J Neurosurg Pediatrics 13:

Transcription

1 J Neurosurg Pediatrics 13: , 2014 AANS, 2014 Computed tomography characteristics in pediatric versus adult traumatic brain injury Clinical article Korak Sarkar, M.D., 1 Krista Keachie, M.D., 2 UyenThao Nguyen, M.S., 3 J. Paul Muizelaar, M.D., Ph.D., 2 Marike Zwienenberg-Lee, M.D., 2 and Kiarash Shahlaie, M.D., Ph.D. 2 Departments of 1 Neurology and 2 Neurological Surgery; and 3 Clinical and Translational Science Center, University of California, Davis, School of Medicine, Sacramento, California Object. Traumatic brain injury (TBI) is a leading cause of injury, hospitalization, and death among pediatric patients. Admission CT scans play an important role in classifying TBI and directing clinical care, but little is known about the differences in CT findings between pediatric and adult patients. The aim of this study was to determine if radiographic differences exist between adult and pediatric TBI. Methods. The authors retrospectively analyzed TBI registry data from 1206 consecutive patients with nonpenetrating TBI treated at a Level 1 adult and pediatric trauma center over a 30-month period. Results. The distribution of sex, race, and Glasgow Coma Scale (GCS) score was not significantly different between the adult and pediatric populations; however, the distribution of CT findings was significantly different. Pediatric patients with TBI were more likely to have skull fractures (OR 3.21, p < 0.01) and epidural hematomas (OR 1.96, p < 0.01). Pediatric TBI was less likely to be associated with contusion, subdural hematoma, subarachnoid hemorrhage, or compression of the basal cisterns (p < 0.05). Rotterdam CT scores were significantly lower in the pediatric population (2.3 vs 2.6, p < 0.001). Conclusions. There are significant differences in the CT findings in pediatric versus adult TBI, despite statistical similarities with regard to clinical severity of injury as measured by the GCS. These differences may be due to anatomical characteristics, the biomechanics of injury, and/or differences in injury mechanisms between pediatric and adult patients. The unique characteristics of pediatric TBI warrant consideration when formulating a clinical trial design or predicting functional outcome using prognostic models developed from adult TBI data. ( Key Words traumatic brain injury computed tomography pediatric adult trauma Traumatic brain injury (TBI) is a leading cause of death and disability among pediatric patients worldwide. In the US alone, over 475,000 children and adolescents sustain a TBI each year, resulting in 37,000 hospital admissions, 2600 deaths, and over $20 billion in health care costs. 2,14,18,31 Despite its significant impact, pediatric TBI is frequently excluded completely or grouped with adult TBI without an adequate understanding of the differences between these distinct patient populations. Although specific guidelines for the neurocritical management of pediatric TBI have been developed and Abbreviations used in this paper: EDH = epidural hematoma; FDR = false discovery rate; GCS = Glasgow Coma Scale; GOS = Glasgow Outcome Scale; IPH = intraparenchymal hematoma; IVH = intraventricular hemorrhage; SAH = subarachnoid hemorrhage; SDH = subdural hematoma; TBI = traumatic brain injury. J Neurosurg: Pediatrics / Volume 13 / March 2014 recently updated, 12 urgent decision making in the acute period is largely guided by adult protocols. For example, surgical guidelines are based on adult studies, 3 and acute outcome prediction models are almost entirely derived from adult TBI data. 21,30,32 Furthermore, most clinical trials on the acute management of TBI do not include pediatric patients. 22 Clinical outcomes for pediatric TBI have not significantly improved over the past decade, suggesting that targeted research is necessary. 22,25,27 Admission CT scans play a critical role in the acute management of both adult and pediatric TBI. Computed tomography findings are used to classify injury patterns, predict survival and functional outcome, and guide surgical and neurocritical care interventions. Previous studies have shown that the biomechanics of TBI are significantly different between children and adults, 13,23 as are the anatomical relationships between the brain and its surrounding supportive structures. 7,29 Despite these differences, 307

