Received February 23, 1990; accepted after revision May 3, 1990.

1 Department of Radiology, Box 3808, Duke University Medical Center, Durham, NC 27710. Address reprint requests to R. M. vandemark.

AJR 155:465-472, September 1990 0361 -803X/90/1 553-0465 American Roentgen Ray Society

465

Perspective

Radiology of the Cervical Spine in Trauma Patients:Practice Pitfalls and Recommendations for ImprovingEfficiency and CommunicationRobert M. Vandernark1

Trauma constitutes a significant portion of emergency depart-ment practice. Such patients often have suspected cervical spineinjury necessitating cervical spine radiographs. The importanceof detecting cervical spine injury is obvious because failure to doso can lead to tragic consequences for patient and physicianalike. Although most cervical spine radiographs are justified,poorly indicated and unnecessary examinations are unfortunatelycommonplace. Indiscriminate ordering of cervical spine exami-nations can easily exceed radiology resources assigned to theemergency department. Rational ordering practices are thereforeessential for efficient patient management. A risk-tailored ap-proach to performing these examinations, which can improveefficiency, is presented. Once obtained, cervical spine radio-graphs are presumed to detect injury with consistently highsensitivity. Prevailing conditions of emergency department prac-tice that may lower the sensitivity of cervical spine radiographsare reviewed. Overreliance on the initial radiologic examinationmay lead to inappropriate haste in the evaluation of suspectedcervical region injury as exemplified by the commonly voicedmandate to clear the cervical spine of injury. This approach isdiscouraged in patients with significant trauma in favor of acareful, progressive evaluation of the potentially injured cervicalspine. Periodic review of these complex issues and close co-operation between clinical services are emphasized.

Evaluation of the patient with suspected cervical spineinjury remains a constant challenge for radiologists and emer-

gency physicians alike. No single component of the traumaevaluation so consistently produces frustration, anxiety, andmiscommunication. These problems are accentuated in cer-vical spine trauma because of the common association withpermanent, severely disabling spinal cord injuries. Implicit inthe trauma resuscitation is the need to protect the spinal cordfrom iatrogenic injury.

In addressing these concerns, referring physicians relyheavily on cervical spine radiographs (CSR). This approachcan lead to indiscriminate ordering practices, especially whenradiographs are requested by protocol rather than with regardto the actual risk of injury in a particular patient. Unselectiveordering practices can severely strain radiology resourcesallocated to the emergency department. Even when theseexaminations are successfully performed, problems of com-munication between radiologist and referring physician oftenemerge because of discordant expectations of the role ofCSR in trauma.

The Flood of Cervical Spine Requests: Cause and Effect

In the past decade, the number of patients with suspectedcervical spine injury has increased steadily, paralleled by asurge in demand for CSR. Factors fueling these trends includethe increased frequency of traumatic injuries, improved phy-

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

466 VANDEMARK AJR:i55, September i990

sician education in the diagnosis of spine trauma, and theever-present fear of misdiagnosis or mismanagement of theseinjuries. These trends only partly explain the increased num-ber of cervical spine examinations requested today.

Perhaps the primary source of these examinations is thewidespread use of spinal immobilization protocols by emer-gency medical service personnel at the accident scene. With-out question, these prehospital care protocols have preventedor minimized many serious spinal injuries. Nonetheless, as aconsequence of this universal practice, the presence of acervical spine collar carries little, if any, predictive value forthe presence of a spinal injury. The index of suspicion forcervical spine injury has become so high that virtually allpatients with blunt trauma or injuries above shoulder level willundergo CSR. From the treating physicians perspective,liberal indications for CSR seem justified by the costly medicaland legal consequences of a missed spinal injury.

