Prognostic Model for Mortality in Pediatric Pulmonary Hypertension

Newborn in ICU
Newborn in ICU
The Pediatric Index Pulmonary Hypertension Intensive Care Mortality model was compared with the Pediatric Risk of Mortality 2 and 3 models to determine the best model for predicting mortality.

The Pediatric Index of Pulmonary Hypertension Intensive Care Mortality (PIPHM) score is a parsimonious model that outperformed the Pediatric Risk of Mortality 3 (PRISM3) and Pediatric Index of Mortality 2 (PIM2) scores in predicting risk for mortality in a multicenter cohort of patients with pediatric pulmonary hypertension (PH) admitted to pediatric intensive care units (PICUs). The results of the retrospective post hoc analysis of prospectively collected multicenter pediatric critical care data were published in Pediatric Critical Care Medicine.

A total of 143 centers submitted data to the Virtual Pediatric Systems database between January 1, 2009, and December 31, 2015. All patients were age <21 and had a diagnosis of PH. Patient age at admission was stratified into 1 of 6 categories: age 0 to 6 months, >6 to 12 months, >12 months to 2 years, >2 to 5 years; >5 to 16 years; and >16.

There were 21 demographic, diagnostic, and physiologic variables obtained within 12 hours of PICU admission that were evaluated for inclusion in the analysis. Receiver operating characteristic (ROC) curves were used to compare the PIPHM score with the PRISM3 and PIM2 scores. Overall, 14,268 admissions with a diagnosis of PH were included in the study.

The primary study outcome was PICU mortality. A total of 14 variables were chosen for the final model: age, bradycardia, systolic hypotension, tachypnea, pH, fraction of inspired oxygen, hemoglobin, creatinine, blood urea nitrogen, mechanical ventilation, nonelective admission, PICU admission due to nonsurgical heart disease, prior PICU admission, and cardiac arrest immediately prior to PICU admission.

The ROC curve for the PIPHM model (area under the curve [AUC], 0.77) performed significantly better than the ROC curves for PRISM3 (AUC, 0.71; P <.001) and for PIM2 (AUC, 0.69; P <.001).

Investigators concluded that applying the PIPHM model to patients with PH in the PICU might facilitate earlier identification of individuals at high risk for mortality, thus leading to more rapid initiation of appropriate treatment, improved counseling and prognostication for patients and their families, and a potentially useful clinical research tool.


Balkin EM, Zinter MS, Rajagopal SK, Keller RL, Fineman JR, Steurer MA. Intensive care mortality prognostic model for pediatric pulmonary hypertension [published online June 19, 2018]. Pediatr Crit Care Med. doi:10.1097/PCC.0000000000001636