Validation of the Pediatric Index of Mortality 2 (PIM2) in Argentina: a prospective, multicenter, observational study

Abstract

Introduction: The Pediatric Index of Mortality 2 (PIM2) is one of the most commonly used scoring systems to predict mortality in patients admitted to pediatric intensive care units (PICU) in Argentina. The objective of this study was to validate the PIM2 score in PICUs participating in the Quality of Care Program promoted by the Argentine Society of Intensive Care.

Population and methods: Multicenter, prospective, observational, cross-sectional study. All patients between 1 month and 16 years old admitted to participating PICUs between January 1st, 2009 and December 31st, 2009 were included. The discrimination and calibration of the PIM2 score were assessed in the entire population and in different subgroups (risk of mortality, age, diagnoses on admission).

Results: Two thousand, eight hundred and thirty-two patients were included. PIM2 predicted 246 deaths; however, 297 patients died (p < 0.01). The standardized mortality ratio was 1.20 (95% confidence interval [CI]: 1.01-1.43). The area under the ROC curve was 0.84 (95% CI: 0.82-0.86). Statistically significant differences were detected between the observed and the predicted mortality for the entire population and for the different risk intervals (χ2: 71.02, df: 8, p < 0.001). Statistically significant differences were also found between observed and predicted mortality in adolescent patients (37/22, p = 0.03) and in those hospitalized due to respiratory disease (105/81, p = 0.03).

Conclusions: The PIM2 score adequately discriminates survivors from non-survivors. However, it underscores the overall risk of death, especially in adolescent patients and those hospitalized due to respiratory disease. It is critical to take such differences into account when interpreting results.