Cervical Spine Trauma Associated with Moderate and Severe Head Injury

~&U•••I••II••••II•••••II••l!l*II.MAAli,I.!llii1¥••'I"II;!1••iliff,;'••••III•••••••a.

)oi, et aJ.

mOQ"ranh,. :" It Surg'(B~; J Neurosurg (Spine 31 96:285-291, 2002

ld :esults, in urles, ed 1.

c dura] tears: 3-446, 1984 Cervical spine trauma associated with moderate and . of myeloa. .) 14:59--64. severe head injury: incidence, risk factors, and injury characteristics c resonance It Surg (BrJ

transfers in LANGSTON T. HOLLY, M.D., DANIEL F. KELLY, M.D., GEORGE J. COUNELIS, M.D., reconstruc· THANE BLINMAN, M.D., DAVID L. McARTHUR, PH.D., M.P.H., AND H. GILL CRYER, M.D. 3 Saunders, Division ofNeurosurgery, Departments of Surgery and Epidemiology, University of California 1 Alnot JY at Los Angeles Center for Health Sciences, and the UCLA Brain Injury Research Center. juries, ell 1: Los Angeles; Division ofNeurosurgery, Harbor-UCLA Medical Center, Torrance, California; ;)-64 and Department ofNeurosurgery, University ofPennsylvania, Philadelphia, Pennsylvania II investiga_ iiagnosis of y intradural Object. Diagnosing and managing cervical spine trauma in head-injured patients is problematic due to an altered 531,1998 level of consciousness in such individuals. The reported incidence of cervical spine trauma in head-injured patients has tic value of generally ranged from 4 to 8%. In this retrospective study the authors sought to define the incidence of cervical injury lrg (lBr) 19: in association with moderate or severe brain injury, emphasizing the identification of high-risk patients. Methods. The study included 447 consecutive moderately (209 cases) or severely (238 cases) head injured patients laging tech­ who underwent evaluation at two Levell trauma centers over a 40-month period. Of the 447 patients, 24 (5.4%) suf­ -553,1995 fered a cervical spine injury (17 men and seven women; mean age 39 years; median Glasgow Coma Scale rGCS] score yelographic of 6, range 3-14). Of these 24 patients, 14 (58.3%) sustained spinal cord injuries (SCIs), 14 sustained injuries in the ns: discrep­ occiput-C3 region. and 10 underwent a stabilization procedure. Of the 14 patients with scrs, nine experienced an early Nerve 17: hypotensive andJor hypoxic insult. Regarding the mechanism of injury, cervical injuries occurred in 21 (8.2%) of 256 patients involved in motor vehicle accidents (MVAs), either as passengers or pedestrians, compared with three (1.6%) lries of the of 191 patients with non-MVA-associated trauma (p < 0.01). In the subset of 131 MVA passengers, 13 (9.9%) sus­ iO, 1968 tained cervical injuries. Patients with an initial GCS score less than or equal to 8 were more likely to sustain a cervi­ cal injury than those with a score higher than 8 (odds ratio [OR] 2.77. 95% confidence interval [CI] = 1.11-7.73) and were more likely to sustain a cervical SCI (OR 5.5, 95% CI 1.22-24.85). At 6 months or more postinjury, functional neurological recovery had occurred in nine patients (37.5%) and eight (33.3%) had died. Conclusions. Head-injured patients sustaining MVA-related trauma and those with an initial GCS score less than or )artment of equal to 8 are at highest risk for concomitant cervical spine injury. A disproportionate number of these patients sustain mogo862­ high cervical injuries, the majority of which are mechanically unstable and involve an SCI. The development of safer Jimac@ca. and more rapid means of determining cervical spine integrity should remain a high priority in the care of head-injured patients.

KEyWORDS cervical spine injury traumatic brain injury spinal stabilization

NDIVIDUALs who sustain traumatic brain injury are at severity, as measured by the GCS, has been associated increased risk of sustaining cervical spine injury com­ with a higher rate of cervical injury, few other risk factors 13 c4 I pared with victims of nonhead-related blunt trauma in this patient population have been identified. • Iden­ I injury. jJ The reported incidence of cervical spine trauma tifying such high-risk patients may be particularly useful i after clinically significant head injury has generally given that diagnosing and managing cervical spine in­ ranged from 4 to 8%.11,13.19c4 Although an increasing injury juries in head-injured patients is problematic due to an altered level of consciousness and the overall critical na­ ture of their injuries. Despite advances in imaging tech­ Abbreviations used in this paper: CI = confidence interval; CT = nologies and screening protocols, cervical spine injury I computerized tomography; GCS = Glasgow Coma Scale; GOS= ; Glasgow Outcome Scale; GSW == gunshot wound; MR == magnetic may go undetected even in optimum circumstances. resonance; MVA = motor vehicle accident; OR == odds ratio; SCI = Failure to recognize unstable cervical injuries during ini­ spinal cord injury; UCLA = University of California at Los An­ tial evaluation can result in serious neurological worsen­ geles. ing.' Increased neurological deficits have been reported to ril,2002 1. Neurosllrg: Spine / Volume 96/April. 2002 285 L. T. Holly, et al, :'ervical spine traum,

