Abstract
To the Editor:
The letter to the editor from Dr Zafren et al raises some excellent points worthy of an expanded conversation. We welcome the opportunity to respond in the hope that this discussion will provide added clarity to our Wilderness Medical Society practice guidelines for spine immobilization in the austere environment. 1
Manual cervical traction is the standard technique for moving patients with known spine trauma in the hospital setting. That is done in an effort to keep the spine in the anatomic position and to prevent distortion of the spine that might occur otherwise. Traction is often used for stabilization and reduction of unstable spine injuries. In the monitored hospital setting, up to 150 pounds of cervical traction has been used safely in the reduction of unstable spine injuries. 2 Although it is true that traction can be dangerous in case of a grossly unstable spine injury, and therefore excessive traction should be avoided in the unmonitored setting, light to moderate traction when returning a cervical spine to the anatomic position and transferring a patient is preferred.
We are unaware of any evidence that would preclude transportation in the lateral decubitus position. Spine-injured patients are frequently placed in the lateral decubitus position when hospitalized without ill effect.
For those readers unfamiliar with guidelines, it is important to understand what they are meant to be. A systematic review is a review article that summarizes the best available evidence on a given topic. Guidelines combine the best available evidence with expert opinion in an effort to provide clinical recommendations when the evidence itself fails to adequately do so. Our manuscript aims to provide 3 “levels” of guidance: 1) to review the pertinent evidence for each topic; 2) to provide recommendations that are either implicit in the evidence or combined with expert opinion to be clinically meaningful; and 3) to provide a sample algorithm that represents an attempt to combine the above in a practical format for general field practice across a broad range of disciplines.
The algorithm is not meant to provide the definitive answer to spine care in the field. It is meant as a template open to modification by skilled and knowledgeable providers. As Dr Zafren et al point out, the evidence fails to support the use of immobilization altogether, and in that light, there is no need for the algorithm. We realize, however, that the levels of evidence currently available, although likely accurate, are not high level. That, combined with the fact that many will consider the very notion of discarding immobilization in its entirety “heresy,” makes our algorithm a reasonable transition to a new paradigm while allowing (and, it is hoped, promoting) further study to improve our understanding of spine injury, spinal protection, and the quality of evidence on which to base further recommendations.
We have already received several suggestions for changing our algorithm, none of which materially or substantially change the spirit of the existing one. No algorithm will address every clinical situation or be accepted by every medical provider, and situational use will likely differ. Adaptations to the algorithm that might improve the fit to alpine rescue may adversely affect the fit to swift water or cave rescue, and so forth. Such adaptations are encouraged at the field level when implemented by practiced and knowledgeable providers.
Regarding the National Emergency X-Radiography Utilization Study (NEXUS) criteria and Canadian C-spine Rule (CCR), these have been “validated” but not “optimized” (particularly for the dangerous rescue environment). A meta-analysis on clearance of the asymptomatic cervical spine showed a sensitivity of 98.1% and a negative predictive value of 99.8%; however, specificity was 35.4% and positive predictive value a mere 3.7%. 3 The good news is that we are not missing those with cervical spine injury; however, we are still immobilizing far too many unnecessarily (96.4%). How many of the true positives were serious injuries is not documented but, no doubt, these represent an even smaller subgroup.
The 3 most controversial (or least validated) components of the algorithm would be these: 1) discarding mechanism of action as a stand-alone qualifier; 2) requiring significant spine pain or tenderness at a level ≥7/10; and 3) requiring voluntary flexion, extension, and rotation 45 degrees in each plane. We believe the rationale for mechanism of action was discussed adequately in the manuscript. Requiring significant spine pain or tenderness at a level ≥7/10 was chosen in an attempt to separate inconsequential spine injury (muscular, fracture of spinous/transverse process, and so forth). Regarding range of motion testing, we do concede that 30 degrees of motion in all planes is more appropriate for the thoracolumbar spine, and the future algorithm will be changed accordingly. The premise is based on the well-validated use of flexion/extension cervical spine radiographs to clear a cervical spine. For years (before magnetic resonance imaging), this procedure served as the “gold standard” used to definitively clear the cervical spine, based on the knowledge that a standard lateral c-spine radiograph may appear normal in the presence of significant soft tissue injury with underlying spine instability. Flexion/extension cervical spine radiographs have been routinely performed under the direct volition of the patient under the premise that an alert patient will not cause themselves neurologic harm in the presence of an injury with the capacity to do so. To our knowledge, no adverse patient reaction has been reported after many years of use. The ability to perform the maneuver, and the extent to which range of motion should occur, should be left entirely to the alert patient; pain alone should not be used as a disqualifier to interrupt the maneuver.
Regarding the letter by Mr Nicolazzo, we believe the same responses apply, particularly regarding mechanism of injury (MOI) and NEXUS/CCR. Unfortunately, the inclusion of some subjectivity in this type of algorithm is inescapable, whether referring to MOI, tenderness, or interpretation of pain. Risk calculation, too, is highly subjective and will likely vary substantially by individual and team providers and scene. We believe that adding this to the algorithm is likely to contribute confusion. Additionally, he has added other patient assessment steps and principles not related directly to the spine and a “new” concept of the “spine as a long bone principle.” The principle is not a bad one, but does interject a new concept. Those concerns aside, we believe his algorithm represents a good example of how the template provided in our guidelines can be modified and adapted to better satisfy the needs of individual operations without changing the spirit of the principles.
We thank Dr Zafren et al and Mr Nicolazzo for their thoughtful comments and hope that the expanded conversation is helpful to our readers.