PTS 2014 PLENARY PAPER
Utility of magnetic resonance imaging in diagnosing cervical spine injury in children with severe traumatic brain injury David Qualls, BS, Jeffrey R. Leonard, MD, Martin Keller, MD, Jose Pineda, MD, and Julie C. Leonard, MD, MPH, Columbus, Ohio
Evaluation of children for cervical spine injuries (CSIs) after blunt trauma is complicated, particularly if the patient is unresponsive because of severe traumatic brain injury. Plain radiography and computed tomography (CT) are commonly used, but CT combined with magnetic resonance imaging (MRI) is still considered the gold standard in CSI detection. However, MRI is expensive and can delay cervical clearance. The purpose of this study is to determine the added benefit of MRI as an adjunct to CT in the clearance of children with severe head trauma. METHODS: We performed a retrospective chart review of pediatric head trauma patients admitted to the pediatric intensive care unit at St. Louis Children’s Hospital from 2002 to 2012. Patients who received both cervical spine CT and MRI and presented with a Glasgow Coma Scale score of 8 or lower were included in the study. Imaging was analyzed by two pediatric trauma subspecialists and classified as demonstrating ‘‘no injury,’’ ‘‘stable injury,’’ or ‘‘unstable injury.’’ Results were compared, and discrepancies between CT and MRI findings were noted. RESULTS: A total of 1,196 head-injured children were admitted to the pediatric intensive care unit between January 2002 and December 2012. Sixty-three children underwent CT and MRI and met Glasgow Coma Scale criteria. Seven children were identified with negative CT and positive MRI findings, but none of these injuries were considered unstable by our criteria. Five children were determined to have unstable injuries, and all were detected on CT. CONCLUSION: The results of this study suggest that MRI does not detect unstable CSIs in the setting of negative CT imaging. Given the limited patient population for this study, further and more extensive studies investigating the utility of MRI in the head-injured pediatric patient are warranted. (J Trauma Acute Care Surg. 2015;78: 1122Y1128. Copyright * 2015 Wolters Kluwer Health, Inc. All rights reserved.) LEVEL OF EVIDENCE: Diagnostic and care management study, level IV. KEY WORDS: Pediatric; cervical spine injury; magnetic resonance imaging; traumatic brain injury; cervical clearance. BACKGROUND:
BACKGROUND Cervical spine injuries (CSIs) in the pediatric population are rare, occurring in less than 1% of children experiencing blunt trauma; however, CSIs can have devastating consequences, including death or life-altering neurologic deficits.1,2 As a result, many children presenting to the emergency department after blunt trauma are evaluated for CSIs, a process referred to as ‘‘cervical spine clearance.’’ One principle of patient management in cervical spine clearance is the minimization of cervical spine imaging. Unnecessary imaging can be associated with complications in children and interferes with optimal care. A major concern is neck irradiation, which has been shown to be associated with malignancies in children; therefore, plain radiographs and Submitted: September 10, 2014, Revised: February 12, 2015, Accepted: February 24, 2015. From the Washington University in St. Louis School of Medicine (D.Q., J.R.L., M.K., J.P., J.C.L.) St. Louis, Missouri; St. Louis Children’s Hospital (M.K., J.P.), St. Louis, Missouri; Nationwide Children’s Hospital and the Ohio State University College of Medicine (J.R.L., J.C.L.), Columbus, Ohio. This study was presented at the 1st annual meeting of the Pediatric Trauma Society, November 14-15, 2014, in Chicago, Illinois. Address for reprints: Julie C. Leonard, MD, MPH, Nationwide Children’s Hospital and the Ohio State University College of Medicine 700 Children’s Drive, RB III, Columbus, OH 43205-2664; email: [email protected]. DOI: 10.1097/TA.0000000000000646
particularly computed tomography (CT) are avoided when possible.3,4 In patients without altered mental status, cervical spine clearance can be performed clinically by following guidelines that have been developed for both the adult5Y7 and pediatric8,9 populations. When a higher index of suspicion is noted but a clinical examination is still possible, plain radiographs have been shown to be highly sensitive for CSI and, thus, CT and associated radiation can be avoided.10 However, in cases of severe traumatic brain injury, patients are comatose, precluding a thorough neurologic examination to evaluate for focal deficits or physical examination for midline neck tenderness or torticollis. The incidence of CSI in comatose patients is higher than the general trauma population, approximately 8%.11 In addition, in children with altered mental status, cervical spine CT has been demonstrated to have greater sensitivity for axial spine injuries than plain radiographs.12 Given these circumstances, cervical spine CT is often preferred to plain radiography in evaluation of comatose children for CSI, despite the increased radiation. Many institutions now use magnetic resonance imaging (MRI) in addition to plain radiography and CT in the evaluation of patients with severe traumatic brain injury.13Y15 The combination of cervical spine CT and MRI is often considered the gold standard given the sensitivity to osseous injury provided by CT and sensitivity for soft tissue and ligamentous injury J Trauma Acute Care Surg Volume 78, Number 6
1122
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
provided by MRI. However, the efficacy of MRI in cervical spine clearance has recently been brought into question. Magnetic resonance imaging use is expensive and difficult to perform in unstable patients, thus delaying cervical spine clearance and increasing the duration of time bed-ridden children spend constrained in a cervical collar. Prolonged cervical collar use reduces respiratory capacity and leads to increased occurrence of pain, pressure ulceration, and tissue necrosis.16Y19 There are also inherent risks involved with transport to and from the site of MRI, and extended periods are spent away from the critical care unit. In recent years, improvements in CT have prompted many studies investigating the utility of MRI in detecting CSI in adult trauma patients. Results are conflicting, with some studies suggesting that MRI provides no additional clinically significant findings after normal CT.20Y25 Others indicate that MRI does detect clinically significant CSI missed by the combination of plain radiography and CT.13Y15,26,27 Most of these studies focus on adult populations, and there are few exclusively addressing the pediatric population. The paucity of evidence evaluating the benefit of MRI in the evaluation of children for CSI is important because pediatric CSIs are not the same as adult CSIs because of differences in both anatomy and mechanisms of injury.28Y31 Although there are emerging clinical guidelines for the evaluation of potential CSI in children, these guidelines stress the increased risk of CSI among children with altered mental status.8,9 For children who are obtunded or comatose, there is no general consensus regarding the appropriate imaging for cervical clearance.32 The purpose of this study is to determine the added benefit of MRI as an adjunct to CT in cervical spine clearance of children with severe traumatic brain injury.
METHODS Study Group The institutional review board at Washington University in St. Louis approved this retrospective cohort study. The trauma registry at St. Louis Children’s Hospital, a freestanding academic children’s hospital with Level 1 trauma certification, was queried to identify all children with severe traumatic brain injury admitted to the hospital from January 1, 2002, to December 31, 2012. Radiology databases were reviewed to identify patients who had received cervical spine MRI. Children who had received cervical spine MRI but had a history of previous CSI were
excluded. The remaining children with cervical spine MRI were assessed for Glasgow Coma Scale (GCS) at three points: admission to the ED, initial neurosurgery evaluation, and on admission to the pediatric intensive care unit (PICU). Those children demonstrating a GCS score higher than 8 at all three points were excluded. The radiology database was again reviewed to determine whether the remaining children had received cervical spine CT. Those receiving cervical spine CTwere included in the final study group.
Institutional Protocol During the period of this study, our institution followed the cervical spine evaluation protocol for pediatric blunt trauma patients with significant altered mental status or intubation detailed in Figure 1 Axial cervical spine CT was obtained for all children immediately on presentation. Cervical spine precautions were continued for children with normal CT. If mental status and intubation normalized within 72 hours, children were cleared clinically. Otherwise, MRI was obtained if intubation or altered mental status persisted for longer than 72 hours. If MRI was normal, the patient was cleared and cervical spine precautions were discontinued.
Data Analysis The cervical spine imaging and medical records of children who met inclusion criteria were reviewed for unstable CSI. An unstable injury was defined as an injury that resulted in a neurologic deficit localized to the cervical spinal cord, operative stabilization, halo placement, or cervical immobilization of 3 months or greater. Patient records, including follow-up appointments by the neurosurgery service, were reviewed to determine the duration of cervical immobilization. Presence of CSI on CT or MRI was determined by review of all imaging reports by a pediatric neurosurgeon (J.R.L.) and a pediatric emergency physician (J.C.L). The physicians determined whether the patient had a CSI, whether the injury was unstable, and which imaging modalities were able to detect the injuries seen. The institutional protocol for evaluating children with severe traumatic brain injury did not require plain radiography of the cervical spine. However, if patients underwent plain radiography, this imaging was evaluated in the same manner as CT and MRI and determined to demonstrate evidence of injury or no evidence of injury.
