Highlights: Demonstrating gaps in current knowledge and research in mild traumatic brain injury and sports-related concussion that are opportunities for new research endeavors and providing links to essential resources advocated by the National Institutes of Health, termed Common Data Elements (CDEs), for research in mTBI that attempt to standardize clinical data acquisition, data collection, neuroimaging, and outcome metrics to enable better comparison of research studies and multicenter collaboration.

Talk Title: Pediatric TBI and Sports-Related Concussion: Common Data Elements (CDEs) to Inform Diagnosis, Neuroimaging, and Outcome Metrics.

Target Audience: Neuroradiologists, Neurologists, and Researchers who have an interset in mild traumatic brain injury and sports-related concussion, specifically diagnosis, treatment, management, and outcome metrics, are the target audience.

Introduction: Most pediatricians treat concussion with a period of rest for the first few days followed by a gradual return to the normal activities of daily living, which is a pivotal concern for patients, parents, and schools, because there is no consensus on the optimal duration and type of rest during recovery, as there is no consensus on the optimal duration and type of rest during recovery. Most concussion studies in pediatric patients are limited by small sample sizes, retrospective study design, and inability to account for variability of symptoms or heterogeneity of injury mechanism, lack of standardization of neuroimaging protocols, particularly advanced neuroimaging MR, and variable outcome metrics and treatment management. Currently, it is difficult for clinicians to reach consensus on best practices for the treatment and management of mild traumatic brain injury and even definitions of mTBI including clinical severity given a paucity of high quality, longitudinal, prospective research studies.

Even the definition of what constitutes mTBI in children has evolved. The International Consensus statement on concussion in sport, states mTBI is a “complex pathophysiologic process affecting the brain induced by biomechanical forces” including one or more of the following five clinical domains: symptoms that can be somatic (headache), cognitive (foggy feeling), and/or emotional (labile); physical signs (loss of consciousness, persistent amnesia), behavioral (irritability), and sleep disturbances (1,2). The American Academy of Neurology defines mTBI as a loss of consciousness (LOC) lasting less than 30 minutes, initial Glasgow Coma Scale (GCS) of 13 or above, and post-traumatic amnesia lasting less than 24 hours (3). The Centers for Disease Control (CDC) defines mTBI as an injury to the head resulting from blunt trauma or translational forces in children with a GCS of 14 or above, with one or more of the following related signs: transient confusion, disorientation, impaired consciousness, amnesia, LOC less than 30 minutes, and signs of neurological dysfunction including headache (4). However, on the NIH website for CDEs, no specific definition for mTBI is provided.

Common Data Elements (CDEs) is an initiative of the National Institute of Health (NIH) to standardize data acquisition, data collection, definitions, and outcome metrics in pediatric mTBI. CDEs may make it easier to compare research studies performed at different institutions in different settings, so that consensus recommendations can be achieved. The purpose of this review is to inform neuroradiologists and researchers about the development of CDEs for pediatric mTBI, which can inform longitudinal, prospective studies that will provide the data for much needed evidence-based guidelines in the management of pediatric mTBI.

Common Data Elements (CDEs):

The Common Data Elements (CDEs) are the National Institute of Health’s interagency efforts to standardize naming, definitions, and data structure for clinical research variables (5). Without standardization, it is difficult to compare research studies or perform clinical trials, especially in an emerging field of neuroscience such as mild Traumatic Brain Injury (mTBI). Working groups of experts in the field of mTBI have developed recommendations for demographic data, clinical data such as symptoms, imaging parameters, and neuropsychological testing and other metrics to evaluate clinical symptoms during and after mTBI, which are specific to the pediatric population (5-7). CDEs are currently available on line at the National Institute of Neurological Disorders and Stroke (NINDS) website, which encourages the use of CDEs (http://www.commondataelements.ninds.nih.gov) in its grant funding announcements. According to the National Institute of Health, the goal of the CDEs are “to dissemble standards for the collection of data from participants enrolled in studies of neurological diseases; create accessible data-collection tools for investigators that are ready to use ‘off-the-shelf’; encourage focused and simplified data collection to reduce the burden on investigators and practice-based clinicians to increase clinical research participation; reduce costs of data entry, cleaning, and analysis by providing uniform data descriptions and tools across NINDS-funded clinical studies of treatment for neurological disorders” (5). There are core, supplemental, and emerging components to the CDEs. Core elements are essential measures needed in any pediatric mTBI study. Supplemental components enable more detailed research, and emerging components are those where consensus is developing on its utility (i.e. advanced MR pulse sequences). This is a work in progress in pediatric TBI (6-8), and there are numerous supplemental and emerging data elements due to a need for greater depth of data analysis (“higher data granularity”) (5).

