Pediatric Trauma Scoring Tools That Incorporate Neurological Status
Marina L Reppucci1, Emily Cooper2, Maxene Meier2, Jenny Stevens1, Ryan Phillips1, Shannon N Acker1, Steven L Moulton1, Denis Bensard1
1Children's Hospital Colorado, Aurora, CO;2University of Colorado School of Medicine, Aurora, CO
Background: Two novel pediatric trauma scoring tools, SIPAB+ and rSIG, which combine neurological status with pediatric adjusted shock index (SIPA), have been shown to predict mortality better than SIPA alone. We sought to compare the sensitivity and specificity of each scoring tool, to determine if one serves as a more accurate bedside triage tool.
Methods: Patients 1-18 years old from the 2014-2018 Pediatric TQIP database were included. SIPAB+ was defined as an elevated SIPA and Glasgow Coma Scale (GCS) ≤ 8 and rSIG was calculated as reverse SIPA times GCS. Abnormal rSIG cutoffs were defined as previously published. Sensitivity and specificity for SIPAB+ and rSIG cutoffs were calculated for predicting blood transfusion at 4 hours and mortality and compared using a McNemar’s test.
Results: Of the 604,931 patients who met criteria for analysis, 34.4% (207,900) were classified as SIPAB+ and 6.4% (38,602) had an abnormal rSIG. The average age of the overall cohort was 11.1 years and the average Injury Severity Score was 7.6. The blood transfusion and mortality rates for all patients were 3.4% and 0.9%, respectively. SIPAB+ was more sensitive than rSIG at predicting blood transfusion (64.6% vs 55.2%, p<0.001), but there was no difference in sensitivity for mortality (46.6% vs, 47.9%, p=0.32). rSIG had a higher specificity for blood transfusion (66.3% vs 94.8%, p<0.001) and mortality (65.7% vs 93.8%, p<0.001).
Conclusions: rSIG is a highly specific scoring tool. Though SIPAB+ has a higher sensitivity, both scoring tools have low overall sensitivities, which may result in under triage.
Back to 2021 Abstracts