Return to Play Guidelines After Cervical Spine Injuries in American Football Athletes

A Literature-Based Review

Peter R. Swiatek, MD; Tejas S. Nandurkar, MS; Joseph C. Maroon, MD; Robert C. Cantu, MD; Henry Feuer, MD; Julian E. Bailes, MD; Wellington K. Hsu, MD


Spine. 2021;46(13):886-892. 

In This Article


The diagnosis, management, and decision-making process surrounding cervical spinal injuries in American football athletes can be complex and must incorporate many different considerations—both short- and long-term. Even though players have been treated for these conditions for over a century, only recently has some evidence-based literature emerged to help guide health care practitioners with this challenging clinical problem. We believe that large-scale retrospective cohort, case series, and even case report studies can be incorporated with expert opinion to set forth guidelines to practitioners and players who are afflicted by cervical spinal injuries.

Cervical Strains and Sprains

Many of the cervical spine injuries in NFL athletes involve the soft tissue supporting the cervical spine and may be classified as strains (muscular), or strains (ligamentous). Patients typically present with symptoms of neck pain with pain limiting range of motion and absence of distal numbness, tingling, or paresthesias. For these injuries, despite absence of a direct neurologic insult, medical evaluation is necessary to rule out cervical instability from ligamentous disruption or occult fractures. For example, although not necessarily a contraindication to RTP, discovery of a bipartite atlas has been reported on standard workup of a collegiate football player presenting with a sprain/strain.[29] Guidelines for return to play after a cervical strain/sprain include pain-free full range-of-motion, normal strength, and the dearth of subluxation or abnormal curvature on cervical radiographs.[9,30] Torg[9] recommends that any NFL athletes who fail one of the aforementioned criteria should be protected in a cervical collar for 2 weeks and then evaluated with either flexion/extension radiographs or magnetic resonance imaging (MRI) to evaluate for intervertebral disc pathology if clinically warranted.

Stingers and Burners

Stingers or burners are defined as a type of brachial plexopathy manifesting as transient sensory and/or motor deficits in the ipsilateral upper extremity resulting from compression- or traction-type trauma.[31,32] Given the typical mechanisms of action, the upper and middle trunks arising from C5–C7 are more likely to be injured than C8–T1, which may be injured with forceful abduction and external rotation of the arm.[31] For athletes with recurrent injuries, cervical foraminal stenosis is commonly an associated pathology, as the dorsal root ganglion takes up the majority of space in the neural foramen and may be more susceptible to injury during a traumatic compression-type event.[33] In our anecdotal experience, the incidence of recurrence is also increased when players are returned to full contact before complete resolution of symptoms.

For these types of injuries, initial evaluation must rule out a more significant structural or spinal cord injury. For athletes with residual symptoms after injury such as limited ROM or neck pain, RTP should be postponed. After a first stinger with significant physical findings, evaluation should include cervical spine radiographs and MRI evaluation. CT and/or EMG (for chronic symptoms) can be considered in case-by-case instances.[30] In a modified Delphi consensus study that included 48 spine surgeons, of which 12 were either NCAA or NFL team physicians, Schroeder et al[34] found that 84.5% of surgeons would allow RTP if symptoms lasted less than 5 minutes and would require additional evaluation for an episode lasting longer. With complete resolution of symptoms, guidelines for RTP after injury are the same as for a strain/sprain.

Transient Quadriplegia and Cervical Stenosis

Transient quadriplegia from a cervical cord concussion is defined as a temporary loss of motor or sensory function in the arms and legs that can last from seconds to days following excessive compression, flexion, or extension injury of the cervical spine.[3,11,35–37] The presence of pre-existing cervical stenosis greatly reduces functional reserve, which increases the incidence of these symptoms with hyperextension or axial compression forces.[38] The severity of the episode also deserves consideration in the individualized return-to-play algorithm. The more severe and longer duration of symptoms, the greater ramifications and risk of neurologic sequelae with a repeat injury.

There is a significant amount of controversy as it pertains to the diagnosis of cervical stenosis and the proper and safe return-to-play criteria[39] for athletes of all sports. Based on the available literature to date, we believe that practitioners should rely upon four categories of information for the management of these conditions: clinical symptoms, physical examination, sport and position played, and imaging characteristics.[40] In the setting of a player without any residual symptoms and normal strength and neurologic examination, individualized recommendations should rely upon the type of demands of the player's respective sport as well as what the advanced imaging (e.g., MRI) studies demonstrate. The categorization of collision (i.e., American football, rugby, hockey, etc.) versus contact sports (basketball, soccer, lacrosse, etc.) is important because of the relative risk of repetitive and traumatic hyperextension and axial compression forces to the cervical spine. In our opinion, collision sports such as American football represent a significantly higher risk of re-injury in players with stenosis than contact sports.

