Examine evidentiary support of patient immobilization;
Evaluate potential harm of using cervical collars in cases suspected spine injury;
Propose alternative packaging options for the neurologically compromised patient.
You’re cruising down the road when you get a call for a motor vehicle accident on the highway. As you flip your lights and siren on, you mentally prepare for what to do when you arrive: Assess the airway and stabilize the c-spine, then breathing and circulation. You mentally note the location of the equipment you’ll need: monitors, airway box, rigid cervical collar, dressing supplies for hemorrhaging wounds.
On scene you find two cars blocking the off-ramp. Neither has rolled over, but there’s damage to the driver’s side of one of the vehicles. The driver of this vehicle is in his mid-40s and still in his car, splinting his left arm and complaining of pain in his left leg. He still has his seat belt on. You confirm he can talk to you and then place him in a c-collar. He asks why you’re doing this—his neck doesn’t hurt, just his arm and leg. You tell him it’s the protocol for all trauma patients and continue to assist in extricating him from the vehicle.
His question stays with you, though, and you wonder why you put awake and alert trauma patients without neck pain in c-collars. You wonder if there’s any evidence to support it, or if you could be putting fewer people in collars.
We Have Questions
The packaging of a trauma patient for delivery to an emergency department is not complete without placement of a cervical collar. This has been hammered into all prehospital and ED providers for years. Immobilization of the cervical spine in trauma has been deemed so critical that the Advanced Trauma Life Support manual includes cervical spine immobilization along with airway in the ABCs of trauma. This is to prevent the unrecognized unstable spinal fracture that may cause feared complications like spinal cord damage during manipulation by providers in the trauma assessment.
The traditional trauma dogma regarding potential cervical spine injuries is that everyone should be immobilized until cleared by a hospital provider. While the idea behind immobilization of the spine patient makes intuitive sense (if the spine can no longer protect the spinal cord, external protection must be applied), there is a lack of evidence supporting its use. In fact, no randomized controlled trial has ever been conducted comparing neurologic outcomes of patients with and without c-collars.
Recent evidence regarding spinal immobilization with backboards has shown limitations to their usefulness in preventing neurologic injury, and several papers have demonstrated harm in the form of more frequent pressure ulcers, decreased pulmonary function, and greater pain for patients.1–4 Because of these findings, many EMS protocols have shifted away from routine use of backboards for anything other than extrication. While this change is progressive and shows a reasonable response to the literature, it took decades to occur. The evidence against cervical collar use is similarly mounting, yet there is little sign practice recommendations are changing.
Since the late 1990s authors have questioned the utility of the cervical collar and its effectiveness in immobilizing the spine during trauma. Despite evidence suggesting its ineffectiveness, few changes to trauma protocols have been made, and the American College of Surgeons still unequivocally recommends c-collars’ use.5 This is largely due to providers’ fear of missing—or, even worse, inducing—a spinal cord injury in a trauma patient. Spinal cord injuries can cause significant disability, paraplegia, and quadriplegia, so it’s no wonder malpractice lawsuits involving spinal cord injuries in trauma are exceedingly expensive, with compensations averaging $3 million.
But questions remain: Are cervical collars good at preventing the injuries providers fear? Is there potential harm in using cervical collars too frequently? Is there a better protocol for prehospital providers to determine when placement of c-collar is appropriate, rather than placing almost every trauma patient in one?
Evidence to Date
At a basic level cervical collars are used in the prehospital setting because we don’t know if a patient has a spinal cord injury or unstable spine fracture that could lead to one, and we assume immobilizing the cervical spine will result in improved neurologic outcomes for patients who do. Additionally, providers wish to prevent secondary spinal cord injury caused by manipulation of an unstable spine.
Data on the frequency of secondary spinal cord injuries is lacking, with estimates ranging from 3%–25%.6 However, these figures do not account for the expected neurologic worsening that may result from traumatic injuries unrelated to spinal manipulation, like hematomas, edema, or inflammation. There is no data estimating how many secondary spinal cord injuries have been prevented by cervical collars, and there is similarly sparse data on the true efficacy of cervical collars in limiting spinal movement in actual trauma patients.7 Indeed, biomechanical studies performed on both cadavers and healthy volunteers have shown contradictory results regarding the motion control provided by cervical collars: Some show limitations in spinal movement, others do not. The heterogeneity of results suggests the efficacy of cervical collars in doing what they claim to do (i.e., immobilize the spine) has not been proven.