2 K. Sarkar et al. however, it is not known if CT findings after TBI significantly differ in pediatric patients. For example, little is known about the patterns of intracranial injury that tend to occur in pediatric patients, and how those findings may influence clinical management and outcome. A more thorough understanding of pediatric TBI CT findings is necessary to improve our clinical approach to diagnosis and treatment and may prove invaluable in designing future clinical trials. Methods Patient Population This is a retrospective analysis of data collected from all patients with TBI who had been treated at the University of California, Davis, Level 1 adult and pediatric trauma center between August 2008 and January 2011 as part of an institutional TBI registry. Patients included in this registry met at least 1 of the following 2 criteria that prompted neurosurgical consultation: 1) TBI suspected due to clinical history, clinical symptoms, or signs of neurological deficits on physical examination; or 2) abnormal CT findings after trauma. Patients with penetrating head injury were excluded from this study. Pediatric patients with suspected nonaccidental trauma were also excluded from this study and underwent a detailed evaluation by our Child Abuse Program. The pediatric age group was defined as patients having an age less than or equal to 16 years. Clinical Data Data were collected from an electronic medical record system, with paper chart review performed when necessary. Data points consisted of demographic, clinical, and radiographic variables, including age, sex, race, ethnicity, mechanism of injury, initial vital signs, and admission laboratory values; postresuscitation Glasgow Coma Scale (GCS) score; and pupil examination, as determined by the neurosurgery consultant at the time of initial evaluation. Glasgow Outcome Scale (GOS) scores were obtained at 3 and 6 months after injury via clinical examination or telephone interview by an experienced clinical research nurse. Radiographic Data The initial postinjury CT scan was reviewed by neurosurgery house staff at the time of initial consultation; for patients who were transferred from an outside hospital, the scan obtained at the transferring institution was reviewed as the initial scan. As part of a separate study, a subset of 550 CT scans were rescored by a blinded neurosurgery attending (K.S.) to confirm adequate interrater reliability with each neurosurgery house staff consultant. 4 This study demonstrated reasonable interrater reliability for those radiographic findings used to calculate the Rotterdam score with an average kappa greater than 0.5 (17). An electronic data collection form was completed at the time of initial evaluation to document the presence or absence of the following intracranial hemorrhage patterns: epidural hematoma (EDH), subdural hematoma (SDH), intraparenchymal hematoma (IPH, 1 cm), intraparenchymal contusion (< 1 cm), intraventricular hemorrhage (IVH), and subarachnoid hemorrhage (SAH). Skull fractures were classified as absent, nondepressed, or depressed. Basal cisterns were classified as open, compressed, or absent, and midline shift was classified as < 5 or 5 mm. These variables were used to calculate a Rotterdam CT Score for each patient, as previously described. 19 Statistical Analysis Differences in the proportion of patients with any particular CT finding were tested using a chi-square analysis. Patients with at least 1 CT finding were then evaluated using a 2-sample proportion test to identify a difference in the proportion of patients with each finding. To determine if certain combinations of CT findings tended to co-occur in pediatric versus adult TBI, 2-sample proportion tests were used. False discovery rates were calculated to account for multiple testing. Parametric data are presented as the means ± standard error, and nonparametric data are presented as the median/range, when appropriate. Statistical analyses were conducted using R statistical computing software. Significance was defined as p < Results Between August 2008 and January 2011, 1258 patients were enrolled in the prospective TBI database at UC Davis Medical Center. Thirty adult and 4 pediatric patients were excluded from analysis because of a penetrating injury mechanism, and 18 pediatric patients were excluded because of suspected nonaccidental trauma/ abuse. These exclusions resulted in a study population of 1206 patients: 336 (27.9%) were classified as pediatric (age < 16 years) and 870 (72.1%) were classified as adult. Clinical Characteristics The