Introduction
The diagnosis of concussion is a critical step in the appropriate management of patients following minor head trauma.Reference Naunheim, Matero and Fucetola 1 – Reference McCrory, Meeuwisse and Echemendia 4 The International Conference on Concussion in Sport defines concussion as a “complex pathophysiologic process affecting the brain, induced by biomechanical forces” due to either a direct or indirect blow, resulting in the impairment of neurologic function with clinical symptoms.Reference McCrory, Meeuwisse and Aubry 5 The diagnosis of concussion is particularly challenging because acute symptoms, which are likely caused by functional disturbances rather than gross structural injury, are associated with normal standard structural neuroimaging (e.g., computed tomography scan).Reference McCrory, Meeuwisse and Aubry 5 Symptoms of concussion may overlap with more severe pathologic traumatic brain injury or mimic other medical conditions, such as posttraumatic stress disorder, depression, and headache syndromes.Reference Ayr, Yeates and Taylor 6 – Reference Kashluba, Paniak and Casey 10 The identification of concussion is critical to deliver appropriate care and support to patients and their families.Reference Taylor, Dietrich and Nuss 11 – Reference Ganesalingam, Yeates and Ginn 13 Persistent symptoms following concussion can have a devastating impact.Reference Carroll and Rosner 14 , Reference Yeates, Kaizar and Rusin 15 Consequences of concussion can include school absenteeism, depression, loss of social activities, and lower quality of life.Reference Scorza, Raleigh and O’Connor 16 Children and adolescents may miss weeks or even months of their school year, which can affect marks and jeopardize a promotion to the next grade level.Reference Meehan and Bachur 17 , Reference Meehan and Mannix 18 Slowed cognition, decreased attention, and memory impairment may make schoolwork a challenge upon return and require educational interventions.Reference Babikian, Satz and Zaucha 12 , Reference Yeates, Kaizar and Rusin 15 , Reference Forsyth and Kirkham 19 , Reference Korinthenberg, Schreck and Weser 20 Given the serious short- and long-term effects following mild traumatic brain injury, the application of appropriate initial management practices is essential.Reference Marchie and Cusimano 21 – Reference Swaine, Tremblay and Platt 26
Numerous guidelines for the management of concussion have been released from specialty groups (including sports medicine, neurology, pediatrics, and family medicine) and from large medical associations. There is a collective agreement across all guidelines that the initial management should include removal from play, physical and cognitive rest, and graduated return to normal activities following medical clearance.Reference McCrory, Meeuwisse and Aubry 5 , 27 – Reference Marshall, Bayley and McCullagh 32 The 2012 Zurich Consensus statement (of the 4th International Conference on Concussion in Sport) concluded that there should be no return to play on the day of injury for the child or adolescent athlete, and it recommended a stepwise progression to return to play. Cognitive rest was described as a need to modify school attendance and activities.Reference McCrory, Meeuwisse and Aubry 5 This consensus statement noted that it is appropriate to have an extended time of asymptomatic rest in children and adolescents. Similarly, the American Academy of Neurology guidelines recommend a more conservative return to play in children and adolescents than in adults, but the guidelines do not specify cognitive rest.Reference Giza, Kutcher and Ashwal 30
The incidence of pediatric concussion has increased by over 50% in the past decade, resulting in a proportional increase in pediatric emergency department (ED) visits. There are approximately 250,000 pediatric concussion ED visits annually in the United States.Reference Gilchrist, Thomas and Xu 33 The objective of this study was to examine the ability of Canadian pediatric emergency physicians to diagnose a concussion and evaluate the scope of practice variation for recommendations of physical and cognitive rest. The authors hypothesized that wide practice variation exists among pediatric emergency medicine physicians in the application of physical and cognitive rest recommendations following an acute concussion.
Materials and Methods
Study design
A survey of Canadian pediatric emergency physicians was carried out to assess the knowledge of pediatric concussion and initial management, specifically regarding the application of physical and cognitive rest.
Instrument development
A multidisciplinary team of physicians with expertise in concussion diagnosis and management, pediatric emergency medicine, neuropsychology, allied health, assessment tool design, and medical education research established the questionnaire’s length, format, and content. Four domains were identified: 1) knowledge of concussion diagnosis; 2) initial management using return-to-play/school/work recommendations (including subgroups B1: same-day return-to-play, B2: graduated return-to-play, and B3: cognitive rest [e.g., school attendance, homework, reading, screen time]); 3) knowledge of existing concussion guidelines or recommendations; and 4) perceived barriers to guideline use.
