Performance of the PRISM I, PIM2, PELOD-2 and PRISM IV scoring systems in western China: a multicenter prospective study.

BACKGROUND: The aim of this study was to evaluate the performance of the four scoring tools in predicting mortality in pediatric intensive care units (PICUs) in western China.

METHODS: This was a multicenter, prospective, cohort study conducted in six PICUs in western China. The performances of the scoring systems were evaluated based on both discrimination and calibration. Discrimination was assessed by calculating the area under the receiver operating characteristic curve (AUC) for each model. Calibration was measured across defined groups based on mortality risk using the Hosmer-Lemeshow goodness-of-fit test.

RESULTS: A total of 2034 patients were included in this study, of whom 127 (6.2%) died. For the entire cohort, AUCs for Pediatric Risk of Mortality Score (PRISM) I, Pediatric Index of Mortality 2 (PIM2), Pediatric Logistic Organ Dysfunction Score-2 (PELOD-2) and PRISM IV were 0.88 [95% confidence interval (CI) 0.85-0.92], 0.84 (95% CI 0.80-0.88), 0.80 (95% CI 0.75-0.85), and 0.91 (95% CI 0.88-0.94), respectively. The Hosmer-Lemeshow goodness-of-fit Chi-square value was 12.71 (P = 0.12) for PRISM I, 4.70 (P = 0.79) for PIM2, 205.98 (P < 0.001) for PELOD-2, and 7.50 (P = 0.48) for PRISM IV [degree of freedom (df) = 8]. The standardized mortality ratios obtained with the PRISM I, PIM2, PELOD-2, and PRISM IV models were 0.87 (95% CI, 0.75-1.01), 0.97 (95% CI, 0.85-1.12), 1.74 (95% CI, 1.58-1.92), and 1.05 (95% CI, 0.92-1.21), respectively.

CONCLUSIONS: PRISM IV performed best and can be used as a prediction tool in PICUs in Western China. However, PRISM IV needs to be further validated in NICUs.