Podcast 63 – A Pain in the Neck – Part I

Cervical Spine Injuries in the ED

In this episode, I discuss the diagnosis of c-spine injuries. I argue that we should not send patients to imaging unless we have used the NEXUS rule and then added the Canadian C-spine Rule to the sequence. If we are imaging, it should be with a 3-view reconstructed CT scan. And even after that is done, you still need a clearance exam before removing the collar.

 The Fine Print of the NEXUS rule

You Need to Read Your Footnotes

The folks from Virginia think (J Trauma. 2011 Apr;70(4):829-31. & J Trauma2011;70(4):829-831) Nexus can’t be used, but I think if you follow my advice in the podcast, you are probably going to come as close to 100% as a rule can provide. The Canadians also showed less than 100% Sens when using NEXUS (N Engl J Med. 2003 Dec 25;349(26):2510-8), but I would make the same argument–did they really do it the same as the NEXUS study advocates? Do you do it the same? If not, you may be missing injuries.

Then add the Canadian C-Spine Rule if there is Midline Tenderness, but no other NEXUS Criteria

Click on the Image for the Whole Algorithm

Plain Films Suck!

Want the evidence, check out the Spinal Cord Injury chapter at CrashingPatient

Injuries Missed on CT scan

Cervical spine magnetic resonance imaging in alert, neurologically intact trauma patients with persistent midline tenderness and negative computed tomography results. (Ann Emerg Med. 2011 Dec;58(6):521-30)

Computed tomography alone for cervical spine clearance in the unreliable patient–are we there yet? (J Trauma. 2008;64:898 –904.)
Update: this article is hot off of the presses: JAMA Surg. 2014 Sep;149(9):934-9. doi: 10.1001/jamasurg.2014.867. Clinical relevance of magnetic resonance imaging in cervical spine clearance: a prospective study.


Check out the c-spine guidelines from the Eastern Assoc of Surgeons for Trauma (EAST)

And now to the podcast…

You finished the 'cast,
Now get CME credit

Already an EMCrit CME Subscriber?
Click Here to Get CME Credit for the Episode

Not a subcriber yet? Why the heck not?
By subscribing, you can...

  • Get CME hours
  • Support the show
  • Write it off on your taxes or get reimbursed by your department

Sign Up Today!


Subscribe Now

If you enjoyed this post, you will almost certainly enjoy our others. Subscribe to our email list to keep informed on all of the ED Critical Care goodness.

This Post was by , MD, published 4 years ago. We never spam; we hate spammers! Spammers probably work for the Joint Commission.


  1. smoggindakrak says

    I am not a true blue believer in the xray as the perfect modality. However if used judiciously it still has place, I believe, in screening and initial evaluation of neck injury. The material that you refer to in your links seems to support the fact that. I do not think there would be too many people arguing that in a patient with significantly depressed mental status or one with depressed mental status. Again the paper in your reference seems to support the same fact – there is insufficient evidence that CT scan should replace xray in patients with low risk…
    I suppose the trick is to identify the suitable patients… I tend to think that in a younger, slender built patient with longer neck and low risk an xray would be appropriate…
    I hope I am not going to face this dilemma tonight on my night shift…
    On the other note – you do a sterling job Scott, I can not tell you how many times on my shifts I used your pearls of wisdom…

    • says

      Well that is really the question–what is a low risk patient? I would argue if you fail Canadian, you are not low risk by definition.

      In a blog post later this week, I’ll give my take on the existing evidence; let me know what you think.


  2. Greg says

    I know a lot of people are saying that plain films are dead but I hear some of the NEXUS authors claim otherwise. Their argument is that although CT will show fractures not seen on plain films, those plain films are not completely normal or are inadequate. They say that in patients who fail NEXUS but are still at relatively low risk, a completely NORMAL and ADEQUATE plain film is enough.

    It seems to me this discussion of plain films v. CT really needs to include a description of which patients we are talking about. I don’t see much major trauma but I see a lot of minor trauma and I find it hard to avoid x-rays in some despite applying NESUS and sometimes the CCR. CTs in most of those patients would be overkill (literally if you believe the radiation risk).

