Objective To determine the frequency of clinically significant cervical spine injuries missed by multidetector computed tomography (CT) in obtunded blunt trauma victims.
Subjects and Methods We retrospectively identified 590 consecutive patients over 2 years with blunt trauma and decreased level of consciousness who had undergone emergency department cervical spine CT. CT results were compared with a composite reference standard derived from both medical record review and imaging reports, including institutionally mandated upright cervical spine radiographs. Additional outcomes studied included resource utilization and time in cervical collar.
Results A total of 367 subjects were eligible, of whom 75.5% were male, with a mean age of 40.2 years. The mean Injury Severity Score was 24.5 and the mean Glasgow Coma Score was 5.9. There was a single missed, nonoperative ligamentous cervical injury, resulting in negative predictive value of 366 of 367 (99.7%). The mean time in a cervical collar following a normal cervical spine CT interpretation was 2.7 days, with 42% (138 of 328) in collar over 48 hours. A total of 459 additional cervical spine imaging studies were performed on subjects with a normal initial CT, including 367 upright cervical radiographs, 60 supine cervical spine radiographs (16%), 3 CT (1%), and 28 MRI (8%).
Conclusion Given the extremely low yield and substantial cost of further evaluation, it is reasonable to consider clearing cervical spines of obtunded patients with concurrent head trauma with a single negative cervical CT. This should allow earlier removal of cervical collars and diminished subsequent use of imaging resources.