mean age of the pediatric and adult groups was 5.01 ± 0.26 years and ± 0.76 years, respectively (Table 1). The pediatric group was composed of 64.3% males, and the adult group included 69.9% males, a difference that was not statistically significant (p > 0.05). Both groups were predominantly white. Pediatric patients were almost twice as likely as their adult counterparts to have been transferred from an outside medical facility (50.3% vs 28.8%, p < 0.001). The CT scan obtained at the outside facility was available for review at our institution in 100% of cases. The most common mechanisms of injury in pediatric patients were falls (46.1%), motor vehicle crashes (16.7%), and sports-related injuries (16.7%). In adults, the most common injury mechanisms were falls (30.3%) and assaults (19.2%). Mechanism of injury was significantly different between the 2 groups. Postresuscitation GCS scores did not significantly differ between the pediatric and adult patient populations (c 2 = , df = 2, p = 0.365). The respective mean and median GCS scores in the pediatric population were 308 J Neurosurg: Pediatrics / Volume 13 / March 2014

3 Computed tomography findings in pediatric TBI TABLE 1: Clinical characteristics of study populations* Parameter Pediatric Adult p Value no. of patients average age in yrs male sex 64.3% 69.9% white 81.0% 79.4% NS transfer from OMF 50.3% 28.8% <0.001 mechanism of injury fall 46.1% 30.3% <0.001 motor vehicle crash 16.7% 7.24% <0.001 assault 2.08% 19.2% <0.001 sporting event 16.7% 7.2% <0.001 auto vs pedestrian 13.3% 7.8% other 5.06% 28.6% <0.001 average GCS score GCS score category mild 65.5% 63.2% moderate 19.6% 19.3% severe 14.9% 17.4% * NS = not significant; OMF = outside medical facility ± 0.2 and 15 (range 3 15); in the adult group, the respective mean and median scores were 12.1 ± 0.1 and 15 (range 3 15). Radiographic Characteristics Skull Fracture. Skull fractures were present in 62.2% of the pediatric patients and 25.8% of the adult patients, a difference that was statistically significant (p < 0.001; Table 2). Nondepressed skull fractures were 3.21 times more likely in pediatric than in adult TBI (c 2 = 69.31, false discovery rate (FDR)-adjusted p < 0.01), and depressed skull fractures were 5.25 times more likely in pediatric patients (c 2 = 48.41, FDR-adjusted p < 0.01). Intracranial Hemorrhage. Epidural hematoma was significantly more common in pediatric patients (18.9% vs 10.6%, OR = 1.96, c 2 = 12.81, FDR-adjusted p < 0.01), whereas SDH was more likely to occur in adult TBI (44.2% vs 29.3%, OR = 0.52, c 2 = 19.84, FDR-adjusted p < 0.01). Intraparenchymal hemorrhage was more common in adult versus pediatric TBI. Contusions (< 1 cm diameter) were present in 12.4% of pediatric patients and 22.4% of adult patients (OR 0.49, c 2 = 13.38, FDR-adjusted p < 0.01). Intraparenchymal hematoma ( 1 cm diameter) was present in 30.7% of the adults and 19.2% of the pediatric patients (OR 0.54, c 2 = 14.13, FDR-adjusted p < 0.01). Subarachnoid hemorrhage was also more common in adult TBI, occurring in 43.1% versus 24.4% of pediatric patients (OR 0.43, c 2 = 31.98, FDR-adjusted p < 0.01). The TABLE 2: Admission CT results in pediatric versus adult TBI* CT Finding Peds Group Adult Group OR (peds:adult) p Value FDR-Adjusted p Value no abnormality 7.3% 4.0% 0.03 <0.01 contusion 12.4% 22.4% 0.49 <0.01 <0.01 IPH 19.2% 30.7% 0.54 <0.01 <0.01 SDH 29.3% 44.2% 0.52 <0.01 <0.01 EDH 18.9% 10.6% 1.96 <0.01 <0.01 IVH 3.3% 3.3% traumatic SAH 24.4% 43.1% 0.43 <0.01 <0.01 MLS >5 mm 5.5% 10.2% cistern compression 7.5% 8.1% cistern absent 2.6% 2.1% depressed fracture 15.0% 3.3% 5.25 <0.01 <0.01 nondisplaced fracture 48.2% 22.5% 3.21 <0.01 <0.01 * MLS = midline shift; peds = pediatric. J Neurosurg: Pediatrics / Volume 13 / March

4 K. Sarkar et al. incidence of IVH did not differ between age groups (3.3% vs 3.3%). Mass Effect. Compression of the basal cisterns was present in 7.5% of pediatric and 8.1% of adult patients, and the complete absence of the cisterns was noted in 2.6% and 2.1%, respectively. These differences were not statistically significant. However, a midline shift > 5 cm was more common in adult versus pediatric TBI (10.2% vs 5.5%, c 2 = 5.46, FDR-adjusted p = 0.03). Multiple CT Findings. At least 1 CT abnormality was present in 96.0% of adult and 92.8% of pediatric patients with TBI, a difference that was statistically significant (c 2 = , df = 2, p = 0.03; Fig. 1 and Table 3). The presence