  3. Mike Jasumback says

    One wonders what the cost of this quest for perfection will be. In lives, cancer and dollars. Why bother with exam or judgement, anyone who ever has fallen down should immediately go and have ct followed by mri. Of course then the quuestion will be, how do we diagnose the missed injuries from those two studies (I assure you, there will be some)

    I personally use nexus as written with the exception of waiting for tox studies and feel that i am within the standard of care.

    Mike jasumback

    • says

      Mike, not sure what you mean. I thought I had stressed repeatedly that the clinical judgment comes in by clearing the patients without imaging whenever possible. This means using NEXUS and I would argue Canadian as well.

      If your patient doesn’t clear by NEXUS secondary to midline tenderness, and you get plain films and they are inadequate, you then order a CT. This seems to me the place where the patient is getting over-radiated if they would have cleared by Canadian.

      I think if you use NEXUS, have perfect plain films, the tenderness is not at C1 or C7, and you have the patient rotate after negative PERFECT plain films, then you are performing good care. If the tenderness is at the cranio-cervical or cervico-thoracic interfaces, I think plain films are inadequate.

  4. Nilesh Patel says

    Excellent piece. As always, something I can take to my practice and use.
    -One comment about the use of clinical decision rules:
    -You mention that clinical decision rules should be objective and that clinical judgement often comes into play. Also, you mention that clinical judgement is the pretest probability that should bring us to use a rule. I don’t disagree.
    However, one thing I like about clinical decision rules and encourage my residents to do is use gestalt in conjunction with the rule. I actually think that using judgement with a rule is the correct way to utilize CDRs. No rule can encompass all clinical variables so gestalt must be a part of using rules, whether gestalt is built into the rule (such as with Well’s) or you are asked to use gestalt with the rule (such as PERC). Also my take is rather than use clinical judgement to get you to using the rule in the first place, use the rule to determine your pretest probability of disease with your gestalt. In reality, we would like clinical decision rules to be completely objective but they never will, nor should they be. NEXUS is pretty close, though, to being completely objective (if you compare it to other rules).


  5. Will Fleischman says

    The other week in Journal Club we reviewed Duane et al’s validation study of the Canadian rule where they applied the rule to all trauma team activations, and found the CDR was positive in all but 18 of 3201 patients screened.

    Now, it’s a little difficult to apply/extrapolate this to other places since the sparsely-written paper doesn’t describe the parameters for trauma team activation at the study center. Still, would it be reasonable to say that with any fair amount of trauma it would be a pointless trying to apply the currently existing CDR’s since they cast such a wide net in the moderate-significant trauma patient group?


  6. says

    Great podcast Scott. A personal favourite.

    This is a topic I’ve ranted about for years and it’s only in the last couple of years that people are finally giving up on the plain films when imaging is actually warranted for all the reasons you outlined.

    The clearance of the obtunded ICU trauma pt with a normal CT C-Spine is a prickly one and at my institution we are still undecided on the best way to clear. If you’re waiting for clear sensorium for a clinical clearance you’ll be waiting for weeks, and after 12 hours in a collar (even a Miami J) people aften have sore stiff necks and may not even want to give you 45 degrees. So I’m leaning towards early MRI for clearance, however the logistical difficulties are not insignificant.

    A further problem we’ve hit is some saying that an MRI after 48 hours can miss ligamentous injury as the initial oedema has subsided. This means that if one team is hoping to clinically clear, misses out on the early window for MRI then realises the patient will be unassessable to clear the neck for weeks, it becomes tricky.

    This is partly triggered by this ancient paper: Emery S, Pathria M, Wilber R, Masaryk T, Bohlman H. Magnetic resonance imaging of posttraumatic spinal ligament injury. Journal of Spinal Disorders 2: 229-233, 1989, where 2/19 ligamentous injuries were missed in MRI; the paper is flawed in many ways and 1 of the 2 had the injury missed on day 1 (so not relevant); the other was a C5/6 fracture dislocation missed at day 40 – I can’t understand how that could be missed. Do you know of any other better literature discussing the limitations of delayed MRI in this context?

    The other problem with MRI is the false positive rate, which we know is present but how high it is, is harder to quantify. This means that some head injured patients with a false positive MRI have to be woken up with a collar on when they probably have a normal neck, which doesn’t help with airway, ICP, pressure areas etc.

