Background: Glasgow Coma Scale (GCS) is a validated assessment of neurologic state. Assessment of the eye and verbal components is difficult to reliably obtain in children. We hypothesized that an abnormal Glasgow motor scale (GMS) score alone will reliably identify children with serious traumatic brain injury (TBI).

Methods: We reviewed all children with a diagnosis of TBI from 2002 to 2011 at two urban Level I pediatric trauma centers. We used logistic regression to model GCS, GMS, Glasgow verbal scale (GVS), and Glasgow eye scale (GES) for seven outcomes: need for craniotomy, intracranial pressure monitoring, admission to the intensive care unit, hospital stay of 5 days or longer, discharge to rehabilitation, dependence on caretakers at follow-up, and survival to hospital discharge. We then used three measures of fit analysis to determine which scale offered the best fit for each of the outcomes.

Results: A total of 2,341 patients (mean [SD] age, 6.9 [5.8] years; 64.7% male) with TBI and GCS data available were identified. The median GCS on presentation was 15 (interquartile range [IQR], 8-15); the median GMS on presentation was 6 (IQR, 4-6). The median GVS was 5 (IQR, 1-5), and the median GES was 4 (IQR, 2-4). GCS as a whole offered the best fit for the data in predicting need for intensive care unit admission, need for intracranial pressure monitoring, prolonged hospital length of stay, and discharge to rehabilitation but was equivalent to GMS in predicting need for craniotomy, survival to hospital discharge, or dependence on a caretaker at follow-up. Further analysis revealed that GMS was more predictive of these outcomes than GVS + GES, indicating that GMS provides the greatest contribution to the predictive ability of the GCS.

Conclusion: GMS score alone and GCS do not differ in identifying children with serious TBI. Eliminating the eye and verbal components of GCS does not adversely affect the accuracy of this tool to identify children at risk for serious TBI.

Level of evidence: Prognostic study, level III.