Based on the most recent literature, we believe that the measurement of the cervical spinal canal diameter on sagittal T2 MRI provides the most specific information in the diagnosis of this condition. Although historically cervical spinal stenosis has been defined as a space available for the cord less than 14 mm on plain radiographs,[41] more recent and pertinent data suggest that this threshold may be too conservative when it comes to the risk of a spinal cord injury in the American football population. The data from Schroeder et al[42] suggests that 10-mm sagittal canal diameter is compatible with successful NFL careers without increased risk of neurologic sequelae despite a small sample size. Furthermore, Aebli et al[15] demonstrated that sagittal canal diameter of less than 8 mm on MRI increases the risk of acute spinal cord injury after minor trauma to the cervical spine. Although there are certainly other data that utilizes other measures of spinal canal space,[43] most of the evidence-based literature points towards this range as a potential threshold for return-to-play criteria when treating these athletes. In a systematic review of clinical studies involving high-energy contact athletes, Dailey et al[3] strongly recommended that those without stenosis can return to play and weakly recommend that those with stenosis not return to play after one episode of transient cervical cord neurapraxia.

In players with cervical stenosis, absolute contraindications of RTP[4] include persistent neurological symptoms, cervical pain, loss of ROM, os odontoideum, basilar invagination, atlanto-occipital instability, Chiari malformation, functional spinal stenosis, cervical instability due to fracture or ligamentous disruption, symptoms lasting more than 36 hours, or repeat episodes of transient quadriplegia.[4,34,44] RTP recommendations in NFL athletes who demonstrate persistent cord hyperintensity despite surgical treatment is controversial.[45] While the presence of residual edema in the spinal cord may represent smaller reserve to withstand future forces on the cervical spine, Maroon et al[12,20] recommends that cord hyperintensity after surgery may not preclude RTP in NFL athletes. Tempel et al[12] also reported persistent T2 hyperintensity in three professional athletes, all of whom returned to play without re-injury. In their modified Delphi analysis, Schroeder et al[33] found strong consensus among physician regarding RTP in asymptomatic patients with no signal changes and canal diameter more than 10 mm or resolved signal changes and canal diameter more than 13 mm. Those with no history of signal changes and canal diameters less than 10 mm should be considered on case-by-case basis and those with resolved signal changes and canal diameters more than 10 mm should not RTP. No consensus was reached for patients with persistent signal changes.

In summary, RTP indications for patients with a history of congenital cervical stenosis or transient quadriplegia are controversial. Imaging studies should be considered in context with the athlete's history, mechanism of injury, physical examination, and discussion of specific RTP risks with the athletes.[13,46,47]

Cervical Disc Herniation

CDH, which is a career-threatening injury, is more common in NFL and other contact athletes relative to the general population.[1,48] The injury is associated with upper extremity radiculopathy, pain with range of motion, and/or coordination difficulties. Players without significant nerve root compression may be treated conservatively and return to play upon regaining full cervical motion and upper extremity strength.[9,49]

Both nonoperative and operative treatments of CDH have been proven to be successful, which indicates that players with this condition in the absence of neurologic deficits or myelopathy can heal without aggressive treatment. For example, in their review of 16 NFL athletes with CDH, Meredith et al[49] describes eight players who were treated nonoperatively with therapy, NSAIDs, or epidural steroid injections and successfully returned to play. Hsu[18] reviewed 99 NFL athletes with CDH and found that although operative management was associated with higher RTP compared with nonoperative management, performance scores and numbers of games started was not different between treatment groups. As with cases of CDH in non-football athletes, initial management of CDH should include a trial of nonoperative management which could include epidural injections and anti-neuropathic agents.