One problem faced by the referring physician is that themajority of patients immobilized in a cervical collar, whetherinjured or not, volunteer some degree of neck discomfort,pain, or tenderness on physical examination. Once suchsymptoms are elicited, it becomes difficult to omit the radio-graphs because posttraumatic neck pain is the most widelyaccepted indication for CSR. Paradoxically, prolonged im-mobilization in a collar sometimes produces symptoms ofneck pain that were not present when the collar was applied.The examining physician in this situation may be confusedabout the need for CSR, given the concomitant history oftrauma and the patients complaint of neck discomfort. Logi-cally, one would assume that, in the neurologically normalpatient, a careful physical examination of the neck mighteliminate the need for radiographs. This incorrect assumptionis an extrapolation from the observation that many extremityfractures can be anticipated by careful physical examinationbefore obtaining radiographs. Unfortunately, physical exami-nation of the neck differs from that of the extremities in severalimportant ways. First, the vertebral column is a well-protectedstructure, deep within the neck. Except for the neurologicallyinsignificant spinous processes, it is not readily accessible topalpation. Second, unlike physical examination of a limb,motion of the neck during the examination is prohibited.Despite these drawbacks, physical examination of the neckshould still be performed but has limited value in determiningthe need for GSA. These limitations were highlighted in arecent study that demonstrated that the ability of some refer-ring physicians to predict spinal injury on the basis of historyor physical examination is poor, with only half of cervicalinjuries identified prospectively [1]. Acknowledging theseproblems, many physicians omit this important part of thetrauma evaluation and rely almost exclusively on the radio-logic examination instead.

Another reality of emergency department practice, whichoften goes unrecognized, is the pressure from patients them-selves to have radiographs taken. The patient may be merelyseeking reassurance that their X-rays are normal. Some-times, however, the treating physician is coerced into orderingthese radiographs by patients whose litigious nature surfacesduring the emergency department evaluation.

Whether radiographs are indicated or not, the likelihood ofpatients eventually undergoing CSR is high, as noted byMcNamara et al. [2] in a study of posttraumatic neck pain.These authors note that of patients who did not undergo CSRat their initial emergency department visit, 52% subsequentlyhad radiographs taken within 6 months of their injury. Inaddition, 66% of patients were pursuing litigation. Undoubt-edly, some trauma patients legitimately suffer from chronicposttraumatic cervical pain, accounting for a portion of thedelayed examinations. Recognizing these clinical and behav-ioral factors, some experienced referring physicians feel thata baseline cervical spine series is justified.

As if there were not enough reasons for the clinician torequest a cervical spine series, there is also the myth of thepainless or asymptomatic cervical spine fracture, fosteredin the literature by several disputed case reports [3-5]. Thethoroughness or validity of the clinical examination has beenchallenged in most of these cases [6]. Typically, the patientspain symptoms are not elicited or are ignored in the haste ofresuscitation, blunted by altered mental status or masked byother, more painful, injuries [7]. Significant cervical spineinjuries can be occult, escaping initial clinical and radiologicscreening, but a truly painless injury in a mentally unimpairedpatient has not been reported [8]. Despite doubts cast onthese reports, the myth of the painless cervical spine injurystill persists in some quarters as a justification for CSR.

On the basis of these clinical and social factors, frequentuse of CSR is unavoidable. A liberal approach to CSR shouldnot, however, be construed as tolerance for indiscriminateuse of the examination. CSR should be recognized as anecessary but low-yield examination, with less than 2% ofcervical spine series in trauma patients revealing a potentiallysignificant finding or definite abnormality. The impact of un-necessary or overly exhaustive cervical spine examinationson patient throughput in the emergency department shouldnot be trivialized, and indiscriminate ordering practices mustbe prevented. Unfortunately, it is not unusual for some phy-sicians to adopt a protocol approach to trauma in whichevery patient, regardless of symptoms or mechanism of injury,undergoes CSR [9].

Another common source of unnecessary CSR occurs wheninexperienced clinicians misconstrue the relationship betweenhead injury and cervical spine injury. Although it is well doc-umented that patients with significant head injuries can haveassociated cervical spine injuries, this is not true in patientswith normal mental status and minor head injury. In themajority of these patients, CSR, at least on clinical grounds,is not indicated. Experienced clinicians recognize this as ano-risk patient and do not order radiographs.

