Article Text

Download PDFPDF

The risk of injury associated with body checking among Pee Wee ice hockey players: an evaluation of Hockey Canada’s national body checking policy change
  1. Amanda M Black1,2,3,
  2. Brent E Hagel1,3,4,5,6,
  3. Luz Palacios-Derflingher1,5,
  4. Kathryn J Schneider1,2,3,
  5. Carolyn A Emery1,2,3,4,5,6
  1. 1 Sport Injury Prevention Research Centre, Faculty of Kinesiology, University of Calgary, Calgary, Alberta, Canada
  2. 2 Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
  3. 3 Alberta Children’s Hospital Research Institute, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
  4. 4 O’Brien Institute for Public Health, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
  5. 5 Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
  6. 6 Department of Pediatrics, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
  1. Correspondence to Amanda M Black, Sport Injury Research Prevention Centre, Faculty of Kinesiology, 2500 University Drive NW, Calgary, Alberta T2N 1N4, Canada; ablack{at}


Background In 2013, Hockey Canada introduced an evidence-informed policy change delaying the earliest age of introduction to body checking in ice hockey until Bantam (ages 13–14) nationwide.

Objective To determine if the risk of injury, including concussions, changes for Pee Wee (11–12 years) ice hockey players in the season following a national policy change disallowing body checking.

Methods In a historical cohort study, Pee Wee players were recruited from teams in all divisions of play in 2011–2012 prior to the rule change and in 2013–2014 following the change. Baseline information, injury and exposure data for both cohorts were collected using validated injury surveillance.

Results Pee Wee players were recruited from 59 teams in Calgary, Alberta (n=883) in 2011–2012 and from 73 teams in 2013–2014 (n=618). There were 163 game-related injuries (incidence rate (IR)=4.37/1000 game-hours) and 104 concussions (IR=2.79/1000 game-hours) in Alberta prior to the rule change, and 48 injuries (IR=2.16/1000 game-hours) and 25 concussions (IR=1.12/1000 game-hours) after the rule change. Based on multivariable Poisson regression with exposure hours as an offset, the adjusted incidence rate ratio associated with the national policy change disallowing body checking was 0.50 for all game-related injuries (95% CI 0.33 to 0.75) and 0.36 for concussion specifically (95% CI 0.22 to 0.58).

Conclusions Introduction of the 2013 national body checking policy change disallowing body checking in Pee Wee resulted in a 50% relative reduction in injury rate and a 64% reduction in concussion rate in 11-year-old and 12-year-old hockey players in Alberta.

  • Ice hockey
  • concussion
  • injury
  • body checking
  • policy change
View Full Text

Statistics from


Youth ice hockey is a popular winter sport in North America with approximately half a million youth in Canada and the USA registered annually.1 2 While hockey has the benefits of many team sports including assisting with physical, motor and social development, it also carries a high risk of injury.3–7 One of the most common injuries resulting in time loss in youth hockey is concussion, accounting for 18%–66% of all injuries and raising significant health concerns.5–11 Systematic reviews examining the risk factors for injury and concussion in hockey have identified the tactical strategy of body checking as one of the most consistent risk factors for injury.12 13 Body checking is a defensive technique in which a player makes contact with another player moving in the opposite direction in an attempt to remove the player from the puck; it is often heralded for bringing excitement to the game at the professional level.14 15 While advocates of body checking believe that it is fundamental to the game, the age and level at which checking is introduced remain contentious.8 Several studies comparing provincial policy identified a twofold to fourfold increased risk of injury and concussion among Pee Wee players in leagues that introduce body checking at the Pee Wee age group (11–12 years) compared with leagues that introduce body checking later at both elite and non-elite levels of play.5 6 The strong research evidence along with meetings between researchers and members of local and national hockey associations contributed to USA Hockey’s decision to delay the introduction of body checking until Bantam (13–14 years) in 2012.16 17 Hockey Canada introduced a similar policy change in 2013.18 Previous reports examining the effect of body checking policy changes in Pee Wee using prospective community surveillance data have relied on comparing injury rates between provinces.5 6 The objective of this study is to determine if the risk of injury, including concussion, differs for Pee Wee (11–12 years) ice hockey players in the season following a national policy change disallowing body checking (2013–2014) compared with a season when body checking was allowed (2011–2012) in the same province. A secondary objective was to examine how mechanisms of injury differ between a cohort (2011–2012) where body checking was permitted and a cohort when it was not permitted (2013–2014).


