Evaluation of the C-spine in Trauma

Journal Club Synopsis-Evaluation of the Cervical Spine in Trauma-September 2014

Many thanks to Trushar and Rupal Naik for hosting, and to Adam, Stephen, Theresa, Natalie K, John P and John M for their insightful analyses.

We see a lot of patients with neck pain after MVCs.  Many can be safely cleared clinically using NEXUS or the Canadian C-spine decision instruments.  When imaging is performed, it’s usually CT rather than Xrays.  We’ve all seen trauma patients hang out for hours/days in c-collars after CT, awaiting either re-assessment when alert, or an MRI.  C-collars cause complications, and this can be a resource utilization challenge.  So, the question on the table...is CT enough?

1. Resnick S, et al:  Clinical Relevance of Magnetic Resonance Imaging in Cervical Spine Clearance A Prospective Study. JAMA Surg 2014; doi:10.1001/jamasurg.2014.867 Published online July 30, 2014.

Just saying, this is the best trial of the three.  It’s a prospective observational study from USC/LAC of 830 awake/alert adults after blunt trauma who had midline tenderness and/or focal neuro deficits (so can’t be cleared with NEXUS) and a negative C-spine CT.  Kenji Inaba of EM:RAP fame is second author.  Primary outcome was clinically significant C-spine (CS) injuries, defined as requiring surgery or a halo.  Overall, 164 (20%) CS injuries were diagnosed, and 23 (3%) were clinically significant.  All clinically significant CS injuries were detected by CT.  CT missed 9% of injuries, i.e., 15 of 681 patients (2.2%) had normal CT but new finding on MRI.  However, none of these injuries required surgery, halo, or change in management based on MRI.  For detecting any CS injury, CT sensitivity was 91% and specificity was 100%.  For clinically significant CS injury, CT sensitivity and specificity were both 100%.  Patients with distracting injuries were excluded.

  MRIs were ordered at discretion of attending surgeon or neurosurgeon.  Follow-up was only until day of discharge.  They used advanced 64-slice CT scanners.  Of the 15 patients with normal CT/abnormal MRI, only 6 had neurologic symptoms, and these were all sensory deficits.  Clinically significant injury  in patients with motor deficits were all identified on CT, leading authors to state that MRI “may be indicated” in patients with motor deficit and normal CT scan.

2. Russin JJ, et. al.: Computed tomography for clearance of cervical spine injury in the unevaluable patient. World Neurosurg. 2013;80(3-4):405-413.  

Ok, what about the “unevaluable” blunt trauma patient, defined as GCS < 15, distracting injuries, and/or altered/intoxicated?  This is a higher risk population for cervical spine injury, and there’s a lot of low quality literature out there about this group of patients, with conflicting recommendations.  This was an analysis of 13 articles including patients with negative C-spine CT who also underwent MRI, however only 9 studies included patient management data.  In these 9 studies, 115 of 855 patients with abnormal MRI + negative C-spine CT had change in management.  Three of 855 (0.35%) patients required surgical stabilization (NNT=285), and 57/855 (7%) received “extended time in C-collar.”  

Limitations:  Why were patients chosen to receive an MRI...who knows?  All three of the patients requiring surgical stabilization were from 2 studies by the same author....hmmm.   What’s the meaning/significance of “extended time in C-collar?”  “Missed injuries” had variable treatment depending on the study. Several studies used older generation CT scanners, several were retrospective. There was limited followup/clinical information available for many of these studies. This is a descriptive review, and does not include the statistical methodology of a meta-analysis.  It’s also always imperative to remember the garbage in/garbage out philosophy...if the included studies
(even in a meta-analysis) are of poor quality, then conclusions are dubious. Authors conclude that
CT alone is inadequate in this patient population.  The room was not impressed.


3. Panczykowski DM, et. al.: Comparative effectiveness of using computed tomography alone to exclude cervical spine injuries in obtunded or intubated patients: meta-analysis of 14,327 patients with blunt trauma. J Neurosurg. 2011;115(3):541-549.  

Another attempt to look at the literature on the “unevaluable patient”, this time specifically blunt trauma patients who were obtunded or intubated.  This was a meta-analysis of 17 cohort trials with 14,000+ patients, 12,700+ with negative CT.  Compared to MRI, CT missed 7 unstable injuries (0.05%).  One patient received a halo, 3 received surgical stabilization, 3 were treated with collars (NNT=3000), sensitivity/specificity of CT both >99.9% (95% CI 0.99-1.00).  Much better methodology than Russin, and included only studies with modern CT slice thickness and those reporting patient outcomes.  

Limitations:  Included 10 retrospective trials, and missing significant amount of descriptive demographic data.  There were a variety of reasons listed for ordering MRI, and 1,573 patients didn’t receive MRI.


Final thoughts...in the room, no love for Flex/Ex films.  Reasonable agreement based on discussion, JC articles and background articles that even after negative CT, MRI is important in patients with motor symptoms, and should be considered in the elderly, especially as advancing age is associated with cervical spondylosis.  Other high risk groups include patients with rheumatoid arthritis/ankylosing spondylitis, as these arthritides predispose to cervical spine fracture, and also to ligamentous injury such as atlantoaxial instability. Mechanism and degree of point tenderness will always play a role in imaging decisions.  Evidence for the management of the “unevaluable patient” is still inconclusive, although the third article (Panczykowski) is reassuring.