    I thought this paper was interesting, but as you say, probably too few patients:
    Where did you get the 1:3000 from?

    Looking forward to c-spine pudding!

  7. Daryl Pudney says

    Hello Scott,
    Thankyou for this podcast. You have helped me once again. The great points that you make about the specifics of clearing the C spine and using NEXUS prompted me to buy the Medcalc Pro App (I am in no way affiliated with anyone smart enough to develop an iphone app). I have been using the standard free medcalc app for a long time, but the Pro App allowed me to add in the info section for NEXUS and Canadian C-Spine all of the good gen from your podcast on the specifics.
    I wanted to post this for your listeners who muttered ‘damn that was good stuff but I will forget it before my next shift….where can I write this down’
    Warm regards,

  8. BF says


    Great podcast as always. The hardest patients are the ones with a minor MOI but have “persistent mid-line tenderness” or are tender to light touch but have no neuro deficits and have full ROM in their neck. In the past I would get a CT and when its negative, confirm my clinical gestalt that they were fine. What would you do, would you send all of these people home with Philly Collar or Miami J with FU at PCP for potential MRI or do you reserve that for the obtunded etc…

  9. says

    To what extent are we prepared to directly equate spinal injury (as evidenced by clinical findings, or on imaging, or whatever you will) with clinically significant and/or unstable spinal injury?

    That is, how many of the potential “misses” (say, from a less rigorous clearance process) actually matter? All of them? Less than all?

  10. Muhammad Umer Shehzad says

    This was an awesome podcast Scott.
    you ,for once, actually CLARIFIED how to examine for ‘midline’ tenderness. Pretty much everyone has got ‘neck tenderness’ However, its not like ‘everyone who is tender has got a C-spine fracture’ and you were very clear on that.
    And I fully agree that X-rays are pretty much waste of time.
    And another thing while we are at it: with a HUGE geriatric population, no radiologist is gonna ‘clear’ the C-spine X-rays because of age-related changes. At least in this patient group , CT spine is the way to go coupled with clinical assessment for who needs it.

  11. Anthony says

    Scott-great podcast- quick question. If you determine someone needs a CT scan and it turns out to be negative, you then go back and try to clinically clear them one more time per the podcast. Are there any studies that say giving narcotic pain medicine alters your exam after negative CT? Does that change anything from your standpoint, if a patient gets narcotics, in terms of clearing C-spine after a negative CT?

    • says

      Great question. AFAIK, no studies have been done. Every other area that this has been looked at in, pain meds make the exam more not less reliable. Assuming you did not shlog the patient to the point of heroin stupor. I think if you documented the pt was awake, alert, interactive, etc. you are in good shape.

  12. Chris P says

    Scott – I’m all on board with this way of approaching possible c-spine injuries. I have one question though in your algorithm. What about putting “and no dangerous mechanism” in the first box with the other “nexus” rules. My thinking here is that if you don’t, then you could potentially clear someone with a “dangerous mechanism” going straight down if they don’t have midline tenderness. This would violate Canadian rules. If that part goes in the first box, you’re golden. Love the blog, going into CC fellowship next year after I finish up my EM residency, can’t wait!


  13. Derek says


    Thank you for clarification on this subject. I definitely learned some pearls on the finer points of the NEXUS criteria.

    Even with the sensitivity listed of plain films, I still believe they have a role in the lower risk patient. We know there are some patients in a low risk MVC for example, who will just not move their neck or complain of pain no matter how we touch them. I cannot see doing a CT scan in the 20 year old who was rear-ended, there was no damage to the car, and c/o diffuse and non-specific neck pain (knowing there may be ulterior motives for the pain).

    I think that the neck x-ray should be used as a screening test just as the ECG is used as a screening test in the low risk CP (21 year old for example).

    How about this analogy:

    CT c-spine=CTA chest

    You just have to use them correctly

  14. Cherinor Sillah says

    Hi Scott ,
    Great podcast. Do you take your patients off the spine board after your primary survey or ct c-spine? If ct c spine negative and suspect ligamentous injury after reassessing and considering MRI do u still keep them on the spine board or take them off and leave c-collar on? Do u have good reference on how long a trauma patient should stay a spine board? I understanding taking off in 30 mins reduce the risk of pressure sore but can they go up to 2 hours.