The team developed three case vignettes featuring common pediatric minor head injury scenarios of varying complexity (Appendix 1). The first vignette described a preadolescent male who developed a headache, dizziness, visual disturbances, and mild cognitive changes after an indirect blow to the head caused by an accidental collision to his chest during a hockey game. His symptoms partially improved over the 20 minutes subsequent to the collision. The purpose of this vignette was to determine whether a concussion would be recognized without a direct blow to the head and to determine the knowledge of same-day return-to-play contraindications. The second vignette described an adolescent female who developed amnesia, dizziness, and nausea after a nonsports-related fall at school in which her head struck a wall in a stairwell. The purpose of this vignette was to determine recognition of a nonsports concussion and to ensure that recommendations were made for cessation of physical exertion (even noncontact such as running) until the child was asymptomatic. The final vignette described an adolescent male who had a football helmet-to-helmet collision resulting in a 1-minute loss of consciousness, amnesia, headache, vision changes, and nausea. This classic case was included to correspond with the more constrained pre-2002 definition of a concussion.Reference McCrory, Meeuwisse and Echemendia 4 All three vignettes exceeded the lowest threshold for the diagnosis of a concussion.Reference McCrory, Meeuwisse and Echemendia 4 , Reference McCrory, Meeuwisse and Aubry 5 To assess the respondents’ knowledge of concussion diagnosis and management, four closed-ended questions were posed for each of the case vignettes. The first question inquired whether the child suffered a concussion (correct answer: yes). Subsequent questions queried whether same-day return to play was possible (correct answer: no) or whether athletic competitions were possible in the next 2 to 5 days (correct answer: no), and suggested cognitive rest options (choices included school absence; reduction of schoolwork; and avoidance of computer use, text messaging, playing video games, and watching the television). Five demographic questions were also included. Prior to distribution, the questionnaire was pilot tested for clarity and appropriateness by 10 emergency physicians and members of the Concussions Ontario: Diagnosis and Early Education working group (which included allied health and physician representatives) and revised as necessary.Reference Burns, Duffett and Kho 34 All reviewers were excluded from the participant population.
Outcomes
The primary outcome was the application of the Zurich Consensus definition to diagnose a concussion. Secondary outcomes included practice variation for physical rest recommendations (no same-day return-to-sport and graduated return-to-play) and practice variation for cognitive rest.
Participants
The authors enrolled pediatric emergency medicine physicians who are members of the Pediatric Emergency Research Canada (PERC). PERC is a network of 15 pediatric or teaching hospitals across Canada with diverse geographic membership: 38% from Ontario, 21% from Alberta, 17% from Quebec, 8% from Nova Scotia, 5% from British Columbia, 4% from Manitoba, 3% from Saskatchewan, and 3% from Newfoundland. After the exclusion of study investigators and pilot survey participants (n=6), the questionnaire was distributed to the remaining 176 PERC members. Survey respondents who were not attending physicians or did not regularly diagnose or manage children and adolescents with concussions were excluded.Reference Burns, Duffett and Kho 34
Ethical considerations
Prior to survey distribution, the study was approved by the PERC Executive Committee and the Research Ethics Board of the Children’s Hospital of Eastern Ontario. This project was funded by a grant from the Ontario Neurotrauma Foundation, Toronto, Ontario, Canada.
Data collection procedure
The questionnaire and accompanying information letter were distributed electronically from November 15, 2012, to January 15, 2013, using FluidSurveys (Fluidware, Ottawa, ON), a Canadian Web-based questionnaire system. 35 Based on a modified version of Dillman’s tailored design method,Reference Dillman 36 two reminder emails were sent in an attempt to maximize the response rate. All questionnaire responses were kept confidential, and only aggregated data were reported.
Data analysis
Analyses of the questionnaire data were performed using the statistical software tool, SPSS (Version 20). 37 Standard descriptive statistics were calculated, with 95% confidence intervals (CI) using the Wilson method when appropriate.
Results
Characteristics of respondents
In total, 115 health care providers (65%) responded to the questionnaire; 109 providers replied to questions for all three case vignettes; and 108 answered all questions for the three vignettes. Of those, 105 provided demographic information and identified their highest level of postgraduate training in pediatric emergency medicine (n=76), pediatrics (n=16), emergency medicine (n=11), family medicine (n=1), and sports medicine (n=1). Three respondents declined to provide demographic information. The great majority of respondents (89% [95% CI: 0.81, 0.93]) reported having diagnosed greater than 20 concussions per year. Demographic characteristics of the study participants are provided in Table 1.
