Point-Counterpoint: Cervical Spine Imaging in Children

Emily MacNeill, MD
Chair-Elect and Co-Editor, Newsletter
Carolinas HealthCare System
Charlotte, NC

Marjorie Lee White, MD, MPPM, MA
Section Secretary and Co-Editor, Newsletter
University of Alabama at Birmingham School of Medicine
Birmingham, AB

Authors’ note: This article is intended to spark thought and discussion on a controversial topic. It is not intended to change guidelines. Note that the available evidence on this topic, despite great work, is not definitive.

Point: Ordering plain films of the cervical spine
Marjorie Lee White

A 2 year-old female is brought by EMS from the scene of an accident, “packaged” and screaming. You examine the patient, quickly assessing that the patient has an intact airway, is in no respiratory distress and is warm, well-perfused with normal circulation. EMS states that they aren’t sure about the damage to the car and they don’t know if the patient was restrained. Your patient is moving all her extremities and grabbing her c-collar. She has no other obvious injuries. Her grandmother who rode in the ambulance with her after rushing to the scene insists that the reason why she is crying is because of the collar on her neck and wants you to take it off. You are wondering about imaging of the cervical spine.

You decide to obtain plain x-rays. Your thinking is that most algorithms around the country utilize plain films in the evaluation of cervical spine injury. Your plan is to review for obvious bony abnormalities and get additional imaging if any are found. Given the normal physical exam, you believe the patient is at low risk for serious cervical spine injury, but believe imaging is indicated because all information about the mechanism and whether the child was restrained is not available. Your hope is that you’ll be able to use the combination of plain x-rays and a clinical exam when the grandmother can calm your patient down. Because the incidence of cervical spine injury is low, the negative predictive value of a normal plain film is high.

You know that many adult facilities are going straight to CT scanning with 3-D reconstruction, but you remain concerned about the reported increase in thyroid cancer and you want to stick with a tiered approach. You are also pretty sure that the latest evidence indicates that MRI imaging should be the next step if ligamentous injury is the concern as it won’t be evident on CT anyway. You are also feeling pressure to be “cost-conscious” in the Affordable Care Act era.

Your colleague notices that you are ordering plain x-rays for a patient in a cervical collar and says they never do that …
Counter-point: Why are we still ordering plain films of the cervical spine?
Emily MacNeill

There are two questions we must ask ourselves when we see a child with a potential cervical spine injury: Do they need imaging, and what imaging do they need?

The first question is difficult in and of itself, especially when you deal with non-verbal or minimally verbal children. This is a point at which many specialties disagree. The 2013 guidelines for cervical spine clearance, authored by the American Academy of Neurological Surgeons/Congress of Neurological Surgeons, suggest that any non-verbal child involved in an motor vehicle collision of any kind, requires imaging of some kind to clear them from a cervical collar.1 This plan makes sense if you see the sickest of this population, but no neurosurgeons that I know are involved in clearing the cervical spine of a 2 year-old in a fender bender who comes to the ED in a collar. Do all of these children require imaging? When the incidence of c-spine injury in pediatric trauma patients (already a sicker subset) is 1-2%? I like the PECARN rule that takes mechanism into account, but keep in mind that they only had about ~500 patients in their rule derivation and most of the children were older than 3 years of age.2 When you look at the large studies of pediatric trauma patients, a large number of children <3 years old are still cleared clinically. So, what the guidelines suggest and what is happening in the real world are clearly disparate. If a non-verbal child comes in, in their car seat, moving all extremities without a mark on them, why don’t we rely on our clinical exam? I’m not talking about the kid with a GCS of 12 or a femur fracture. I’m talking about the kid who looks good. What do I do? If the kid looks great, I hold them still and take off the collar. I palpate their neck. If they don’t cry or wince, I release them to watch to see if they move it. If they move it all around, and they’re not in pain, they’re done. Otherwise, I put it back on and image the spine.

If you decide that the non-verbal child needs imaging, why are you wasting time and money on a plain film? The sensitivity isn’t great (50-95% and especially watch those confidence intervals). The most recent data from PECARN looking at children with confirmed cervical spine injuries showed a sensitivity of plain films to be 90% (95% CI, 85-94%).3 Some of you may be complaining right now about radiation. The Journal of Trauma in 2012 published an article that said that a single cervical spine CT in a pediatric patient increases the relative risk of thyroid cancer by 25% in females and 13% in males.4 Doesn’t that sound scary? OK, ok, the absolute risk increase is 5.2 to 5.87/100,000 in males and 15.2 to 19/100,000 in females. Does that still sound scarier than a potential missed cervical spine injury? Also, have you spoken to a pediatric radiologist recently? These studies are already out of date by the time they hit the journals because most of them are using scanning technology that is 10 years old (the Journal of Trauma article looked at data from 2004-2009). (If you think your home computer goes out of date quickly, think about your CT scanner. The future, per our radiology team, coming soon to a hospital near you, is a CT scanner that can obtain a cervical spine CT with the same radiation dose of 3-view plain film series.)

Still want to get plain films?

This is a conversation that is going to continue to evolve and there will never be a solution that ensures that we only image children with a high likelihood of injury with a perfect, cheap study that does not expose our pediatric patients to risk. We’re going to have to get comfortable with it.

My bias is image fewer kids and image with the right study that will give you the answer you’re looking for.

1. Rozelle CJ, et al. Management of pediatric cervical spine and spinal cord injuries. Neurosurgery. 2013; 72: 205-226.
2. Leonard JE, et al. Factors associated with cervical spine injury in children after blunt trauma. Ann Emerg Med. 2011; 58: 145-155.
3. Nigrovic LE, et al. Utility of plain radiographs in detecting traumatic injuries of the cervical spin in children. Pediatr Emer Care. 2012; 28: 1-7.
4. Muchos RD, et al. Theoretical increase of thyroid cancer induction from cervical spine multidetector computed tomography in pediatric trauma patients. J Trauma. 2013; 72: 403-409.

Leave a Reply

Your email address will not be published.