    Thanks again in advance for great education thoughts.

  15. Andy Bourgeois says

    What is the evidence for waiting to clear the c-spine of an intoxicated patient after a normal CT until they are sober? My personal practice is to wait until they are sober, but a colleague does not wait, which I think is dangerous. I could not, however, find any official recommendation for waiting until sober in the literature.

    • says

      Evidence is the same as for clearing an uncon with CT alone. UK believes CT is enough; most of the US believes it is not. I think the inebriated pt has more likelihood to splint than an uncon, so it is probably even safer. I could not say your colleague is wrong.

  16. Anand Senthi says

    Scott – great podcasts series. I do have some concerns about your recommended sequential use of the Nexus and Canadian C-spine rules. They were never designed to be used that way and to do so reduces their statistical value – you end up with a test that has no proven increase in sensitivity but almost certainly very low specificity. Because these are ultra high sensitivity and only low/moderate specificity. It doesn’t matter if they fail one rule – as long as they can pass 1 rule you maintain the high sensitivity of the rule. Each rule can and should be used individually and if done properly there is no need to tweak them.
    Also why do you say if there is persistent midline tenderness you can’t clear them post CT? In the Canadian Rules you can have midline tenderness (persistent or otherwise) and still be cleared with 100% sensitivity.
    I’m puzzled why if you are not comfortable with the Nexus rule alone (since you don’t use it properly by appending the Canadian rules) why you don’t then just use the Canadian rules instead.

    • says


      I believe you are missing how the algorithm above actually functions. If you run the following imaginary patients through each of the rules individually vs. the algorithm, I think it will make more sense.

      • A normal functional patient with a low speed MVC and mild midline tenderness
      • A 70 y/o with a fall from standing height c/o neck pain, no midline tenderness
      • An inebriated patient who fell off his barstool, no neck pain. Lac on his forehead.

      Run each of these three patients through NEXUS, CCR, and then the algorithm and see what happens.

      • Anand Senthi says

        yep so, assuming all non-stated information was normal/negative
        1.Nexus – image, CCR-clear,Algo-clear
        2. Nexus – clear, CCR – image, Algo- clear
        3. Nexus – image, CCR – clear, Algo image

        Is that correct? If so then this shows the Algo imaged 1 out of 3 patients when if you had used either one of the Nexus or CCR you could have cleared all 3 patients.
        The fundamental problem is that the Algo compared with the CCR is that it adds an additional overlay of intoxication/altered mental state/distracting injury that are not required so this reduces the specificity without any evidence that it increases the sensitivity (esp given it is already essentially 100%). Patient 3 above demonstrates this problem. Additionally if you add distracting injury, intoxication or altered mental state to patient 1, you would then have to image them with the algo but could still clear them with CCR.
        That’s the point of ultra high sensitivity/low-mod specificity clinical decision rules – they are only designed to work with the set of criteria provided. Adding more criteria adds nothing to sensitivity but drops specificity.

        Also what about my 2nd question re “persistent” midline tenderness – any thoughts?

        thanks Scott


        • says

          That is correct, so you are using the algo already when you evaluate the pt and pick between the two rules. This is the purpose of the algorithm. Most people will not commit 2 rules to memory and pick and choose based on the patient; screening for the variety of factors that make 1 rule capable of clearance while the other is not.

          NEXUS is by far the predominant rule in the States. Therefore people are unnecessarily getting imaging solely b/c of midline tenderness. By branching off at that point to CCR we sig. decrease imaging. You could do the same thing in a predominantly CCR country by branching off at age, but v. few people in the States use CCR at all, much less as their first choice.

          The only difference between selective choice between the two rules and the algo is that potential CCR patients are having the additional screening of ETOH intox. Altered mental status was part of the inclusion criteria of the CCR (before the rule was even applied), so you should already be screening for that or the rule is not applicable. CCR did not include patients with major trauma resuscitation. The patient writhing in pain from their broken femur is not really appropriate for the rule. The intoxication is the only sig. difference that leads to decreased specificity vs. choosing CCR outright. Now when you talk to Dr. Stiell about this exact issue, his response is if the patient doesn’t appear too drunk to have a good interaction, then he’ll clear them. That doesn’t sound like an inebriated patient to me. I am not excluding patients who had a couple of drinks, I am excluding acutely inebriated patients.

