(1755) Performance of Day 1 Pediatric Logistic Organ Dysfunction-2 Score in a Pediatric Trauma Population

Introduction: Scoring systems for multiorgan dysfunction, such as The Pediatric Logistic Organ Dysfunction-2 (PELOD-2) score, show good discrimination and calibration in a general pediatric intensive care unit (PICU) population for predicting mortality. The performance of the admission (day 1) PELOD-2 score in predicting mortality for the subpopulation of critically ill pediatric trauma patients is unknown.

Methods: This single center retrospective study included all pediatric trauma patients admitted to a tertiary care center PICU from January 2021-March 2024. PELOD-2 scores were calculated via the electronic medical record on admission. The primary outcome was mortality during hospital admission. Performance of the PELOD-2 score was analyzed by discrimination via the area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (AUPRC). The precision recall curve has been utilized in the context of imbalanced data and is determined by the positive predictive value (precision) on the Y axis and the sensitivity (recall) on the X axis. Calibration was determined via the Hosmer–Lemeshow test.

Results: Of 589 pediatric trauma patients, the median age was 3.17 (interquartile range [IQR] 0.33 – 11.7) years and hospital mortality was 5.7% (34/589). Day 1 PELOD-2 scores were higher in non-survivors as compared to survivors (21.0 [18.0-23.8] vs 2.0 [2.0 – 7.0], p < 0.001). Day 1 PELOD-2 score had an AUROC of 0.91 (95% confidence interval, 0.85 – 0.98) and AUPRC of 0.61 (0.47-0.84). Calibration resulted in a Chi-square of 4.17 and p= 0.24.

Conclusions: Overall, day 1 PELOD-2 score has good discrimination and acceptable calibration, indicating that this initial score is highly predictive of mortality in the critically ill pediatric trauma population. These results suggest the PELOD-2 score on day 1 may be considered as a surrogate for severity of injury in this population. Validation of these results are required in large multicenter populations as well as further evaluation of the performance of the PELOD-2 score in predicting mortality throughout subsequent ICU admission days.