CDEs in the Context of Clinical mTBI Diagnosis:

Accurate characterization of pediatric mTBI in its initial presentation is difficult. Many published research articles on pediatric mTBI are on sports-related concussions but over half of mTBI in children are related to activities of daily living and recreational play (9). Reliance on patient memory regarding loss of consciousness or amnesia and symptoms affects the diagnostic accuracy of mTBI by introducing selection bias (10, 11), particularly in cases where mTBI is unwitnessed or when reporting symptoms may be influenced by secondary gain from a desire to return-to-play in athletes or a desire to proceed with litigation related to the accident that caused the mTBI (12). There is inconsistency in the initial evaluation of mTBI including imaging protocols and non-standardized symptom assessments, which undermine scientific comparisons (13). CDEs have been created to help document the clinical presentation and core patient demographics including age, gender, ethnicity, race, number of years of education, parental ethnicity and race, and socioeconomic status. For pediatric TBI, primary care giver’s highest grade or level of education is noted. For subjects, additional core elements include date of injury, age of injury, head circumference, height and weight. For family and medical history, no core data elements are required. Medical history, including medications and/or drug use, and family history are obtained in almost all mTBI clinical research studies, but medical history is often the least reliable data collected.

Additional core data elements include the type and mechanism of injury, which is highly relevant given the heterogeneity of injuries that cause mTBI. The cause of TBI, type of TBI (i.e. closed, penetrating, etc), mechanism (i.e. fall, contact sport), likelihood that the TBI caused the injury and evidence of prior brain injury are core elements in children. For initial assessments of concussion, the Glasgow Coma Scale (GCS) has been traditionally used, which is not ideally suited to mTBI in children, in whom symptoms and severity may fluctuate more than in adults. A pediatric GCS (14) is used, which is part of the CDE core elements. The pediatric GCS like the adults measures the domains of eye movements, verbal responses, and motor responses, but the major differences are the evaluation in younger children, which scales from normal response (coos, babbles) followed by irritable crying, then crying in response to painful stimuli, then moaning in response to painful stimuli and no response, which is the most abnormal. Additional core elements include a neurologic assessment for loss of consciousness (length of time, verification by witness), post-traumatic amnesia (length of time, witnessed), and alteration of consciousness (length of time, witnessed), and presence of absence of symptoms commonly witnessed in mTBI including headache, nausea, emesis, fatigue, balance problems, difficulty concentrating, difficulty remembering, sadness, more emotional, drowsiness, and sensitivity to light and noise.

CDEs for Neuroimaging and mTBI Diagnosis:

