Available online at www.sciencedirect.com

Surgical Neurology 70 (2008) 53–58www.surgicalneurology-online.com

Trauma

Is the lateral cervical spine x-ray obsolete during the initial evaluation ofpatients with acute trauma?

Andrea Fisher, BS, William F. Young, MD⁎

Fort Wayne Center, Indiana University Medical School, Fort Wayne, IN 46805, USA

Received 19 May 2007; accepted 1 August 2007

Abstract Background: Clearing the cervical spine is a vital part of the treatment of trauma patients, and the

Abbreviations: LCtomography; CSCT, c

⁎ Corresponding aIN 46805, USA. Tel.:

E-mail address: w

0090-3019/$ – see frodoi:10.1016/j.surneu.2

failure to accurately diagnose an injury to the cervical spine can result in paralysis and even death.For decades, plain film imaging, primarily LCSX, was the standard imaging method used to initiallyevaluate the cervical spine, with CSCT used as an adjunct. With advancements in CSCT over the pastdecade, it is generally accepted that CSCT should be used as the screening method for clearance ofthe cervical spine in patients with trauma.In this study our goal was to determine whether lateral cervical spine radiographs (LCSX) arewarranted in the initial evaluation of trauma patients or whether they should be eliminatedcompletely in favor of CSCT scans as the initial method of evaluating the cervical spine in traumapatients.Methods: This is a retrospective study using a prospectively maintained computerized database ofall trauma admissions to a level II American College of Surgeons verified trauma center. Patientswho were identified to have both LCSX and CSCT on admission were analyzed. Radiology readings(LCSX and CSCT) of the selected patients were reviewed and comparisons were made to determinethe number of patients for which the LCSX was inconclusive or unsatisfactory. It was alsodetermined whether, in selected cases, there were injuries or abnormal findings that were detected inone imaging modality but were not detected in the other.Results: A total of 895 trauma admissions were reviewed; 177 had both LCSX and CSCT. Theradiological results of the 177 patients were analyzed. Fifty-one (28%) patients were determined tohave inadequate LCSX in which further scans were required for diagnosis or clearance of thecervical spine. Thirty-six (20%) patients had fractures that were undetected by LCSX.Conclusions: Our research supports previous studies demonstrating the greater accuracy of CT inevaluating the cervical spine in acute trauma patients. Moreover, with spiral CT scanning, the lengthof time required to obtain images has been eliminated as an issue. We conclude that LCSX should beeliminated from trauma protocols and that CSCT should be the sole imaging modality used in theinitial evaluation of the cervical spine after trauma.© 2008 Elsevier Inc. All rights reserved.

Keywords: Cervical; Spine; X-ray; Trauma

1. Introduction

Clearing the cervical spine is of paramount importancewhen evaluating trauma patients. The failure to accurately

SX, lateral cervical spine x-ray; CT, computedervical spine computed tomography.uthor. Fort Wayne Neurological Center, Fort Wayne,+1 260 460 3100; fax: +1 260 460 3130.fymd@comcast.net (W.F. Young).

nt matter © 2008 Elsevier Inc. All rights reserved.007.08.011

diagnose a cervical spine injury can have devastatingconsequences. Clearance of the cervical spine can typicallybe accomplished by thorough physical examination in thosepatients who are alert and cooperative and who are withoutneurologic deficit or neck pain [9]. However, the matter ofhow best to clear the cervical spine when physicalexamination is not sufficient, especially in patients who areunresponsive or have multiple trauma, has been an issue ofcontroversy [1,9]. Radiography, primarily LCSX, was used

54 A. Fisher, W.F. Young / Surgical Neurology 70 (2008) 53–58

to clear the cervical spine for decades. With the introductionof CT, CT scans came to be used as an accessory form ofimaging when deemed necessary for better visualization ofthe cervical spine [15].

It is widely accepted that a significant number of injuriesare missed with LCSX alone, especially those injuriesoccurring at the craniocervical and cervicothoracic junctions[14]. Injuries most often go undiagnosed in patients who areuncooperative or have severe injury [15]. It is also acceptedthat LCSX is an insufficient method when used alonebecause of the high incidence of false positives and falsenegatives [14]. LCSX is often inadequate because of poorvisual quality [14].