Selective use of CSR in patients with head injuries issupported in the literature [1 0]. Even among trauma patientswhose clinical history suggests a genuine possibility of cervi-cal spine injury, it has been shown that CSR is usuallyunnecessary when an alert, mentally unimpaired, neurologi-cally intact patient denies neck pain [8, 1 1 , 1 2]. Normal mentalstatus implies absence of drug, alcohol, intoxicants, or psy-chiatric disturbance. Clearly, certain trauma patients do notrequire CSR, and protocol ordering should be discouraged.

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

AJR:i55, September 1990 CERVICAL SPINE RADIOLOGY 467

Occasionally, such practices are detrimental to care of thepatient when the treating physicians attention becomes fixedon the cervical spine evaluation rather than on life-threateninginjuries, which are statistically more common in these pa-tients.

Managing the Demand for CSR: A Risk-TailoredApproach

Our approach to these complicated issues is based on anopen dialogue with our referring physicians. Requesting phy-sicians are motivated to stratify each cervical spine traumacase by the perceived risk of significant cervical spine injury.We consider patients to fall into four possible risk categoriesfor cervical spine injury: category 1 , no risk; category 2, lowrisk; category 3, medium risk; category 4, high risk.

General guidelines for assessment of a patients risk andthe characteristics of the high-risk patient are noted in Tables1 and 2. Appropriately trained and experienced physiciansusually have little difficulty in estimating a patients risk. Inaddition, knowledge of the physicians assessment of injuryrisk is very helpful to the radiologist reviewing the case andoften alters his recommendation for additional or follow-upstudies. Assessment of risk can facilitate CSR because theclinician is prospectively challenged to justify the allocation ofradiology resources required for the examination. Patientsclassified into low-, medium-, and high-risk groups undergo

TABLE 1: Clinical Risk Categories in Cervical Trauma

Clinical Risk DescriptionNone Collar placed owing to protocol; no historical

or physical findings to suggest a neck in-jury

Low Mechanism/history of injury unlikely to haveexceeded physiologic range of motion

Medium Mechanism/history of injury sufficient tohave exceeded physiologic range of mo-tion

High Mechanism/history of injury very likely tohave exceeded physiologic range of mo-tion

TABLE 2: Clinical and Historical Characteristics of the High-Risk Patient

1 . High-velocity blunt trauma2. Significant motor vehicle accident3. Direct cervical region injury4. Altered mental status at the time of trauma and/or during emer-

gency department evaluation (includes alcohol, drugs, intoxi-cants, loss of consciousness, and mental illness)

5. FaIls/diving injuries6. Significant head/facial injury7. Abnormal neurologic examination8. Prominent neck pain or tenderness9. Thoracic or lumbar spine fracture

io. Rigid spine (ankylosing spondylitis, etc.)

TABLE 3: Risk-Tailored Cervical Spine Radiologic Examinationin Trauma Patients

Category Chnical Radiologic Examination

1 None Radiographs unneces-sary

-

2 Low Erect three-view senesaLateralAnteropostenor

Odontoid

10 mm

3 Medium Five-view supine-erectseries

Supine lateral ( swim-mers)

Supine odontoidSupine anteroposterior

Radiologist/clinician mustapprove

Erect obliques

20 mm

4 High Five-view supine-only so-ries

Lateral ( swimmers)AnteroposteriorOdontoidSupine trauma

obliques

30 mm

a The three-view examination can be performed with the patient supine when

necessary.

different radiologic examinations requiring different time com-mitments.

As shown in Table 3, patients categorized in the low-riskgroup undergo a brief, erect, three-view examination, whichtakes approximately 1 0 mm. Occasionally, the examination isperformed with the patient supine when, for example, anotherinjury prevents radiographs from being obtained with thepatient sitting or standing. Placement of a cervical spine collarin the emergency department is discouraged in this groupuntil the radiographs have been completed and reviewed.This improves the quality of radiographs, which are unaffectedby collar-engendered artifacts. Even if a patient with mediumor high risk were inadvertently classified as a low-risk patient,we consider a high-quality, three-view cervical spine seriessufficient for a primary cervical spine evaluation.