Of course, what we truly care about is not whether cervical collars immobilize the spine, but if they prevent neurologic disability and secondary spinal cord injuries. While there has never been a randomized controlled trial comparing neurologic outcomes, a retrospective chart review published in 1998 compared the neurologic outcomes of patients with known spinal injury treated either in the United States, where all patients were placed in cervical collars prior to arrival at the hospital, or in Malaysia, where immobilization of the spine was not common in the prehospital setting.1 The authors included data from all types of spinal injuries, but when limited to only cervical spine injuries, they found no difference in the neurologic outcomes of patients receiving immobilization compared to no immobilization.
This study has several limitations, however. First, it’s more than two decades old, and there have likely been significant advances in medical practice that render some of its findings obsolete. Second, the vastly different areas of practice (U.S. vs. Malaysia) mean it is unlikely all other variables besides cervical spine immobilization were controlled for. Lastly, the sample size for patients with cervical spine injuries was relatively small, with 113 patients in the immobilized pool and 40 patients in the nonimmobilized pool.
Despite these limitations, the study strongly suggests there is no benefit from cervical spine immobilization in trauma patients with regard to neurologic disability.
A distinction should be made with regard to the conscious versus unconscious patient. The forces required to injure the spine and cause spinal cord injury are significant, and this level of force is unlikely to be worsened by smaller forces generated by manipulation of the spine in transport by a provider. Additionally, the physical response to injury is pain and muscle spasm, which naturally protects against movement that would cause further injury. In the awake and conscious patient where this reflex is not impaired, this is likely sufficient self-immobilization to prevent further injury.8
Even if there is no compelling evidence suggesting cervical collars improve neurologic outcomes in patients with potential spinal cord injuries, trauma protocols are unlikely to change without demonstration of harm by their use. Recent studies have shown this may in fact be the case. C-collars are routinely poorly fitted to the patients to whom they’re applied, and a poorly fitted c-collar can cause myriad problems, including failure to properly control neck motion, lateral displacement of the head and neck, and hyperextension of the neck.8 All these can worsen the neurologic condition of the patient with an unstable injury.
Cervical collars may raise intracranial pressure through compression of the jugular vein.9 Traumatic brain injuries are commonly associated with spine injuries, and management of a TBI requires normalization of ICP for best outcomes. Therefore, it is reasonable to conclude that the placement of cervical collars in patients with TBI raises ICP and works antagonistically to the goals of care for that patient.
Cervical collars may make endotracheal intubation more challenging. In a comparison of healthy volunteers who were intubated with and without cervical collars, there was significantly more difficulty in successfully intubating patients with collars, resulting in longer intubation times, more airway trauma, and greater likelihood of requiring backup maneuvers like a bougie.10
In clinical practice the placement of cervical collars in the prehospital setting greatly influences the management of patients once they arrive at the ED. Some providers feel paralyzed by cervical collars and are terrified to remove them, even from patients who have no indication for immobilization. Patients with cervical collars placed in the prehospital setting are significantly more likely to undergo radiographic imaging to clear their spine, which exposes them to unnecessary radiation. This is especially concerning in children, for whom every effort should be made to reduce radiation exposure.11
Lastly, the use of cervical collars in penetrating trauma is associated with higher morbidity and mortality. This may be from restriction of airway management, missed wounds beneath the collar, or delays in examination. A retrospective analysis from the National Trauma Data Bank demonstrated that penetrating-trauma patients placed in cervical collars were twice as likely to die as patients who were not immobilized.12 In that study only 0.01% of the patients had spinal cord injury. To potentially benefit one patient by immobilizing the spine, the number needed to treat was 1,032. To potentially contribute to the death of one patient by immobilizing the spine, the number needed to harm was 66. These figures strikingly show the damage cervical collars in penetrating trauma can cause.
All this data prompts the question: Is there a better way? Cultural change is slow, and cervical collars are unlikely to be phased out of protocols any time soon. Too many providers still fear litigation for missed spinal cord injuries, and until legal precedent is established by confirming that failure to place a cervical collar is not the sole determinant of a patient’s worsening neurologic status, providers will generally continue to place every trauma patient in a cervical collar. However, it seems prudent to address low-hanging fruit that can easily be changed in prehospital trauma protocols.