Because of advancements in CT over the past decade it isnow used routinely for visualizing the cervical spine intrauma patients. These factors call into question whetherLCSX should be eliminated during the initial evaluation oftrauma patients whose cervical spines cannot be cleared byphysical examination alone. This study was undertaken toconfirm in our own institution whether CT is superior toLCSX as an imaging modality to clear the cervical spine andwhether traditional LCSX should be eliminated in the initialevaluation of trauma patients.

Table 1Classification of findings by mechanism of injury

Mechanismof injury

Number Missedinjury

Poor Visualization onx-ray/recommendfurther scans

Incidentalfindings

Motor vehicleaccident

98 22 32 10

Motor/moped 23 4 6 10All-terrain vehicle 8 2 3 7Fall 37 7 4 3Other 11 1 6 2Total 177 36 51 32

2. Methods

This is a retrospective study that involved traumapatients who were admitted to Parkview Hospital fromJanuary 1, 2005, through December 30, 2005. ParkviewHospital is a 573-bed institution in Fort Wayne, Indiana. Itis a level II trauma center verified by the AmericanCollege of Surgeons. The catchment area for the hospitalincludes more than 35 surrounding counties, includingsome in the neighboring states of Ohio and Michigan. Thetrauma center serves an estimated population of approxi-mately 2.4 million people in this mixed urban and ruralarea. An institutional review board evaluated this projectbefore the beginning of research.

Institutional databases were queried retrospectively toidentify all trauma patients during the study period. Patientswho were identified to have both LCSX and CSCT onadmission were analyzed. As a part of its trauma verification,Parkview Hospital maintains a prospective computerizedtrauma registry for all admissions. This database has ademonstrated accuracy of more than 95% using multipleindicators (TraumaBase software, Windows version 6.0CDM). Demographic information was collected for thesepatients including name, date of birth, age, sex, GlasgowComa Scale score, and mechanism of injury. Injury SeverityScores were assigned retrospectively; hence, this informa-tion was not available for the patients in this study. All CTscans were performed on a GE light speed helical scanner(United Kingdom). The average time for acquisition ofimages was less than a minute. All patients had both standardaxial images with 2.5-mm slices as well as sagittal andcoronal reconstructed images. Each patient's results for both

LCSX and CSCT were collected and analyzed for severalpossible outcomes. These included (1) reports in which theconclusions between scans were the same, either bothpositive for an injury or finding or both negative for an injuryor finding; (2) reports in which there was an injury that wasidentified by one type of scan but missed by the other,classified as missed injuries; (3) reports in which it wasmentioned that the cervical spine was poorly visualized or inwhich it was determined that further x-ray scans would berequired to be diagnostic; and (4) reports in which a findingwas cited that was separate from those specifically involvingthe cervical spine, classified as an incidental finding.

3. Results

Of the 895 trauma admissions analyzed, 180 weredetermined to have had both an LCSX and an CSCT aspart of their trauma care. The wide discrepancy between totaladmissions and those who received both LCSX and CSCT isdue to many factors including the following: (1) manypatients received CSCT alone, (2) the total includes patientswho died before completing the imaging evaluation, and (3)the total admissions include those patients who weredetermined clinically not to require radiologic imaging forclearance of the cervical spine (eg, gunshot wounds) Ofthose 180 patients, 3 were excluded because of insufficientdata regarding their scans. The remaining 177 individualshad a mean age of 35.48 years. The mean Glasgow ComaScale score was 12.31.

Table 1 classifies the 177 cases and their associatedfindings in terms of the mechanism of injury. Mechanisms ofinjury were divided into (1) motor vehicle accidents, (2)accidents involving a motorcycle or a moped, (3) all-terrainvehicle accidents, (4) falls, and (5) those categorized asother, which included 2 buggy accidents, a tubing accident, atree falling on an individual, 3 pedestrians struck by anautomobile, 2 bicycle accidents, and 2 diving accidents.

Our study found that in 51 (28%) of the 177 casesreviewed, the LCSX was determined to be inadequate by theradiologist and further scans were recommended, eitherbecause the quality of the LCSX was poor or because thecervical spine was insufficiently visualized. In only 1 (0.5%)

Table 2Levels of undetected fractures

Vertebra No. of fractures (%)

C1 12 (33)C2 8 (22)C3-5 5 (13)C6 5 (13)C7 6 (16)

Table 4Location/nature of incidental findings

Incidental finding Number

Temporal bone fracture 8Thyroid cysts 2Rib fractures 10Thoracic vertebra fractures 6Skull base fibrous dysplasia 2Basilar/occipital fractures 2Dissection of internal carotid artery/thrombus 2

55A. Fisher, W.F. Young / Surgical Neurology 70 (2008) 53–58

of the 177 cases CSCT was determined to be insufficient fordiagnosis. This was because the patient was uncooperativeand a great deal of motion artifact was found on the images.The difference between LCSX and CSCT regarding qualityof imaging studies was statistically significant (P b .05).