occur in 3 to 10% in patients with SCI after arriving at the Patients. with .combined craniocervical inj~ries Were hospital. 13 managed m the mtensIve care umt after stabIlIzation Or The goals of this study were to determine the incidence after surgical evacuation of an intracranial hematoma. In­ and characteristics of cervical spine injuries in patients divid~als with defini~ive or presumed SCI received methyl_ who sustained moderate or severe head trauma. We also predmsolone, as mdIcated by the SCI protocol outlined in Level of Injury sought to identify risk factors for combined craniocervical the National Acute Spinal Cord Injury Study II.' Central trauma to identify more effectively the high-risk patient venous catheters were placed, and hemodynamic sUpport Oc-Cl Fx-SL early in their hospitalization. with intravenous fluids and vasopressor agents were used Oc-Cl SL to maintain adequate blood pressure. Depending on the nu­ Col Fx ture of the cervical injury, severity of head injury. and car­ Col Fx Clinical Material and Methods CI-2 SL diopulmonary stability, patients were initially muintained CI-2 SL Patient Population in a cervical collar, placed in cervical traction in which CI-2 SL Gardner-Wells tongs were used, placed in a halo vest, or Between December 1992 and March 1996,464 patients C2-3 Fx-SL taken to the operating room to undergo open reduction and C-2 Fx with moderate or severe head injury (postresuscitation GCS internal cervical fusion. In all cases of severe head injury C-2Fx scores of 3-12 or > 12 with abnormal findings demon­ (GCS Scores 3-8) and in some cases of moderate injury C2-3 Fx-SL strated on head CT scans) were admitted to either the UCLA (GCS Scores 9-14) in which CT scanning demonstrated C3-4 Fx-SL or Harbor-UCLA Medical Centers. Of these patients, 287 C3-4 Fx-SL evidence of mass effect, intracranial pressure ICP monitor­ were prospectively emolled into the UCLA Brain Injury C3-S Fx-SL ing was performed. In such patients, goals of head injury Research Center program. Informed consent was obtained C-4 Fx management included maintenance of intracranial pressure CS-6 Fx-SL from family members of all participants in the study. An less than 20 mm Hg and cerebral perfusion pressure higher CS-6 Fx-SL additional 160 patients were recruited into the program but than 70 mm Hg. Global neurological outcome was deter­ C-S Fx were not emolled. Because the overall age, sex, and injury mined at 6 months postinjury by the GOS. Favorable out­ C6-7 Fx characteristics of the emolled and nonemolled patients did C-6 Fx come was defined as good recovery or moderate disability not differ substantially, these two patient groups were com­ C6-7 Fx-SL and poor outcome as severe disability, persistent vegetative bined for the purposes of this study, yielding a total of 447 e-7 Fx state, or death. 15 C-7 Fx moderately or severely head injured patients evaluated dur­ C7-Tl SL ing this time period. An additional 17 patients were exclud­ Statistical Analysis ed from this study because data regarding characteristics of * Patients are pres, Data were managed by Paradox (version 7; Corel Corp., juries involving leve their head injuries were insufficient. injUry with punctate To identify the head-injured patients with concomitant Ottowa, Ontario, Canada) and analyzed using SPSS (ver­ HT = hypotensive in cervical spine injury, the UCLA Brain Injury Research sion 8, SPSS Inc., Cary, NC) and PEPI (version 3; devel­ mod = moderate; M' Center database was reviewed and retrospective chart re­ oped by 1. H. Abramson, The Hebrew University of Jeru­ ery; SAH = subaracJ views were performed. All head-injured patients with doc­ salem, Jerusalem, Israel; accessible at www.myatt.demon. umented cervical spinal cord and/or spinal column trau­ co.uk). Statistics for tabular data included chi-square and ma were included. The GCS scores, Abbreviated Injury Fisher exact tests; a probability value of 0.05 was used for 1 Scores, and Injury Severity Scores were documented in all significance. Odds ratio statistics with 95% CIs were used I patients. I Potential risk factors and injury-related character­ to compare the difference in risk of cervical spine injury lOt d . h istics were documented including age, sex, mechanism of due to MVA- and non-MVA-related causes of injury. and in ,live, an fractures wIt ou injury, location of injury, GCS score, presence of neurolog­ those with initial GCS scores of 8 or less or greater than 8. lThe injuries were first doc ical deficits, and occurrence of hypotension or hypoxia l:!diographs in 18 patients identified in the prehospital setting or within 24 hours of llwo of the six patients, ce !'lrated on the original pIa hospitalization. Early hypotension or hypoxia was defined Results as a systolic blood pressure of less than 90 mm Hg or PaO, licothoracic junction was less than 60 mm Hg (or apnea), respectively. Injury Demographics !:Iuding one patient with a IJll0ther with a C-7 spinous Of the 447 patients in the cohort (238 with an initial Patient Management ,tour patients all underwl GCS score of :s 8 and 209 with an initial GCS score of j'evealed nondisplaced late On admission to the emergency room after initial car­ 9-14), 24 (5.4%) suffered a cervical spine injury. Their hatients, C-4 in one, and ( diopulmonary stabilization, cervical spine radiographs demographic and injury characteIistics are summarized III l!ng in these four patients were obtained (cross-table lateral, anteroposterior, and Table 1. The mean age of these patients was 39 years :lOn of SCI in two patier open-mouth [odontoid] views). Computerized tomography (range 18-64 years), the mean Injury Severity Score was Jocumented areas in twc 32 ± 14, and 17 were men and seven were women. In 18 scanning was performed to assess the occiput-C2 area and titre first detected during I the cervicothoracic junction in cases in which these areas (75%) of 24 patients the initial GCS score was less than or ;:ation except in one patier t Th~re were poorly visualized on plain radiographs. Other levels equal to 8 (median GCS score of 6, range 3-1 ,). r,hich was documented on of obvious or suspected abnormality on plain radiographs were no significant differences in terms of age or sex d.r s­ The most frequently ir tribution between the 24 patients with combined eramo­ I were further investigated by acquiring thin-cut CT scans, i~ciput and C-3, demons cervical injury and the 423 head-injured patients \1'ithout including reformatted sagittal and coronal scans when clin­ lents. A neurological deB ically indicated. Since 1995. obtunded or comatose pa­ cervical injury. Las found in 14 patients: tients, in whom nOlllial findings are demonstrated on plain I J Were incomplete SCIs cervical radiographs and cervical CT scans, have under­ ervical spine injuries we gone bedside lateral flexion-extension radiography to rule Characteristics of Cervical Spine Injury ~e 14 cases in which hi! out ligamentous instability. typically within 2 to 3 days of pre~:::nt Hi Cervical spine fracture-subluxations were d.ocumented. Ten pc injury. nine patients, subluxations without fractures oeC:;:'ed rn ,Jden,,'ent stabilization p I l 286 200:: J. Neurosurg: Spine / Volume 96/April. ­ r;; !