Demographic characteristics of study participants
Knowledge of concussion diagnosis and understanding of initial management using return-to-play guidelines
The great majority of respondents (98 out of 109; 90% [95% CI: 0.83, 0.94]) were able to properly diagnose a concussion correctly in all three clinical vignettes. Similarly, 102 of 108 (94% [95% CI: 0.87, 0.97]) of respondents correctly applied the recommended guidelines to disallow return to play on the same day of injury (domain B1).Reference McCrory, Meeuwisse and Aubry 5 , Reference Giza, Kutcher and Ashwal 30 Respondents demonstrated a larger practice variation in the application of graduated return-to-play recommendations in a conservative stepwise progression, defined as 24 hours between successive steps. Correct application would result in an asymptomatic patient requiring 1 week to proceed through the six stepsReference McCrory, Meeuwisse and Aubry 5 (domain B2). Only 69 of 108 respondents (64% [95% CI: 0.54, 0.72]) applied a conservative graduated return-to-play recommendation in all three vignettes when queried about the possibility of return to full-contact sport competition at day four. For example, for vignette one, 33% of respondents would have possibly allowed a child to return to a hockey tournament 4 days after the initial concussion. In vignette two, 12% would have possibly allowed the adolescent to participate in a 5-km running race in 2 days. For vignette three, only 7% would have allowed participation in a hockey game in 4 days. Permission (or possible permission) to participate in these activities was deemed to be an incorrect application of conservative stepwise progression.
Practice variation of recommendations for cognitive rest
Respondents showed wide practice variation in recommendations for cognitive rest (domain B3, Table 2). Of note, 35% of respondents (95% CI: 0.27, 0.45) did not recommend a reduction of screen time (aggregate of computer, text messaging, video game, and television reductions). In total, only 40 of 108 (37%, 95% CI: 0.29, 0.46) recommended thorough cognitive rest, defined as a reduction of school, schoolwork, and all forms of screen time (Appendix 2 [Supplementary table of the remaining recommendations]).
Participant recommendations for cognitive rest
Use of existing concussion protocols
Only 22% (95% CI: 0.15, 0.30) of respondents reported having used common concussion scoring scales frequently or always (Table 3); 70% (95% CI: 0.61, 0.78) of respondents reported having used balance-testing tools frequently or always. In addition, 100% (95% CI: 0.97, 1.00) of respondents reported that they do not use computerized testing tools. Most (80% [95% CI: 0.72, 0.87]) reported that they use published guidelines frequently or always to guide clinical decisions regarding concussion, with only one person (1% [95% CI: 0.00, 0.05]) reporting having never used guidelines. Table 4 summarizes the reported use of commonly known concussion guidelines/policies.
Concussion scoring scales used “frequently” or “Always”
Reported Use of Concussion Guidelines/Policies
Conclusions
Only 18% (95% CI: 0.12, 0.26) of respondents reported that they frequently or always refer patients to concussion specialists. Respondents who refer to concussion specialists most often refer to sports medicine specialists (67% [95% CI: 0.57, 0.75]) and neurologists (50% [95% CI: 0.41, 0.60]). The most common barrier reported by those who never refer to concussion specialists (3% [95% CI: 0.01, 0.08]) is the lack of concussion specialists in their region (67% [95% CI: 0.21, 0.94]).
Discussion
Summary
This is the first broad Canadian analysis of pediatric emergency medicine physicians’ knowledge of concussion diagnosis and initial management practice patterns. The results in this study indicate that knowledge of concussion diagnosis is adequate for pediatric emergency providers in Canada. However, a wide-practice variation for the appropriate use of physical rest (return-to-play) recommendations and cognitive rest advice was found.