Patients with medium risk for cervical spine injury aregenerally classified by exclusion from either the low-risk orhigh-risk categories. Patients in the medium-risk categoryhave a trauma history that suggests sufficient force to pro-duce a spine injury but high-risk features are absent. Theadvantage of placement in the medium-risk group is thatthese examinations, which begin with the patient supine, canbe terminated with the patient sitting or erect. This decreasesexamination time and improves the diagnostic quality of theoblique radiographs, which, when obtained with the patientsupine, show inherent geometric distortion. Again, even if ahigh-risk patient were inadvertently categorized into the me-dium-risk group, a review of the supine lateral radiographs bya physician, as noted in Table 3, should allow detection ofsignificant cervical spine injury before any erect views areobtained.

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

468 VANDEMARK AJR:i55, September 1990

Common Sources of Miscommunication

The interpretation of the radiographs can lead to severalareas of miscommunication between the radiologist and re-femng physician. These problems involve the interrelatedissues of accuracy, speed of diagnosis, and structural stabilityof the traumatized cervical spine.

Sensitivity of GSA

When CSRs are considered, two misconceptions areshared by many radiologists and referring physicians con-cerning the presumed high sensitivity of CSR for detection ofinjury. Both groups tend to forget the two major componentsof the examination that contribute to its capacity to detectinjury: quality of radiographs and experience of the interpreter.

With a complete trauma series of well-positioned and opti-mally exposed radiographs of the cervical spine, recognitionof significant injuries is quite high, with a sensitivity of 95%under ideal conditions. Unfortunately, a complete, high-qualityexamination is frequently impractical or impossible to obtainbecause of time constraints imposed by other trauma priori-ties, lack of cooperation from patients, and physical limitationsimposed by supine positioning of the patients. Significantly,those patients at highest risk of injury are most likely to havea technically compromised cervical spine examination. lnad-equate radiographs are a common source of misdiagnosis incervical spine trauma, especially with regard to visualizationof the cervicothoracic and atlantoaxial regions [1 3]. Singlelateral portable radiographs of the spine have been shown tobe inadequate in up to 25% of examinations [1 4]. Clearlythen, compromised examinations, particularly in the high-riskgroups, should be anticipated. Increasingly, CT has been usedto overcome these problems and has improved fracture de-tection. Acheson et al. [15] note that approximately 53% ofcervical spine fractures were missed on the initial traumaradiographs, but subsequently were detected on CT scans.Despite this observation, the authors support a near-perfectsensitivity for CSR in the detection of cervical spine injuries,because all patients in their series with fractures proved withCT had initially abnormal or suspicious trauma radiographs.Nevertheless, the retrospective nature of the study, a 63%false-positive rate, and a high degree of interpretive experi-ence are factors that generally do not prevail in the acutetrauma setting. Missed cervical spine injury rates of 10-20%are probably more representative of the typical trauma setting,in which reliance on incomplete or technically compromisedexaminations is not uncommon. The propensity for certaincervical spine fractures to escape detection on radiographswas recently reported by Clark et al. [1 6], who found delayedradiologic diagnosis in 23% of odontoid, 1 6% of teardrop,1 4% of facet, and 1 0% of hangmans fractures. Decreasedsensitivity of the initial trauma series should be expectedwhen radiographs are suboptimal, with provisions made forrepeated radiographs or additional studies when the patientscondition permits.

Precisely how many films constitute a complete examina-tion varies widely. The three-view series is the minimumcomplete examination. Five views are desirable in medium-

and high-risk patients, although the value of supine obliqueradiographs has recently been questioned [1 7]. As it is oftendifficult to obtain high-quality films, and the experience of theinterpreter may be limited, the number of radiographs shouldnot be overly restricted.

