Pediatric Index of Mortality 3 score as a predictor for the outcomes of critically ill patients

Abstract

Background The ability to predict mortality in critically ill patients is important for assessing patient prognosis, evaluating therapy, and assessing intensive care unit quality. The Pediatric Index of Mortality (PIM) 3 is a scoring system to predict outcomes in order to assist clinical decision-making. Objective To assess the ability of PIM 3 to predict outcomes of critically ill PICU patients.Methods This prospective cohort study included 150 children aged 1 month to 18 years who were admitted to the pediatric intensive care unit (PICU), Sanglah Hospital, Denpasar, Bali. Subjects were grouped into two based on ROC curve PIM score ≥48 and <48. The PIM 3 score was consisted of 10 variables, with a re-diagnosis classification of the PIM 2 score. Bivariate analysis was conducted to both groups to find the distribution of mortality in both groups, followed by homogenity test on variables gender, age, nutritional status, lenght of stay and mechanical ventilation. Variables which made the cut on bivariate test were included in multivariate analysis.Results The optimal PIM 3 score limit in predicting mortality was ≥48, with area under the curve (AUC) 76% (95%CI 0.69 to 0.85). Multivariate analysis revealed a 2.48 times increased risk to mortality in patients with PIM 3 score ≥48 (95%CI 1.6 to 3.7). In addition, PICU length of stay ≤7 days was a significant risk factor for mortality. Conclusion The PIM 3 has a good ability to predict the outcome of critically ill PICU patients. Critically ill patients with PIM 3 score ≥48 have a higher risk of mortality compared to those with PIM 3 < 48.