Figure 1. Institutional protocol for cervical spine evaluation in children with severe traumatic brain injury. * 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
1123
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
The charts of all included children were assessed for multiple other variables including age, sex, Injury Severity Score (ISS), Pediatric Trauma Score, mechanism of injury, neurologic condition at discharge, and any interventions performed.
Statistical Analysis Medians with ranges were calculated for all continuous variables. For categoric variables, frequencies and 95% confidence intervals (CIs) were calculated. To compare the use of MRI or CTalone with the combined approach of CT followed by MRI, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with CIs for MRI and CT were calculated using the MedCalc Diagnostic Test Calculator.33 In calculating these values, a positive screening was considered any finding on imaging, whereas negative screening was any imaging that did not demonstrate an abnormality. All other calculations were performed using Microsoft Excel.
RESULTS A total of 1,196 children with traumatic brain injury were admitted to the PICU at St. Louis Children’s Hospital from January 1, 2002, to December 31, 2012. Sixty-four children meeting GCS criteria received both cervical spine CT and MRI. One patient was excluded because of a previous CSI. The final study population consisted of 63 children (Fig. 2). Table 1 presents the characteristics of the study cohort. Of the 63 children in the study group, the median age was 9.6 years (range, 0.1Y17.8 years) and 33 (52.4%; CI, 40.3Y64.2%) were male. Similar GCS findings were recorded at the three time points, with a median GCS score at ED admission of 5 (range, 3Y13), at Neurosurgical Consult 6 (range, 3Y13), and at PICU admission 6 (range, 3Y14). Median ISS was 30 (range, 11Y75), and children spent a median of 13 days in the PICU (range, 2Y34 days). The most common mechanism of injury was motor vehicle accident, which accounted for 47.6% of patient injuries, followed by nonaccidental trauma, with 20.6% of injuries.
Table 2 presents the children who had positive findings on CT and/or MRI. Fifty-one children were determined to have a negative CT of the cervical spine. Of these 51 children, 7 had positive findings on MRI: 3 epidural hematomas, 2 soft tissue injuries, 1 ligamentous injury, and 1 unilateral facet dislocation. None of these patients had unstable CSIs. Of the 12 children with positive findings on CT, 5 were considered to have unstable CSIs. These consisted of an atlanto-occipital dislocation resulting in death, ligamentous injury requiring posterior wiring, type III odontoid fracture requiring c-collar stabilization greater than 3 months, type I odontoid fracture with ligamentous injury requiring c-collar use longer than 3 months, and C6-C7 transverse fracture with associated spinal cord injury. The one patient with spinal cord injury had localizing neurologic findings on presentation that would have prompted obtaining an MRI in addition to the CT scan. Using the results determined by physician reviewers, sensitivity and NPV of CT and MRI for unstable injury were calculated. CT was determined to have a sensitivity of 100% (CI, 48Y100%) and an NPV of 100% (CI, 93Y100%) for the unstable CSIs in our cohort. MRI sensitivity for unstable CSI was 80% (CI, 29Y97%), and the NPV was 98% (CI, 89Y100%), with MRI having had no positive findings for the patient with type III odontoid fracture. PPV and specificity were also calculated for CT and MRI in assessing for unstable CSI. CT was determined to have a specificity of 84.5% (CI, 73Y93%) and PPV of 35.7% (CI, 13Y65%) for unstable CSI. MRI had a specificity of 81% (CI, 69Y90%) and a PPV of 26.7% (CI, 8Y55%). Similar values were calculated reviewing the sensitivity and NPVof CT and MRI for any CSI, including those that were not considered unstable. CT was determined to have a sensitivity of 63.2% (CI, 38Y84%) and an NPV of 86.3% (CI, 74Y94%) for CSIs. MRI was 68.4% (CI, 43Y87%) sensitive, with an NPV of 88.0% (CI, 76Y95%), for all CSIs. Only 38% of our study population had adjunctive plain radiographs. Five patients in our cohort underwent single
Figure 2. Flowchart detailing patient inclusion. 1124
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
TABLE 1. Characteristics of Children Included in the Study Cohort Demographics
Total Study Group (n = 63) Unstable CSI Patients (n = 5) Stable CSI Patients (n = 14) Non-CSI Patients (n = 44)
Median age (range), y Male, n (%; 95% CI) Median ED GCS (range) Intubated in ED: n (%; 95% CI) Median Neurosurgical Consult GCS (range) Intubated at Consult: n (%; 95% CI) Median PICU GCS (range) Intubated in PICU: n (%; 95% CI) Median ISS (range) Median Pediatric Trauma Score (range) Median hospital days (range) Median PICU days (range) Mechanism of injury: n (%; 95% CI) MVC occupant Nonaccidental trauma Other motorized transport crash Pedestrian vs. automobile Fall from elevation
9.6 (0.1Y17.8) 33 (52.4; 40.3Y64.2) 5 (3Y13) 57 (90.5; 80.7Y95.6) 6 (3Y13) 61 (96.8; 89.2Y99.1) 6 (3Y14) 62 (98.4; 91.5Y99.7) 30 (11Y75) 3 (j3 to 10) 34 (5Y129) 13 (2Y34)
9.6 (3.4Y17.8) 2 (40.0; 11.8Y76.9) 3.0 (3Y6) 5 (100; 56.6Y100) 5 (3Y11) 5 (100; 56.6Y100) 6.0 (3Y7) 5 (100; 56.6Y100) 45 (17Y75) 0 (j1 to 2) 61.0 (7Y107) 23.0 (7Y31)
30 (47.6; 35.8Y59.7) 13 (20.6; 12.5Y32.2) 9 (14.3; 7.7Y25.0) 9 (14.3; 7.7Y25.0) 2 (3.2; 0.9Y10.9)
lateral views, and 19 underwent 2- or 3-view portable radiographs during the time between ED admission and MRI. Table 2 depicts the plain radiography findings for these patients. No patient had findings on plain radiography, which were undetected on CT and MRI. Three patients in the unstable injury group had at least single lateral films: the patient with type I odontoid fracture had lucency at the dens on plain films, whereas the patient with C6 + C7 transverse process fractures and patient with ligamentous and spinal cord injuries had no findings. Four patients in the stable injury group had plain radiographs. Two of these patients (diagnosed as having occipital condyle fracture and soft tissue injury) demonstrated prevertebral soft tissue swelling, whereas two others (C6 unilateral facet fracture and nondisplaced lateral mass fracture) had no findings.
