Clinical examination and its reliability in identifying cervical spine fractures.

BACKGROUND The Eastern Association for the Surgery of Trauma (EAST) guidelines recommend that cervical spine (c-spine) radiographic evaluation is unnecessary in the awake, alert blunt trauma patient who is not intoxicated, has no distracting injuries, and demonstrates no tenderness over the c-spine or neurologic deficits. The purpose of this study was to compare the reliability of the clinical examination (CE) with that of computed tomography in identifying the presence of c-spine fractures. METHODS We prospectively evaluated 534 blunt trauma patients between February 2004 and January 2005. Positive CE was defined as complaints of neck pain, external trauma of the c-spine or neurologic deficit, tenderness or abnormalities to palpation over the cervical spine. Computed tomography was used to define the accuracy of CE. RESULTS There were 52 patients with, and 482 patients without, c-spine fractures. Forty of the 52 patients with fractures were accurately identified by CE for a sensitivity of 76.9% and a negative predictive value (NPV) of 95.7%. In the group with an initial Glasgow Coma Score of 15, 16 of 24 patients with fractures were accurately identified for a sensitivity of 66.7% and an NPV of 96.5%. In the subset of patients who by EAST guidelines would not require any radiographic evaluation, there were 17 fractures and 10 were accurately identified by clinical examination. The sensitivity in this group was 58.8% with an NPV of 96.4%. Four of the seven missed injuries required intervention. CONCLUSIONS This trial suggests that with a normal Glasgow Coma Score, CE cannot be relied upon to rule out c-spine fracture. CE is unreliable to diagnose or exclude a cervical spine fracture.