Pediatric Index of Mortality 2 as a predictor of death risk in children admitted to pediatric intensive care units in Latin America: A prospective, multicenter study

Abstract

PurposeThe purpose of this study is to asses the performance of the Pediatric Index of Mortality 2 (PIM2) score in pediatric intensive care units (PICUs) in Latin America. Materials and methodsThis is a prospective, observational, multicenter study. We included patients aged 1 month to 16 years old admitted consecutively during 1 year to 34 PICUs in 9 Latin American countries. Discrimination and calibration tests were performed to validate the performance of PIM2 in the entire sample and in different subgroups. ResultsA total of 7391 patients were analyzed. Pediatric Index of Mortality 2 predicted 573 deaths, whereas the observed deaths were 663 (P < .001). The area under the receiver operating characteristic curve for the entire population was 0.817 (95% confidence interval, 0.808-0.825). The score showed good discrimination. Instead, calibration was inadequate. The difference between observed and predicted deaths for the entire population and across different risk intervals was statistically significant (χ2 = 121.87; df = 8; P < .001). Pediatric Index of Mortality 2 did not predict mortality correctly in different diagnostic categories (injury, postoperative, and miscellaneous), in children younger than 12 months, adolescents, and patients with chronic complex conditions. ConclusionsPediatric Index of Mortality 2 showed good discrimination, but calibration was inadequate. To use PIM2 for monitoring PICU performance in Latin America, it might be necessary to recalibrate the score locally.