of 2 or more CT findings occurred in 52.4% of adults and 45.9% of pediatric patients (c 2 = 3.58, p = 0.058). The presence of 3 or more and 4 or more findings occurred in 25.5% and 11.8% of adult patients and 22.1% and 7.3% of pediatric patients, respectively (c 2 = 1.276, p = 0.259; and c 2 = 4.425, p = 0.035, respectively). In pediatric TBI, the most common combination of CT findings was fracture and EDH, which occurred in 15.8% of cases as compared with 7.8% of adults (c 2 = 7.22, FDR-adjusted p = 0.038). The most common combination of CT findings in adult TBI was SDH and traumatic SAH, which occurred in 28.0% of cases versus 15.8% of pediatric TBI cases (c 2 = 8.335, FDR-adjusted p = 0.024). Rotterdam CT Score. The distribution of Rotterdam TABLE 3: Presence of multiple CT findings after TBI No. of CT Findings Peds Group Adult Group p Value % 96.0% % 52.4% % 25.5% % 11.8% 0.04 CT scores was significantly different between pediatric and adult patients (c 2 = , df = 4, p = ; Fig. 2 and Table 4). Rotterdam CT scores of 1 and 2 were more common in pediatric patients (11.2% vs 5.5% and 61.0% vs 45.4%, respectively, p = and p < 0.001), whereas a Rotterdam CT score of 3 was more common in adults (40.2% vs 21.2%, p < 0.001). There were no significant differences in the incidence of Rotterdam CT Scores 4, 5, or 6 between age groups. The mean Rotterdam CT scores in pediatric versus adult patients was 2.3 ± 0.04 versus 2.6 ± 0.03, respectively. The difference between these scores was statistically significant (p < 0.001). Direct Transport Versus Secondary Transfer The significant difference in the number of pediatric patients transferred from an outside facility prompted a subanalysis of the pediatric population comparing direct transport with secondary transfer. Of the 336 pediatric patients in this study, 166 (49.4%) were transferred. When comparing patients who were directly transported to our institution with those who were transferred from another hospital, we found no significant differences in mean age (6.3 vs 5.6 years, respectively, c 2 = 1.333, FDR-adjusted p = 0.245), sex distribution (c 2 = 0.421, FDR-adjusted p = 0.516), ethnicity (c 2 = 2.30, FDR-adjusted p = 0.245), or postresuscitation GCS score (c 2 = 1.437, FDR-adjusted p = 0.715). A comparison of CT findings between these 2 populations demonstrated no significant differences in the distribution of specific CT findings (Table 5) or Rotterdam CT score (c 2 = 2.525, FDR-adjusted p = 0.715). Clinical Outcomes Clinical outcomes at 6 months were measured using the GOS. Loss to follow-up occurred in 11.5% of pediat- Fig. 1. Common patterns of CT findings in pediatric (left) and adult (right) patients after nonpenetrating traumatic TBI. Pediatric patients had thinner skull anatomy and were more prone to skull fracture and EDH, whereas adult patients were more likely to demonstrate complex patterns such as subdural and intraparenchymal hemorrhages with > 5 mm midline shift. Fig. 2. Plot of adult and pediatric Rotterdam CT score (RS) distribution. Rotterdam scores of 1 and 2 were significantly more common in the pediatric population, while a Rotterdam score of 3 was significantly more common in adults. The distribution of more severe Rotterdam scores was not significantly different between the 2 patient populations. 310 J Neurosurg: Pediatrics / Volume 13 / March 2014

5 Computed tomography findings in pediatric TBI TABLE 4: Rotterdam CT Score distributions in pediatric versus adult TBI Rotterdam Score Peds Group Adult Group p Value 1 37 (11.0) 46 (5.3) (60.1) 378 (43.4) < (20.8) 335 (38.5) < (4.5) 52 (6.0) (2.1) 22 (2.5) ric cases and 11.9% of the adult cases (p = 0.932). Favorable outcomes, defined as good recovery or moderate disability, occurred in 76.5% of pediatric and 57.3% of adult patients (p < 0.001). The 6-month mortality rate was 3.3% for the pediatric age group and 11.2% for the adult group. Discussion In this study, we found significant differences in CT findings between pediatric and adult TBI, despite the fact that both groups were statistically similar with regard to clinical severity of injury as measured using the GCS score. In general, pediatric patients were more likely to experience extraaxial injuries (skull fracture, EDH) and less likely to demonstrate multiple CT abnormalities or significant midline shift. In contrast, adult patients were more prone to multiple intraaxial hemorrhage patterns and significant mass effect with