DISCUSSION This study retrospectively reviewed all children from 2002 to 2012 presenting to a freestanding academic pediatric trauma center who had severe traumatic brain injury and who received both cervical CT and MRI. Charts were reviewed for evidence of unstable CSI, and imaging was reviewed to determine whether CT and MRI demonstrated evidence of these injuries. Of the 63 children included in our final study group, 5 were determined to have unstable CSIs. All of these injuries were detected on CT, and 4 of the 5 were detected on MRI. Seven children had negative CT imaging and subsequent positive findings on MRI, but none of these children were determined to have unstable CSI. These results suggest that, for the population examined in this study, axial cervical spine CT with reconstruction is sufficient to detect unstable CSI. Three other studies have investigated the efficacy of CT and MRI in pediatric cervical spine clearance, with varying results. Two of these studies, by Flynn et al.34 and Henry
4 (80.0; 37.6Y96.4) 0 (0; 0Y43.5) 1 (20.0; 3.6Y62.5) 0 (0; 0Y43.5) 0 (0; 0Y43.5)
5.4 (0.2Y16.6) 6 (42.9; 21.4Y67.4) 5 (3Y8) 11 (78.6; 52.4Y92.4) 6 (3Y13)
11.2 (0.1Y17.6) 25 (56.8; 42.2Y70.3) 4.5 (3Y13) 41 (93.2; 81.8Y97.7) 6 (3Y10)
12 (85.7; 60.1Y96.0) 5.5 (3Y14) 13 (92.9; 68.5Y98.7) 32 (16Y50) 4 (0Y10) 29.5 (5Y129) 13 (3Y27)
44 (100; 92.0Y100) 6 (3Y8) 44 (100; 92.0Y100) 29 (11Y54) 3 (j3 to 8) 33.5 (5Y117) 12.5 (2Y34)
6 (42.9; 21.4Y67.4) 5 (35.7; 16.3Y61.2) 0 (0.0; 0.0Y21.5) 2 (14.3; 4.0Y40.0) 1 (7.1; 1.3Y31.5)
20 (45.5; 31.7Y59.9) 8 (18.2; 9.5Y32.0) 8 (18.2; 9.5Y32.0) 7 (15.9; 7.9Y29.4) 1 (2.3; 0.4Y11.8)
et al.,35 investigated MRI in cervical clearance of the pediatric population, with both concluding that MRI provides valuable insight in the evaluation of the pediatric cervical spine. Flynn et al.34 evaluated ‘‘high-risk’’ patients with MRI, concluding that 31% of those who underwent MRI had findings that altered their care. Henry et al.35 specifically evaluated MRI and CT in the detection of CSI, determining the sensitivity of each modality to soft tissue and osseous injuries. MRI demonstrated 100% sensitivity to osseous and ligamentous injuries, whereas CT had poor sensitivity to ligamentous injuries. Gargas et al.24 performed a retrospective study on pediatric trauma patients, describing the cervical spine MRI findings in children with previous normal CT. Their study group was separated into two populations, an ‘‘early’’ group that was imaged with single-slice CTI and a ‘‘late’’ group imaged with a GE 64slice Lightspeeed VCT. Whereas five children with negative CT were found to have unstable CSIs in the early group, none were found in the late group, suggesting modern high-resolution CT was sufficiently sensitive to detect all unstable CSIs. Significant differences exist between those previously performed studies and our current study. Flynn et al.34 did not specifically design their study to investigate the efficacy of MRI in the setting of negative CT. Henry et al.35 investigated the sensitivity of these modalities to all injuries, not specifically clinically significant injuries. This study was also not limited to children with severe traumatic brain injury, as in our study, but instead investigated all pediatric trauma patients who underwent both CT and MRI. The retrospective study performed by Gargas et al.24 was most similar to our own study and demonstrated similar results. However, there are significant differences in study population and data gathered. Our study limited the study group to children with severe traumatic brain injury, whereas Gargas et al.24 included all trauma patients without GCS restriction. In addition, our study does not take into account the type of CT or MRI machine used. There may also be variability between
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
1125
1126
Mechanism of Injury
Female Occupant of motor vehicle crash Male Occupant of motor vehicle crash
Sex
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. Negative Negative Negative Negative
Negative Negative N/A N/A
Prevertebral soft tissue swelling (C3) N/A
Occipital condyle fracture
Negative
N/A N/A
16.1 Female Occupant of motor vehicle crash 16.6 Female Occupant of motor vehicle crash
Multilevel unilateral facet joint edema
C6 vertebral body fracture, Negative C6 laminar injury, C7 transverse process injury C5-C6 posterior ligamentous Injury Negative
N/A
Nondisplaced lateral mass fracture Negative
Negative
Occipital condyle fracture
Prevertebral soft tissue swelling N/A Negative
Occipital condyle fracture
C6 unilateral facet fracture
Prevertebral soft tissue swelling C2-C4 prevertebral soft tissue swelling Prevertebral soft tissue swelling FM-C4 Negative
C6-T1 posterior ligamentous injury Posterior epidural hematoma
Epidural hematoma C6-T2
C5 epidural blood
C6-T1 spinal cord injury
Negative
Negative
C2 body fracture, nondisplaced
Negative
Negative Negative
MRI Findings
Specific Injury
Treatment
Posterior ligamentous injury (C5-C6) Unilateral facet fracture/dislocation
Nondisplaced lateral mass fracture (C1) Vertebral body fracture, other (C6)
Occipital condyle fracture
Occipital condyle fracture
Unilateral facet fracture (C6)
Ligamentous injury without fracture (C6-T1) Posterior epidural hematoma (T1-T2) Prevertebral soft tissue swelling (C1-C4) Prevertebral soft tissue swelling (C2-C4) Vertebral body fracture, other (C2)
Epidural hematoma (C6-T2)
Transverse process fractures (C6 + C7) Epidural hematoma (C5)
Type III odontoid fracture
Yes Yes
Rigid collar 93 mo Rigid collar 93 mo
Yes Yes Yes
Yes Yes
Rigid collar 91 mo Rigid collar 91 mo
Rigid collar 91 mo Rigid collar 93 mo
Yes
Yes
Yes
None
None
None
Rigid collar 93 mo
Yes
Yes
Rigid collar 91 mo
None
Yes
None
Yes
No
Rigid collar 93 mo None
No
No
No
No
Stable Injury?
Rigid collar 93 mo
Type 1 odontoid fracture; posterior C1-C3 posterior ligamentous Type I odontoid fracture, ligamentous in- Rigid collar 93 mo ligamentous injury at C1-C2 injury; epidural blood jury without fracture (C1-C2) Widening of C7-T1 interspace Spinal cord injury Ligamentous injury Posterior wiring of (posterior ligament disruption) C6-T2, posterior without fracture (C7-T1) C7-T1 ligament disruption Atlanto-occipital dislocation Atlanto-occipital dislocation, Ligamentous injury Patient died before spinal cord injury C2 without fracture (C1-C2), management atlanto-occipital dislocation
CT Findings
Type 3 odontoid fracture and lateral mass fracture C6 + C7 transverse process fracture Negative
N/A
N/A
Subtle lucency of dens Negative
Plain Radiography Findings
15.6 Female Occupant of motor vehicle crash
Female Other motorized transport crash (ATV, 4-wheeler, motorcycle, etc.) 17 Female Occupant of motor vehicle crash 17.8 Male Occupant of motor vehicle crash 0.2 Female Nonaccidental trauma 0.2 Male Nonaccidental trauma 0.4 Male Nonaccidental trauma 1 Male Nonaccidental trauma 1.4 Female Nonaccidental trauma 2 Female Occupant of motor vehicle crash 4.3 Female Occupant of motor vehicle crash 6.4 Female Pedestrian hit by automobile 8.1 Male Occupant of motor vehicle crash 10.7 Male Pedestrian hit by automobile 13 Male Fall from elevation
9.6
6.1
3.4
Age, y
TABLE 2. Children With Severe Head Injury and Comorbid Injury Findings on Cervical Spine Imaging
Qualls et al. J Trauma Acute Care Surg Volume 78, Number 6
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
institutions regarding what constitutes a ‘‘clinically significant’’ CSI. These differences illustrate the need for a multicenter investigation of the efficacy of MRI in cervical clearance of the pediatric trauma patient population.
inform practice on the relative sensitivity of newer high-resolution CT machines or specific imaging protocols. A multicenter study with a larger population and more specific information on imaging parameters is necessary to gain greater insight into the utility of MRI in cervical clearance for this population.
Implications The results of this study suggest that CT alone is sufficient to rule out unstable CSI in children with severe traumatic brain injury, and MRI may not be necessary in these children. By clearing the cervical spine by CT only, unnecessary cost and significant complications associated with MRI use can be avoided. The need to obtain an MRI prolongs cervical collar use, which increases the risk of pressure ulcers and tissue necrosis, a major concern in traumatic head injury.16,18,19 There are also issues associated with obtaining MRI studies in critically ill children, including complications associated with prolonged transport out of the PICU and expense of the actual study. There is much debate concerning the use of CT scans in the pediatric population, given the documented increased risk for radiation-induced malignancies in children. Recent evidence suggests that CT scans of the cervical spine in children are associated with a significant risk of future cancer, with particular concern for the development of thyroid malignancy.3,4 Therefore, it is important to minimize the use of CT scanning in children, and CT-based protocols are discouraged. Most children who are evaluated after trauma can be cleared of CSI based on clinical evaluation alone. For those who have concerning findings, plain radiographs have been demonstrated to have a high sensitivity for CSI in children without altered mental status. For pediatric patients with severe traumatic brain injury, the incidence of CSI is substantially higher and CT has been shown to be more sensitive in the detection of axial CSI.10,12 Hannon et al.36 recently performed a study comparing the costs and benefits of cervical CT in pediatric patients with blunt trauma and determined that the frequency of CSI in a specific pediatric population must exceed 24.9% for the benefits of CT evaluation to outweigh future radiation risk. In our study group, 19 (30%) of 63 patients were found to have CSIs. Given the high risk of CSI in this specific patient population and the superior sensitivity of CT for CSI, the benefits of a CTbased protocol in detecting and ruling out CSI may outweigh the radiation-associated risks in this specific population. The consequences of untreated unstable CSIs may be devastating and can result in paralysis or death. This study suggests that CT provides adequate evaluation of unstable CSI in children with severe traumatic brain injury and avoids the negative aspects associated with MRI. Because of interinstitutional variation, further investigation in a multicenter trial is necessary to determine the appropriate imaging techniques for clearing children with severe traumatic brain injury of CSI.