It was also determined that 36 (20%) of the 177 patientshad cervical spine fractures that were undiagnosed by LCSX.Twenty-three of these were determined to be clinicallysignificant and required further treatment. In addition, in 23of the 36 cases, the LCSX was mentioned to be of poorquality. In the remaining 13 cases, LCSX was found to be ofsufficient quality by the radiologist. However, although theLCSX was of sufficient quality it did not demonstrate thefracture seen on CSCT.

Of the 36 fractures undetected by LCSX alone, 20 (55%)were located at the C1-2 vertebral level, 5 (13%) at the C3-5midcervical level, and 11 (30%) at the C6-7 vertebral level(Table 2). Sixteen (44%) fractures involved lateral aspects ofthe vertebrae, including the transverse process and the lateralmass. Only 5 (14%) involved the vertebral body (Table 3).

There was one incidence of a false positive detected byLCSX. An anterolisthesis of the C6-7 vertebral level wasreported on LCSX but not on CSCT. It was later determinedwith further imaging including repeat plain x-rays and CSCTthat there was not a subluxation and that the original plain x-rays had been misinterpreted. Thus, no cervical spinefractures or subluxations were detected by LCSX that werenot also detected by CSCT.

In addition to findings of injury to the cervical spine, thecases were reviewed for incidental findings, or those injuriesor findings not specifically involving the cervical spine(Table 4). Incidental findings were present in 32 (18%) of the177 cases. Of these 32, 31 were present only in the CSCTreport. One rib fracture was noted on the LCSX report, butnot on the CSCT report. Upon further review it was

Table 3Locations of undetected fractures

Location No. of fractures (%)

Occipital articular process 4 (11)C1 arch 2 (5)Dens 4 (11)Vertebral body 5 (14)Transverse process or vertebral foramina 12 (33)Lateral mass 4 (11)Lamina 5 (13)

determined that the rib fracture was not in the field of viewof the CSCT.

4. Discussion

When determining how to adequately evaluate thecervical spine of a trauma patient, one wants the fastestand most accurate scanning method while keeping patientcost at a minimum.

There is no debate as to imaging superiority of CTscans ascompared to plain film radiography. During the early use ofCSCT in trauma patients it was necessary to perform plain x-rays because of the high incidence of missing subluxations onaxial CT images. However, with the advent of high-resolutionCT imaging in conjunction with sagittal and coronalreconstructed images CSCT is at least as accurate indiagnosing subluxations as conventional plain x-ray. Numer-ous authors of recent literature have concluded that CT scansare the more precise form of imaging, resulting in fewermissed injuries and improving their detection [9,16]. In 1993,Woodring and Lee [16] found that 23% of cervical injurieswere undiagnosed when conventional radiography alone wasused and that, of those injuries, half were unstable. Thesefindings have also been confirmed by other authors [5,8,13].

Our study had very similar findings, with 20% (n = 36)undiagnosed with only LCSX. Furthermore, in 36% (n = 13)of the undetected injuries, the radiologist's reading of theLCSX indicated that the image was of adequate quality butstill did not demonstrate the fracture seen on CSCT.Consequently, it seems impossible to conclude simplyfrom the quality of the LCSX whether it is sufficientdiagnostically. The superior quality of the image ensures thatimprecise or unclear scans and false positives are uncommonwith CSCT [9]. CT scans allow the reconstruction of theentire cervical spine in the coronal and sagittal planes,maximizing its visualization [11]. Moreover, because oftechnological advances, CSCT can be performed expedi-tiously, especially when coupled with head or abdominal CTscans. This minimizes both cost and time investment andreduces movement of potentially unstable patients [3].

Recent literature has shown that most injuries missed byLCSX occur at the craniocervical and cervicothoracicjunctions [2,4,10]. In addition, Link et al [12] showed ahigh percentage of injuries to the occipital condyles. Mostattribute this unequal distribution to the fact that the extreme

Fig. 2. Coronal CT image of same patient showing fracture of right lateralmass of C1.