midline shift. These differences in CT findings yielded a lower average Rotterdam score in pediatric than in adult TBI. Anatomy of Pediatric Versus Adult TBI Differences in CT findings between pediatric and adult patients may be attributable, in large part, to anatomical differences that influence the biomechanics of TBI. The cranial anatomy of infants and children is distinct from that of adults, and there is emerging evidence of a unique age-dependent response following TBI. 10,26 Traumatic brain injury is caused by the imposition of linear and rotational forces on the skull and its internal components. The mechanical effects of these linear and rotational forces are influenced by material properties different for pediatric and adult patients. For example, linear forces are much more likely to cause fractures in the thinner pediatric skull, 23 and the cranial vault is more susceptible to fracture than other craniomaxillofacial bones. 29 Moreover, the neck is weaker and the head proportionally larger in the pediatric population than in adults, which may also contribute to the different injury patterns following TBI. 23 The rate of intracranial fracture has been reported to decrease with age, presumably as the strength of the cranial vault increases through fusion, skull thickening, and increased prominence of the facial bones. 7,29 These differences may explain the increased incidence of skull fracture in the present study. They may also explain the increased incidence of EDH in pediatric TBI, which often results from damage to middle meningeal artery branches in the dura mater and/or bleeding from skull fracture bone edges into the epidural space. Since skull fractures dissipate and absorb impact forces rather than transmitting them to the intracranial vault, the pediatric brain may be less prone to intraaxial damage as a result of increased skull fractures. In general, angular and rotational forces from TBI are thought to be more damaging than translational forces. These forces can result in acute SDH, intraparenchymal contusions, and diffuse axonal injury 8,23,26 and have been proven to correlate with brain mass; that is, the larger the brain, the larger the effect of rotational forces. 8,23 Thus, larger adult anatomy may be more susceptible to these rotational forces and resultant patterns of intraaxial injury, which include diffuse axonal injury, contusions, IPH, and SAH. Rotational forces are also thought to be responsible for the development of SDH, due to tearing of the bridging veins. The smaller mass of the pediatric skull-brain makes it less susceptible to these rotational forces, which would explain the lower incidence of these injury patterns seen in our series. It is also important to note that the ratio of brain volume/csf volume is much lower in adult patients, which results in more relative movement of the brain within the cranial vault at the time of impact. 5 This increases the TABLE 5: Summary of CT findings in pediatric TBI patients directly transported versus those in pediatric TBI patients secondarily transferred from another institution CT Finding Direct Transport Secondary Transfer p Value FDR-Adjusted p Value contusion IPH SDH EDH IVH SAH MLS >5mm cistern compression cistern absent depressed fracture nondisplaced fracture J Neurosurg: Pediatrics / Volume 13 / March

6 K. Sarkar et al. likelihood of damage to bridging veins that are tethered between the dural sinuses and the brain surface and also increases the chance of subarachnoid and intraparenchymal hemorrhage from collisions between the brain and bony components of the intracranial vault. As noted above, the adult skull is less likely to fracture and dissipate force vectors, and this may further amplify the magnitude of intracranial brain movement with closed head injury. This decreased ratio of brain/csf volume may also explain the increased incidence of significant midline shift ( 5 mm) in adult patients, as there is more relative space surrounding the brain to accommodate brain shift. The net effect of these differences may be an increased likelihood of more complex patterns of hemorrhage in adults, a significant observation in the present study. Adults were more prone to intraaxial injuries, multiple intracranial injuries, and midline shift of the brain, all of which predict a worse clinical outcome. 