Limitations This study is limited by its retrospective single-institution model as well as its small sample size. Results of imaging were determined through unblinded reviews, which may have been biased by previous imaging, therefore falsely inflating the sensitivity of certain modalities. Data on specific CT machines and imaging specifications were not recorded, so this study cannot
CONCLUSIONS We retrospectively reviewed the charts of all children with severe traumatic brain injury admitted to the PICU who received both cervical CT and MRI and assessed the radiology reviews to determine the presence of unstable CSIs and whether each modality detected such injuries. The results of this study suggest that, in the setting of a negative CT imaging, MRI does not detect additional unstable CSIs. Given the negative consequences associated with prolonged cervical collar use and the risks associated with MRI in critically ill children, a CT-based cervical spine clearance protocol may lead to improved hospital courses for children with severe traumatic brain injury without missing unstable CSIs. AUTHORSHIP J.R.L., M.K., J.P., and J.C.L. developed and implemented the institutional CSI evaluation protocol under investigation. D.Q., J.R.L., and J.C.L. designed the study. D.Q. performed literature review, chart review, collected and analyzed data, and wrote the first draft of the manuscript. J.R.L. and J.C.L. served as advisors throughout the research and writing process, analyzed the imaging to determine the injury status of patients, and reviewed and edited the manuscript. M.K. and J.P. reviewed and edited the manuscript.
ACKNOWLEDGMENT There was no funding received for this study.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR, Hoffman JR, NEXUS Group. A prospective multicenter study of cervical spine injury in children. Pediatrics. 2001;108(2):E20. 2. Leonard JR, Jaffe DM, Kuppermann N, Olsen CS, Leonard JC. Pediatric Emergency Care Applied Research Network (PECARN) Cervical Spine Study Group. Cervical spine injury patterns in children. Pediatrics. 2014;133(5):1179Y1188. 3. Berrington de Gonza´lez A, Mahesh M, Kim KP, Bhargavan M, Lewis R, Mettler F, Land C. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med. 2009; 169(22):2071Y2077. 4. Jimenez RR, Deguzman MA, Shiran S, Karrellas A, Lorenzo RL. CT versus plain radiographs for evaluation of c-spine injury in young children: do benefits outweigh risks? Pediatr Radiol. 2008;38(6):635Y644. 5. Stiell IG, Wells GA, Vandemheen KL, Clement CM, Lesiuk H, De Maio VJ, et al. The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA. 2001;286(15):1841Y1848. 6. Stiell IG, Clement CM, Grimshaw J, Brison RJ, Rowe BH, Schull MJ, et al. Implementation of the Canadian C-Spine Rule: prospective 12 centre cluster randomised trial. BMJ. 2009;339:b4146. 7. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National Emergency X-Radiography Utilization Study Group. N Engl J Med. 2000;343(2):94Y99.
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
1127
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
8. Jaffe DM, Binns H, Radkowski MA, Barthel MJ, Engelhard HH. Developing a clinical algorithm for early management of cervical spine injury in child trauma victims. Ann Emerg Med. 1987;16(3):270Y276. 9. Leonard JC, Kuppermann N, Olsen C, Babcock-Cimpello L, Brown K, Mahajan P, et al. Factors associated with cervical spine injury in children after blunt trauma. Ann Emerg Med. 2011;58(2):145Y155. 10. Nigrovic LE, Rogers AJ, Adelgais KM, Olsen CS, Leonard JR, Jaffe DM, et al. Utility of plain radiographs in detecting traumatic injuries of the cervical spine in children. Pediatr Emerg Care. 2012;28(5):426Y432. 11. Piatt JH Jr. Detected and overlooked cervical spine injury among comatose trauma patients: from the Pennsylvania Trauma Outcomes Study. Neurosurg Focus. 2005;19(4):E6. 12. Schenarts PJ, Diaz J, Kaiser C, Carrillo Y, Eddy V, Morris JA. Prospective comparison of admission computed tomographic scan and plain films of the upper cervical spine in trauma patients with altered mental status. J Trauma. 2001;51(4):663Y668. 13. Muchow RD, Resnick DK, Abdel MP, Munoz A, Anderson PA. Magnetic resonance imaging (MRI) in the clearance of the cervical spine in blunt trauma: a meta-analysis. J Trauma. 2008;64(1):179Y189. 14. D’Alise MD, Benzel EC, Hart BL. Magnetic resonance imaging evaluation of the cervical spine in the comatose or obtunded trauma patient. J Neurosurg. 1999;91(Suppl 1):54Y59. 15. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW, Espinosa MC. Magnetic resonance imaging for the evaluation of patients with occult cervical spine injury. J Neurosurg. 1996;85(5):824Y829. 16. March JA, Ausband SC, Brown LH. Changes in physical examination caused by use of spinal immobilization. Prehosp Emerg Care. 2002;6(4):421Y424. 17. Schafermeyer RW, Ribbeck BM, Gaskins J, Thomason S, Harlan M, Attkisson A. Respiratory effects of spinal immobilization in children. Ann Emerg Med. 1991;20(9):1017Y1019. 18. Linares HA, Mawson AR, Suarez E, Biundo JJ. Association between pressure sores and immobilization in the immediate post-injury period. Orthopedics. 1987;10(4):571Y573. 19. Chan D, Goldberg R, Tascone A, Harmon S, Chan L. The effect of spinal immobilization on healthy volunteers. Ann Emerg Med. 1994;23(1):48Y51. 20. Stelfox HT, Velmahos GC, Gettings E, Bigatello LM, Schmidt U. Computed tomography for early and safe discontinuation of cervical spine immobilization in obtunded multiply injured patients. J Trauma. 2007; 63(3):630Y636. 21. Tomycz ND, Chew BG, Chang YF, Darby JM, Gunn SR, Nicholas DH, et al. MRI is unnecessary to clear the cervical spine in obtunded/comatose trauma patients: the four-year experience of a level I trauma center. J Trauma. 2008;64(5):1258Y1263.
1128
22. Sanchez B, Waxman K, Jones T, Conner S, Chung R, Becerra S. Cervical spine clearance in blunt trauma: evaluation of a computed tomographybased protocol. J Trauma. 2005;59(1):179Y183. 23. Harris TJ, Blackmore CC, Mirza SK, Jurkovich GJ. Clearing the cervical spine in obtunded patients. Spine (Phila Pa 1976). 2008;33(14):1547Y1553. 24. Gargas J, Yaszay B, Kruk P, Bastrom T, Shellington D, Khanna S. An analysis of cervical spine magnetic resonance imaging findings after normal computed tomographic imaging findings in pediatric trauma patients: ten-year experience of a level I pediatric trauma center. J Trauma Acute Care Surg. 2013;74(4):1102Y1107. 25. Vanguri P, Young AJ, Weber WF, Katzen J, Han J, Wolfe LG, Duane TM. Computed tomographic scan: it’s not just about the fracture. J Trauma Acute Care Surg. 2014;77(4):604Y607. 26. Diaz JJ Jr, Aulino JM, Collier B, Roman C, May AK, Miller RS, et al. The early work-up for isolated ligamentous injury of the cervical spine: does computed tomography scan have a role? J Trauma. 2005;59(4):897Y903. 27. Kaiser ML, Whealon MD, Barrios C, Kong AP, Lekawa ME, Dolich MO. The current role of magnetic resonance imaging for diagnosing cervical spine injury in blunt trauma patients with negative computed tomography scan. Am Surg. 2012;78(10):1156Y1160. 28. Fesmire FM, Luten RC. The pediatric cervical spine: developmental anatomy and clinical aspects. J Emerg Med. 1989;7(2):133Y142. 29. Kokoska ER, Keller MS, Rallo MC, Weber TR. Characteristics of pediatric cervical spine injuries. J Pediatr Surg. 2001;36(1):100Y105. 30. Brown RL, Brunn MA, Garcia VF. Cervical spine injuries in children: a review of 103 patients treated consecutively at a level 1 pediatric trauma center. J Pediatr Surg. 2001;36(8):1107Y1114. 31. Hadley MN, Zabramski JM, Browner CM, Rekate H, Sonntag VK. Pediatric spinal trauma. Review of 122 cases of spinal cord and vertebral column injuries. J Neurosurg. 1988;68(1):18Y24. 32. Brockmeyer DL, Ragel BT, Kestle JR. The pediatric cervical spine instability study. A pilot study assessing the prognostic value of four imaging modalities in clearing the cervical spine for children with severe traumatic injuries. Childs Nerv Syst. 2012;28(5):699Y705. 33. MedCalc Diagnostic Test Calculator . Available at: http://www.medcalc.org/ calc/diagnostic_test.php2014. Accessed May 3, 2014. 34. Flynn JM, Closkey RF, Mahboubi S, Dormans JP. Role of magnetic resonance imaging in the assessment of pediatric cervical spine injuries. J Pediatr Orthop. 2002;22(5):573Y577. 35. Henry M, Riesenburger RI, Kryzanski J, Jea A, Hwang SW. A retrospective comparison of CT and MRI in detecting pediatric cervical spine injury. Childs Nerv Syst. 2013;29(8):1333Y1338. 36. Hannon M, Mannix R, Dorney K, Mooney D, Hennelly K. Pediatric cervical spine injury evaluation after blunt trauma: a clinical decision analysis. Ann Emerg Med. 2015;65(3):239Y247.