\eurosurg:'r Spine / Volu --l - .. lolly, et al. ,'eoica1 spine trauma after head injury njuries wep lbilization o~ TABLE 1 :matoma. In. Demographic and injury characteristics in 24 patients with craniocervical trauma" ived methyl. )1 outlined in Unstable Age \yrs). GOS 2 Level of Injury SCI Injury Sex Mechanism HTrr-IP Head CT GCS Category (Score) 1 II. Central LmlC SUppon Oc-C1 Fx-SL comp yes 64,F MVA HT/HP SAH.IVH 3 died (1) ts were Used Oc-C1 SL comp yes 37, M MVA HTIHP IVH 3 died (1) 19 on the na. C-1 Fx incomp no 26,F MVA HT DS, SAH 4 mod dis (4) ury, and car­ C-1 Fx no no 25. M AP HT SAH 3 severe dis (3) , maintained Cl-2 SL no yes 28, M AS none DS, SDH 10 died (1) Cl-2 SL incomp yes 18,M AP HP DS, SAH, IVH 6 vegetative (2) In in which Cl-2 SL no yes 23, M MVA(e) HP DS, nSDH 3 vegetative (2) lalo vest, or C2-3 Fx-SL incomp yes 51, F MVA HP SAH 7 died (1) ~duction and C-2 Fx incomp no 39, M MVA none normal 6 good rec (5) head injury C-2 Fx incomp yes 33,F MVA HT DS. SAH. IVH 3 vegetative (2) lerate injury C2-3 Fx-SL incomp yes 19,M AP none DS,IVH 7 mod dis (4) emonstrated C3--4 Fx-sL no yes 49,M MVA none contusion 14 mod dis (4) C3--4 Fx-SL comp yes 54.F MVA HP DS, SAH 7 died (1) ::::P monitor- C3-5 Fx-SL incomp yes 41, M AP none nSDH. SAH 3 severe dis (3) head injury C-4 Fx no no 28. M MVA(e) HT/HP DS. SAH 3 died (1) lial pressure C5-6 Fx-SL incomp yes 27. M MCA none SAH. MC 10 mod dis (4) ssure higher C5-6 Fx-SL incomp yes 56.M AP HT DS, SAH 3 died (I) ~ was deter­ C-5 Fx no no 44. F MVA(e) none SAH 14 mod dis (4) vorable out­ C6-7 Fx no no 31,M fall HP DS. EDH 14 mod dis (4) 54,M te disability C-6 Fx no no MVA(e) none DS, SDH, MC 6 severe dis (31 C6-7 Fx-SL incomp yes 81,M AP HT/HP DI,IVH 9 vegetative (2) t vegetative C-7 Fx no no 24,M AS none EDH, SDH,MC 3 mod dis (4) C-7 Fx no no 50,M AP none DS. nSDH, MC 6 died (I) C7-T1 SL incomp yes 32,F MVA none DS, EDH 8 mod dis (4) * Patients are presented in descending order of cervical injury level, including 14 patients with occiput-C3 injuries and 10 with in­ :orel COrp., juries involving levels C-4 and below. Abbreviations: AP =auto versus pedestrian; AS =assault; comp =complete; Dr =diffuse brain SPSS (ver­ injury with punctate contusions; dis =disability; DS =diffuse swelling; EDH =epidural hematoma: Fx =fracture: HP =hypoxic insult; HT =hypotensive insult; inc =incomplete: IVH =intraventricular hemorrhage: MC =multiple contusions; MCS =motorcycle accident; III 3; devel­ mod =moderate; MVA(e) =ejected from vehicle during MVA; nSDH =nonevacuated subdural hematoma; Oc =occiput: rec =recov­ ity of Jeru­ ery; SAH =subarachnoid hemorrhage; SL =subluxation. yart.demon. square and 'as used for ,were used pine injury jury, and in five, and fractures without subluxations were seen in 10. and four by placement in a halo ring and vest. Of five 3.ter than 8. iThe injuries were first documented on standard three-view patients with unstable injuries in whom stabilization was radiographs in 18 patients and CT scans in six patients. In not performed, four died during the acute injury period two of the six patients, cervical injuries were not demon­ and one (an 81-year-old man) remained in a persistent veg­ strated on the original plain radiographs because the cer­ etative state; he underwent rigid collar therapy for man­ vicothoracic junction was not adequately visualized, in­ agement of his C6-7 fracture-subluxation. cluding one patient with a C7-Tl ligamentous injury and another with a C-7 spinous process fracture. The remaining Characteristics of Brain Injury I an initial four patients all underwent cervical CT scanning that S score of The most common findings in head CT scans among revealed nondisplaced lateral mass fractures at C-7 in two ury. Their the 24 patients with combined craniocervical traumatic patients, C-4 in one, and C-2 in another. Use of CT scan­ marized in injury were diffuse swelling in 13 patients and subarach­ : 39 years ning in these four patients was prompted by clinical suspi­ noid hemorrhage in 12 patients. Evacuation of an epidural cion of SCI in two patients and suspicious radiography­ Score was hematoma was required in three patients, evacuation of a nen. In 18 documented areas in two patients. All cervical injuries subdural hematoma was required in three, and multiple Were fIrst detected during the initial emergency room eval­ ~ss than or contusions were present in four (Table 1). l4). There uation except in one patient with a stable C-l ring fracture, Which was documented on the day following admission. )r sex dis­ Neurological Outcome The most frequently injured region was between the ~d cranio­ ts without occiput and C-3, demonstrated in 14 (58.3%) of 24 pa­ At a minimum of 6 months postinjury, good recovery or tients. A neurological deficit secondary to cervical injury moderate disability (GOS criteria) was demonstrated in Was found in 14 patients: three cases were complete and nine (37.5%) of 24 patients and eight (33.3%) had died. !l were incomplete SCIs. In 15 of 24 patients unstable Of the eight deaths, two patients died of causes related to .cervical spine injuries were determined, including II of upper cervical SCI within 24 hours of hospital admission, Jresent in ,the 14 cases in which high cervical (occiput-C3) injury three died of delayed SCI-related complications, and three :curred in \Vas documented. Ten patients with cervical instability. died primarily of severe brain injury. Thirteen (54.2%) of Underwent stabilization procedures, six by open fusion 24 patients suffered a hypotensive and/or hypoxic insult in uri!, 