Background/aim The risk of injury among Pee Wee (ages 11–12 years) ice hockey players in leagues that allow body checking is threefold greater than in leagues that do not allow body checking. We estimated the cost–effectiveness of a no body checking policy versus a policy that allows body checking in Pee Wee ice hockey.
Methods Cost–effectiveness analysis alongside a prospective cohort study during the 2007–2008 season, including players in Quebec (n=1046), where policy did not allow body checking, and in Alberta (n=1108), where body checking was allowed. Injury incidence rates (injuries/1000 player-hours) and incidence proportions (injuries/100 players), adjusted for cluster using Poisson regression, allowed for standardised comparisons and meaningful translation to community stakeholders. Based on Alberta fee schedules, direct healthcare costs (physician visits, imaging, procedures) were adjusted for cluster using bootstrapping. We examined uncertainty in our estimates using cost–effectiveness planes.
Results Associated with significantly higher injury rates, healthcare costs where policy allowed body checking were over 2.5 times higher than where policy disallowed body checking ($C473/1000 player-hours (95% CI $C358 to $C603) vs $C184/1000 player-hours (95% CI $C120 to $C257)). The difference in costs between provinces was $C289/1000 player-hours (95% CI $C153 to $C432). Projecting results onto Alberta Pee Wee players registered in the 2011–2012 season, an estimated 1273 injuries and $C213 280 in healthcare costs would be avoided during just one season with the policy change.
Conclusion Our study suggests that a policy disallowing body checking in Pee Wee ice hockey is cost-saving (associated with fewer injuries and lower costs) compared to a policy allowing body checking. As we did not account for long-term outcomes, our results underestimate the economic impact of these injuries.
- Ice hockey
- Children's injuries
- Contact sports
Statistics from Altmetric.com
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.
In Canada, sport injury accounts for 66% of youth injuries.1 Ice hockey has one of the highest injury rates among youth sports in Canada,2–5 and a strong body of research suggests that body checking increases the risk of injury.5–12 After years of debate, Hockey Canada and USA Hockey have recently raised the age at which body checking is allowed from the Pee Wee level (ages 11–12 years) to the Bantam level (ages 13–14 years). Prior to this policy change in Canada, it was demonstrated that ice hockey players in Alberta, where body checking was allowed at the Pee Wee level, were at a threefold greater risk of injury and fourfold greater risk of concussion than players in Quebec, where body checking was not allowed.7 Hockey Quebec disallowed body checking in Pee Wee over three decades ago in the light of high injury risk and disparity in player size.7
Treatment for injuries is costly given the immediate and long-term consequences, including increased risk of future injuries, cognitive deficits and other health burdens such as osteoarthritis.13–15 In 2004, the total cost of unintentional injuries, including those related to sports, amounted to $C16 billion in Canada.16 This places a large economic burden on parents of players participating. Further, ice hockey injuries are associated with opportunity costs by consuming limited healthcare resources that are paid for by the publicly funded healthcare system. Given that children enrolled in ice hockey are voluntarily exposed to excess risk of injury, are the arguably unnecessary costs associated with body checking placing an unfair burden on the healthcare system and taxpayers who fund it?
Using health economics to evaluate injury prevention strategies in terms of costs and effectiveness (ie, health gains) is a valuable tool for policymakers responsible for the allocation of healthcare resources and funds.17 Understanding the economic implications associated with body checking in youth hockey will place a monetary value on the newly implemented policies and inform policy in leagues where body checking is currently allowed.
The objective of this study was to examine the cost–effectiveness of disallowing body checking in Pee Wee ice hockey in Alberta (Canada). Standardised injury rates and healthcare costs associated with all injuries sustained throughout one season of Pee Wee ice hockey in Alberta, where policy allowed body checking, were compared to those in Quebec, where body checking was not allowed.