For patients with persistent pain, significant extremity weakness, or cervical stenosis are present, operative management has demonstrated successful RTP. While anterior cervical discectomy and fusion is the most common operation offered to these players, posterior foraminotomies can also be successful for appropriate pathology. Recent data suggest that players who undergo a foraminotomy need to be counseled as to the significant increased risk of reoperation compared with a fusion, but depending on the goals of the player and medical team, this can be a reasonable option. An athlete who has undergone a single-level anterior or posterior fusion who is asymptomatic, neurologically intact, without pain or ROM restriction, and evidence of radiographic fusion, can be cleared to return to sport;[49] however, the athlete may be at increased risk of repeated herniation above or below the fused level.[50] This recommendation is supported by Schroeder et al[33] in their Delphi analysis. However, when considering asymptomatic players without symptoms, no T2-signal changes, and solid 2-level ACDF fusions, the authors found no consensus among team physicians regarding RTP. Similarly, they found only weak consensus among experts regarding the case-by-case consideration needed for athletes undergoing corpectomy or posterior cervical surgery.[33]

Relative contraindications to RTP include athletes with an asymptomatic pseudarthrosis, 2- and 3-level fusion, and/or persistent signal intensity of T2 imaging.[9] Absolute contraindications to RTP include persistent neurologic findings, ROM limitations, pain, 4+ level anterior, or posterior fusions.[9]

Cervical Fractures

Cervical fractures can occur when the area between the vertebral body and decelerating head undergoes rapid compression, requiring as little as 150 ft. lbs of force to fracture the cervical vertebrae.[29,51,52] Certain fracture patterns involving the spinous process or unilateral lamina do not typically cause significant instability and require only hard collar immobilization for a period of 6 to 12 weeks as definitive treatment.[44] Stability on flexion-extension radiographs after healing and maintenance of cervical lordosis are required for RTP.[29,51] Unstable fractures are best treated with surgical fixation,[27,29,47] and in general, the literature supports return to play for players who are treated with a one-level subaxial anterior or posterior fusion if there is no pain, full ROM, normal muscle strength, complete fusion on imaging, and a normal neurological exam.[13]

The decision to allow an athlete to RTP is highly dependent upon surgeon and pattern of injury. For example, in their case-based survey study, France et al[39] found that 12% of spine surgeons would not allow a collegiate football player with a nonoperatively healed C5 burst fracture to return to football. In their case-based questionnaire of 113 sports and spine physicians, Morganti et al[8] found similar disagreement in management of certain subaxial cervical spine injuries treated with one-level fusion. Schroeder et al[33] found unanimous consensus among an expert panel of spine surgeons regarding RTP in athletes with solid fusion after burst, compression, or facet fracture with no instability, no T2 signal change, and solid fusion.

Despite some disagreements, there appears to be strong consensus regarding certain absolute contraindications, including any patient with a fracture with residual neurological deficits, underlying congenital stenosis, cervical laminectomy, and any procedures that require occipital, C1–C2, C2–C3, or 3+ level fusions.[29,53,54] Consideration for NFL athlete RTP should involve imaging demonstrating solid fusion in the area of injury, lack of instability on flexion and extension radiographs, maintenance of normal cervical lordosis, full ROM with no associated pain, and preinjury muscle strength.[29,44,53] Additionally, sagittal misalignment more than 11° when compared with other levels for isolated burst fractures and fractures that disrupt the posterior bony or ligamentous tension band may preclude RTP.[44,53]

Our ability to develop generalizable recommendations for RTP in American football athletes after cervical spine injury is with substantial limitation. The majority of primary sources used for developing RTP recommendations were cases studies, panel discussions, or review of general injury data that is limited by lack of clinical granularity. Although certain recommendations are strongly supported by the literature (athletes with persistent neurological symptoms should not RTP), others are less robust or even controversial. In these cases, individual surgeons must rely upon experience and patient-specific factors for making RTP decisions. Given the lack of high-level evidence directing RTP after cervical spine injury, there is a theoretical risk that this consensus-based approach to managing cervical spine injuries may become more dogma than guideline. However, we acknowledge that the purpose of this review is to generate guiding principles rather than rules for the management of cervical spine injuries in athletes and to encourage future research efforts aimed at improving the evidence for RTP.

The development of strong recommendations for RTP and cervical spine injuries in football athletes requires a strong study design, prospective collection, and outcomes data both on and off the field. Major professional sport organizations all have this mechanism in place, however, the approach and type of allowable data to be collected varies widely.[55] We believe that if medical staff of professional sports are allowed to prospectively collect data on spine injuries in athletes and assess posttreatment outcomes through patient-reported outcomes and statistical performance then the controversial questions in this population that spine physicians struggle with can be answered. While some major North American sport leagues currently allow for this, many critical ones do not.