2002 '1. Neurosurg: Spine / Volume 96/ April, 2002 287 ~ i i L. T Holly, et al. kervical ~ the prehospital setting or within 24 hours of admission. Of Glasgow Outcome Scale in Relation to the 14 patients with incomplete or complete SCIs, nine Spinal Cord Injury experienced an early hypotensive or hypoxic insult. Early t~;::::=== hypotension or hypoxia occurred in two (22%) of nine pa­ tients with favorable neurological outcomes (GOS score 6 (J) ~---- of 4 or 5), but occurred in II (73%) of 15 patients with : c 5 '\IVA poor neurological outcomes (GOS scores of 1-3) (p < -Ql :;:; Iauto-pedestriar 0.05). All three patients with complete SCIs died. All lG 4 III Death 0. Hall three of these patients had sustained high cervical injuries, losw 0 3 OPVS/SD hypoxic insults, and severe brain injuries (GCS score < 8) -... Jassault Ql Iii1lGR/rvlD (Table 1 and Fig. 1). ' ..c 2 motorcycle E I~ 1other Mechanism ofInjury, GCS Score, and Relative Risk of ::::l 1 1vehicular-relate C Cen'ical Spine Injury 0 1 (MVA, auto­ non vehicular in The most common mechanism of injury was MVA, Complete Incomplete No SCI llfall, GSW, ass: which occurred in 131 patients (29.3%), followed by auto­ SCI SCI j •p < 0.01 fo mobile versus pedestrian accidents (20.6%) (Table 2). The FIG. 1. Bar graph showing GOS category in relation to SCI. Only remainder included motorcycle accidents, falls, assaults, one (4.2%) of 24 patients achieved a level of good recovery, and the and GSWs. Of the 24 patients with combined craniocer­ overall functional recovery rate (good recovery or moderate disabil­ vical injuries, 13 were passengers in MVAs, seven were ity) was 37.5%. GR = good recovery; MD = moderate disability; PVS =persistent vegetative state; SD =severe disability. pedestrians struck by vehicles, two were assaulted, one was j a motorcyclist who did not wear a helmet, and one suffered a fall. Thus, 21 (87,5%) of 24 patients sustained vehicular­ jmechanical related injuries, Patients with vehicular-related head trauma ling of head were significantly more likely to have sustained cervical I Two aspe l spine injury than those who suffered nonvehicular-related the present study.11.I3.l9.24 As we found, Hills and Deanne ] ,> compound tI trauma (8.2% compared with 1.6%, respectively; p < and Williams, et al.,24 have also shown an association be­ vical injury \ 0.01). In none of the 64 patients who suffered a cranial tween GCS score-related head injury severity and the risk ltion with i~, GSW was cervical spine injury documented. In Table 3, an of concomitant cervical injury. In their report of almost .severe cervll OR analysis demonstrates the relative risks of cervical 8300 trauma victims, Hills and Deanne also showed that J genic shock spine injury in MVAs compared with other mechanisms of head-injured patients had a significantly higher risk of cer­ Iterious impa injury. vical spine injury (4,5%) than those without head injury .! hypotensive The postresuscitation GCS score was also predictive (Ll %). In their study of more than 5000 trauma patients ;insults were of the likelihood of concomitant cervical injury, Eighteen Williams, et al.,24 however, found no association between .outcome in t (7.6%) of 238 patients with a GCS score less than or equal to head injury and cervical injury. Itained comI 8 sustained a cervical injury overall, compared with six In addition to low GCS score, the mechanism of injury iearly hypote (2.9%) of 209 patients with a GCS score greater than 8 (OR emerged as an important risk factor in cases of combined ;part attributa 2.77,95% CI LlI-7.73); additionally, 12 (5%) of 238 pa­ craniocervical injury in this study. Vehicular accidents (au­ ian SCI madt tients with a GCS score of less than or equal to 8 sustained a tomobile passenger, automobile-pedestrian, and motorcy­ j eight deaths cervical SCI compared with two (l %) of 209 patients with a cle accidents) accounted for 88% of the cervical injuries, ;plications fn GCS score greater than 8 (OR 5.5, 95% CI =1.22-24,85). with the highest rate of nearly 10% demonstrated in victims :potension ar of MVAs. Given the increasing availability of airbags and ;cerebral and improved restraint devices in automobiles, this high rate of Iyadvocatec brain and sp Discussion cervical injuries is hopefully declining. In contrast, victims of GSWs comprised nearly 15% of this head injury cohort Incidence and Risk Factors of Combined Craniocervical but none sustained a cervical spine injury. Similarly, Kaups Cervical Spil l6 Injurv and Davis found no evidence of cervical spine injury in a A survey t Sir Geoffery Jefferson first described the association be­ series of 202 cranial GSW victims and concluded that cer­ Conducted il tween head trauma and cervical spine injury in 1920. 21 He vical spine clearance could be abandoned in these patients. ;Surgery of 1 detennined that a sufficient axial load directed to the vertex Unless a GSW victim's neurological status or the mjecto­ Idiagnostic p of the skull may result in disruption of the atlas. Since that ry of the bullet suggests the likelihood of direct spinal llQtegrity afte time other authors have studied the relationship between injury, foregoing a cervical spine radiological evaluC!!ion IS ':esponded to head trauma and cervical spine injury, and the results have probably reasonable in such patients. itandard thrl varied. Buch?lz, et al,,3 perfonned postmortem radiography 'v ere obtaine and autopsy ll1spections in 100 consecutive victims of traf­ Importance of Cervical Injury Locatio/1 and Impact Oil i:he three-vie' fic accidents; they found a cervical spine injury rate of Neurological Outcome Unction if tn 24%. More recently, Eda, et aL 14 determined that in a series We found that the most frequently injured cervic2.1 are J :-ray films. ~ of 188 cervical spine injury patients, 35% had sustained a was between the occiput and C-3, which accounted f?! 'ormed flexi, moderate or severe head injury. 58% of the injuries, as has been reported by oth::;;·s.~ ,'adiography In contrast, in trauma patients with known head injury, The unique anatomical relationship between the ~ppeJ ! ~gs. Based ( the incidence of simultaneous cervical spine injury has gen­ two cervical vel1ebrae and the occiput appears to ::._::d!5­ ~ssociation erally ranged from 4 to 8%, which is similar to the rate in pose the upper cervical spine to rela,ively more -,,:;ve!'t ;'le2.:-ance C