The second component of sensitivity is the quality of theinterpretation of radiographs. Many articles discussing sen-sitivity of CSR fail to clarify who has interpreted the cervicalspine radiographs. Studies advocating a high sensitivity forthis examination have used attending radiologists or experi-enced traumatologists. This high degree of interpretive ex-perience rarely exists in the majority of hospitals, especiallyduring peaks in trauma caseloads. This may explain theexperience of Reid et al. [1 8], who noted that 8% of cervicalspine injuries were missed on initial review of adequate radio-graphs. Notably, one third of injuries missed because ofinadequate clinical evaluation or misinterpretation of radio-graphs occurred in a tertiary care hospital. The majority ofreviewing physicians probably do not have sufficient trainingor experience with these injuries to attain a desirable 95%sensitivity. Furthermore, the emergency department is a tra-ditional training ground for most physicians, including radiol-ogists, emergency physicians, and general surgeons. Undersuch circumstances, caution is advised in attributing highsensitivity to examinations interpreted at this level of experi-ence. Less than ideal circumstances may explain Bohlmansobservation that fully one third of 300 cervical spine injuriesescaped initial clinical and/or radiologic detection [1 9]. Closesupervision of house officers in training, regardless of spe-cialty, and review of all trauma cases by a radiologist isessential in maximizing sensitivity.

Clearing the C-SpineProbably the most frustrating aspect of cervical spine

trauma for both the treating physician and the radiologist isthe differing expectations these physicians may have con-cerning the role of the radiographs in the management of thepatient.

Perhaps the most detrimental concept to have evolved isthe imperative to radiologically clear the c-spine. The radiol-ogist is often uncomfortable responding to this directive witha yes or no response. This concern is justified, becausemost physicians lack a firm understanding of the term. In itsmost common (but incorrect) usage, clearing the spine istaken to mean that if no abnormality is identified on theradiographs then the patient is free of cervical spine injury. Inthis syntax, cleared c-spine means a normal c-spine. Thisdefinition is seriously flawed by its presumption of 100%sensitivity for CSR in the detection of injury. Reliance on aperfect sensitivity for any screening examination is unwise. Inaddition, even if we assume an optimistic 95% sensitivity,most physicians would not be comfortable with a missedinjury (false-negative) rate of 5%. Therefore, CSR cannot beconsidered the gold standard for cervical spine injury.

Physicians who function under the illusion that a normalcervical spine series has eliminated cervical region injury arelikely to miss significant injuries. As shown in Figure 1 , pa-tients can have serious cervical region injuries without obvious

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

Fig. 1.-Spectrum of cervical spine injuries: any cervical injury can beascribed to one or more injury subsets a, b, or c. Cross-hatched zone isexemplified by a fracture-dislocation with quadriplegia. Radiographs aresensitive to detection of most injuries in subset b, many in a, but few in c(e.g., traumatic disk hemiation or anterior and central spinal cord syn-dromes).

Fig. 2.-Case 1. 45-year-old man with neck pain after a motor vehicle accident.A, Initial lateral radiograph, in a neurologically normal trauma patient, shows prevertebral soft-tissue swelling only.B, A second exposure with improved positioning of patient shows unusual lucency through craniocervical junction but no fracture.C, Lateral tomogram verified an unstable ligamentous injury; vertical atlantooccipital dislocation.

AJR:i55, September 1990 CERVICAL SPINE RADIOLOGY 469

evidence of fracture or subluxation on the cervical spineradiographs. In addition to soft-tissue injuries, frequently in-apparent on the radiographs, some fractures and destabilizingligamentous injuries can elude detection on a routine traumasurvey. Delayed detection of spinal injuries is not uncommonand should not be surprising [1 9]. Cervical spine examinationsthat are precipitously cleared may later turn up significantinjuries that were either overlooked or incompletely visualizedin the initial trauma survey. Figures 2 and 3 show two unstableinjuriesthat might have been erroneously cleared on the basisof a single lateralradiograph.

A minority of experienced trauma physicians use theexpression clear the c-spine in a more limited and cautiouscontext. In this definition,the cervical spine radiographs are

evaluated primarily for the purpose of detecting unstablecervical spine injury. Careful examination of one lateral cervi-cal spine radiograph alone can eliminate the presence ofunstable cervical spine injuries in a high percentage of cases.It should be emphasized that in this context, a spine could beclassified as stable but any number of soft-tissue or boneinjuries could be present.