56 A. Fisher, W.F. Young / Surgical Neurology 70 (2008) 53–58

proximal and distal portions of the cervical spine arefrequently poorly visualized on LCSX. Our study showedsimilar findings, as only 13% (n = 5) of the missed injuriesoccurred in the midcervical range. Fifty-five percent (n = 20)of the injuries occurred at the C1-2 vertebral levels and 30%(n = 11) occurred at the C6-7 vertebral levels. Thus, it seemsthat LCSX is inferior in the detection of injuries to theproximal and distal portions of the cervical spine. Inaddition, it was found that most missed injuries involvedthe lateral aspects of the vertebrae, specifically the transverseprocesses and lateral masses. Similar findings were dis-cussed by Nunez et al [15].

In our study we only analyzed LCSX because this is theinitial x-ray view used in severely injured patients who maynot be hemodynamically stable enough to get additionalviews. It is true that these additional views (eg. open mouth)may improve the chances of diagnosing abnormalitiesparticularly at C1-2. However, there is a trade-off for usingthem with regard to the additional time required for theiracquisition. It is also recognized that the additional plainx-ray views are also often inadequate to diagnose abnorm-alities at C1-2 when compared with CSCT (see Figs. 1-3).

Overall, in our study, 28% (n = 51) of LCSX weredetermined by the radiologist to be inadequate for diagnosis

Fig. 1. Lateral cervical spine x-ray not showing any abnormalities.

of injury of the cervical spine or for its clearance. In suchcases, further scans are required, typically either CSCT oradditional x-rays. The additional images lead to the loss ofvaluable time and delays in diagnosis. Moving patients foradditional studies subjects the patients to additionaldiscomfort. Also, excessive scans result in added costs forthe patient. The initial use of CSCT prevents suchburdensome procedures in most cases.

Not only is CSCT superior for the detection of injury tothe cervical spine, but as was seen in our study, CT scanshave the added benefit of detecting findings within the

Fig. 3. Open-mouth x-ray view of the same patient. The fracture of C1 is notseen because of Mach bands. These artifactual lines are commonly seen withthis x-ray view.

57A. Fisher, W.F. Young / Surgical Neurology 70 (2008) 53–58

cervical region that may be clinically significant. In ourstudy, we referred to these as incidental findings. There were32 incidental findings discovered in our study; 31 werediscovered by CSCT alone. CSCT is superior for visualizingsoft tissues, and use of contrast allows the detection ofinjuries to vascular structures, especially the internal carotidarteries and the vertebral arteries.

In our study, 2 patients were determined, by CSCT, tohave a dissection of the internal carotid artery. Carotid injuryhas the potential to cause a stroke if not detected and treated.Hence, CSCT has the benefit of detecting findings of interestbeyond the bony structures of the cervical spine.

Despite the benefits of CSCT, there are some potentialdrawbacks that have been cited as criticisms of its routineuse in trauma situations. Disadvantages include cost, timerequired, and radiation risk. At our institution, the totalcost of a CSCT is about 10 times that of an LCSX. Forthis reason, and because of the growing concern abouthealth care costs, many physicians hesitate to use CTtechnology. However, studies have found that when allcosts are taken into account, CT technology may actuallybe the more cost-effective screening method [3]. In 1999,Blakemore et al [4] analyzed the total cost of all cervicalspine radiography compared with the total cost of CSCT.They concluded that when all costs were considered,including the short-term costs of extended hospital staysand the long-term costs associated with paralysis resultingfrom missed injuries, CSCT was actually the lessexpensive method and hence, the preferred method formoderate- to high-risk trauma patients [4]. Furthermore, ifthe CSCT is performed at the same time as head orabdomen CT scans, little extra cost is incurred.

The time it takes to clear the cervical spine of a traumapatient is vitally important and any delay in this process canbe devastating. Daffner [6] concluded that the averageexamination time for cervical radiography (6 views) was22 minutes, whereas the average examination time for CSCTwas 11 minutes. Examination times included both theexposure time to radiation, which is brief for plain x-rayscompared to CSCT, as well as the time required to positionthe patient for the study, which is much longer for plainx-rays [7]. Examination times averaged longer because 79%of those receiving radiographic examinations requiredrepeated scans [6,7]. Repeated scans waste valuable timein a patient's treatment. Again, it has been suggested that ifother CT scans are necessary, CSCT can be performed at thesame time, thus eliminating the time consumed by radio-graphy [3]. Moreover, the latest generation CT scanners havereduced the average examination time even further.