19 Clinical outcomes are generally better in younger patients after TBI a finding that was confirmed in the present study, and one that may be due, at least in part, to significant differences in patterns of CT findings between pediatric and adult populations. Characterizing Injury Severity: GCS Score Versus Rotterdam CT Score The GCS score was first described by Teasdale and Jennett over 28 years ago 33 and remains the most commonly used tool for classifying the clinical severity of TBI. However, one major limitation of the GCS score is its insensitivity to the pattern and severity of intracranial injuries detected on CT scanning, many of which have been shown to predict clinical course and outcome. 19,21 For example, a patient with an isolated extraaxial injury may present with the same GCS score as a patient with multiple intraaxial hemorrhages involving deep brain structures and the brainstem; clearly, these patients have experienced very different brain injuries and will likely have unique treatment needs and clinical outcomes. In this study, 2 large patient populations were found to have many significant differences in specific CT findings, but the overall GCS scores were not different, highlighting an important limitation of the GCS scoring system. Growing evidence demonstrates that radiographic scales such as the Rotterdam and Marshall scales, as well as clinical scales such as the APACHE II (Acute Physiology and Chronic Health Evaluation II) and Injury Severity Scale, may be better tools than the GCS in predicting mortality and functional outcome. 11,26 This may be particularly true for pediatric patients, 16,17 despite efforts to use modified GCS criteria that are more age appropriate. 6,24 The Marshall CT classification 20 was described in 1991 and represented the first major effort to characterize TBI based on CT findings. In 2005, this scoring system was modified to a version with greater predictive value: the Rotterdam CT score. 19 Like the Marshall scale, the Rotterdam system is sensitive to the presence of mass effect and midline shift but also takes into account 2 important clinical facts: patients with isolated EDH have improved outcomes, and those with SAH or IVH have significantly more morbidity. Neither the Rotterdam CT scoring system nor the GCS take into account patient age, which is known to be a strong independent predictor of outcome following TBI. The current study demonstrates that adult and pediatric patients experience different patterns of intracranial injury following TBI, which results in a small but significant difference in the Rotterdam CT score (2.3 vs 2.6). However, this difference in the Rotterdam score alone does not account for the significant differences in survival and clinical outcome following TBI, both of which were significantly higher in pediatric populations. Since similar CT findings can have very different consequences in adult versus pediatric TBI, an age-specific CT classification system should be developed. Study Strengths and Limitations The results of this study are derived from a large registry database maintained at a Level 1 trauma center for both adult and pediatric TBI. One limitation of our study was the significantly different mechanisms of injury between the 2 patient groups. This resulted in a degree of heterogeneity between the 2 study populations and may have contributed to the differences in neuroradiographic findings. Children with suspected abuse from nonaccidental trauma were excluded from this study, since they represent a unique subgroup of patients with TBI that is more likely to experience repeat TBI or TBI associated with superimposed hypoxia and/or ischemia. 15,28 It is important to note that children with nonaccidental trauma often have worse clinical outcomes and that their pattern of intracranial findings is significantly different, so inclusion may have affected some of the results of this study. For example, Foerster and colleagues 9 reviewed CT and MRI findings from 57 children who had experienced abuse and found no cases of EDH (the most common findings in our study) but a significant incidence of SDH (86%), diffuse brain ischemia (25.5%), and SAH (17%). Intraparenchymal hemorrhages (6.4%) and diffuse axonal injury (2.1%) were relatively less common, and the majority of children were found to have extracranial injuries such as retinal hemorrhage (84.9%) and orthopedic fractures (54.4%). In the present study, only 5.4% of the pediatric patients experienced nonaccidental trauma, so including them in the analysis would not have