Rapid evaluation of the trauma radiographs for obvioussigns of unstable injury is important to the trauma physicianfor patient management. Should neck motion be necessitatedby endotracheal intubation, no neurologic sequelae would beexpected in the patient with a stable cervical spine. In theinitial trauma resuscitation, this limited goal of CSR is usuallysufficient for patient management. Prudence dictates thateven when a spine is thought to be stable on the basis of theinitial radiologic evaluation, unnecessary manipulation of theneck should be avoided. The exclusion of injuries that wouldbe considered stable is a low priority and can be postponeduntil the patients condition allows a more definitive clinicaland radiologic examination of the spine. Although clearing thec-spine in this more limited context is a valid approach, theuse of the term clear as a verb inappropriately promotesthe concept that the initial trauma radiographs can be ex-pected to eliminate all unstable and stable cervical spineinjuries. To avoid miscommunication, radiologists and traumaphysicians in each institution should come to a commonunderstanding of the usage of these terms.

Reliable communication of CSR results is vital becausetracheal/esophageal intubation and many routine aspects ofpatient care can involve motion of the neck. Physicians,nurses, technologists, and other care providers must have anunambiguous report of the status of the trauma patientsspine, which encompasses the results of CSR and clinicalfindings or concerns. An incomplete series of radiographs isespecially problematic for care providers and is not unusualinmedium- and high-risk patients when other pressing traumaprioritiesdo not allow sufficienttime for a definitiveevaluation.

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

470 VANDEMARK AJR:155, September 1990

Fig. 3.-Neurologically intact 18-year-old manwith neck pain after a motorcycle accident.

A, Single lateral radiograph shows preverte-bral soft-tissue swelling but no unstable injury.

B, Odontoid view from complete five-view su-pine series is deceptively normal with exceptionof a vague zone of lucency at base of dens.

C and 0, Anteroposterlor and sagfttal refor-matted CT scans show an oblique undisplacedodontoid fracture.

This may require secondary radiologic examinations such asconventional tomography or CT, flexion-extension views, CT-myelography, or MR imaging. A rating system that plainlycommunicates the status of the cervical spine evaluationbased on initial CSR is suggested. CSR A is an adequateexamination with no evidence of instability, fracture, or soft-tissue injury. CSR B is a bothersome finding on radiographsthat requires repeated CSR, flexion-extension views, or asecondary examination. CSR C is a compromised or tech-nically inadequate examination that must be repeated. CSRD is a definite injury with fracture and/or instability detectedby CSR. The rating of radiographs is sometimes overruled byclinical concerns or findings. For example, a patient couldhave a CSR rating of A and yet have a neurologic findingthat requires further evaluation. Another patient with suspi-cious findings on radiographs (CSR B) might be clinicallyasymptomatic. Likewise, patients with high-risk characteris-tics may require secondary examinations despite apparentlynormal findings on initial CSR. The importance of clinicalparameters in the interpretation of initial CSR should not beunderestimated.

Stability/Instability

Referring physicians often expect the radiologist to renderan opinion about the stability of a detected injury on the basisof his examination of the trauma radiographs. The radiologistis sometimes appropriately uncomfortable in rendering thisopinion on the basis of routine trauma radiographs alone. Asno universally accepted definition of cervical spine stabilityexists, misunderstanding and miscommunication often result.

White et al. [20] have proposed a definition of cervical spineinstability. Their approach is by no means complete or entirelysatisfactory for the radiologic analysis of instability. However,it is based on sound biomechanical research and provides astarting place for this analysis. According to these authors, aclinically unstable spine, subjected to physiologic loads, isunable to maintain normal anatomic relations between spinalsegments so that current (or subsequent) spinal cord, nerveroot, or painful spine deformity results [20]. For practicalpurposes, an unstable cervical spine injury is consideredpresent if critical aspects of structural integrity of the spinehave been compromised, so that a neurologic injury could

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

Fig. 4.-Bone and soft-tissue structures responsible for spinal stability.White et al [20] suggest that for a margin of clinical safety, an unstableinjury is considered present when there is structural failure of eitheranterior or posterior spinal elements.

AJR:i55, September 1990 CERVICAL SPINE RADIOLOGY 471

occur if the cervical spine moved through a normal range ofmotion.