Sample, design and data acquisition
A cost–effectiveness analysis (CEA) was conducted alongside a prospective cohort study.7 Cohorts were selected based on exposure to a league that allowed body checking, using the following inclusion criteria: players aged 11–12 years during one season of play (October 2007 to March 2008); male or female players; written informed consent to participate; players registered with Hockey Calgary, Hockey Edmonton or Hockey Quebec; players participating in the Pee Wee age group only; players in the top 60% by level of play; agreement of the head coach to participate; and agreement of a team designate (coach, safety manager or other team parent) to collect information on individual player participation. Details regarding each cohort, recruitment and injury surveillance (including the preseason baseline medical questionnaire, weekly exposure sheet and injury report form (IRF)) were previously reported.7 Based on the primary research question examining injury risk, a sample size of 1944 (972 from each province) was deemed necessary for a minimally important incidence rate ratio of 2 or greater based on the expected concussion rate of 1/1000 player-hours in the Alberta cohort, adjusting for cluster and an anticipated dropout rate of 10% (two-sided test, α=0.05, β=0.2).7
Throughout one season of play, ice hockey injuries that required medical attention, resulted in the inability to complete a session or resulted in time loss from hockey were recorded and followed up by the research coordinators for up to 1 year following the end of the season. An IRF was completed by a team therapist for each injury to document the mechanism of injury, session type (game or practice), time loss (from sports, hockey, and work for player guardians) and medical follow-up as well as therapist and physician assessments.
Ethical approval (ID: 20252) was obtained from the University of Calgary, University of Alberta, McGill University, Université de Montréal and Laval University.
Effectiveness was measured using injury incidence rates (IRs), calculated as the number of injuries/1000 player-hours, to enable standardised comparisons between cohorts. Incidence proportion, calculated as the number of injuries/100 players, was used to provide meaningful results from the perspective of coaches, players, parents and policymakers.
Healthcare resource use and costs
Healthcare resources consumed by each injured player, including visits to healthcare professionals, treatments, services, supplies and equipment, were recorded on the IRF on an item-by-item basis.
The cost analysis was performed from both the partial societal and healthcare system perspectives. The partial societal perspective included direct costs to the publicly funded healthcare system and direct out-of-pocket medical costs required for private services and equipment not borne by the public healthcare system. Public and private costs were estimated separately to enable multiple audiences to extract relevant results.
Alberta fee schedules and unit costs were applied to the healthcare resources used by injured players in Alberta and Quebec in order to achieve a direct comparison of costs. Current (2012) unit costs and rules from the Alberta Health Services (AHS) Calgary Zone were applied to value healthcare system resources, including visits to physicians, specialists and emergency departments (EDs), surgery, X-rays, scans and fibreglass casts. Unit costs for out-of-pocket healthcare costs borne by families of players, including fees for physiotherapy, athletic therapy, massage therapy, reflexology, chiropractic services, ambulance rides, splints, braces, crutches and tensors, were obtained by contacting various providers in Calgary. Any non-current costs were adjusted to 2011 Canadian dollars using the Consumer Price Index 2009 medical basket (annual (2002=100), geography=Canada, commodities and commodity groups=healthcare).18 Transfer costs, such as goods and services tax, were not included in cost calculations. Unit costs, sources and relevant billing codes of all healthcare resources used by injured players, broken down by costs covered by the publicly funded healthcare system and out-of-pocket private healthcare costs, are available as Web Only tables.
Healthcare costs were estimated as the cost of injuries/1000 player-hours and cost of injuries/100 players, by province. The total cost for each cohort was calculated by summing the costs of all healthcare resources documented by injured players. The mean cost per injury was calculated as the total cost divided by the total number of injuries in each cohort.
To compare the joint distribution of costs and injury rates for Quebec incremental to Alberta, we calculated incremental cost–effectiveness ratios (ICERs): 1 2
If the Quebec cohort demonstrated lower costs and lower injury rates, the intervention was considered ‘cost-saving’ and an ICER was not calculated. We examined joint uncertainty around the estimates of costs and injury rates using cost–effectiveness planes, generated by bootstrapping point estimates of the ICER 10 000 times.
