Article Text

PDF

A systematic video analysis of National Hockey League (NHL) concussions, part II: how concussions occur in the NHL
  1. Michael G Hutchison1,2,
  2. Paul Comper1,3,
  3. Willem H Meeuwisse4,
  4. Ruben J Echemendia5,6
  1. 1David L. MacIntosh Sport Medicine Clinic, Faculty of Kinesiology & Physical Education, University of Toronto, Toronto, Ontario, Canada
  2. 2Injury Prevention Research Office, Keenan Research Centre, St. Michael's Hospital, Toronto, Ontario, Canada
  3. 3Graduate Department of Rehabilitation Science, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada
  4. 4Faculty of Kinesiology, Hotchkiss Brain Institute, Sport Injury Prevention Research Centre, University of Calgary, Calgary, Alberta, Canada
  5. 5Psychological and Neurobehavioral Associates, Inc., State College, Pennsylvania, USA
  6. 6University of Missouri—Kansas City, Kansas City, Missouri, USA
  1. Correspondence to Dr Michael Hutchison, David L. MacIntosh Sport Medicine Clinic, Faculty of Kinesiology & Physical Education, University of Toronto, 55 Harbord Street, Toronto, Ontario, Canada M5S 2W6; michael.hutchison{at}utoronto.ca

Abstract

Background Concussions in sports are a growing cause of concern, as these injuries can have debilitating short-term effects and little is known about the potential long-term consequences. This work aims to describe how concussions occur in the National Hockey League.

Methods Case series of medically diagnosed concussions for regular season games over a 3.5-year period during the 2006–2010 seasons. Digital video records were coded and analysed using a standardised protocol.

Results 88% (n=174/197) of concussions involved player-to-opponent contact. 16 diagnosed concussions were a result of fighting. Of the 158 concussions that involved player-to-opponent body contact, the most common mechanisms were direct contact to the head initiated by the shoulder 42% of the time (n=66/158), by the elbow 15% (n=24/158) and by gloves in 5% of cases (n=8/158). When the results of anatomical location are combined with initial contact, almost half of these events (n=74/158) were classified as direct contact to the lateral aspect of the head.

Conclusions The predominant mechanism of concussion was consistently characterised by player-to-opponent contact, typically directed to the head by the shoulder, elbow or gloves. Also, several important characteristics were apparent: (1) contact was often to the lateral aspect of the head; (2) the player who suffered a concussion was often not in possession of the puck and (3) no penalty was called on the play.

  • Concussion
  • Contact sports
  • Head injuries
  • Injury Prevention
  • Ice hockey

Statistics from Altmetric.com

Introduction

In the past decade, there has been a considerable attention on the issue of concussions in hockey, as it is one of the most common injuries in the sport, occurring at all levels of play and age groups.1–7

In a previous paper, we described ‘the four W's’ of concussions sustained by National Hockey League (NHL) players, namely: Who, When, Where and What?8 Building on our previous work, the purpose of the present paper was to describe how concussions occur in the NHL. In addition to identify the principal mechanisms associated with concussions, two specific contextual issues (ie, puck possession and penalty) will be addressed.

Methods

NHL players suspected of having a concussion were initially identified by team athletic trainers or therapists and then evaluated by team physicians who diagnosed concussion according to the Zurich (2008) Concussion in Sport definition.9 Digital video records of diagnosed concussions from the NHL injury surveillance system were provided to the researchers by NHL Operations for regular season games occurring from 4 October 2006 to 31 December 2009. The study was approved by Ethics Review Board of the Office of Research Services at the University of Toronto (protocol reference #23882).

Heads up checklist

The Heads Up Checklist (HUC) provides a standardised framework for coding injury mechanisms of concussion using digital video records.10 The HUC contains 17 specific factors of interest organised into three broad sections: (1) Event, (2) Game Situation and (3) Equipment. The majority of this paper will focus on the analysis of factors related to the Event section. See table 1 for summary of operational definitions from the HUC Event Section. Two additional HUC factors of interest and relevance to the present paper were puck possession and penalty. Puck possession refers to the situation in which a player had control of—although was not necessarily in direct contact with—the puck while attempting a hockey skill (eg, skating or passing). Consequently, a player deemed to be in possession of the puck at the time of contact with another player was coded as ‘puck possession—yes’. A player deemed not to have possession of the puck at the time of contact was coded as, ‘puck possession—no’. The situation in which a player completed a pass or a shot following the possession of the puck was coded as ‘just released’. Time to contact (ms) was documented post hoc in cases where puck possession was classified as just released.