significantly affected the population outcomes. Another important potential limitation of our study is the higher proportion of pediatric patients transferred from an outside institution, which may have introduced a selection bias onto our results. The California Emergency Medical Services Authority recommends the transfer of pediatric patients with TBI when the injury is accompanied by a penetrating wound, CSF leak, open head injury, depressed skull fracture, or associated injuries requiring intensive care. Despite these recommendations, there is considerable variation in the utilization pattern of trauma specialty care in the pediatric population, particularly as regards transfers, because California lacks an integrated statewide emergency medical services system. 1,34 In the present study, we did not find any significant differences between transfer patients with regard to age, sex, ethnic- 312 J Neurosurg: Pediatrics / Volume 13 / March 2014

7 Computed tomography findings in pediatric TBI ity, GCS score, or CT findings including the Rotterdam CT score. This suggests that transfer status is unlikely to have significantly affected the results of this study. Conclusions Following nonpenetrating TBI, adult and pediatric patients tend to demonstrate very different patterns of intracranial injury. Pediatric patients are more prone to skull fractures and EDH, whereas adult patients have a higher incidence of intraaxial hemorrhage and brain shift and are more likely to demonstrate multiple patterns of intracranial injuries. These differences may be attributable to variations in the mechanisms, anatomy, and biomechanics of TBI between pediatric and adult patients. Future injury classification systems that include age, clinical examination, and CT findings may be more accurate and useful than currently available scoring systems. Acknowledgments We thank Sandra Taylor and the Clinical and Translational Science Center (CTSC) at the University of California, Davis. Disclosure This project was supported by the National Center for Advancing Translational Sciences (NIH Grant No. TR000002). Author contributions to the study and manuscript preparation include the following. Conception and design: Shahlaie, Keachie, Zwienenberg-Lee. Acquisition of data: Shahlaie, Sarkar. Analysis and interpretation of data: Shahlaie, Sarkar, Keachie, Nguyen, Zwienenberg-Lee. Drafting the article: Sarkar, Keachie. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Shahlaie. Statistical analysis: Shahlaie, Sarkar, Keachie, Nguyen. Study supervision: Shahlaie, Zwienenberg-Lee. References 1. Acosta CD, Kit Delgado M, Gisondi MA, Raghunathan A, D Souza PA, Gilbert G, et al: Characteristics of pediatric trauma transfers to a level I trauma center: implications for developing a regionalized pediatric trauma system in California. Acad Emerg Med 17: , Brener I, Harman JS, Kelleher KJ, Yeates KO: Medical costs of mild to moderate traumatic brain injury in children. J Head Trauma Rehabil 19: , Bullock MR, Povlishock JT: Guidelines for the management of severe traumatic brain injury. Editor s commentary. J Neurotrauma 24 Suppl 1:vii viii, 2007 (Erratum in J Neurotrauma 25: , 2008) 4. Cadena G, Lewis B, Bobinski M, Nidecker A, Rudisill N, Gurkoff GG, et al: CT scan interpretation in TBI: implications for resident training. J Neurotrauma 29:A-41, 2012 (Abstract) 5. Courchesne E, Chisum HJ, Townsend J, Cowles A, Covington J, Egaas B, et al: Normal brain development and aging: quantitative analysis at in vivo MR imaging in healthy volunteers. Radiology 216: , Dietrich AM, Bowman MJ, Ginn-Pease ME, Kosnik E, King DR: Pediatric head injuries: can clinical factors reliably predict an abnormality on computed tomography? Ann Emerg Med 22: , Eggensperger Wymann NM, Hölzle A, Zachariou Z, Iizuka T: Pediatric craniofacial trauma. J Oral Maxillofac Surg 66: 58 64, 2008 J Neurosurg: Pediatrics / Volume 13 / March El Sayed T, Mota A, Fraternali F, Ortiz M: Biomechanics of traumatic brain injury. Comput Methods Appl Mech Eng 197: , Foerster BR, Petrou M, Lin D, Thurnher MM, Carlson MD, Strouse PJ, et al: Neuroimaging evaluation of non-accidental head trauma with correlation to clinical outcomes: a review of 57 cases. J Pediatr 154: , Huh JW, Raghupathi R: New concepts in treatment of pediatric traumatic brain injury. Anesthesiol Clin 27: , Katsnelson M, Mackenzie L, Frangos S, Oddo M, Levine JM, Pukenas B, et al: Are initial radiographic and clinical scales associated