The gold standard for imaging in acute mTBI in children remains a non-contrast head CT. The decision to image in the setting of concussion often is driven by concerns over symptoms such as loss of consciousness and headache, anxieties of referring clinicians and/or parents, as well as concerns about the mechanism of injury (15). Most neuroimaging studies are normal in children with concussion (16). The Pediatric Emergency Care Applied Research Network (PECARN) traumatic brain injury prediction rules were developed to identify children at low risk for clinically significant traumatic brain injury (17-19) and who do not need to undergo imaging acutely. Rapid MR exams are proposed as an alternative to head CT (20, 21). However, there are obvious hurdles to increased MR use in the ER setting including lack of 24/7 technical coverage, lack of MR access, and higher costs related to MR and potential needs for sedation with its risks. Most research studies continue to define mTBI as having negative CT and conventional MR scans, but whether advanced neuroimaging can enhance the diagnosis or risk stratify pediatric patients with mTBI or prognosticate outcomes is unknown. Neuroimaging CDEs (6, 7) are needed to inform research and clinical trials and to facilitate reliable analysis and comparisons of research data. On the NINDS website, there is a detailed imaging appendix for CDEs for CT and MR. The MRI core data gives parameters for each sequence at both 1.5T and 3.0T including T1W, T2W, FLAIR, SWI or Gradient Echo, DWI, and volumetric 1W and T2W sequences. Repetition time (TR), echo time (TE), inversion time (TI), flip angle, field of view (FOV), matrix size, slice thickness, and gap, voxel size, number of phase encoding steps or NEX, phase encoding direction, fat suppression options, acceleration factors, bandwidth, echo train length, flow compensation when needed, and b-values for DWI are all listed for each MR acquisition. More importantly, the site provides a suggestion for “best practices” regarding the MR parameters that generate the most reliable and high quality imaging. Beyond the core MR protocols there are levels II through IV, which are commonly used for research studies. Level II includes diffusion tensor imaging (DTI), Level III includes perfusion-weighted imaging (PWI), MR angiography (MRA), and arterial spin labeling (ASL), and Level IV includes spectroscopic imaging. These sequences are more advanced, not readily available on all scanners, and quantitative (6, 7).

For CT parameters, there are tables available on-line which detail image acquisition mode (i.e. helical, pitch 0.8-1.1); gantry rotation; and kVp and mA values. Levels II through IV detail advanced CT techniques, which can generate great image quality using lower kVp and mA values and gantry rotation speeds of 0.4-0.8 seconds per gantry rotation and with different pitches. CT perfusion (CTP) techniques are listed in Level IV (6, 7). Additional CT parameters listed include coverage and slice thickness, slice orientation, use of contrast material and contrast volume, injection rate, and IV access (i.e. 18-20 gauge IV line or CTP) (6, 7).

Imaging CDEs ask researchers to keep track of the MR and CT vendor, software platform, and field strength (i.e. 3.0T), and imaging sequences acquired. The neuroimaging CDE includes a detailed checklist of common abnormalities observed in TBI that includes skull fracture, extra-axial hemorrhage (epidural, subdural), subarachnoid hemorrhage, vascular injury (dissection, aneurysm), venous sinus injury, contusion (bland and/or hemorrhagic), midline shift, cisternal effacement, intraventricular hemorrhage, traumatic axonal injury/diffuse axonal injury, edema (i.e. cytotoxic, vasogenic, etc), brain swelling, and ischemia/infarction (i.e. hypoxia, watershed zones, non-watershed infarct). There are over 30 imaging features listed (6, 7). Supplemental CDEs include estimations of size or volume of all abnormalities detected.

CDEs for Outcome Metrics in Pediatric mTBI

For adults, the recovery from concussion is short, with 80-90% of patients experiencing symptom resolution and a return to normal within 10 days (22, 23). In children and adolescents prolonged recovery times are observed. Epidemiological studies reveal that about 59% of children remain symptomatic at one month, 11% at three months, and 2.3% at 1 year (15, 24). Concussion adversely affects children and adolescents in a variety of ways including diminished academic and athletic performance, difficulties at home and school, and problems with social relationships (25). Decline in school attendance and performance have been reported in up to 30% of children postconcussion (26). Impairment of memory, slow information processing, and executive dysfunction is commonly uncovered on neuropsychological testing within the first two weeks following injury (27).