Table 1

Operational definitions of terms from the Heads Up Checklist—Event section

Procedure

Two raters independently viewed the digital video records of events associated with physician-diagnosed concussions. Raters viewed each event using Quicktime Player Pro V.7.6.6 software residing on an Apple MacBook Pro 4.1 (operating software Mac OS X V.10.5.8).11 Each category of the HUC was completed for each distinct concussion event. Raters were allowed to view the event as many times as required, at any playback speed deemed necessary to complete all categories on the HUC. In cases where there was initial disagreement between raters, video records of events were reviewed to determine consensus. If consensus was not reached between the two raters, a third rater forced the consensus.

Outcome measures

Digital video records of concussions were coded using the HUC. The following factors from the Event section of the HUC were analysed: scenario (eg, opponent, team-mate, fall, other); initial contact with (eg, shoulder, elbow, stick, etc); region (eg, head, neck, torso or below waist); anatomical location (eg, anterior, posterior or lateral); and acceleration of head (eg, sagittal or coronal, transverse or multiplane). The two additional factors analysed from the HUC were puck possession (eg, yes, no or just released) and penalty (eg, yes or no).

Statistical analyses

Descriptive statistics of the outcome measures of interest were limited to coded events leading to concussion (n=197) or the subset of concussions that involved contact with an opponent, but did not due to fighting (n=158). All analyses were performed using Statistical Analysis Software, V.9.2.

Results

A total of 260 concussions occurred in 4299 NHL regular season games from the beginning of the 2006–2007 season to the end of December 2009 (6.05/100 games). Seventy-six per cent (197 of the 260 concussions) were identified on video and analysed for the present study. Of the events that were not coded by the raters, 19 were excluded because of poor image quality or lack of appropriate camera angles. The remaining events (n=44) were not included in the present study because the video record could not be matched to the concussion event, or the concussion occurred outside of game play, for example, during a warm-up or a practice setting.

Scenario

The greatest proportion of events associated with concussion involved contact with an opponent (n=174/197, 88%). Of the 174 events involving contact with an opponent, 9.2% (n=16/174) were due to fighting. With respect to the 16 events classified as fights, all involved direct contact to the head and 75% (n=12/16) involved secondary contact, most often by contact to the head by an opponent's fist or by the player's head striking the ice surface. Those concussions involving contact with an opponent, but not involving fighting (174 – 16(fights)=158), were used for the majority of the analyses that follow.

Initial contact with

In order to describe the mechanisms of injury leading to concussion, each event was analysed according to its component elements. The first step was to describe the body part of the ‘hitter’ that initiated contact with the concussed player, referred to as initial contact with. As the aggregate data illustrate in figure 1, the most common initial contact to any part of the player's body leading to a concussion was by a hitter's shoulder (55%), followed by the elbow (21%) and gloves (12%), for a total of 88%. The remaining 12% of initial contacts were classified as stick, puck, helmet, knee or inconclusive.

Figure 1

Initial contact with: first body part of the hitter that initiated contact with the concussed player.

Region

The next stage in the analysis was to identify the region or part of the body on the concussed player that was initially contacted by the hitter, broadly classified as head, neck, torso or below the waist. The most common body part initially contacted was the player's head, 68% (n=108/158), while the torso region accounted for approximately 28% (n=44/158). The remaining percentage (4%, n=6) was classified as below the waist, below the neck or inconclusive.

With respect to contacts that occurred by hitters directly to a player's head, the contact was initiated by the shoulder 42% of the time (n=66/158), by the elbow 15% (n=24/158) and by gloves in 5% of cases (n=8/158; see figure 1). In summary, 62% (n=98) of the concussions in the sample of 158 were directly attributed to shoulders, elbows or gloves delivered by the hitter to the head of an opponent during the flow of the game.

Body location and acceleration of head

A further refinement was made to identify the specific anatomical aspect (ie, anterior, posterior or lateral) of the region of body of the concussed player first contacted. For example, if a player was struck directly in the mouth by an opponent's elbow, then the body location would be classified as ‘anterior.’ Table 2 provides a breakdown of events identified as contact to the torso or head, stratified by body location (anterior, posterior or lateral), from the subset of 158 events involving an opponent. The lateral aspect of the head or torso was the most common area of initial contact (n=92/158, 58%), while initial contact of the anterior and posterior aspects of the head or torso accounted for 22% (34/158) and 16% (n=25/158), respectively.