As there is no universally accepted definition of instability,designation of a particular injury as unstable can vary amongdifferent authors. Hams [21 } has presented a concise, butnonexclusive listof injuriesthat can be recognized as unstableon routine CSR. Injuries defined as unstable include: bilateralinterfacetal dislocation, flexion teardrop fracture, extensionteardrop fracture (unstable in extension), hangmans fracture,

Jefferson fracture, and hyperextension fracture-dislocation.When such patterns are not obvious, White et al. [20] haveidentified several findings on radiographs that, in combination,strongly suggest instability: loss of anterior structural integrity,loss of posterior structural integrity, subluxation greater than3.5 mm, kyphosis greater than 1 1 degrees, and traumaticdisk space widening or narrowing. The anterior and posteriorstructures responsible for stability of the cervical spine aredepicted in Figure 4. Although kyphosis and subluxation areusually obvious, traumatic loss of anterior or posterior bindingelements may be difficult to assess on radiographs. Never-theless, this approach is valuable when no obvious instabilitypattern is recognized on radiographs. Although less specific,when thefindings on cervical spine radiographs are apparentlynormal, the presence of a neurologic deficit, unusual degreeof neck pain, or marked prevertebral soft-tissue swellingshould alert the radiologist to the possibility of an occult,unstable injury. Caution is advised in these patients until athorough evaluation of the cervical spine can be accom-plished.

Stating with confidence that a structure such as the cervicalspine is stable often requires a two-phase analysis with bothstatic and dynamic testing. Sometimes the initial supinetrauma examination will detect obvious signs of unstablecervical spine injury. For practical use and patients safety,we would consider identification of any of the criteria notedabove as suspicious for unstable cervical spine injury, so thatspinal consultation is obtained. Based on this consultation,further imaging, including flexion-extension views, can beperformed to make a final determination.

Occasionally, when the findings on supine cervical exami-nation appear normal, it may be necessary in high-risk orsymptomatic patients to proceed to a dynamic flexion-exten-sion series for a definitive assessment of stability. In ourinstitution, the need for and timing of flexion-extension ex-aminations are based on consultation between the spineconsultant and the radiologist. For patients who are beingdischarged from the emergency department with a cervicalcollar, we favor a delayed flexion-extension examination in 10to 14 days, because of the inability of many patients with

Fig. 5.-Neurologically intact 45-year-old manafter a motor vehicle accident.

A, Radiograph of supine patient showsmarked prevertebral soft-tissue swelling but nounstable injury. Predental space is 3 mm, upperlimit of normal.

B, A subsequent flexion maneuver revealedan unstable, traumatic atlantoaxial subluxation.

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed

472 VANDEMARK AJR:155, September 1990

acute symptoms to perform flexion-extension maneuvers sat-isfactorily. For high-risk patients, an erect flexion-extensionseries is almost always impossible or impractical to performin multisystem trauma patients. Under rare circumstances,the spinal consultant and radiologist may jointly perform afluoroscopically monitored flexion-extension examination ofthe cervical spine.

Clearly, certain unstable injury patterns can be recognizedon static supine radiographs, whereas others may not be-come apparent until repeated radiographs are obtained withphysiologic loading, as in erect flexion-extension films. Figure5 demonstrates this principle.

Physicians must appreciate the limitations of a purely staticsupine cervical spine evaluation in trauma. The expectationthat the radiologist can exclude unstable cervical spine injuryin all cases, without benefit of a dynamic examination, iserroneous.

Conclusions

Radiologists must be knowledgeable in the complex medi-cal and social issues that surround the subject of cervicalspine trauma. Referring physicians are subject to multipleforces that favor liberal use of cervical spine radiographs andheavy reliance on the interpretation of radiographs. It is un-likely that this trend will be reversed in the near future;therefore, steps must be taken to accommodate the increaseddemand for these studies. A risk-tailored approach to CSR,which can improve efficiency in performance of these studiesby expediting the examination of medium- and low-risk pa-tients and encouraging the elimination of radiographs in no-risk patients, is suggested. The role of CSR in the clinicalmanagement of medium- and high-risk trauma patients isparticularly liable to lead to misunderstandings. Improvedcommunication and insight into the limitation of the initialtrauma radiographs in these patients will ameliorate much ofthe problem. In patients with significant risk of injury, a stagedevaluation, which may include repeated radiographs or sec-ondary examinations, is appropriate and will minimize delayeddetection or misdiagnosis of these injuries. Finally, clinicalparameters are extremely important in guiding our evaluationof cervical spine trauma and highlight the need for frequentconsultation between radiologist and spine injury consultant.A team approach, which involves radiology and the clinicalservices in a periodic review of these complex issues, is the

best approach in caring for patients with suspected cervicalspine injury.

