The objective of this study was to determine the predictive value of the Glasgow Coma Scale (GCS) and the Glasgow Motor Component (GMC) for overall mortality, death on arrival, and major injury and the relationship between GCS and length of stay (LOS) in the emergency department (ED) and hospital.
Records from the American College of Surgeons National Trauma Data Base from 2007 to 2009 were extracted. Patients 0 to 18 years old transported from a trauma scene with complete initial scene data were included. Statistical analysis, including construction of receiver-operator curves, determined the correlation between GCS, GMC, and the clinical outcomes of interest.
There were 104,035 records with complete data for analysis, including 3946 deaths. Mean patient age was 12.6 (SD, 5.5) years. Glasgow Coma Scale was predictive of overall mortality, with area under the receiver-operator curve (AUC) of 0.946 (95% confidence interval [CI], 0.941–0.951); death on arrival, with AUC of 0.958 (95% CI, 0.953–0.963); and risk of major injury, with AUC of 0.720 (0.715–0.724). Lower GCS scores were associated with shorter ED LOS and longer hospital stays (P <0.001, analysis of variance) except GCS 3, associated with shorter hospitalizations. For predicting overall mortality, the AUC for GMC was 0.940 (95% CI, 0.935–0.945), and for predicting major injury, the AUC was 0.681 (95% CI, 0.677–0.686).
For pediatric trauma victims, the GCS is predictive of mortality and injury outcomes, as well as both ED and hospital LOS, and has excellent prognostic accuracy. The GMC has predictive value for injury and mortality that is nearly equivalent to the full GCS.
From the *Department of Pediatrics, Yale-New Haven Hospital, Yale School of Medicine, New Haven, CT; and †Department of Pediatrics, Kosair Children’s Hospital, University of Louisville, Louisville, KY.
Disclosure: The authors declare no conflict of interest.
Reprints: Mark X. Cicero, MD, Department of Pediatrics, Yale-New Haven Hospital, Yale School of Medicine, New Haven, CT, 100 York St, Ste 1F, New Haven, CT 06511 (e-mail: email@example.com).