Although several previous studies demonstrate that knowledge gaps exist in the diagnosis and management of concussion by athletes, coaches, and trainers,Reference Provvidenza, Engebretsen and Tator 38 – Reference Sye, Sullivan and McCrory 40 few have examined practicing physicians’ knowledge of concussion diagnosis and early management. The findings in this study are consistent with other literature, suggesting that concussion knowledge and management deficiencies are pervasive. Recently, Zonfrillo et al. examined the self-reported knowledge and attitudes of 145 pediatricians and pediatric emergency physicians affiliated with a single institution at an urban, academic, tertiary-care centre in the United States, and reported that inadequate knowledge existed of pediatric concussion diagnosis and initial management.Reference Zonfrillo, Master and Grady 41 Boggild and Tator reported that gaps in general knowledge of concussion (e.g., concussion definition, signs/symptoms, initial management, and sequelae) existed in a survey of graduating medical students and postgraduate trainees in neurology and neurosurgery.Reference Boggild and Tator 42 Lebrun et al. found that family physicians in Alberta and North Dakota rarely recommended cognitive rest (47.5% and 28%, respectively), despite existing guidelines.Reference Lebrun, Mrazik and Prasad 43 Bazarian et al. examined pediatric, family, and emergency providers in Rochester, New York, on their knowledge of concussion grading and similarly found that the correct application of return-to-play guidelines was below 57%. That study examined knowledge of the pre-2002 definition of concussion.Reference Bazarian, Veenema and Brayer 44 Giebel et al. similarly noted a low rate of guideline use (23%) in a survey of emergency providers in Kalamazoo, Michigan.Reference Giebel, Kothari and Koestner 45
Given strong recommendations by the Zurich Consensus and American Academy of Neurology for a more conservative application of return-to-sport guidelines in children and adolescents,Reference McCrory, Meeuwisse and Aubry 5 , Reference Giza, Kutcher and Ashwal 30 it seems clear that further knowledge translation is required to improve clinician application of guidelines. The development and implementation of easy-to-use tools and management pathways have been demonstrated to be effective in aligning clinical practice with the best available evidence. Integrated knowledge translation with key end-users and stakeholders is essential to permit broad implementation of future guidelines for concussion.
This study reveals wide practice variation regarding cognitive rest advice. There is little evidence beyond expert opinion for the ideal cognitive and physical recommendations, and the observed wide practice variation may reflect the lack of high level (level A or B) evidence in this area.Reference DeMatteo, Hanna and Mahoney 46 A comparative trial is needed for both return-to-activity and return-to-school guidelines to determine the ideal components to improve outcomes after concussion.
This study employed a broad distribution in an effort to obtain a baseline understanding of how concussion is currently being diagnosed and managed. The majority of respondents indicated that they diagnose more than 20 concussions per year; therefore, our results reflect the practice for in excess of 2000 acute pediatric concussions across Canada annually. This study is the first examining practicing pediatric emergency providers’ practice variation of cognitive rest. Zonfrillo et al. examined the barriers to providing education to families following a pediatric concussion across pediatric providers at the Children’s Hospital of Philadelphia, but they did not describe the practice variation of cognitive rest instructions provided.Reference Zonfrillo, Master and Grady 41 The development of our questionnaire was systematic, and it strengthened by involving a group of experts from various fields.
Limitations
The nature of this study involved the potential for selection bias, specifically that those who chose to respond may have been more likely to have a greater concussion knowledge or familiarity. It has been found that in surveys of the health care professionals, response rates are often influenced by respondents’ characteristics and high workload demands.Reference Hill, Fahrney and Wheeless 47 Other limitations are similarly inherent to the use of surveys examining practice variation. For example, respondents may not practice in the manner that they indicate in the survey. Recall that bias may also have been a factor in this study’s findings (e.g., physicians may under- or over-report the number of concussion patients treated per year). The potential for variation in the interpretation of graduated scale questions also exists (e.g., different respondents may have different interpretations of the words sometimes or frequently). A prospective examination of discharge instructions by providers would provide a more accurate representation of practice variation. The lack of a single gold standard guideline for pediatric concussion is also a limitation, which is further limited by the large proportion of level C (expert opinion) evidence related to return-to-play and cognitive rest recommendations. There is a need for a large comparative trial examining the benefit of applying gradations of rest. The authors recognize that it is likely that the majority of children in Canada with a concussion are treated by general practitioners and at non-tertiary EDs. A recent study estimated the incidence of concussion to be approximately 6 of 1000.Reference Ryu, Feinstein and Colantonio 48 The authors’ team carried out a similar study of family doctors, general emergency physicians, nurse practitioners, and general pediatricians practicing in Ontario and found that similar gaps in knowledge existed across all provider types.Reference Zemek, Eady and Moreau 49
Conclusion
Gaps in knowledge and broad practice variation in recommending graduated return to play and cognitive rest following concussion exist among Canadian pediatric emergency physicians. The development and implementation of easy-to-use pediatric-specific concussion guidelines and management pathways have the potential to standardize clinical practice based on best-available evidence. Integrated knowledge translation with key end-users and stakeholders is essential to encourage the broader implementation of future pediatric concussion guidelines. Future research should address how best to overcome barriers to knowledge uptake and protocol application for the diagnosis and management of pediatric concussion.
Competing interests: None declared.
Supplementary material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/cem.2014.38