First, patients with penetrating trauma should never be placed in a cervical collar. The evidence is unequivocal in this regard, and we cause more harm than good by immobilizing these patients. Second, the placement of a cervical collar in the prehospital setting greatly influences the care that patient receives in the hospital, including potentially unnecessary imaging and greater discomfort from long periods wearing the collar.
In the awake and cooperative patient, a clinical decision rule should be used to determine if the patient needs to be immobilized in the prehospital setting. In the hospital providers use the NEXUS criteria and Canadian C-Spine Rule to determine if a patient warrants imaging of the cervical spine. These criteria have been extensively validated, with the Canadian C-Spine Rule (CCR) slightly outperforming the NEXUS criteria. The CCR is more sensitive (99.4% vs. 90.7%) and specific (45.1% vs. 36.8%) at identifying clinically significant cervical spine injuries.13 While the CCR was originally built for identifying patients in need of radiographic examination, it can be employed by the prehospital provider to identify patients who do not require spinal immobilization prior to transport. Figure 1 outlines the decision tree used by the CCR, adapted for use in the prehospital setting.
In time more aggressive protocols may be implemented to reduce the number of cervical collars placed in the prehospital setting. For now employing the Canadian C-Spine Rule in the awake and cooperative patient as part of EMS trauma protocols can help minimize unnecessary spinal immobilization and prevent the harm that can be caused by their placement.
1. Hauswald M, Ong G, Tandberg D, Omar Z. Out-of-hospital spinal immobilization: its effect on neurologic injury. Acad Emerg Med, 1998; 5(3): 214–9.
2. Ham W, et al. Pressure Ulcers From Spinal Immobilization in Trauma Patients: A Systematic Review. J Trauma Acute Care Surg, 2014; 76(4): 1,131–41.
3. March J, et al. Changes In Physical Examination Caused by Use of Spinal Immobilization. Prehosp Emerg Care, 2002; 6(4): 421–4.
4. Totten VY, et al. Respiratory Effects of Spinal Immobilization. Prehosp Emerg Care, 1999; 3(4): 347–52.
5. American College of Surgeons Committee on Trauma. Advanced Trauma Life Support, 7th ed. Chicago: American College of Surgeons, 2007.
6. Sundstrøm T, Asbjørnsen H, Habiba S, Sunde GA, Wester K. Prehospital use of cervical collars in trauma patients: a critical review. J Neurotrauma, 2014; 31(6): 531–40.
7. Deasy C, Cameron P. Routine application of cervical collars—what is the evidence? Injury, 2011; 42(9): 841–2.
8. Benger J, Blackham J. Why do we put cervical collars on conscious trauma patients? Scand J Trauma Resusc Emerg Med, 2009; 17: 44.
9. Mobbs RJ, Stoodley MA, Fuller J. Effect of cervical hard collar on intracranial pressure after head injury. ANZ J Surg, 2002; 72(6): 389–91.
10. Durga P et al. Effect of Rigid Cervical Collar on Tracheal Intubation Using Airtraq. Indian J Anaesth, 2014; 58(4): 416–422.
11. Leonard JC, Mao J, Jaffe DM. Potential adverse effects of spinal immobilization in children. Prehosp Emerg Care, 2012; 16(4): 513–8.
12. Haut ER, Kalish BT, Efron DT, et al. Spine immobilization in penetrating trauma: more harm than good? J Trauma, 2010; 68(1): 115–20.
13. Stiell IG, Clement CM, Mcknight RD, et al. The Canadian C-spine rule versus the NEXUS low-risk criteria in patients with trauma. N Engl J Med, 2003; 349(26): 2,510–8.
Evan Conner, MD, is an emergency medicine resident at Washington University in St. Louis.
Hawnwan Philip Moy, MD, is an assistant medical director of the St. Louis City Fire Department and emergency medicine clinical instructor and core faculty in the EMS Section of the Division of Emergency Medicine at Washington University in St. Louis. He completed his emergency medicine residency at Barnes Jewish Hospital/Washington University in St. Louis and his EMS fellowship at the University of North Carolina in Chapel Hill.