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Utility of magnetic resonance imaging in diagnosing cervical spine injury in children with severe traumatic brain injury David Qualls, BS, Jeffrey R. Leonard, MD, Martin Keller, MD, Jose Pineda, MD, and Julie C. Leonard, MD, MPH, Columbus, Ohio
Evaluation of children for cervical spine injuries (CSIs) after blunt trauma is complicated, particularly if the patient is unresponsive because of severe traumatic brain injury. Plain radiography and computed tomography (CT) are commonly used, but CT combined with magnetic resonance imaging (MRI) is still considered the gold standard in CSI detection. However, MRI is expensive and can delay cervical clearance. The purpose of this study is to determine the added benefit of MRI as an adjunct to CT in the clearance of children with severe head trauma. METHODS: We performed a retrospective chart review of pediatric head trauma patients admitted to the pediatric intensive care unit at St. Louis Children’s Hospital from 2002 to 2012. Patients who received both cervical spine CT and MRI and presented with a Glasgow Coma Scale score of 8 or lower were included in the study. Imaging was analyzed by two pediatric trauma subspecialists and classified as demonstrating ‘‘no injury,’’ ‘‘stable injury,’’ or ‘‘unstable injury.’’ Results were compared, and discrepancies between CT and MRI findings were noted. RESULTS: A total of 1,196 head-injured children were admitted to the pediatric intensive care unit between January 2002 and December 2012. Sixty-three children underwent CT and MRI and met Glasgow Coma Scale criteria. Seven children were identified with negative CT and positive MRI findings, but none of these injuries were considered unstable by our criteria. Five children were determined to have unstable injuries, and all were detected on CT. CONCLUSION: The results of this study suggest that MRI does not detect unstable CSIs in the setting of negative CT imaging. Given the limited patient population for this study, further and more extensive studies investigating the utility of MRI in the head-injured pediatric patient are warranted. (J Trauma Acute Care Surg. 2015;78: 1122Y1128. Copyright * 2015 Wolters Kluwer Health, Inc. All rights reserved.) LEVEL OF EVIDENCE: Diagnostic and care management study, level IV. KEY WORDS: Pediatric; cervical spine injury; magnetic resonance imaging; traumatic brain injury; cervical clearance. BACKGROUND:
BACKGROUND Cervical spine injuries (CSIs) in the pediatric population are rare, occurring in less than 1% of children experiencing blunt trauma; however, CSIs can have devastating consequences, including death or life-altering neurologic deficits.1,2 As a result, many children presenting to the emergency department after blunt trauma are evaluated for CSIs, a process referred to as ‘‘cervical spine clearance.’’ One principle of patient management in cervical spine clearance is the minimization of cervical spine imaging. Unnecessary imaging can be associated with complications in children and interferes with optimal care. A major concern is neck irradiation, which has been shown to be associated with malignancies in children; therefore, plain radiographs and Submitted: September 10, 2014, Revised: February 12, 2015, Accepted: February 24, 2015. From the Washington University in St. Louis School of Medicine (D.Q., J.R.L., M.K., J.P., J.C.L.) St. Louis, Missouri; St. Louis Children’s Hospital (M.K., J.P.), St. Louis, Missouri; Nationwide Children’s Hospital and the Ohio State University College of Medicine (J.R.L., J.C.L.), Columbus, Ohio. This study was presented at the 1st annual meeting of the Pediatric Trauma Society, November 14-15, 2014, in Chicago, Illinois. Address for reprints: Julie C. Leonard, MD, MPH, Nationwide Children’s Hospital and the Ohio State University College of Medicine 700 Children’s Drive, RB III, Columbus, OH 43205-2664; email: [email protected]. DOI: 10.1097/TA.0000000000000646
particularly computed tomography (CT) are avoided when possible.3,4 In patients without altered mental status, cervical spine clearance can be performed clinically by following guidelines that have been developed for both the adult5Y7 and pediatric8,9 populations. When a higher index of suspicion is noted but a clinical examination is still possible, plain radiographs have been shown to be highly sensitive for CSI and, thus, CT and associated radiation can be avoided.10 However, in cases of severe traumatic brain injury, patients are comatose, precluding a thorough neurologic examination to evaluate for focal deficits or physical examination for midline neck tenderness or torticollis. The incidence of CSI in comatose patients is higher than the general trauma population, approximately 8%.11 In addition, in children with altered mental status, cervical spine CT has been demonstrated to have greater sensitivity for axial spine injuries than plain radiographs.12 Given these circumstances, cervical spine CT is often preferred to plain radiography in evaluation of comatose children for CSI, despite the increased radiation. Many institutions now use magnetic resonance imaging (MRI) in addition to plain radiography and CT in the evaluation of patients with severe traumatic brain injury.13Y15 The combination of cervical spine CT and MRI is often considered the gold standard given the sensitivity to osseous injury provided by CT and sensitivity for soft tissue and ligamentous injury J Trauma Acute Care Surg Volume 78, Number 6
1122
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
provided by MRI. However, the efficacy of MRI in cervical spine clearance has recently been brought into question. Magnetic resonance imaging use is expensive and difficult to perform in unstable patients, thus delaying cervical spine clearance and increasing the duration of time bed-ridden children spend constrained in a cervical collar. Prolonged cervical collar use reduces respiratory capacity and leads to increased occurrence of pain, pressure ulceration, and tissue necrosis.16Y19 There are also inherent risks involved with transport to and from the site of MRI, and extended periods are spent away from the critical care unit. In recent years, improvements in CT have prompted many studies investigating the utility of MRI in detecting CSI in adult trauma patients. Results are conflicting, with some studies suggesting that MRI provides no additional clinically significant findings after normal CT.20Y25 Others indicate that MRI does detect clinically significant CSI missed by the combination of plain radiography and CT.13Y15,26,27 Most of these studies focus on adult populations, and there are few exclusively addressing the pediatric population. The paucity of evidence evaluating the benefit of MRI in the evaluation of children for CSI is important because pediatric CSIs are not the same as adult CSIs because of differences in both anatomy and mechanisms of injury.28Y31 Although there are emerging clinical guidelines for the evaluation of potential CSI in children, these guidelines stress the increased risk of CSI among children with altered mental status.8,9 For children who are obtunded or comatose, there is no general consensus regarding the appropriate imaging for cervical clearance.32 The purpose of this study is to determine the added benefit of MRI as an adjunct to CT in cervical spine clearance of children with severe traumatic brain injury.
METHODS Study Group The institutional review board at Washington University in St. Louis approved this retrospective cohort study. The trauma registry at St. Louis Children’s Hospital, a freestanding academic children’s hospital with Level 1 trauma certification, was queried to identify all children with severe traumatic brain injury admitted to the hospital from January 1, 2002, to December 31, 2012. Radiology databases were reviewed to identify patients who had received cervical spine MRI. Children who had received cervical spine MRI but had a history of previous CSI were
excluded. The remaining children with cervical spine MRI were assessed for Glasgow Coma Scale (GCS) at three points: admission to the ED, initial neurosurgery evaluation, and on admission to the pediatric intensive care unit (PICU). Those children demonstrating a GCS score higher than 8 at all three points were excluded. The radiology database was again reviewed to determine whether the remaining children had received cervical spine CT. Those receiving cervical spine CTwere included in the final study group.
Institutional Protocol During the period of this study, our institution followed the cervical spine evaluation protocol for pediatric blunt trauma patients with significant altered mental status or intubation detailed in Figure 1 Axial cervical spine CT was obtained for all children immediately on presentation. Cervical spine precautions were continued for children with normal CT. If mental status and intubation normalized within 72 hours, children were cleared clinically. Otherwise, MRI was obtained if intubation or altered mental status persisted for longer than 72 hours. If MRI was normal, the patient was cleared and cervical spine precautions were discontinued.
Data Analysis The cervical spine imaging and medical records of children who met inclusion criteria were reviewed for unstable CSI. An unstable injury was defined as an injury that resulted in a neurologic deficit localized to the cervical spinal cord, operative stabilization, halo placement, or cervical immobilization of 3 months or greater. Patient records, including follow-up appointments by the neurosurgery service, were reviewed to determine the duration of cervical immobilization. Presence of CSI on CT or MRI was determined by review of all imaging reports by a pediatric neurosurgeon (J.R.L.) and a pediatric emergency physician (J.C.L). The physicians determined whether the patient had a CSI, whether the injury was unstable, and which imaging modalities were able to detect the injuries seen. The institutional protocol for evaluating children with severe traumatic brain injury did not require plain radiography of the cervical spine. However, if patients underwent plain radiography, this imaging was evaluated in the same manner as CT and MRI and determined to demonstrate evidence of injury or no evidence of injury.
