Poster Discussion Abstracts
Aims & Objectives:
Trauma is the leading cause of death worldwide in children >1 year. Roughly, 10% of all pediatric intensive care units (PICU) admissions are related to this cause. Methods to benchmark PICU performance are critical for quality improvement. Whether trauma or critical illness specific scores are better to predict outcomes in this population remains controversial.
Retrospective review of Hospital de Niños Ricardo Gutiérrez PICU trauma registry, which collects deidentified data from all admitted trauma patients between 1995 and 2017. We tested pediatric trauma score (PTS), injury severity score (ISS), new injury severity score (NISS), revised trauma score (RTS), trauma and injury severity score (TRISS) and pediatric index of mortality 2 (PIM2) to predict risk of death in the PICU.
We included 360 children (58% males). Median age was 60 months (IQR 30–116). 73% had intracranial injuries, 26% involved extremities,19% thorax, 14% abdomen, 6% pelvis and 5% spine. 43% had injuries in one body region, 29% in two and 28% in more than two. The most frequent combination was head and extremities (16%). Main causes of injury were falls (42%), pedestrians (20%) and car occupants (7%). Overall mortality was 5.3%.
The areas under ROC curve (AUC and 95%CI) were: NISS 0.749 (0.63–0.86), ISS 0.79 (0.69–0.89), PTS 0.90 (0.84–0.96), RTS 0.912 (0.84–0.98), TRISS 0.930 (0.86–0.99) and PIM2 0.977 (0.93–1.0).
In this sample from a single institution, PIM2 had higher discrimination capability than trauma scores when using death as the outcome variable. These results need to be tested in a bigger population.