To compare the test performance of plain radiography and computed tomography (CT) in the detection of patients with cervical spine injuries following blunt traumatic events among those patients determined to require screening radiography.

We conducted a MEDLINE search for articles published from January 1995 through June 2004, manually reviewed bibliographies, and hand searched four journals. Studies were included if they contained data on the performance of both plain radiography and CT in the detection of patients with blunt cervical spine injuries. Both authors screened titles and abstracts identified by the search and seven of the 712 articles met all inclusion criteria. Both authors independently abstracted data from these seven studies and disagreements were resolved by mutual agreement.

Patient entry criteria were highly variable for each study and there were no randomized controlled trials. For identifying patients with cervical spine injury, the pooled sensitivity for cervical spine plain radiography was 52% (95% CI 47, 56%) and for CT was 98% (95% CI 96, 99%). The test for heterogeneity suggests that significant differences exist between studies in the measurement of the sensitivity for plain radiography (p = 0.07). Due to limitations of the gold standard tests in each study, a calculation of a combined specificity was not possible.

Despite the absence of a randomized controlled trial, ample evidence exists that CT significantly outperforms plain radiography as a screening test for patients at very high risk of cervical spine injury and thus CT should be the initial screening test in those patients with a significantly depressed mental status. There is insufficient evidence to suggest that cervical spine CT should replace plain radiography as the initial screening test for less injured patients who are at low risk for cervical spine injury but still require a screening radiographic examination.