Design, sample and data acquisition

A historical cohort study was conducted using data from two prospective cohort studies examining injury rates and mechanisms of injury in youth ice hockey.6 The first cohort included injury data collected from Pee Wee teams in all divisions of play in Calgary during the 2011–2012 season when body checking was permitted. The second cohort included data on Pee Wee teams and individual players from all divisions of play in Calgary collected during the 2013–2014 season when body checking was disallowed.

An inclusive sampling strategy was used for both cohorts; all Pee Wee hockey associations affiliated with Hockey Calgary were invited to participate in the study. Recruitment started at the association level. If the association agreed to allow researchers to contact team managers/coaches, the teams were invited to participate. Teams were included if they identified a parent volunteer (team designate) who could report weekly participation and initiate an injury report form (IRF). Only male teams were included in this analysis because female teams have never been permitted to body check. However, women playing on male teams were included. Teams are entitled to a maximum of 19 players though the number per team enrolled in the study varied depending on individual player assent and parent/guardian consent.15 In the 2013–2014 cohort, individuals on teams that did not have a team designate could also be included if they agreed to track their weekly participation and report injuries. Players who were concussed at cohort time zero were excluded.

Data collection procedures were based on a previously validated injury surveillance system described in detail in previous studies.5–7 11 19 Players completed a Preseason Baseline Questionnaire to collect information on participant characteristics and potential risk factors for injury (sex, height, weight, year of play, division of play, prior injury experience, self-reported lifetime concussion history and other injury within the last year) and attitude toward body checking.20 Trained research assistants also administered the Sport Concussion Assessment Tool21 22 (most current version 2 or 3 depending on year) to all participants. Throughout the season, team or individual designates recorded individual level weekly exposure to hockey sessions and injuries. Injuries were defined as any incident occurring during hockey that required medical attention and resulted in the inability to complete the session of activity and/or the inability to participate in sporting activities for 1 day. Data regarding how the injury occurred, date of injury, whether the injury occurred in game or practice, mechanisms, injury location, injury type and time loss from sport were reported on an IRF and validated by a study certified athletic therapist or physiotherapist. Injury mechanism categories included: (1) body checking; (2) other intentional player contact (elbowing, slashing, tripping, roughing, cross-checking, head contact); (3) incidental (unintentional) contact with another player or their equipment; (4) contact with the environment (puck, board, net); (5) no contact or (6) unknown. In cases where head contact or contact with the environment occurred as a result of a body check, the injury mechanism was coded as body checking. Injuries resulting in more than 7 days of time loss from sport were defined as severe injuries and referred to the study therapist for assessment. Any player with a suspected concussion was referred to a study sport medicine physician trained to diagnose concussion based on the most current consensus guidelines on concussion in sport.21 23 If the participant did not visit the study physician, a suspected concussion would be identified by the study therapist if the IRF indicated a mechanism that involved a force to the head or body resulting in symptoms of concussion in accordance with the most current consensus guidelines.21 23 To remain consistent with previous studies examining policy changes in youth hockey, concussions resulting in more than 10 days of time loss were also identified.5 6 11 24

Because of the nature of the historical cohort design, no a priori sample size calculation was performed. However, based on previous sample size calculations to investigate differences in injury rate associated with body checking, a sample size of 30 teams per cohort was determined as required to identify a difference between the two cohorts (powered based on an incidence rate ratio (IRR)=3.00; concussion rate=1.47/1000 player-hours in Alberta, adjusting for cluster and anticipated 10% dropout rate (α=0.05, β=0.20)).6