Figure 1. Institutional protocol for cervical spine evaluation in children with severe traumatic brain injury. * 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
1123
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
The charts of all included children were assessed for multiple other variables including age, sex, Injury Severity Score (ISS), Pediatric Trauma Score, mechanism of injury, neurologic condition at discharge, and any interventions performed.
Statistical Analysis Medians with ranges were calculated for all continuous variables. For categoric variables, frequencies and 95% confidence intervals (CIs) were calculated. To compare the use of MRI or CTalone with the combined approach of CT followed by MRI, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with CIs for MRI and CT were calculated using the MedCalc Diagnostic Test Calculator.33 In calculating these values, a positive screening was considered any finding on imaging, whereas negative screening was any imaging that did not demonstrate an abnormality. All other calculations were performed using Microsoft Excel.
RESULTS A total of 1,196 children with traumatic brain injury were admitted to the PICU at St. Louis Children’s Hospital from January 1, 2002, to December 31, 2012. Sixty-four children meeting GCS criteria received both cervical spine CT and MRI. One patient was excluded because of a previous CSI. The final study population consisted of 63 children (Fig. 2). Table 1 presents the characteristics of the study cohort. Of the 63 children in the study group, the median age was 9.6 years (range, 0.1Y17.8 years) and 33 (52.4%; CI, 40.3Y64.2%) were male. Similar GCS findings were recorded at the three time points, with a median GCS score at ED admission of 5 (range, 3Y13), at Neurosurgical Consult 6 (range, 3Y13), and at PICU admission 6 (range, 3Y14). Median ISS was 30 (range, 11Y75), and children spent a median of 13 days in the PICU (range, 2Y34 days). The most common mechanism of injury was motor vehicle accident, which accounted for 47.6% of patient injuries, followed by nonaccidental trauma, with 20.6% of injuries.
Table 2 presents the children who had positive findings on CT and/or MRI. Fifty-one children were determined to have a negative CT of the cervical spine. Of these 51 children, 7 had positive findings on MRI: 3 epidural hematomas, 2 soft tissue injuries, 1 ligamentous injury, and 1 unilateral facet dislocation. None of these patients had unstable CSIs. Of the 12 children with positive findings on CT, 5 were considered to have unstable CSIs. These consisted of an atlanto-occipital dislocation resulting in death, ligamentous injury requiring posterior wiring, type III odontoid fracture requiring c-collar stabilization greater than 3 months, type I odontoid fracture with ligamentous injury requiring c-collar use longer than 3 months, and C6-C7 transverse fracture with associated spinal cord injury. The one patient with spinal cord injury had localizing neurologic findings on presentation that would have prompted obtaining an MRI in addition to the CT scan. Using the results determined by physician reviewers, sensitivity and NPV of CT and MRI for unstable injury were calculated. CT was determined to have a sensitivity of 100% (CI, 48Y100%) and an NPV of 100% (CI, 93Y100%) for the unstable CSIs in our cohort. MRI sensitivity for unstable CSI was 80% (CI, 29Y97%), and the NPV was 98% (CI, 89Y100%), with MRI having had no positive findings for the patient with type III odontoid fracture. PPV and specificity were also calculated for CT and MRI in assessing for unstable CSI. CT was determined to have a specificity of 84.5% (CI, 73Y93%) and PPV of 35.7% (CI, 13Y65%) for unstable CSI. MRI had a specificity of 81% (CI, 69Y90%) and a PPV of 26.7% (CI, 8Y55%). Similar values were calculated reviewing the sensitivity and NPVof CT and MRI for any CSI, including those that were not considered unstable. CT was determined to have a sensitivity of 63.2% (CI, 38Y84%) and an NPV of 86.3% (CI, 74Y94%) for CSIs. MRI was 68.4% (CI, 43Y87%) sensitive, with an NPV of 88.0% (CI, 76Y95%), for all CSIs. Only 38% of our study population had adjunctive plain radiographs. Five patients in our cohort underwent single
Figure 2. Flowchart detailing patient inclusion. 1124
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
TABLE 1. Characteristics of Children Included in the Study Cohort Demographics
Total Study Group (n = 63) Unstable CSI Patients (n = 5) Stable CSI Patients (n = 14) Non-CSI Patients (n = 44)
Median age (range), y Male, n (%; 95% CI) Median ED GCS (range) Intubated in ED: n (%; 95% CI) Median Neurosurgical Consult GCS (range) Intubated at Consult: n (%; 95% CI) Median PICU GCS (range) Intubated in PICU: n (%; 95% CI) Median ISS (range) Median Pediatric Trauma Score (range) Median hospital days (range) Median PICU days (range) Mechanism of injury: n (%; 95% CI) MVC occupant Nonaccidental trauma Other motorized transport crash Pedestrian vs. automobile Fall from elevation
9.6 (0.1Y17.8) 33 (52.4; 40.3Y64.2) 5 (3Y13) 57 (90.5; 80.7Y95.6) 6 (3Y13) 61 (96.8; 89.2Y99.1) 6 (3Y14) 62 (98.4; 91.5Y99.7) 30 (11Y75) 3 (j3 to 10) 34 (5Y129) 13 (2Y34)
9.6 (3.4Y17.8) 2 (40.0; 11.8Y76.9) 3.0 (3Y6) 5 (100; 56.6Y100) 5 (3Y11) 5 (100; 56.6Y100) 6.0 (3Y7) 5 (100; 56.6Y100) 45 (17Y75) 0 (j1 to 2) 61.0 (7Y107) 23.0 (7Y31)
30 (47.6; 35.8Y59.7) 13 (20.6; 12.5Y32.2) 9 (14.3; 7.7Y25.0) 9 (14.3; 7.7Y25.0) 2 (3.2; 0.9Y10.9)
lateral views, and 19 underwent 2- or 3-view portable radiographs during the time between ED admission and MRI. Table 2 depicts the plain radiography findings for these patients. No patient had findings on plain radiography, which were undetected on CT and MRI. Three patients in the unstable injury group had at least single lateral films: the patient with type I odontoid fracture had lucency at the dens on plain films, whereas the patient with C6 + C7 transverse process fractures and patient with ligamentous and spinal cord injuries had no findings. Four patients in the stable injury group had plain radiographs. Two of these patients (diagnosed as having occipital condyle fracture and soft tissue injury) demonstrated prevertebral soft tissue swelling, whereas two others (C6 unilateral facet fracture and nondisplaced lateral mass fracture) had no findings.