Projected budget impact analysis
A projected budget impact analysis was used to estimate the number of injuries and costs that could have been avoided in the population of hockey players registered in the 2011–2012 Pee Wee ice hockey season in Alberta if policy that did not allow body checking had been implemented. These estimates were obtained by multiplying the mean number of injuries per player and mean cost per player for each province in our study by the number of players registered in the 2011–2012 Alberta Pee Wee ice hockey season (N=9756).19 We took the difference between the projected injuries and costs based on the Alberta and Quebec data to obtain the injuries and costs avoided with a no body checking policy. Although our study only included players from teams in the most elite 60% divisions of play in Alberta, Emery et al4 previously demonstrated that the risk of injury in the lower third divisions of play did not differ from the combined risk among the middle and upper third divisions of play.
Calculations and statistical analyses were performed using Microsoft Office Excel 2007 and Stata/SE V.12.0.20 Injury rates were estimated with 95% CIs using Poisson regression, including player-hours as an offset, accounting for clustering effects by team. As the cost data were not normally distributed, bootstrapping with 10 000 replications was used to estimate percentile-based 95% CIs around the mean costs, accounting for clustering effects by team. Our primary analysis included all injuries (including game-related and practice-related injuries) and associated costs. Given that session type (game or practice) was previously shown to be an effect modifier, we also conducted a stratified analysis by this variable.7
For teams missing occasional weeks of weekly exposure hours, exposure data were imputed based on the mean game and practice hours in the weeks the team had complete exposure data. Given the consistency of ice time distribution for games and practices within a given hockey association and league, this was felt to be an appropriate estimate.
Twelve injuries (4%) were missing details related to healthcare resource use. Since this small proportion of injuries was unlikely to influence our results, we chose to construct a conservative scenario of resource use based on consultation with relevant experts in these instances. Further, we assumed that all ED visits took place at paediatric hospitals, and that players were not admitted to the hospital. We also conservatively assumed the use of upper-extremity fibreglass casts, enabling players to return to play as soon as possible with a cast, and off-the-shelf braces (specific to injury location and type). As details required to estimate costs associated with the use of tape, medication, first aid, heat and ice were not collected, these resources were not included in our analyses.
Of the 183 teams (90 in Alberta; 93 in Quebec) invited to participate in the study, 162 (75 (83.3%) in Alberta; 87 (93.6%) in Quebec) agreed to participate. A total of 1108 players from Alberta (821 in Calgary and 287 in Edmonton) and 1046 players from Quebec (567 in Montreal and 479 in Quebec City) completed the study. Ten teams (1 in Alberta and 9 in Quebec) dropped out of the study. Reasons for nonparticipation or dropping out were primarily at the team level, and were previously addressed.7
The distribution of baseline characteristics (including sex, height, weight, year of play, level of play and position of play) was similar in Alberta and Quebec, although a greater proportion of players in Alberta had experienced previous injury and concussion, and indicated a stronger preference for body checking.7 The median individual total season game-hours and practice-hours were also similar, with both provinces demonstrating significant variability in the IQR.7 The total number of injuries and participation hours observed for each province are presented in table 1.
Healthcare resource use and total costs
The number of healthcare resource units consumed by injured players and their total accumulated direct healthcare costs are presented in table 2. Table 3 provides a breakdown of the total costs by province, payer and session type. The total cost of healthcare resources used by injured players in Alberta was $C40 229 ($C28 406 in public costs and $C11 823 in private costs) versus $C15 111 ($C11 112 in public costs and $C3999 in private costs) in Quebec.
Province-specific IRs, by session type, are presented in table 4. The overall IR (including game-related and practice-related injuries) was higher in Alberta (2.83 injuries/1000 player-hours) compared to Quebec (1.11 injuries/1000 player-hours). The game-related IR was also higher in Alberta (4.20 injuries/1000 player-hours) compared to Quebec (1.37 injuries/1000 player-hours). For game-related and practice-related injuries, the estimated absolute risk reduction, or number of injuries that would be avoided in Alberta if body checking were not permitted, was 1.72 (95% CI 1.27 to 2.18) injuries/1000 player-hours for game-related and practice-related injuries and 2.84 (95% CI 2.18 to 3.49) injuries/1000 player-hours for game-related injuries. There were no significant differences between practice-related IRs. Similar patterns were observed for the number of injuries/100 players.