REFERENCES

1 . Jacobs LM, Schwartz R. Prospective analysis of acute cervical spineinjury: a methodology to predict injury. Ann Emerg Med 1986;15:44-49

2. McNamara RM, OBrien MC, Davidheiser S. Post-traumatic neck pain: aprospective and follow-up study. Ann Emerg Med 1988;1 7:906-911

3. Maull KI, Sachatello CR. Avoiding a pitfall in resuscitation: the painlesscervical fracture. South Med J 1977;70:477-478

4. Bresler MJ, Rich GH. Occult cervical spine fracture in an ambulatorypatient. Ann Emerg Med 1982;1 1:440-442

5. Walter J, Doris PE, Shatter MA. Clinical presentation of patients with acutecervical spine injury. Ann Emerg Med 1984;13:51 2-515

6. Gatrell CB. Asymptomatic cervical injuries: a myth. Am J Emerg Med1985;3:263-264

7. Sumchai AP. Missed cervical spine fractures. Ann Emerg Med 1987;1 6:726-727

8. Fischer RP. Cervical radiographic evaluation of alert patients followingblunt trauma. Ann Emerg Med 1984;13:905-907

9. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN, Militello P.Protocol-driven radiologic evaluation of suspected cervical spine injury:efficacy study. Radio!ogy 1989;170:831-834

10. Gbaanador GBM, Fruin AH, Taylor C. Role of routine emergency cervicalradiography in head trauma. Am J Surg 1986;152:643-648

1 1 . Ringenberg BJ, Fisher AK, LWdaneta LF, Midthun MA. Rational ordering ofcervical spine radiographs following trauma. Ann Emerg Med 1988;17:792-796

12. Roberge RJ, Wears RC, Kelly M, et al. Selective application of cervicalspine radiography in alert victims of blunt trauma: a prospective study.J Trauma 1988;28:784-787

13. Schleehauf K, Ross SE, Civil ID, Schwab CW. Computed tomography inthe initial evaluation of the cervical spine. Ann Emerg Med 1989;18:815-817

14. Ross SE, Schwab CW, David ET, Delong WG, Born CT. Clearing thecervical spine: initial radiologic evaluation. J Trauma 1978;27: 1055-1 060

1 5. Acheson M, Livingston R, Richardson M, Stimax G. High-resolution CTscanning in evaluation of cervical spine fractures: comparison with plainfilm examinations. AJR 1987;148: 1179-1185

1 6. Clark CR, Igram CM, El-Khoury GY, Ehara S. Radiographic evaluation ofcervical spine injuries. Spine 1988;13:742-746

1 7. Freemyer B, Knopp R, Piche J, Wales L, Williams J. Comparison of five-view and three-view cervical spine series in the evaluation of patients withcervical trauma. Ann Emerg Med 1989;18:818-821

18. Reid DC, Henderson R, Saboe L, Miller JDR. Etiology and clinical courseof missed spine fractures. J Trauma 1987;27:980-986

19. Bohlman HH. Acute fractures and dislocations of the cervical spine. J BoneJoint Surg [Am] 1979;61-A:1119-1142

20. White AA, Southwick WO, Panjabi MM. Clinical instability in the lowercervical spine: a review of past and current concepts. Spine 1976;1:15-27

21 . Hams JH Jr. The radio!ogy of acute cervica! spine trauma, 1 st ed. Balti-more, MD: Williams & Wilkins, 1978:41

Dow

nloa

ded

from

ww

w.a

jronli

ne.or

g by 1

24.82

.44.10

on 04

/16/15

from

IP ad

dress

124.8

2.44.1

0. Co

pyrig

ht AR

RS. F

or pe

rsona

l use

only;

all ri

ghts

reserv

ed