AUTHOR=Zhang Lidan , Wu Yuhui , Huang Huimin , Liu Chunyi , Cheng Yucai , Xu Lingling , Tang Wen , Luo Xuequn 

TITLE=Performance of PRISM III, PELOD-2, and P-MODS Scores in Two Pediatric Intensive Care Units in China

JOURNAL=Frontiers in Pediatrics

VOLUME=Volume 9 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/pediatrics/articles/10.3389/fped.2021.626165

DOI=10.3389/fped.2021.626165

ISSN=2296-2360

ABSTRACT=Objective: The performancesof pediatric risk of mortality score Ⅲ(PRISM Ⅲ), pediatric logistic organ dysfunction score-2 (PELOD-2) and pediatric multiple organ dysfunction score (P-MODS) were unclear in Chinese patients.This study was to assess their performances in predicting mortality in critically ill pediatric patients. 
Methods: This retrospective observational study was carried out in two tertiary-care PICUs of teaching hospitals in China.1253 critically ill pediatric patients admitted to Pediatric Intensive Care Unit (PICU) of the First Affiliated Hospital Sun Yat-Sen University and children’s hospital of Shen Zhen from August 2014 to December 2019 were finally analyzed. The indexes of discrimination and calibration were applied to evaluate score performance for these three models (PRISM III, PELOD-2 and P-MODS scores). Receiver operating characteristic (ROC) curve was plotted, the efficiency of PRISM Ⅲ, PELOD-2 and P-MODS for predicting death were evaluated by the area under ROC curve (AUC). Hosmer–Lemeshow goodness-of-fit test was used to evaluate the fitting degree of each scoring system to predict the mortality and the actual mortality.  
Results: A total of 1253 pediatric patients were finally enrolled in this study (median age was 38 months, overall mortality rate 8.9%, median length of PICU stay 8 days). Compared to the survival group, PRISM Ⅲ, PELOD-2 and P-MODS scores in the non-survival group were significantly increased [PRISM Ⅲ : 18 (12, 23) vs. 11 (0, 16), PELOD-2: 8 (4, 10) vs. 4 (0, 6),  P-MODS: 5 (4, 9)vs. 3 (0, 4) , all P < 0.001]. ROC curve analysis showed that the AUCs of PRISM Ⅲ, PELOD-2 and P-MODS for predicting death of critical ill children were 0.858, 0.721 and 0.596, respectively. Furthermore, Hosmer–Lemeshow goodness-of-fit test showed that PRISM Ⅲ was the best calibration between the predicted mortality and the observed mortality (PRISM Ⅲ: χ 2 = 5.667, P = 0.368; PELOD-2: χ 2 =9.582, P = 0.276; P-MODS: χ 2 = 12.449, P = 0.015). 
Conclusions: PRISM Ⅲ and PELOD-2 can discriminate between  survivors and non-survivors well. PRISM Ⅲ had the best calibration between the predicted mortality and the observed mortality, followed by PELOD-2, while P-MODS had poor calibration.