Limitations of cervical radiography in the evaluation of acute cervical trauma.

We retrospectively reviewed the medical records and cervical films, computed tomographic (CT) scans, and tomographic studies of 216 consecutive patients with cervical injuries. A trauma series of roentgenograms--a cross-table lateral (CTL), a supine anteroposterior, and an open-mouth odontoid view--was performed in 100%; CT scanning was performed in 100%; and tomography was done in 9% of cases. We determined what percentage of the patients were asymptomatic initially in the emergency department; the total numbers of fractures, subluxations, and dislocations of the cervical spine in these patients; and what percentage of the cervical injuries were not detected with the plain films. Of the 216 patients in the series, 188 (87%) had known signs or symptoms of cervical injury; however, 28 (13%) of the patients were initially asymptomatic with no neurologic deficit. Of these 28, 17 were intoxicated or had mild closed head injuries; however, in 11 (5%) there was no clinical clue to their cervical injury other than a known injury mechanism. Prospectively, 67% of the fractures and 45% of the subluxations and dislocations were not detected by the CTL films, and 32% of the patients, over half of whom had unstable cervical injuries, were falsely identified as having normal spines. Prospectively, the trauma series improved the sensitivity of plain films for detecting cervical injuries but still did not detect 61% of the fractures and 36% of the subluxations and dislocations, and falsely identified 23% of the patients, half of whom had unstable cervical injuries, as having normal cervical spines.(ABSTRACT TRUNCATED AT 250 WORDS)