DISCUSSION This study retrospectively reviewed all children from 2002 to 2012 presenting to a freestanding academic pediatric trauma center who had severe traumatic brain injury and who received both cervical CT and MRI. Charts were reviewed for evidence of unstable CSI, and imaging was reviewed to determine whether CT and MRI demonstrated evidence of these injuries. Of the 63 children included in our final study group, 5 were determined to have unstable CSIs. All of these injuries were detected on CT, and 4 of the 5 were detected on MRI. Seven children had negative CT imaging and subsequent positive findings on MRI, but none of these children were determined to have unstable CSI. These results suggest that, for the population examined in this study, axial cervical spine CT with reconstruction is sufficient to detect unstable CSI. Three other studies have investigated the efficacy of CT and MRI in pediatric cervical spine clearance, with varying results. Two of these studies, by Flynn et al.34 and Henry
4 (80.0; 37.6Y96.4) 0 (0; 0Y43.5) 1 (20.0; 3.6Y62.5) 0 (0; 0Y43.5) 0 (0; 0Y43.5)
5.4 (0.2Y16.6) 6 (42.9; 21.4Y67.4) 5 (3Y8) 11 (78.6; 52.4Y92.4) 6 (3Y13)
11.2 (0.1Y17.6) 25 (56.8; 42.2Y70.3) 4.5 (3Y13) 41 (93.2; 81.8Y97.7) 6 (3Y10)
12 (85.7; 60.1Y96.0) 5.5 (3Y14) 13 (92.9; 68.5Y98.7) 32 (16Y50) 4 (0Y10) 29.5 (5Y129) 13 (3Y27)
44 (100; 92.0Y100) 6 (3Y8) 44 (100; 92.0Y100) 29 (11Y54) 3 (j3 to 8) 33.5 (5Y117) 12.5 (2Y34)
6 (42.9; 21.4Y67.4) 5 (35.7; 16.3Y61.2) 0 (0.0; 0.0Y21.5) 2 (14.3; 4.0Y40.0) 1 (7.1; 1.3Y31.5)
20 (45.5; 31.7Y59.9) 8 (18.2; 9.5Y32.0) 8 (18.2; 9.5Y32.0) 7 (15.9; 7.9Y29.4) 1 (2.3; 0.4Y11.8)
et al.,35 investigated MRI in cervical clearance of the pediatric population, with both concluding that MRI provides valuable insight in the evaluation of the pediatric cervical spine. Flynn et al.34 evaluated ‘‘high-risk’’ patients with MRI, concluding that 31% of those who underwent MRI had findings that altered their care. Henry et al.35 specifically evaluated MRI and CT in the detection of CSI, determining the sensitivity of each modality to soft tissue and osseous injuries. MRI demonstrated 100% sensitivity to osseous and ligamentous injuries, whereas CT had poor sensitivity to ligamentous injuries. Gargas et al.24 performed a retrospective study on pediatric trauma patients, describing the cervical spine MRI findings in children with previous normal CT. Their study group was separated into two populations, an ‘‘early’’ group that was imaged with single-slice CTI and a ‘‘late’’ group imaged with a GE 64slice Lightspeeed VCT. Whereas five children with negative CT were found to have unstable CSIs in the early group, none were found in the late group, suggesting modern high-resolution CT was sufficiently sensitive to detect all unstable CSIs. Significant differences exist between those previously performed studies and our current study. Flynn et al.34 did not specifically design their study to investigate the efficacy of MRI in the setting of negative CT. Henry et al.35 investigated the sensitivity of these modalities to all injuries, not specifically clinically significant injuries. This study was also not limited to children with severe traumatic brain injury, as in our study, but instead investigated all pediatric trauma patients who underwent both CT and MRI. The retrospective study performed by Gargas et al.24 was most similar to our own study and demonstrated similar results. However, there are significant differences in study population and data gathered. Our study limited the study group to children with severe traumatic brain injury, whereas Gargas et al.24 included all trauma patients without GCS restriction. In addition, our study does not take into account the type of CT or MRI machine used. There may also be variability between
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
1125
1126
Mechanism of Injury
Female Occupant of motor vehicle crash Male Occupant of motor vehicle crash
Sex
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. Negative Negative Negative Negative
Negative Negative N/A N/A
Prevertebral soft tissue swelling (C3) N/A
Occipital condyle fracture
Negative
N/A N/A
16.1 Female Occupant of motor vehicle crash 16.6 Female Occupant of motor vehicle crash
Multilevel unilateral facet joint edema
C6 vertebral body fracture, Negative C6 laminar injury, C7 transverse process injury C5-C6 posterior ligamentous Injury Negative
N/A
Nondisplaced lateral mass fracture Negative
Negative
Occipital condyle fracture
Prevertebral soft tissue swelling N/A Negative
Occipital condyle fracture
C6 unilateral facet fracture
Prevertebral soft tissue swelling C2-C4 prevertebral soft tissue swelling Prevertebral soft tissue swelling FM-C4 Negative
C6-T1 posterior ligamentous injury Posterior epidural hematoma
Epidural hematoma C6-T2
C5 epidural blood
C6-T1 spinal cord injury
Negative
Negative
C2 body fracture, nondisplaced
Negative
Negative Negative
MRI Findings
Specific Injury
Treatment
Posterior ligamentous injury (C5-C6) Unilateral facet fracture/dislocation
Nondisplaced lateral mass fracture (C1) Vertebral body fracture, other (C6)
Occipital condyle fracture
Occipital condyle fracture
Unilateral facet fracture (C6)
Ligamentous injury without fracture (C6-T1) Posterior epidural hematoma (T1-T2) Prevertebral soft tissue swelling (C1-C4) Prevertebral soft tissue swelling (C2-C4) Vertebral body fracture, other (C2)
Epidural hematoma (C6-T2)
Transverse process fractures (C6 + C7) Epidural hematoma (C5)
Type III odontoid fracture
Yes Yes
Rigid collar 93 mo Rigid collar 93 mo
Yes Yes Yes
Yes Yes
Rigid collar 91 mo Rigid collar 91 mo
Rigid collar 91 mo Rigid collar 93 mo
Yes
Yes
Yes
None
None
None
Rigid collar 93 mo
Yes
Yes
Rigid collar 91 mo
None
Yes
None
Yes
No
Rigid collar 93 mo None
No
No
No
No
Stable Injury?
Rigid collar 93 mo
Type 1 odontoid fracture; posterior C1-C3 posterior ligamentous Type I odontoid fracture, ligamentous in- Rigid collar 93 mo ligamentous injury at C1-C2 injury; epidural blood jury without fracture (C1-C2) Widening of C7-T1 interspace Spinal cord injury Ligamentous injury Posterior wiring of (posterior ligament disruption) C6-T2, posterior without fracture (C7-T1) C7-T1 ligament disruption Atlanto-occipital dislocation Atlanto-occipital dislocation, Ligamentous injury Patient died before spinal cord injury C2 without fracture (C1-C2), management atlanto-occipital dislocation
CT Findings
Type 3 odontoid fracture and lateral mass fracture C6 + C7 transverse process fracture Negative
N/A
N/A
Subtle lucency of dens Negative
Plain Radiography Findings
15.6 Female Occupant of motor vehicle crash
Female Other motorized transport crash (ATV, 4-wheeler, motorcycle, etc.) 17 Female Occupant of motor vehicle crash 17.8 Male Occupant of motor vehicle crash 0.2 Female Nonaccidental trauma 0.2 Male Nonaccidental trauma 0.4 Male Nonaccidental trauma 1 Male Nonaccidental trauma 1.4 Female Nonaccidental trauma 2 Female Occupant of motor vehicle crash 4.3 Female Occupant of motor vehicle crash 6.4 Female Pedestrian hit by automobile 8.1 Male Occupant of motor vehicle crash 10.7 Male Pedestrian hit by automobile 13 Male Fall from elevation
9.6
6.1
3.4
Age, y
TABLE 2. Children With Severe Head Injury and Comorbid Injury Findings on Cervical Spine Imaging
Qualls et al. J Trauma Acute Care Surg Volume 78, Number 6
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
institutions regarding what constitutes a ‘‘clinically significant’’ CSI. These differences illustrate the need for a multicenter investigation of the efficacy of MRI in cervical clearance of the pediatric trauma patient population.
inform practice on the relative sensitivity of newer high-resolution CT machines or specific imaging protocols. A multicenter study with a larger population and more specific information on imaging parameters is necessary to gain greater insight into the utility of MRI in cervical clearance for this population.
Implications The results of this study suggest that CT alone is sufficient to rule out unstable CSI in children with severe traumatic brain injury, and MRI may not be necessary in these children. By clearing the cervical spine by CT only, unnecessary cost and significant complications associated with MRI use can be avoided. The need to obtain an MRI prolongs cervical collar use, which increases the risk of pressure ulcers and tissue necrosis, a major concern in traumatic head injury.16,18,19 There are also issues associated with obtaining MRI studies in critically ill children, including complications associated with prolonged transport out of the PICU and expense of the actual study. There is much debate concerning the use of CT scans in the pediatric population, given the documented increased risk for radiation-induced malignancies in children. Recent evidence suggests that CT scans of the cervical spine in children are associated with a significant risk of future cancer, with particular concern for the development of thyroid malignancy.3,4 Therefore, it is important to minimize the use of CT scanning in children, and CT-based protocols are discouraged. Most children who are evaluated after trauma can be cleared of CSI based on clinical evaluation alone. For those who have concerning findings, plain radiographs have been demonstrated to have a high sensitivity for CSI in children without altered mental status. For pediatric patients with severe traumatic brain injury, the incidence of CSI is substantially higher and CT has been shown to be more sensitive in the detection of axial CSI.10,12 Hannon et al.36 recently performed a study comparing the costs and benefits of cervical CT in pediatric patients with blunt trauma and determined that the frequency of CSI in a specific pediatric population must exceed 24.9% for the benefits of CT evaluation to outweigh future radiation risk. In our study group, 19 (30%) of 63 patients were found to have CSIs. Given the high risk of CSI in this specific patient population and the superior sensitivity of CT for CSI, the benefits of a CTbased protocol in detecting and ruling out CSI may outweigh the radiation-associated risks in this specific population. The consequences of untreated unstable CSIs may be devastating and can result in paralysis or death. This study suggests that CT provides adequate evaluation of unstable CSI in children with severe traumatic brain injury and avoids the negative aspects associated with MRI. Because of interinstitutional variation, further investigation in a multicenter trial is necessary to determine the appropriate imaging techniques for clearing children with severe traumatic brain injury of CSI.