with subsequent intracranial pressure and brain oxygen levels after severe traumatic brain injury? Neurosurgery 70: , Kochanek PM, Carney N, Adelson PD, Ashwal S, Bell MJ, Bratton S, et al: Guidelines for the acute medical management of severe traumatic brain injury in infants, children, and adolescents second edition. Pediatr Crit Care Med 13 (Suppl 1): S1 S82, Lang DA, Teasdale GM, Macpherson P, Lawrence A: Diffuse brain swelling after head injury: more often malignant in adults than children? J Neurosurg 80: , Langlois JA, Rutland-Brown W, Wald MM: The epidemiology and impact of traumatic brain injury: a brief overview. J Head Trauma Rehabil 21: , Leventhal JM, Martin KD, Asnes AG: Fractures and traumatic brain injuries: abuse versus accidents in a US database of hospitalized children. Pediatrics 126:e104 e115, Levin HS, Aldrich EF, Saydjari C, Eisenberg HM, Foulkes MA, Bellefleur M, et al: Severe head injury in children: experience of the Traumatic Coma Data Bank. Neurosurgery 31: , Lieh-Lai MW, Theodorou AA, Sarnaik AP, Meert KL, Moylan PM, Canady AI: Limitations of the Glasgow Coma Scale in predicting outcome in children with traumatic brain injury. J Pediatr 120: , Litvack ZN, Selden N: Pediatric traumatic brain injury, in Bhardwaj A, Ellegala DB, Kirsch JR (eds): Acute Brain and Spinal Cord Injury: Evolving Paradigms and Management. New York: Informa Healthcare, 2008, pp Maas AI, Hukkelhoven CW, Marshall LF, Steyerberg EW: Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery 57: , Marshall LF, Marshall SB, Klauber MR, Van Berkum Clark M, Eisenberg HM, Jane JA, et al: A new classification of head injury based on computerized tomography. J Neurosurg 75(1s):S14 S20, Murray GD, Butcher I, McHugh GS, Lu J, Mushkudiani NA, Maas AI, et al: Multivariable prognostic analysis in traumatic brain injury: results from the IMPACT study. J Neurotrauma 24: , Narayan RK, Michel ME, Ansell B, Baethmann A, Biegon A, Bracken MB, et al: Clinical trials in head injury. J Neurotrauma 19: , Ommaya AK, Goldsmith W, Thibault L: Biomechanics and neuropathology of adult and paediatric head injury. Br J Neurosurg 16: , Ong L, Selladurai BM, Dhillon MK, Atan M, Lye MS: The prognostic value of the Glasgow Coma Scale, hypoxia and computerised tomography in outcome prediction of pediatric head injury. Pediatr Neurosurg 24: , Pfenninger J, Santi A: Severe traumatic brain injury in children are the results improving? Swiss Med Wkly 132: , Pinto PS, Meoded A, Poretti A, Tekes A, Huisman TA: The unique features of traumatic brain injury in children. Review of the characteristics of the pediatric skull and brain, mecha- 313

8 K. Sarkar et al. nisms of trauma, patterns of injury, complications, and their imaging findings part 2. J Neuroimaging 22:e18 e41, Raphaely RC, Swedlow DB, Downes JJ, Bruce DA: Management of severe pediatric head trauma. Pediatr Clin North Am 27: , Reece RM: What are we trying to measure? The problems of case ascertainment. Am J Prev Med 34 (4 Suppl):S116 S119, Roeder RA, Thaller S: Unique features of the pediatric craniofacial anatomy. J Craniofac Surg 22: , Roozenbeek B, Lingsma HF, Lecky FE, Lu J, Weir J, Butcher I, et al: Prediction of outcome after moderate and severe traumatic brain injury: external validation of the International Mission on Prognosis and Analysis of Clinical Trials (IMPACT) and Corticoid Randomisation After Significant Head injury (CRASH) prognostic models. Crit Care Med 40: , Rutland-Brown W, Langlois JA, Thomas KE, Xi YL: Incidence of traumatic brain injury in the United States, J Head Trauma Rehabil 21: , Steyerberg EW, Mushkudiani N, Perel P, Butcher I, Lu J, McHugh GS, et al: Predicting outcome after traumatic brain injury: development and international validation of prognostic scores based on admission characteristics. PLoS Med 5:e165, Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81 84, Wang NE, Saynina O, Kuntz-Duriseti K, Mahlow P, Wise PH: Variability in pediatric utilization of trauma facilities in California: 1999 to Ann Emerg Med 52: , 2008 Manuscript submitted May 3, Accepted December 2, Please include this information when citing this paper: published online January 10, 2014; DOI: / PEDS Address correspondence to: Kiarash Shahlaie, M.D., Ph.D., Department of Neurological Surgery, University of California, Davis, School of Medicine, 4860 Y St., Ste. 3740, Sacramento, CA kiarash.shahlaie@ucdmc.ucdavis.edu. 314 J Neurosurg: Pediatrics / Volume 13 / March 2014