The final concern is that of the risk of radiation. Theadded risk of radiation is low for a CSCT when comparedwith other types of CT scans because of the small size of thecervical region [1]. The greatest concern is to childrenbecause of the risk of thyroid carcinoma. The use of CTscansin children may not be warranted and must be decided on acase by case basis. However, increased exposure to radiation

for an adult as the result of CSCT scan is not significantenough to prevent its use in that population.

Although no imaging modality is accurate 100% of thetime, it is the responsibility of the physician to use resourcesthat best serve the patient. The superior quality of CT imagesand the reduction in delayed diagnoses after CT scansindicate that CSCT is more accurate for diagnosing cervicalspine injury. Our research supports the elimination of theLCSX during the initial examination of trauma patients, andwe maintain that CSCT should be the standard initialimaging modality for clearance of the cervical spine aftertrauma in adult patients.

References

[1] Berne JD, Vemahos GC, El-Tawil Q, Demetriades D, Asensio JA,Murray JA, Cornwell EE, Belzberg H, Berne TV. Value of completecervical helical computed tomographic scanning in identifying cervicalspine injury in the unevaluable blunt trauma patient with multipleinjuries: a prospective study. J Trauma 1999;47:896.

[2] Blacksin MF, Lee HJ. Frequency and significance of fractures of theupper cervical spine detected by CT in patients with severe necktrauma. AJR Am J Roentgenol 1995;165:1201-4.

[3] Blakemore CC, Mann FA, Wilson AJ. Helical CT in the primarytrauma evaluation of the cervical spine: an evidence-based approach.Skeletal Radiol 2000;29:632-9.

[4] Blakemore CC, Ramsey SD, Mann FA, Deyo RA. Cervical spinescreening with Uierer G, CT in trauma patients: a cost–effectiveanalysis. Radiology 1999;212:117-25.

[5] Brohi K, Healy M, Fotheringham T, Chan O, Aylwin C, Whitley S,Walsh M. Helical computed tomographic scanning for the evaluationof the cervical spine in the unconscious, intubated trauma patient.J Trauma 2005;48(5):897-901.

[6] Daffner RH. Cervical radiography for trauma patients: a time-effectivetechnique? AJR 2000;175:1309-11.

[7] Daffner RH. Helical CT of the cervical spine for trauma patients. AJRAm J Roentgenol 2001;177:677-9.

[8] Daffner RH, Sciulli RL, Rodriguez A, Protech J. Imaging forevaluation of suspected cervical spine trauma: a 2-year analysis.Injury 2006;37(7):652-8.

[9] Griffen MM, Frykberg ER, Kerwin AJ, Schinco MA, Tepas JJ, RoweKK, Abboud J. Radiographic clearance of blunt cervical spine injury:plain radiograph radiograph or computed tomography scan? J Trauma2003;55:222-7.

[10] Hanson JA, Blakemore CC, Mann FA, Wilson AJ. Cervical spineinjury: accuracy of helical CT used as a screening technique. EmergRadiol 2000;7:31-5.

[11] Lee SL, Sena M, Greenholz SK, Fledderman M. A multidisciplinaryapproach to the development of a cervical spine clearance protocol:process, rationale, and initial results. J Pediatr Surg 2003;38:358-62.

[12] Link TM, Schuierer G, Hufendiek A, Horch C, Peters PE. Substantialhead trauma: value of routine CT examination of cervicocranium.Radiology 1995;196:741-5.

[13] Mathen R, Inaba K, Munera F, Teixeira PG, Rivas L, McKenney M,Lopez P, Ledezma CJ. Prospective evaluation of multislice computedtomography versus plain radiographic cervical spine clearance intrauma patients. J Trauma 2007;62:1427-31.

[14] Nunez Jr DB, Quencer RM. The role of helical CT in the assessment ofcervical spine injuries. AJR Am J Roentgenol 1998;171:951-7.

[15] Nunez Jr DB, Zuluaga A, Fuentes-Bernardo DA, Rivas LA, BecerraJL. Cervical spine trauma: how much more do we learn by routinelyusing helical CT? Radiographics 1996;16:1307-18.

[16] Woodring JH, Lee C. Limitations of cervical radiography in theevaluation of acute cervical trauma. J Trauma 1993;34:32-9.