          You could easily alter the algo to add a box putting intox after midline on the NEXUS side, but I wouldn’t.

          I am not understanding your ? regarding midline tenderness. The ability of the CCR to identify a very low risk cohort despite midline tenderness has no applicability to pts who have screened in for the need for imaging. Beyond which, midline tenderness doesn’t preclude clearance after negative ct, significant tenderness precludes it. The reasons for this are discussed in the later podcasts.

          • Anand Senthi says

            yep ok, I think we are on the same page. I didn’t realise you were coming from a position of NEXUS being predominant and thus trying to stop everyone with midline tenderness being imaged. In Australia, or at least in my state, they are both used and perhaps CCR seems to predominate. That might be in part due to the state health system endorsing them in their official imaging guidelines and not mentioning nexus at all initially (though recently they’ve amended to include both as options). See here:
            Until recently they didn’t have the NEXUS box.
            Also Australia seems always generally eager to adapt Canadian decision rules. Is there resistance in the states?

            Re persistent significant c-spine tenderness – I’ll listen to your later podcasts.

  17. SamG says

    Hey Scott,
    In the above scenario number 3 ” An inebriated patient who fell off his barstool, no neck pain. Lac on his forehead. ” I’m not totally clear as to what you are espousing/practicing with these patients:

    In the podcast, you describe how in the original NEXUS article, the patient shouldn’t even really smell like beer, or you have evidence of intoxication right there, and they don’t pass that one… , but in your above response you seem to be saying that you are not excluding patients who had a couple of drinks, but rather the patients that are acutely inebriated (as Dr. Stiell would, as long as the patient doesn’t appear too drunk to have a good interaction).

    How I personally play this one, is I use the latter approach to the problem. This appears to be adding yet another clinical judgement call to the CDR, or you can just look at it from the perspective that you are basically just completely switching over and using the CCR at that point, since you couldn’t clear with NEXUS.(which is I think essentially what your mixed algorithm does, with this small intoxication caveot).

  18. SamG says

    Totally agree from a medico-legal standpoint. It’s interesting how the whole intoxication thing isn’t really discussed with the CCR- this aspect of the assessment really does become subjective for the clinician.

    The way of thinking I don’t understand with the application of clinical decision rule out criteria is the following: “the patient has a + on the list and therefore we must now image.” Rule out criteria are designed and validated to be highly sensitive. Their specificity is inherently terrible. The specificity of the NEXUS criteria is terrible. This was never the intent. The guy who had a few beers, smells like beer, fell off his barstool and has a forehead lac but no neck pain will never be NEXUS negative for this visit. (and for the purpose of this point, let’s assume there is no such thing as CCR) All that means is that you can’t use the criteria at face-value and expect the super high sensitivity reported in the study. Which is totally fine- but the common notion that this automatically= CT Scan I think is a joke.

  19. Nathan says

    I work in a major US trauma center. Most people here use NEXUS, but I like using CCR first for the same reason you include it: the ability to clear patient with midline tenderness who at very low risk for significant injury. I kinda branch out to NEXUS in the case of the clinically intoxicated patient though; I do not routinely clear intoxicated patients with any degree of suspicion for injury.

  20. J Kim says

    New study has been published from USC trauma center. Clinical Relevance of Magnetic Resonance Imaging in Cervical Spine Clearance, JAMA surgery 2014 article argues that CT is 100% sensitive for detecting ‘clinically significant’ cervical spine injury and they argue against MRI for persistent midline tenderness or sensory deficit expressed by Pt after normal CT. Now, I am not arguing against MRI based on this single study. There are lots of limitations here but I just wanted to throw this article in.

    • says

      I think i mentioned it somewhere in the podcast, but based on the internal QA data of the trauma center, MRI will find a clinically sig. injury missed by CT 1 in 3000 patients. So the JAMA study is obviously severely underpowered for this injury. I have absolutely no problem if as either an institution or a country we say that is way too much wasted resources to detect such a rare injury. I don’t think the individual clinician should have to be exposed to that risk. I think the clever way to go is to make a policy for the trauma service in your hospital and if that policy is clearance by CT alone, I think that is well supported.


Speak Your Mind (Along with your name, job, and affiliation)