The spinal immobilization of trauma patients suspected of having spinal injury has been a cornerstone of prehospital care for decades. Current practices are based on the belief that a patient with an injured spinal column can deteriorate neurologically without immobilization. This concern has ballooned to include large numbers of patients with little or no chance of such an injury and caregivers with little appreciation for the complications caused by use of the cervical collar and spinal board. Somewhere between 1 million and 5 million patients receive spinal immobilization each year in the United States.1,2
The injury of concern is not the cervical spine fracture but the unstable cervical fracture with the potential for further neurological deficits.3 It is clear that among severely traumatized patients admitted to hospitals, the rate of cervical spine fractures is 2%–5% and the rate of unstable cervical fractures is 1%–2%.4–6 For patients with head injuries, the rate of cervical spine injuries increases substantially.7 Among patients with known unstable cervical spine fractures, half in one study demonstrated neurological deficits upon hospital arrival.8 Most clinicians would agree that this high-risk group would benefit from spinal immobilization, and we are truly concerned about that 0.5%–1% with unstable cervical spine fractures and intact spinal cords.
It is logical that among patients with lesser mechanisms of injury, the potential for unstable cervical spine fractures is much smaller. It is with this group that we must consider the trade-offs with the complications of cervical spine immobilization. Several studies have examined the rate of cervical fracture among generic blunt-trauma patients, whose mechanisms included MVCs, falls from standing, falls from heights and assaults. In these commonly encountered patients, the rate of cervical fracture is 1.2%–3.3%,1,9–12 and the rate of cervical spinal cord injury is 0.4%–0.7%.13,14
One of the larger studies of blunt-trauma patients with high-energy mechanisms had clear inclusion criteria and used a well-defined endpoint of clinically important cervical spine injury (essentially an unstable cervical spine fracture). In this Canadian system, patients with blunt assaults and falls from standing are generally not assessed for cervical spine injury. Among this cohort of patients with high-energy mechanisms, the rate of clinically important cervical spine injury was 0.6%.1,15 This study outlined a clear method (the Canadian C-Spine Rule) for evaluating patients with normal GCS and determining by exam those who do not have clinically important cervical spine injuries. This method has been validated in the field.15 Other criteria have also been well studied to safely discriminate a subgroup without risk of cervical spine fracture.10 Many EMS systems have incorporated these methods of clinical clearance.
Trauma expert Peter Rhee, MD, and colleagues did a retrospective study of 4,390 blunt-assault patients and noted a cervical spine fracture rate of 0.4% and cervical spinal cord injury rate of 0.14%.6 Only 4 (0.03%) of 51 patients with fractures were considered to be unstable. There has been no study that specifically examines patients who fall from standing.
The subgroup that has been most studied is those who have penetrating trauma. One recent study led by Johns Hopkins’ Elliot Haut, MD, examined the national trauma registry for such patients.16 The authors demonstrated a doubling of mortality among patients who received cervical spine immobilization. It is unclear whether this implies causality or is a proxy for more severe injury. From more than 30,000 patients with penetrating trauma, 443 (1.43%) had spine fractures, and 116 (0.38%) had unstable spine fractures. Of those with unstable spine fractures, 86 (74%) had completed spinal injuries prior to immobilization. The authors concluded that in order to potentially benefit one person with spinal immobilization, 1,032 people would have to be immobilized. But in order potentially harm/contribute to one death, just 66 would have to be.