Concussion management in children centers on the concept of rest, both brain rest and physical rest until acute symptoms resolve followed by a gradual return to school and activity (28), as well as education and reassurance to patient and family (29-32). Animal models have shown a theoretic basis for rest during the first few days following a concussion. Rat models show decreased molecular markers of plasticity, neuroinflammation, and worse cognitive outcome when exposed to exercise following acute TBI (33-35). A study by Burke MJ et al, identified 71 different clinical trials on mTBI and/or concussion studying a wide range of treatments that were mostly behavioral/cognitive (28%) or related to medications (28%), but only a few centered on the more pressing clinical problem of rest/return-to-activity (1.4%) (13). It is hard for the best-intentioned educators and clinicians to support the recovery of children from concussion when little empirical research or evidence-based medicine exists regarding this treatment including duration and type of rest or potential therapies to speed recovery or reduce symptoms (29, 36-38).

Given recent research, the strategy of brain rest is under intense scrutiny. In a randomized study of pediatric patients with concussion, children were either treated with strict brain did worse, as symptoms lasted longer (39). In a case series, active rehabilitation including home exercise in 16 children and adolescents with persistent post-concussive symptoms resulted in decreased post-concussive symptoms and a successful return to activities of daily living including sports (40). In a recent cohort study, adolescent patients who had higher levels of activity had shorter symptom duration following concussion (41). In another study, adolescents with mTBI who used a web-based intervention, which incorporated anticipatory cognitive and physical activities, had fewer postconcussion symptoms (42). Prolonged periods of reduced activity may delay recovery following mTBI (26, 37, 40, 43, 44), so it may not be prudent or feasible to place pediatric patients on strict rest until all symptoms resolve.

Prospective, controlled, longitudinal studies are a top priority for the NIH, and CDEs provide an opportunity to researchers by listing quantitative, reliable, and accurate outcome metrics that can be used to assess symptoms and characterize postconcussion recovery. It is beyond the scope of this talk to discuss all listed outcome metrics in detail, but the NIH website does specify a core group of validated outcome metrics (11, 12). For adaptive and daily living skills, the Functional Independent Measure for Children (Wee FIM) and the Pediatric Evaluation of Disability Inventory (PEDI) self-care subscales are recommended. For global outcome, the Glasgow Outcomes Scale-extended Pediatric Revision GOS-E-Peds is considered a core CDE. For language and communication, it is the Wechsler Abbreviated Scale of Intelligence (WASI II) Vocabulary Test. There are 4 core CDEs for the assessment of neuropsychological impairment including the California Verbal Learning Test for Children (CVLT-C), the Delis-Kaplan Executive Function System (D-KEFS) Verbal Fluency, Wechsler Abbreviated Scale of Intelligence (WASI II) 2 subset version, and the WISC-IV/WPP SI-III Processing Speed Index. For physical functioning, the Health and Behavior Inventory (HBI) and the Functional Independence Measure for Children (Wee-FIM) Motor Subscale are considered core elements. Lastly, for perceived generic and disease health-related quality of life assessment includes one core CDE, the Pediatric Quality of Life Inventory: Generic Core. Supplemental outcome metrics, not considered core CDEs, that are often used in the setting of sports-related pediatric mTBI include the CCAT, CNS Vital Signs, the Headminder Concussion Resolution (45-48), and the ImPACT assessment, which is one of the most widely used outcome metrics. ImPACT is a computerized program that takes about 20 minutes to complete. It consists of 6 test modules that measure verbal and visual memory, reaction time, processing speed, and impulse control. This has been perhaps the most extensively used program in the setting of sports concussion (49, 50).

Conclusion:

Common data elements (CDEs) are of critical importance, as they standardize and streamline research efforts for the study of mTBI in adults and children. Diagnosis of mild TBI, identification of the small subset of patients with postconcussion syndrome and prolonged recovery, recommendations for treatment and management, including returning to school or the playing field, and prognostication for long-term outcome and development of evidence-based interventions to lessen postconcussive symptoms and reduce long-term sequelae, which would improve outcomes and reduce costs and morbidity associated with mTBI are key areas where prospective, longitudinal research is needed including. Standardized definitions, data collection, and imaging termed common data elements (CDEs) are needed to expedite and support these research efforts.