley Poss, Colin K Grissom, and Heather T Keenan.
• Primary Children's Medical Center, Salt Lake City, Utah, USA. jill.sweney@hsc.utah.edu
• Disaster Med Public Health Prep. 2012 Jun 1;6(2):126-30.

ObjectiveA pediatric triage tool is needed during times of resource scarcity to optimize critical care utilization. This study compares the modified sequential organ failure assessment score (M-SOFA), the Pediatric Early Warning System (PEWS) score, the Pediatric Risk of Admission Score II (PRISA-II), and physician judgment to predict the need for pediatric intensive care unit (PICU) interventions.MethodsThis retrospective cohort study evaluates three illness severity scores for all non-neonatal pediatric patients transported and admitted to a single center in 2006. The outcome of interest was receipt of a PICU intervention (mechanical ventilation, acute dialysis, depressed consciousness, or persistent hypotension). Predictive ability was assessed using receiver operating curves (ROCs).ResultsOf 752 patients admitted to the hospital, 287 received a PICU intervention. Median scores for all tools were significantly higher for children receiving an intervention than for those who did not. ROCs showed PEWS had the least discriminatory ability, followed by PRISA-II and pediatric M-SOFA. No value of the pediatric M-SOFA produced both positive and negative predictive values better than clinician judgment.ConclusionsNo score had a clinically acceptable discriminate ability to predict patients who required a PICU intervention from those who did not. Physician judgment outperformed all three triage scores.

21 keywords...

hide…