Table 2

Contacts to the head or torso stratified by body location, % (n)

When the results of anatomical aspect are combined with initial contact, almost half of these events (n=74/158) were classified as direct contact to the lateral aspect of the head. Of importance, a large proportion of these were by shoulder, elbow or gloves (n=71/158). Taken together, approximately 35% (n=71/197) of all concussions during the three and half seasons were classified as occurring by direct impact to the lateral aspect of the head by shoulders, elbows or gloves.

In addition to the anatomical aspect, raters also documented observable head motion during events that resulted in diagnosed concussions. Head acceleration frequently occurred in multiple planes, typically in the sagittal and transverse planes (n=62/158). There was a relatively even distribution of events classified as sagittal or coronal plane (n=40/158) and transverse plane (n=42/158). Of note, 12% of the events were deemed to have no observable head motion or were inconclusive for observable head motion (n=18/158).

Puck possession

When considering the events (excluding fighting) that involved an opponent (n=158), the concussed player was in possession of the puck approximately 23% of the time (n=37), while 34% (n=54) of the time the player had no possession of the puck. In 42% (n=67) of the events, the player who suffered a concussion had ‘just released’ the puck; that is, the player had been in possession of the puck during the play, but was no longer in possession of it when the player was contacted by an opponent. Of those classified as ‘just released’, 49 events permitted quantification of time-frames to analyse the time from puck release to contact. In over 70% of these events, the time from puck release to contact was less than 0.5 s. Although not clearly specified in the rules of hockey, this time ‘lag’ for delivering a hit to an opponent is generally considered to be acceptable.

Penalty

A final point of interest was to determine if the on-ice officials called a penalty on the play for the events that resulted in concussions. Of the events evaluated, there was sufficient information on the digital video records to determine whether a penalty was called for 168 events. Of these, 29% (n=49) were considered infractions according to on-ice officials. When looking specifically at direct contact to the head by shoulder, elbow or gloves, 22% (n=37) were deemed to be rule violations. In other words, applying the then-current rules to the player-to-opponent contacts that resulted in concussions, less than one-quarter of the direct contacts to the head were considered penalties by on-ice officials.

Discussion

For three and half NHL regular seasons evaluated as part of this study, the predominant mechanism of concussion was consistently characterised by player-to-opponent-player contact, typically directed to the head by the shoulder, elbow or gloves. When this mechanism was elucidated further, several important characteristics were apparent: (1) contact was often to the lateral aspect of the head; (2) the player who suffered a concussion was often not in possession of the puck and (3) no penalty was called on the play.

With the combination of speed, hard surfaces and frequent collisions in hockey, it is reasonable to expect that a proportion of hockey concussions might result from a collision with a team-mate or by accidental falls or trips. The present data indicate that accidental falls or trips and collisions with team-mates accounted for only 10% of concussions analysed and are therefore a relatively rare occurrence.

More than three quarters of all events involving player-to-player contact occurred when neither player had possession of the puck, suggesting a lack of situational awareness or vulnerability as a significant contributing factor to injury. Consequently, educating athletes and coaches that there is a high probability that concussions can occur even when the player is not handling the puck is imperative.

Applying the rules of play during the study period, on-ice officials did not penalise the majority of the events leading to concussion. The finding of a clear mechanism of concussion identified in the majority of events analysed combined with few on-ice penalties provides evidence for the inclusion of a rule-based approach to help reduce the occurrence of the behaviour and safeguard players from playing situations that could lead to concussions. A rule-based approach to ensure player safety has previously been successful (eg, ‘spearing’ in football, hitting from behind in hockey) and is a common method to define unacceptable behaviour and ensure player safety. Since the study period, the NHL introduced Rule 48 ‘Illegal Check to the Head’, which states that ‘a lateral or blind side hit to an opponent where the head is targeted and/or the principal point of contact is not permitted’.12 This rule—introduced in 2010—was revised in 2011 to encompass targeted contact to the head.

In addition to specific rules and additional sanctions, changes to equipment standards are also potential avenues to reduce the frequency of concussions in hockey. The present study identified a common mechanism of injury and the results may inform future research and area of focus for equipment manufacturers. Helmet manufacturers are attempting to design helmets that incorporate distinct features meant to improve energy attenuation in response to lateral blows.13 Moreover, the high proportion of shoulders and elbows as the principal point of contact initiation may warrant the investigation of equipment improvements to lessen the forces of impact upon contact.