288 1. Neurosurg: Spine I Volume 96 I Ap!'i! ]OO~;:. Neurosurg. :." IIS .._.-I.rtft.__ [oIly, et a; ~ervical spine trauma after head injury

TABLE 2 TABLE 3 Head injury arranged bv mechanism Odds ratios of selected injury mechanisms relative ro !\.IVAs" ::::======PercenI With Cervical Injury No, of Cervical Injury Mechanism Patients (%) (no. of cases) Present Absent Rate RallO 95% CI ------I!VA 131 (29,3) 9.9 MVA 13 118 referent ~to-pedestrian accident 92 (20.6) 7.6 pedestrian 7 85 0.77 0.32-1.85 II Death :dl 74 (16.6) 1.4 assault 2 46 0.42 0.10-1.79 DPVS/SD 0SW 64 (14.3) 0.0 motorcycle I 32 D.31 0.04-2.25 ;ssault 48 (10.7) 4.2 fall I 73 0.14 0.02-1.02 riilGR/MD ,notorcycle 33 (7.4) 3.0 • Examination of the confidence intervals shows no significant differ­ 5(1.1) 0.0 ,ther ences from the null value of I: however, all mechanisms are consistently :ehicular-related injuries 256 (57.3) 8.2* (21) reduced relative to the odds of cervical injury after an MVA. (MVA, auto-pedestrian. motorcycle) ,\onvehicular injuries 191 (42.7) 1.6 (3) :fail, GSW, assault, other) ~:::....=.:::...:....:....::::::..::..:...:..:.:....::..:...:..::..:...._------'" p < 0.0 I for vehicular- compared with non vehicular-related trauma. I to SCI. Only )very, and the mendation that only cervical radiography and CT scan­ lerate disabil­ ning were necessary to clear the cervical spine in patients lte disability; in whom altered mental status was present. The updated ty. 2000 recommendation includes the use of "flexion/exten­ sion lateral cervical spine t1uoroscopy with static images mechanical forces than the lower cervical spine in the set­ obtained at extremes of flexion and extension" if normal ting of head injury. findings are observed on the initial plain x-ray films and Two aspects of cervical spine injury may drastically CT scans. IS d Deanne lJ compound the effects of the head injury. First, upper cer­ Our present cervical spine clearance protocol in head­ Iciation be­ vical injury may result in loss of diaphragmatic innerva­ injured patients at UCLA and Harbor-UCLA Medical md the risk tion with inadequate respiratory effort or apnea. Second, Centers includes 1) plain radiographs; 2) additional CT of almost severe cervical and upper thoracic SCIs may cause neuro­ scans in cases of poorly visualized or suspicious areas lowed that oenic shock from loss of sympathetic outflow. 22 The dele­ ,~erious revealed on plain x-ray films or in cases in which there is risk of cer­ impact of such head injury-induced hypoxic and clinical suspicion of SCI; and 3) flexion-extension radio­ 6 lead injury !hypotensive secondary insults is well appreciated. These graphs if no abnormal findings are visualized on plain l1a patients insults were strongly associated with poor neurological radiographs and CT scans. Flexion-extension fluoroscopy ,n between outcome in the present series. Of the 14 patients who sus­ or cervical MR imaging is also used if plain radiogra­ tained complete or incomplete SCIs, nine suffered an phy or CT scanning demonstrates a nondisplaced cervical J of injury early hypotensive and/or hypoxic insult that was likely in fracture. Notably, in using flexion-extension studies for 6 combined part attributable to SCI. Only two of these 14 patients with years in more than 200 head-injured patients without ab­ idents (au­ Ian SCI made a GaS-rated good recovery, and five of the normal findings demonstrated on initial plain x-ray films l motorcy­ Ieight deaths in the series were directly attributable to com­ and CT scans, we have yet to identify any patients with an 11 injuries, plications from SCI. Rapidly identifying and treating hy­ unstable cervical injury. in victims .potension and hypoxia, as well as maintaining adequate In four recent reports with a total of 588 moderately or lrbags and cerebral and spinal cord perfusion as has been previous­ severely head injured patients, in whom either dynamic igh rate of ly advocated, likely improves the chances for functional cervical fluoroscopy or static plain radiography was per­ st, victims brain and spinal cord recovery.423 formed during passive flexion-extension movements, lig­ try cohort, amentous injuries were discovered in only three patients Cervical Spine Clearance After Head Injury rly, Kaups (0.5%).812.17,20 In two of the series (267 patients total), the njury in a A survey of United States and Canadian trauma centers authors did not diagnose a single patient with an unstable j that cer­ conducted in 1998 by the Eastern Association for the injury by using flexion-extension studies. 12.17 Of the three ~ patients. ·Surgery of Trauma showed substantial variability in the patients in whom ligamentous injuries were diagnosed, ~ trajecto­ diagnostic procedures used to determine cervical spine one was managed by a cervical fusion, one by halo vest ~ct spinal integrity after head injury. 18 In 24 of 31 trauma centers that therapy, and one by cervical collar therapy.8,20 In the Lluation is ·responded to the survey, clinicians routinely performed a largest and most recent of these studies, Davis and col­ standard three-view plain radiography; axial CT scans leaguesS established a diagnosis of unstable injury in two Were obtained through suspicious areas demonstrated on ct on of 301 patients, but neither patient required a cervical ·the three-view x-ray films and through the cervicothoracic fusion. Based on their experience, Davis and colleagues junction if this area was inadequately visualized on plain concluded that "the cervical spine may be cleared after Iical area X-ray films. Staff at only nine of 3J centers routinely per­ a normal cervical spine series (plain radiograph and Jilted for ,fanned flexion-extension radiography if the initial plain computed tomographic scan) as recommended in the 1998 others. 3,14 ,radiography and CT scanning demonstrated normal find­ Eastern Association for the Surgery of Trauma GUIde­ fJ.e upper Iings. Based on their literature review in 2000, the Eastern lines,"8 and thus rejected the routine use of flexion-exten­ o predis­ A.ssociation for the Surgery of Trauma Cervical Spine sion radiographic studies. 'e severe Clearance Committee revised their initial 1998 recom- . Although this controversy is unresolved, the results of

'ri!, 2002 1. Neurosurg: Spine / Volume 96/ April, 2002 289 L. T. Holly, et a1. Cervical spine traum:

these studies show that occult ligamentous cervical in­ roscopy should be considered in high-risk head-injured bedside fluoroscopy to el juries not demonstrated or suspected on initial plain x-ray patients (those with vehicular-related trauma and/or an ini­ trauma patients. J Traun films or thin-cut CT scans are rare, probably occurring in tial GCS score of :S 8) prior to cervical spine clearance. 'l. Shrago GG: Cervical spin fewer than one in 100 head-injured patients. The potential . rna. A review of 50 patieJ Nucl Med 188:670-673, danger of not diagnosing such an injury, however, is ad­ ·'1 Tator CH: Classification mittedly serious. In light of these recent reports and our References " .. logical presentation, in N own experience with flexion--extension studies, we plan to ., . . JT (eds): Neurotrauma. limit their routine use to only those high-risk patients 1. The AbbreVIated Injury Scale, 1990 ReVISIOn. Des Plaines, " 1059-1072 identified in our study - namely, those with an initial GCS IllInOIS: AssoClatlon for the Advancement of AutomotIve Medi- 23. Vale FL, Bums J, Jackson score of 8 or less and/or those involved in vehicular-relat-' CIne, 1990 aical treatment after aeutl 2. Bracken MB, Shepard MJ, Collins WF, et al: A randomized, eective pilot study to asse ed trauma. In total, 22 (92%) of 24 patients with cervical ~uscitation spine injury in the present study met at least one of these controlled trial of methylprednisolone or naloxone in the treat· and blood pre criteria and 14 (93%) of 15 patients with unstable cervical ment of acute spInal-cord InJury. Results of the Second Nation· 239-246 1997 al Acute Spinal Cord Injury Study. N Engl J Med 322: W'lliams'J Jehle D Cotti injuries met these criteria. Because analysis of our data 1405-1411. 1990 I , , suggested that patients with a GCS score greater than 8 3. Bucholz RW, Burkhead WZ, Graham W. et al: Occult cervical and with nonvehicular-related trauma have significantly spine injuries in fatal traffic accidents. J Trauma 19:768-791, lower probability of sustaining an unstable cervical spine 1979 injury, we believe that the practice of obtaining flexion­ 4. Bullock R. Chesnut RM, Clifton G. et al: Guidelines for the extension studies can be abandoned in patients who meet management of severe head injury. Brain Trauma Foundation. both these criteria and in whom cervical radiographs and Eur J Emerg Med 3: 109-127. 1996 CT scans demonstrate normal findings. 5. Chendrasekhar A, Moorman DW, Timberlake GA: An evalua· An important caveat is that flexion--extension radiogra­ tion of the effects of semirigid cervical collars iI, patients with I' phy fails to demonstrate the cervicothoracic junction in up severe closed head injury. Am Surg 7:604-606, 1998 . 2o 6. Chesnut RM, Marshall LF. Klauber MR, et al: The role of sec­ to 30% ofpatients. The use ofMR imaging as an alterna­ ondary brain injury in determining outcome from severe head ' tive method for assessing cervical soft-tissue injury in injury. J Trauma 34:216-222, 1993 I head-injured patients was recently described by D'Alise, et 7. D'Alise MD, Benzel EC, Hart BL: Magnetic resonance imag­ aU In 121 patients, 31 were found to have sustained injury ing evaluation of the cervical spine in the comatose or obtund· to the paravertebral ligaments, disc space, or osseous cervi­ ed trauma patient. J Neurosurg (Spine 1) 91:54-59.1999 cal spine; eight of these patients required surgical stabiliza­ 8. Davis JW, Kaups KL, Cunningham MA. et al: Routine evalua­ tion. The disadvantages of using MR imaging include the tion of the cervical spine in head-injured patients with dynamic I length of study time, patient-related transport difficulties in fluoroscopy: a reappraisal. J Trauma 50:1044-1047, 2001 cases of intubation, and interpreting what the images actu­ 9. Davis JW, Parks SN, Detlefs CL, et al: Clearing the cervical I ally indicate about cervical stability. spine in obtunded patients: the use of dynamic fluoroscopy. J' Trauma 39:435-438, 1995 Regardless of the method, timely assessment of cervical 10. Ferguson J, MardeJ SN, Beattie TF, et al: Cervical collars: a stability after head injury allows earlier removal of the col­ potential risk to the head-injured patient. Injury 24:454-456. lar. Otherwise, collars can cause skin breakdown, limited 1993 5 9 patient mobilization, and suboptimum pulmonary care . In 11. Frye G, Wolfe T, Knopp R, et al: Intracranial hemorrhage as a a series of 116 head-injured patients, Davis and coworkers9 predictor of occult cervical-spine fracture. Ann Emerg Mcd identified cervical decubitus ulcers in 44% of the patients, 23:797-80 I. 1994 particularly in those maintained in collars for over 5 days. 12. Harris OA, Moure FC, Chappell ET: Use of modified EAST Cervical collars may also cause jugular venous obstruction, practice parameters in clearing the cervical spine in the obtund­ with increases in jugular venous pressure and decreases in ed trauma patient: a prospective study. J Neurosurg 94: 413-414A, 2001 (Abstract) cerebral perfusion pressure. JO 13. Hills MM, Deanne SA: Head injury and facial injury: is ther; an increased risk of cervical spine injury? J Trauma 34:549-)54. 1993 Conclusions 14. Iida H, Tachibana S, Kitahara T, et al: Association of head trau, Approximately one in 20 patients with moderate and se­ ma with cervical spine injury, spinal cord injury, or both. J vere head injury will sustain a cervical spine injury, the Trauma 46:450-452, 1999 majority of which are mechanically unstable upper cervical 15. Jennett B. Bond M: Assessment of outcome after severe brain damage. Lancet 1:480-484, 1975 injuries that often involve an SCI. Patients involved in 16. Kaups KL, Davis JW: Patients with gun shot wounds to the vehicular-related accidents and those with an initial GCS head do not require cervical spine inunobilization and evaiLJ3' score of 8 or less are at highest risk for concomitant cervi­ tion. J Trauma 44:865-867, 1998 cal spine injury. Patients with at least one of these risk fac­ 17. Marciano FF. Apostolides PJ, Vishteh AG. et al: Cemcal spJDe tors accounted for 92% of all patients with combined cra­ management in severe, nonpeIietrating closed head injury: a niocervical injuries in this study. Thorough and efficient prospective study. Neurosurgery 41:740-741,1997 (AbstraCl' cervical spine evaluation is mandatory in these patients. 18. Marion D, Domeier R. Dunham CM, et al: (Cervical Spine The results of recent studies in head-injured patients indi­ Clearance Committee): Determination of cervical s;;:ne slll' cate that plain radiography in conjunction with additional bility in trauma patients. (Update of the 1997 EAS! lcasten' Association for the Surgery of Trauma] Cervical Sp:i1~ Cle,," thin-cut CT scanning will diagnose greater than 99% of WJ7 20 ance Document). 2000 cervical spine injuries. . Given the small possibility of 19. Michael DB, Guyot DR. DamlOdy \VR: Coincidence :,f 0 missing unstable ligamentous injUlies, however, lateral cer­ and cervical spin~ injury. J Neurotrauma 6:J77-18;. ;98 • vical flexion--extension radiography with or without fluo­ 20. Sees DW, Rodliguez Cruz LR. Flaherty SF. et a]: ":':iC use 0,

