CT should replace three-view radiographs as the initial screening test in patients at high, moderate, and low risk for blunt cervical spine injury: a prospective comparison.

Background: An estimated 10,000 Americans suffer cervical spine injuries each year. More than 800,000 cervical spine radiographs (CSR) are ordered annually. The human and healthcare costs associated with these injuries are enormous especially when diagnosis is delayed. Controversy exists in the literature concerning the diagnostic accuracy of CSR, with reported sensitivity ranging from 32% to 89%. We sought to compare prospectively the sensitivity of cervical CT (CCT) to CSR in the initial diagnosis of blunt cervical spine injury for patients meeting one or more of the NEXUS criteria.

Methods: The study prospectively compared the diagnostic accuracy of CSR to CCT in consecutive patients evaluated for blunt trauma during 23 months at an urban, public teaching hospital and Level I Trauma Center. Inclusion criteria were adult patient, evaluated for blunt cervical spine injury, meeting one or more of the NEXUS criteria. All patients received both three-view CSR and CCT as part of a standard diagnostic protocol. Each CSR and CCT study was interpreted independently by a different radiology attending who was blinded to the results of the other study. Clinically significant injuries were defined as those requiring one or more of the following interventions: operative procedure, halo application, and/or rigid cervical collar.

Results: Of 1,583 consecutive patients evaluated for blunt cervical spine trauma, 78 (4.9%) patients received only CCT or CSR and were excluded from the study. Of the remaining 1,505 patients, 78 (4.9%) had evidence of a radiographic injury by CSR or CCT. Of these 78 patients with radiographic injury, 50 (3.3%) patients had clinically significant injuries. CCT detected all patients with clinically significant injuries (100% sensitive), whereas CSR detected only 18 (36% sensitive). Of the 50 patients, 15 were at high risk, 19 at moderate risk, and 16 at low risk for cervical spine injury according to previously published risk stratification. CSR detected clinically significant injury in 7 high risk (46% sensitive), 7 moderate risk (37% sensitive), and 4 low risk patients (25% sensitive).

Conclusion: Our results demonstrate the superiority of CCT compared with CSR for the detection of clinically significant cervical spine injury. The improved ability to exclude injury rapidly provides further evidence that CCT should replace CSR for the initial evaluation of blunt cervical spine injury in patients at any risk for injury.