Statistical analysis

STATA V.14 was used to perform all statistical analyses.25 Baseline characteristics of participants were stratified by those who sustained at least one game-related injury throughout the season and described by cohort. Game-related injury rates and IRR with 95% CI were estimated based on Poisson regression adjusted for cluster by team with player exposure game-hours used as an offset. Models for primary injury outcomes (all game injuries and game concussions) were adjusted for previously identified risk factors for injury: year of play (first or second year of Pee Wee based on birth year), weight, previous injury within the last year or any previous concussion, level of play (elite divisions of play 1–3 (top 30%) or non-elite divisions of play 4–12 (lower 70%)), position and attitudes toward body checking (dichotomised at 75th percentile (36/55 items on a body-checking questionnaire) based on a previous sample of 2154 Pee Wee players).5 6 Sex was not considered a covariate because of low numbers. The primary adjusted risk factor analysis comparing cohorts was a complete case analysis where participants with missing covariate data were excluded. Because of low event rates of secondary outcomes, severe injuries and concussions resulting in more than 10 days of time loss, univariate models were reported (adjusting for cluster by team and offset by player-hours). The absolute risk reduction (ARR) was calculated based on the unadjusted injury rates between cohorts. The total population estimates for number of injuries and concussions that were prevented over the course of a season were calculated using the ARR, average game-hours and Pee Wee registration numbers in Alberta and across Canada provided in Hockey Canada’s 2015 annual report.26

A sensitivity analysis was conducted to determine the effect of missing data on the risk factor analysis using multiple chain imputation.27 Logistic regression of each covariate coded as missing or non-missing was used to assess the assumption of data missing at random. Missing weekly exposure hours were replaced for each player with values imputed based on average team level data or level of play-related data if the whole team was missing. An additional sensitivity analysis was conducted repeating the primary risk factor analysis with only physician-diagnosed concussion as an outcome. IRRs with 95% CIs associated with differences between specific mechanisms of injury between the two cohorts were estimated using univariate Poisson regression adjusted for cluster by team with player exposure game-hours used as an offset. Negative binomial regression did not lead to different results. Statistical significance was based on an α level of 5%.


In 2011–2012,  59 of the 115 teams (51.3%) approached participated in the study (n=883). In 2013–2014,  62 of 127 teams (48.8%) approached participated and 13 individuals from 11 teams completed data collection individually and were included in the study (n=618). The primary reason for a team’s refusal to participate was the inability to identify a team designate to collect data. The number of participants enrolled per team ranged from 5 to 19 (median=16) in 2011–2012 and from 2 to 17 (median=11) in 2013–2014. Participant characteristics are displayed in table 1. Baseline characteristics were similar between the two cohorts, with the exception of attitudes toward body checking and previous concussion. Players in the body checking cohort reported a stronger preference toward body checking (mean score=36.99 vs 29.49 for injured participants and 36.03 vs 30.23 for non-injured participants out of 55). A greater proportion of participants in the 2013–2014 cohort reported having previous concussion(s) when compared with the 2011–2012 cohort.

Table 1

Baseline characteristics comparing Pee Wee (11–12 years) hockey players in 2011–2012 and 2013–2014

Game injury rates

Table 2 summarises game outcome measures stratified by cohort. There were a total of 163 game injuries (104 concussions) reported in 2011–2012 and 48 game injuries (25 concussions) in 2013–2014. In 2011–2012, a total of 124 players had one injury, 15 players had two injuries and three players had three injuries. In 2013–2014, a total of 46 players had one injury and one player had two injuries. Injuries that met injury definition included the following: cuts, bruises, joint swelling, ligament sprains, fractures, muscle strains, tendinitis, dislocations and concussions.

Table 2

Summary of game-related outcome variables for Pee Wee ice hockey injuries in Calgary (2011–2012 and 2013–2014)

Risk factors for injury and concussion

Table 3 summarises the complete case risk factor analysis associated with injury outcomes. Policy disallowing body checking (when adjusted for year of play, previous history of injury or concussion in the last year, level of play, position and attitude toward body checking) was significantly associated with a decreased rate of all injuries (IRR=0.50, 95% CI 0.33 to 0.75) and concussions specifically (IRR=0.36, 95% CI 0.22 to 0.58). Previous injury within the last year (IRR=1.70, 95% CI 1.13 to 2.58) and previous concussion (IRR=2.48, 95% CI 1.64 to 3.76) were also identified as independent risk factors for injury and concussion. There were 1134 participants with complete data included in the estimation of the adjusted IRR for injury and 1145 participants for concussion. A sensitivity analysis using multiple chain imputation demonstrated consistent results when comparing the risk associated with policy that removed body checking to policy that permitted it (IRR=0.47, 95% CI 0.29 to 0.65 for injury and IRR=0.35, 95% CI 0.19 to 0.51 for concussion).

Table 3

Risk factor analyses for game-related injury, severe injury, concussion and concussion with more than 10 days of time loss in Pee Wee ice hockey in Alberta 2011–2012 and 2013–2014

Suspected concussions assessed by a physician were 70/104 (67%) in 2011–2012 and 20/25 (80%) in 2013–2014. A sensitivity analysis of risk factors including only concussions diagnosed by a physician yielded a similar adjusted protective effect estimate (IRR=0.46, 95% CI 0.26 to 0.81).

Mechanisms of injury

In 2011–2012, body checking was the most common mechanism of injury whereas in 2013–2014 the majority of injuries were due to incidental contacts. The injury rate associated with each mechanism of injury is displayed in figure 1. There was a significant difference between the rates of all injuries and concussions due to body checking between 2011–2012 and 2013–2014 (IRR=0.16, 95% CI 0.08 to 0.32 and IRR=0.21, 95% CI 0.09 to 0.45, respectively).

Figure 1

Game-related injury rates by injury mechanism and cohort in Pee Wee ice hockey, 2011–2012 and 2013–2014. Error bars indicate 95% CIs.


This is the first study to evaluate the protective effect of the 2013 national policy change by Hockey Canada removing body checking from Pee Wee controlling for player exposure hours. Consistent with provincial comparisons between leagues that permit and do not permit body checking in Pee Wee, this study found a significant reduction in injury and concussion risk associated with the removal of body checking.5 6 Furthermore, a comparison of the mechanisms of injury before and after the rule change suggests there has been a significant reduction in the number of injuries specifically due to body checking. Based on our ARR estimates (2.22 injuries per 1000 player-hours and 1.67 concussions per 1000 player-hours), an estimated 772 total injuries and 581 concussions in Alberta and 6388 total injuries and 4806 concussions across Canada are prevented every year as a result of the national policy disallowing body checking in Pee Wee.

The overall injury rate observed in this study following the body checking policy change (2.16 injuries per player-game-hours) was slightly higher than Ontario injury rates following the rule change removing body checking from non-elite Pee Wee in 2011–2012 (1.60 per 1000 player-game-hours).6 This may be due to the current study that considered all divisions of play including elite and non-elite players while the study in Ontario focused on non-elite players.6

The proportion of suspected concussions assessed by physicians increased in 2013–2014 relative to 2011–2012 (67%–80%). These findings may be the result of significant increases in concussion recognition and emphasis on the importance of seeing a physician.28 Increased recognition may also help explain why this study found a larger proportion of Pee Wee players in 2013–2014 reporting a history of concussion than Pee Wee players in 2011–2012. However, it is also possible that individuals with a previous history of concussion may be more likely to continue playing hockey in a non-checking league.

Pee Wee players playing in the 2013–2014 season when body checking was no longer permitted displayed a lower preference toward body checking in comparison with the cohort from 2011 to 2012. This result is consistent with a previous study describing attitudes toward body checking in Pee Wee.20 Interestingly, the difference in attitudes toward body checking may become more positive once the players reach Bantam and are allowed to body check. A previous study that reported preference toward body checking in Bantam for players who were exposed to body checking in Pee Wee and those who were not exposed reported similar body checking attitudes.11

One of the concerns suggested by those who challenge Hockey Canada’s decision to remove body checking from Pee Wee is that the rule change would lead to decreased interest in the sport. However, according to Hockey Canada’s annual reports, while registration for male Pee Wee players in Alberta may have experienced a slight decrease in numbers in 2013 (the year where the policy was implemented), registration increased in 2014 and 2015. This suggests an increase in the number of players who choose to continue playing in Pee Wee (registration nos 8638 in 2012; 8521 in 2013; 8725 in 2014 and 8768 in 2015).1 26 29 30

While this study was conducted using a Pee Wee sample in Alberta, it is expected that the reduction in injury rates after the policy change removing body checking would be generalisable to other hockey associations with Pee Wee teams. There are also important lessons that can be drawn from these results surrounding policy changes that aim to reduce exposure to key risk factors in other sports.


This study had a low recruitment rate in both cohorts (2011–2012: 51.3% and 2013–2014: 48.8% of teams approached). This may be due to participant’s and team designate’s burden in tracking injuries and exposures for the duration of the season. While it is possible that teams included in this study have increased risk perception and therefore may be more likely to report injuries, a clear injury definition minimises the risk of bias. Further, it is unlikely that this would have differed between the cohorts. Given participant characteristics are self-reported, there is a risk of measurement error where a recall bias or lack of awareness of concussion might lead to an under-reporting of previous history of injury or concussion. This could have led to an underestimation of the proportion of individuals with a history of concussion and resulted in non-differential misclassification bias. Some participants chose not to see a study physician or seek alternative medical assessment. This limits our ability to confirm all suspected cases of concussion. However, the sensitivity analysis including only medically diagnosed concussion suggests a similar reduction in concussion risk associated with the policy change. It is important to recognise that other factors may have changed between 2011–2012 and 2013–2014 beyond the introduction of the policy change, though the authors consulted with Hockey Calgary and there were no primary efforts for prevention in this population during this time.


Hockey Canada’s decision to delay body checking until Bantam was effective at reducing the risk of injury and concussion in Pee Wee ice hockey players. It is estimated that the implementation of this rule change prevented 722 injuries and 581 concussions in Alberta per year. The process in which policy changes were implemented in youth hockey provides promising approaches for injury and concussion prevention in other sports. Hockey remains a collision sport with many features that can increase an athlete’s risk of injury. There is a need to evaluate the effect of body checking policy in older age groups as well as other potentially modifiable risk factors that can reduce the risk of injury.

What new findings?

  • The 2013 national body checking policy change disallowing body checking in Pee Wee resulted in a 50% reduction in injury rate and a 64% reduction in the concussion rate in 11- and 12-year-old hockey players in Alberta.

  • An estimated 772 injuries and 581 concussions in Pee Wee were prevented in Alberta per season following the introduction of the Hockey Canada policy delaying body checking until Bantam.

  • Further research is necessary to evaluate the effect of body checking policy in older age groups.

How might it impact on clinical practice in the future?

  • Ongoing injury surveillance is essential to evaluate injury prevention strategies

  • Team medical staff or team safety designates have an important role in identifying suspected injuries/concussions in youth ice hockey


This research would not have been possible without the hard work of the Sport Injury Prevention Centre research coordinators and research assistants as well as the support of Hockey Calgary, Hockey Alberta, Hockey Canada, study therapists, study physicians, team therapists, team designates, coaches, players and parents.


View Abstract


  • Contributors AMB contributed to all research activities and takes responsibility for the integrity of the data and accuracy of the data analysis. She led the analysis and manuscript preparation. BEH and KJS contributed to the design of the study, acquisition of funding, analysis and interpretation of the results and critically reviewed the manuscript. LP-D contributed to the technical aspects of the data analysis, participated in data interpretation and critically reviewed the manuscript. CAE takes responsibility for the integrity of the data and the accuracy of the data analysis. She also led the design, acquisition of funding and all research activities, and critically reviewed the analysis and manuscript preparation.

  • Funding This study was funded by the Canadian Institutes of Health Research (CIHR), Alberta Children’s Hospital Research Institute and a grant from Alberta Innovates Health Solutions. AMB is supported by the Hotchkiss Brain Institute Dr T Chen Fong Doctoral Scholarship in Neuroscience. CAE is supported by an Alberta Children’s Hospital Foundation Chair in Pediatric Rehabilitation. The University of Calgary Sport Injury Prevention Research Centre is one of the International Research Centres for Prevention of Injury and Protection of Athlete Health supported by the International Olympic Committee. The role of the sponsors and funding bodies was to provide financial support for this research and the translation of the findings to all community partners.

  • Competing interests None declared.

  • Ethics approval Ethics approval for both the 2011–2012 and 2013–2014 cohorts was obtained from the Conjoint Health Research Ethics Board at the University of Calgary (Ethics ID# Nos. 20 252 and 14-0348 respectively). All player participants and parents provided written consent to participate.

  • 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.