290 1. Neurosurg: Spine / Volume 96/ Apri/. ;()(!~ t1. Neurosurg: Spine / Vol L. T. Holly, et 2.L ~enical spine trauma after head injury l in high-risk head-injur", bedside nuoroscopy to evaluate the cervical spine in obtunded ular trauma as a predictor of cervical spme injury. Ann Emerg elated trauma and/or an ir" trauma patients. J Trauma 45:768-77 L 1998 Med 21:719-722,1992 cervical spine clearance. :1. S;wago GG: Cervical spine injuries: association with head trau­ ma. A review of 50 patients. Am J Roentgenol Radium Ther ?'lucl :vIed 188:670-673, 1973 ::. Tator CH: Classification of spinal cord injury based on neuro­ Ices logical presentation, in Narayan RK, Wilberger JE, Povlishock Manuscript received May 21, 200 I. IT (eds): Neurotrauma. New York: McGraw-Hill, 1996, pp Accepted in final fonn November 14.2001 e, 1990 Revision. Des Plaine, J059-1072 The authors received support from National Institutes of Neuro­ 'ancement of Automotive Medi~ ;). Vale FL, Bums J, Jackson AB, et al: Combined medical and sur­ logical Disorders and Stroke (Grant No. NS30308) and Centers for gical treatment after acute spinal cord injury: results of a pro-s­ Disease Control and Prevention lGrant No. R49/CCR903622). Hins WF, et al: A randomized. pective pilot srudy to assess the merits of aggressive medical re­ Address reprint requests to: Daniel F. Kelly. M.D., Division of solone or naloxone in the treat­ suscitation and blood pressure management. J Neurosurg 87: Neurosurgery, University of California at Los Angeles Medical I. Results of the Second Nation_ 239-246, 1997 Center, Room 18-219 NPl. Box 957039. Los Angeles, California Study. N Engl J iVIed 322: ~+. Williams J, Jehle D, Cottington E, et al: Head, facial and clavic­ 90095-7039. email: [email protected] raham W, et al: Occult cervical :idents. J Trauma 19:768-791.