The findings from this study revealed that a particular injury mechanism, that is, a direct blow to the head, accounted for a large number of concussions. However, those familiar with hockey have witnessed collisions that appear to involve significant forces applied directly to a player's head, yet with no adverse outcome (ie, concussion). Meeuwisse14 has suggested that it is important to measure and understand this ‘mechanism of no injury’, to refine the understanding of which component(s) of the apparent mechanism of concussion may be responsible for injury. In addition, capturing and analysing direct hits to the head that do not result in concussion will allow for a more comprehensive understanding of the relative risk of such actions, thereby providing valuable information regarding this potentially debilitating injury.

Limitations

Structured comprehensive video analysis is a promising tool to analyse the mechanisms of specific injury types, successfully applied in a variety of sports.15–18 However, video analysis is not without limitations. Although this study concurs with previous work identifying the possibly significant contribution of rotational acceleration,19–21 the limitations of video recordings did not allow for a detailed biomechanical analysis that quantified linear and rotational accelerations.

Second, the present results were restricted to a dataset arising from video records corresponding to NHL players diagnosed with concussions. Although these data are thought to accurately represent the mechanism of injury in the NHL, they may not be representative of other leagues with players of different ages, skill levels and physical characteristics. Furthermore, it is not clear how these data relate to leagues that have different rules than the NHL.

Conclusion

Professional hockey carries an inherent risk of concussion. To our knowledge, this is the first study that has systematically assessed the antecedent events leading to concussions using video analysis. We were able to identify that the majority of concussions during the study period occurred by direct contacts to the head by the shoulder, elbow or gloves. The results from the present study should be considered in the development and implementation of future prevention measures and strategies. Reducing the frequency of concussions in the NHL will require a multisystems approach that includes a number of education and prevention strategies (eg, rules, education, discipline, etc).

What are the new findings?

  • The study identified a predominant mechanism of concussion in the NHL over a 3.5-year period during the 2006–2010 seasons. The mechanism of concussion was consistently characterised by player-to-opponent-player contact, typically directed to the head by the shoulder, elbow or gloves.

  • When this mechanism was elucidated further, several important characteristics were apparent:

  • Contact was often to the lateral aspect of the head;

  • The player who suffered a concussion was often not in possession of the puck;

  • No penalty was called on the play.

How might it impact on clinical practice in the near future?

  • Medical professionals should be cognizant of the predominant mechanism associated with concussions in professional hockey and thoroughly evaluate and monitor athletes with suspected concussions.

  • The finding of a clear mechanism of concussion identified in the majority of events analysed combined with few on-ice penalties provides evidence for the inclusion of a rule-based approach to help reduce the occurrence of the behaviour and safeguard players from playing situations that could lead to concussions.

Acknowledgments

The authors are grateful to members of the NHL/NHLPA Concussion Working Group: Dr Mark Aubry, Mr Bill Daly, Ms Maria Dennis, Ms Julie Grand, Dr John Rizos, Mr Rob Zamuner and also Mr Jim Ramsay for their insight throughout the research study. We would also like to thank Ms Cynthia Lichter of NHL Operations for her ongoing assistance with the videos. PC and MGH would like to acknowledge Dr Angela Colantonio and Dr Doug Richards of the University of Toronto for their ongoing advice and support through the development of the manuscript.

References

View Abstract

Footnotes

  • Correction notice This article has been corrected since it was published Online First. In the Results section, 'The remaining events (n=43)' has been changed to 'The remaining events (n=44)'.

  • Contributors MGH and PC had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analyses. WHM and RJE were also main contributors to the content of this manuscript and were involved in the study's design. MGH mainly contributed to the statistical analysis. All authors interpreted the results, drafted the manuscript, decided to submit the manuscript and involved in the critical revision of the manuscript.

  • Funding This study received funding by Ontario Neurotrauma Foundation (ONF) and The Pashby Sports Safety Fund. The organisations that contributed funds to the research had no role in the design of the study, the collection, analysis and interpretation of the data.

  • Competing interests All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. PC is a member of the Concussion Working Group and a clinical neuropsychologist consultant with the NHLPA, for which he receives remuneration, but he did not receive any funding from the NHL or the NHLPA to conduct the present research. WHM is a member of the Concussion Working Group and medical consultant to the NHL, for which he receives remuneration, but did not receive any funding from the NHL or NHLPA to conduct the present research. RJE is a consultant to the NHL, MLS and US Soccer Federation for which he receives financial remuneration. No financial remuneration was received for activities specifically related to this research. MGH received graduate training funding from Ontario Neurotrauma Foundation and The Pashby Sports Safety Fund.

  • Ethics approval Ethics Review Board of the Office of Research Services at the University of Toronto.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles