Usefulness of the Pediatric Risk of Mortality Score III as a Predictor of Severity or Mortality for Injured Children

Lee, Sang Hun; Park, Kyung Hye; Park, Deuk Hyun; Cho, Junho; Park, Ha Young; Kwon, In Ho; Yeo, Woon Hyung; Lee, Junyeob; Kim, Yang Weon

Abstract

PURPOSE
The primary prevention and proper initial treatment of childhood injuries is important, as it encompasses a bigger social and economic burden than cancer and ischemic heart disease. The Pediatric Risk of Mortality III (PRISM III) scoring system, used to evaluate the severity or mortality of pediatric patients in critical condition, was investigated for children with injuries in an emergency department (ED).
METHODS
A retrospective analysis included data on 293 injured children (age<16) who visited the ED in two hospitals from March 2010 to February 2012. Physiologic and laboratory data were collected to calculate the PRISM III score and the Injury Severity Score (ISS). The correlation was analyzed between PRISM III scores, the Revised Trauma Scale (RTS), and ISS. The PRISM III score and ISS were assessed for their ability to predict mortality by comparing their receiver operating characteristic (ROC) curves.
RESULTS
The median PRISM III score was 5.0 (Interquartile Range, 5.0-9.0) and correlated with RTS and ISS (the Spearman's rho were -0.19 (p=0.001) and 0.20 (p=0.001), respectively. Five children did not survive after ED admission. The area under the ROC (AUC) was 1.00 for PRISM III (95% confidence interval [CI], 0.99-1.00), and the cutoff value was placed over 20 to predict mortality. The AUC of ISS and RTS was 0.99 (95% CI, 0.98-1.00) and 0.99 (95% CI, 0.98-1.00), respectively.
CONCLUSION
The PRISM III score excellently predicts the mortality of injured children in the ED, and can be used to sort minor pediatric trauma patients in the ED. However, the PRISM III score had no great difference or advantage compared with RTS. The development of other tools for effective prognosis is needed to efficiently predict mortality and severity in the ED.