Limitations This study is limited by its retrospective single-institution model as well as its small sample size. Results of imaging were determined through unblinded reviews, which may have been biased by previous imaging, therefore falsely inflating the sensitivity of certain modalities. Data on specific CT machines and imaging specifications were not recorded, so this study cannot
CONCLUSIONS We retrospectively reviewed the charts of all children with severe traumatic brain injury admitted to the PICU who received both cervical CT and MRI and assessed the radiology reviews to determine the presence of unstable CSIs and whether each modality detected such injuries. The results of this study suggest that, in the setting of a negative CT imaging, MRI does not detect additional unstable CSIs. Given the negative consequences associated with prolonged cervical collar use and the risks associated with MRI in critically ill children, a CT-based cervical spine clearance protocol may lead to improved hospital courses for children with severe traumatic brain injury without missing unstable CSIs. AUTHORSHIP J.R.L., M.K., J.P., and J.C.L. developed and implemented the institutional CSI evaluation protocol under investigation. D.Q., J.R.L., and J.C.L. designed the study. D.Q. performed literature review, chart review, collected and analyzed data, and wrote the first draft of the manuscript. J.R.L. and J.C.L. served as advisors throughout the research and writing process, analyzed the imaging to determine the injury status of patients, and reviewed and edited the manuscript. M.K. and J.P. reviewed and edited the manuscript.
ACKNOWLEDGMENT There was no funding received for this study.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR, Hoffman JR, NEXUS Group. A prospective multicenter study of cervical spine injury in children. Pediatrics. 2001;108(2):E20. 2. Leonard JR, Jaffe DM, Kuppermann N, Olsen CS, Leonard JC. Pediatric Emergency Care Applied Research Network (PECARN) Cervical Spine Study Group. Cervical spine injury patterns in children. Pediatrics. 2014;133(5):1179Y1188. 3. Berrington de Gonza´lez A, Mahesh M, Kim KP, Bhargavan M, Lewis R, Mettler F, Land C. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med. 2009; 169(22):2071Y2077. 4. Jimenez RR, Deguzman MA, Shiran S, Karrellas A, Lorenzo RL. CT versus plain radiographs for evaluation of c-spine injury in young children: do benefits outweigh risks? Pediatr Radiol. 2008;38(6):635Y644. 5. Stiell IG, Wells GA, Vandemheen KL, Clement CM, Lesiuk H, De Maio VJ, et al. The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA. 2001;286(15):1841Y1848. 6. Stiell IG, Clement CM, Grimshaw J, Brison RJ, Rowe BH, Schull MJ, et al. Implementation of the Canadian C-Spine Rule: prospective 12 centre cluster randomised trial. BMJ. 2009;339:b4146. 7. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National Emergency X-Radiography Utilization Study Group. N Engl J Med. 2000;343(2):94Y99.
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
1127
J Trauma Acute Care Surg Volume 78, Number 6
Qualls et al.
8. Jaffe DM, Binns H, Radkowski MA, Barthel MJ, Engelhard HH. Developing a clinical algorithm for early management of cervical spine injury in child trauma victims. Ann Emerg Med. 1987;16(3):270Y276. 9. Leonard JC, Kuppermann N, Olsen C, Babcock-Cimpello L, Brown K, Mahajan P, et al. Factors associated with cervical spine injury in children after blunt trauma. Ann Emerg Med. 2011;58(2):145Y155. 10. Nigrovic LE, Rogers AJ, Adelgais KM, Olsen CS, Leonard JR, Jaffe DM, et al. Utility of plain radiographs in detecting traumatic injuries of the cervical spine in children. Pediatr Emerg Care. 2012;28(5):426Y432. 11. Piatt JH Jr. Detected and overlooked cervical spine injury among comatose trauma patients: from the Pennsylvania Trauma Outcomes Study. Neurosurg Focus. 2005;19(4):E6. 12. Schenarts PJ, Diaz J, Kaiser C, Carrillo Y, Eddy V, Morris JA. Prospective comparison of admission computed tomographic scan and plain films of the upper cervical spine in trauma patients with altered mental status. J Trauma. 2001;51(4):663Y668. 13. Muchow RD, Resnick DK, Abdel MP, Munoz A, Anderson PA. Magnetic resonance imaging (MRI) in the clearance of the cervical spine in blunt trauma: a meta-analysis. J Trauma. 2008;64(1):179Y189. 14. D’Alise MD, Benzel EC, Hart BL. Magnetic resonance imaging evaluation of the cervical spine in the comatose or obtunded trauma patient. J Neurosurg. 1999;91(Suppl 1):54Y59. 15. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW, Espinosa MC. Magnetic resonance imaging for the evaluation of patients with occult cervical spine injury. J Neurosurg. 1996;85(5):824Y829. 16. March JA, Ausband SC, Brown LH. Changes in physical examination caused by use of spinal immobilization. Prehosp Emerg Care. 2002;6(4):421Y424. 17. Schafermeyer RW, Ribbeck BM, Gaskins J, Thomason S, Harlan M, Attkisson A. Respiratory effects of spinal immobilization in children. Ann Emerg Med. 1991;20(9):1017Y1019. 18. Linares HA, Mawson AR, Suarez E, Biundo JJ. Association between pressure sores and immobilization in the immediate post-injury period. Orthopedics. 1987;10(4):571Y573. 19. Chan D, Goldberg R, Tascone A, Harmon S, Chan L. The effect of spinal immobilization on healthy volunteers. Ann Emerg Med. 1994;23(1):48Y51. 20. Stelfox HT, Velmahos GC, Gettings E, Bigatello LM, Schmidt U. Computed tomography for early and safe discontinuation of cervical spine immobilization in obtunded multiply injured patients. J Trauma. 2007; 63(3):630Y636. 21. Tomycz ND, Chew BG, Chang YF, Darby JM, Gunn SR, Nicholas DH, et al. MRI is unnecessary to clear the cervical spine in obtunded/comatose trauma patients: the four-year experience of a level I trauma center. J Trauma. 2008;64(5):1258Y1263.
1128
22. Sanchez B, Waxman K, Jones T, Conner S, Chung R, Becerra S. Cervical spine clearance in blunt trauma: evaluation of a computed tomographybased protocol. J Trauma. 2005;59(1):179Y183. 23. Harris TJ, Blackmore CC, Mirza SK, Jurkovich GJ. Clearing the cervical spine in obtunded patients. Spine (Phila Pa 1976). 2008;33(14):1547Y1553. 24. Gargas J, Yaszay B, Kruk P, Bastrom T, Shellington D, Khanna S. An analysis of cervical spine magnetic resonance imaging findings after normal computed tomographic imaging findings in pediatric trauma patients: ten-year experience of a level I pediatric trauma center. J Trauma Acute Care Surg. 2013;74(4):1102Y1107. 25. Vanguri P, Young AJ, Weber WF, Katzen J, Han J, Wolfe LG, Duane TM. Computed tomographic scan: it’s not just about the fracture. J Trauma Acute Care Surg. 2014;77(4):604Y607. 26. Diaz JJ Jr, Aulino JM, Collier B, Roman C, May AK, Miller RS, et al. The early work-up for isolated ligamentous injury of the cervical spine: does computed tomography scan have a role? J Trauma. 2005;59(4):897Y903. 27. Kaiser ML, Whealon MD, Barrios C, Kong AP, Lekawa ME, Dolich MO. The current role of magnetic resonance imaging for diagnosing cervical spine injury in blunt trauma patients with negative computed tomography scan. Am Surg. 2012;78(10):1156Y1160. 28. Fesmire FM, Luten RC. The pediatric cervical spine: developmental anatomy and clinical aspects. J Emerg Med. 1989;7(2):133Y142. 29. Kokoska ER, Keller MS, Rallo MC, Weber TR. Characteristics of pediatric cervical spine injuries. J Pediatr Surg. 2001;36(1):100Y105. 30. Brown RL, Brunn MA, Garcia VF. Cervical spine injuries in children: a review of 103 patients treated consecutively at a level 1 pediatric trauma center. J Pediatr Surg. 2001;36(8):1107Y1114. 31. Hadley MN, Zabramski JM, Browner CM, Rekate H, Sonntag VK. Pediatric spinal trauma. Review of 122 cases of spinal cord and vertebral column injuries. J Neurosurg. 1988;68(1):18Y24. 32. Brockmeyer DL, Ragel BT, Kestle JR. The pediatric cervical spine instability study. A pilot study assessing the prognostic value of four imaging modalities in clearing the cervical spine for children with severe traumatic injuries. Childs Nerv Syst. 2012;28(5):699Y705. 33. MedCalc Diagnostic Test Calculator . Available at: http://www.medcalc.org/ calc/diagnostic_test.php2014. Accessed May 3, 2014. 34. Flynn JM, Closkey RF, Mahboubi S, Dormans JP. Role of magnetic resonance imaging in the assessment of pediatric cervical spine injuries. J Pediatr Orthop. 2002;22(5):573Y577. 35. Henry M, Riesenburger RI, Kryzanski J, Jea A, Hwang SW. A retrospective comparison of CT and MRI in detecting pediatric cervical spine injury. Childs Nerv Syst. 2013;29(8):1333Y1338. 36. Hannon M, Mannix R, Dorney K, Mooney D, Hennelly K. Pediatric cervical spine injury evaluation after blunt trauma: a clinical decision analysis. Ann Emerg Med. 2015;65(3):239Y247.
* 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