In G, et al: Guidelines for the ury. Brain Trauma Foundation. . 1996 N, Timberlake GA: An evalua­ cervical collars in patients with Surg 7:604-606,1998 lber MR, et al: The role of sec­ ling outcome from severe head 1993 BL: Magnetic resonance imag­ line in the comatose or obtund­ : (Spine 1) 91:54-59,1999 lam MA, et al: Routine evalua­ i-injured patients with dynamic 'auma 50: 1044-1047,2001 :L, et al: Clearing the cervical use of dynamic fluoroscopy. J e TF. et al: Cervical collars: a :d patient. Injury 24:454--456, tl: Intracranial hemorrhage as a ne fracrure. Ann Emerg Med

,II ET: Use of modified EAST he cervical spine in the obtund­ tive study. J Neurosurg 94: ury and facial injury: is there an mjury? J Trauma 34:549-554,

et al: Association of head trau­ spinal cord injury. or both. J t of outcome after severe brain 75 with gun shot wounds to the ne immobilization and evalua­ /98 ishteh AG, et al: Cervical spine letrating closed head injury: a 'Y 41:740-741, 1997 (Abstract) m CM. et al: (Cervical Spwe tination of cervical spine sta­ late of the 1997 EAST [Eastern Trauma] Cervical Spine C!ear-

Jdy WR: Coincidence of head rotrauma 6:177-189.1989 . Flaherty SF, et al: The use ot

~ / Volume 96/ April, 2002 !. Neurosurg: Spine / Volume 96/ April, 2002 291