Thursday, June 18, 2009

Traumatic Cervical Spine Injury: Is CT now the preferred imaging modality?

Leafing through this month's Journal of Trauma, I noticed an article written by the good people at my old stomping ground, Cook County Hospital in Chicago (I refuse to call it Stroger Hospital), that prospectively compares the diagnostic efficacy of CT scan of the cervical spine with the standard of three plain radiographic views of the c-spine. I remember they had just started accumulating the data when I was a chief resident. Over 1500 patients were accrued. Radiographic evidence of cervical spine injury was detected in 78 of the patients, with 50 having clinically significant injuries.

Here's where it gets good. CT scan of the cervical spine detected all 50 injuries (100% sensitivity) while the plain films only identified 18/50. Even in patients with clinically significant injuries, the plain films only had a sensitivity of 46%. The paper concludes by advocating that CT of the cervical spine replace plain c-spine radiographs as the preferred initial test to exclude blunt cervical injury.

This is a classic case of where the data has finally caught up to what actually happens in real life. (Got that, you CER disciples? Science isn't as accomodating or as quick as we would like. The proof of what is already apparent in clinical practice can lag years behind.) I cover trauma at a level II center and I'll be honest; I don't spend a lot of time looking at plain films of the neck. They're always sort of suboptimal and don't consistently show all the vertebrae you need and if there's a question, you're just going to get a CT cervical spine anyway. So I go straight to the CT films. Thanks to the County trauma team, we can now all feel better about doing what has, for years, seemed obvious and intuitive.


Anonymous said...

I dont know, sounds expensive to me... What do they do in England???

Anonymous said...

Good point Buckeye. However, part of the point of the CER is to force us to actually parse out what works. You are describing a case where a study was already taking place comparing the imaging modalities in question. Not to mention the difference in cost between a 200 dollar xray and a 1000 dollar ct is pittance.

Now compare that to lumbar discectomy/ fusions/ x-stop and the ten other commonly done back surgeries. There is a paucity of indication data and CER. Only the crusaders e.g. dartmouth (SPORT trial) are even interested.

Where is the fun in proving that ortho/neurosurgery bread and butter isnt helping?

And who the hell is going to fund a study that hits everyone in the wallet? Nobody, until we put the almighty wrath of CER behind it.

Unknown said...

Arrgghhh... I'm not a doctor, nor an economic analyst, but it seems to me that a big point is being missed: CT scans show a 100% reliabilty in distinguishing "clinically signficant" c-spine injuries. C-spine injuries can change one's life drastically -- is it reasonable to take the chance of misdiagnosing and perhaps making the injury worse by treating for something else?

OK, I'm a quasi-anonymous idealist, but the thought of using a second-level diagnostic for an injury like this horrifies me.

Victor Lazaron said...

CT has been the recommended imaging modality for eval of C-spine in trauma for some time according to the EAST guidelines. It is what I use when imaging is indicated.

radinc said...

in other breaking news, MR is better than standard myleogram at detecting clinically significant disc herniations!


Anonymous said...

America's gone downhill ever since DPL's went out of style...

I know, "Whats a DPL??" last time I did one was 1997, and that was on a dog... but you know how hard it is to keep Fido still in the CT scanner...
Anyway... I'm heavily levereged in Imaging stocks so scan to your heart's content...


Joseph Sucher, MD FACS said...

This is not a CER failure. This is a problem with the failure to "get the word out" and change practice. It is also an example of the well known phenomena of real-life catching up with the multitudes of data. Notice the references that start in 1999. It is known that it takes as much as 15 years for practice patterns to change despite the evidence in the literature. Honestly, I think that until there is unfettered access to literature we will continue to see this. Read below for proof.

I am fortunate to have trained in one of the busiest blunt trauma centers in the world with the famous John "Jack" Harris, Jr. who's incredible ability to read shades of gray on film still astounds me today. Here is a paraphrasing of his comments in the journal of trauma Volume 53(2), August 2002, pp 392-393.
* Be mindful that a “negative” conventional radiographic examination in the presence of an abnormal clinical evaluation of the cervical spine requires
either computed tomography (CT) or magnetic resonance imaging (MRI).

CT provides a more complete
and accurate assessment of the posterior vertebral elements.

The swimmer’s view, always inherently limited by superimposed ribs and clavicles, should be replaced by CT as the definitive imaging examination of
the cervicothoracic junction.

Flexion-extension views should never be used to determine stability or instability of a cervical spine injury.

Even though the incidence of cervical spine injury is less than 6% in multiply injured patients, 1 in those patients in whom the cervical spinal cord
cannot be adequately clinically assessed, an anteroposterior and lateral radiograph of the cervical spine, the latter of which must include from the cervicocranium through the cervicothoracic junction, should be part of the initial patient assessment. These radiographs can be obtained in the patient assessment area with portable equipment. The necessity for cervical spine imaging at this stage of patient evaluation is to preclude causing or aggravating cervical cord injury caused by an unrecognized cervical spine injury during the secondary patient assessment.

In unconscious or multiply injured
patients, the primary attending physician must not accept a lateral radiograph in which the lower cervical vertebrae and the cervicothoracic junction are
obliterated by the density of the shoulders. Such patients require immediate CT of the cervical spine, and it is the responsibility of the primary attending physician to order that CT examination.

This practice was established at the University of Texas - Houston on or about 1999 when helical CT became available to us in the evaluation of patients with severe blunt Trauma.

I thank you for your post.


Joseph Sucher, MD FACS said...

I used to do an average of 3 DPLs every night I was on call (that was during the days when we did every other night call). We did this for everyone that required a Brain CT (for LOC or AMS) or a Chest CT (for widened mediastinum). One rational (albeit not the only) was that scanning more that one body system was time consuming and the scanner would get too hot (this was before we had the nice helical CT). So every night, I could look forward to the occasional bent cath, or the cath that ended up in the properitoneal space, or the more common problem with not getting the fluid back. Not to mention the rare incidence of creating a bowel injury (I was lucky to not have this happen to me). Additionally there is a significant amount of non-information from this test.

So I will quickly and avidly pronounce that I do not miss it for a second. More time than not it was a huge pain in the buttocks. Thank the good Lord for Helical CT and FAST.

Now ... that being said. I actually still use DPL occasionally. It is useful for the rare patient in general surgery with an unknown source of infection, that has free fluid on CT and you need to know if it is bad fluid or just physiologic. Performing a DPL in these cases is a minimally invasive way to get that information.


DKV said...

If someone is getting a CT anyway, I scan the c-spine.

A distinct, and large, population exists in which plain radiographs of the c-spine can be used to exclude significant injury. And to, perhaps more importantly, re-assure the patient and provider.