The mean cost per injury, including game-related and practice-related injuries, was similar in Alberta and Quebec ($C167 (95% CI $C135 to $C200) and $C166 (95% CI $C127 to $C210), respectively). There were no significant differences in the mean cost per injury between provinces for either session type. However, the cost/1000 player-hours and cost/100 players demonstrate that public and private healthcare costs associated with combined game-related and practice-related injuries and game-related injuries only would be significantly reduced if body checking were not permitted (table 4). The estimated total cost/1000 player-hours and total cost/100 players for game-related and practice-related injuries were over 2.5 times higher in Alberta ($C473/1000 player-hours) than in Quebec ($C184/1000 player-hours). For game-related injuries, the estimated costs were over 2.7 times higher in Alberta ($C683/1000 player-hours) compared to Quebec ($C230/1000 player-hours). The difference in costs between Alberta and Quebec was estimated at $C289/1000 player-hours (95% CI $C153 to $C432) for game-related and practice-related injuries, and $C453/1000 player-hours (95% CI $C242 to $C683) for game-related injuries only. The difference in costs was not significant for practice-related injuries ($C70/1000 player-hours (95% CI −$C58 to $C215)). Similar patterns were observed when estimating the cost/100 players.
Our point estimates for each session type demonstrate that policy in Quebec that does not allow body checking is cost-saving (associated with lower costs and lower injury rates) compared to policy in Alberta that allows body checking (table 4); therefore, ICERs were not calculated. In the cost–effectiveness plane for game-related injuries (figure 1A) and game-related and practice-related injuries (figure 1C), 100% of 10 000 bootstrap iterations lie in the southwest quadrant, demonstrating that policy that does not allow body checking in Quebec resulted in lower costs and lower injury rates than policy allowing body checking in Alberta. For practice-related injuries (figure 1B), policy that does not allow body checking resulted in lower costs and injury rates for 74% of the bootstrap iterations. The remaining bootstrap iterations are distributed approximately equally across the three other quadrants, indicating joint uncertainty around the point estimates of costs and injury rates.
Projected budget impact analysis
Table 5 displays the number of injuries and costs that we expect to have occurred during the 2011–2012 Pee Wee hockey season in Alberta when policy allowing body checking continued, compared to the estimated injuries and costs we expect to have occurred if a policy that did not allow body checking were implemented, based on the results of our study and participation rates from the 2011–2012 season. Among 9756 Pee Wee players in Alberta, an estimated 1273 injuries and $C213 280 in healthcare costs ($C146 476 in public costs and $C66 804 in private costs) would have been avoided in just one season. That is, public and private healthcare costs were 2.5 times greater than what would have been expected if body checking had not been allowed.
In addition to a threefold greater risk of injury, healthcare costs incurred by Pee Wee hockey players in Alberta, where policy allowed body checking, were over 2.5-fold greater than costs incurred by players in Quebec, where body checking was not allowed. Our CEA demonstrates that policy that does not allow body checking in Pee Wee hockey is cost-saving (associated with lower costs and fewer injuries) compared to policy allowing body checking. While other studies have examined the risk of injuries associated with allowing body checking for youth ice hockey players,4–11 this is the first study, to the best of our knowledge, to examine both injuries and their associated direct healthcare costs throughout one season of play.
With the limited available healthcare resources and funding, economic analyses have gained recognition as objective decision-making tools for policymakers across many disciplines of medicine.21 However, assessing injuries from an economic perspective is relatively new in the sport medicine field. One of the earliest identified CEAs related to injury prevention interventions examined a proprioceptive training programme to prevent ankle sprains among volleyball players using direct and indirect costs and a sensitivity analysis.22 A more recent economic evaluation examined a similar intervention using cost–effectiveness planes to assess uncertainty around the ICER.23 Krist et al24 also used ICERs and cost–effectiveness planes to examine the cost–effectiveness of an injury prevention programme on adult soccer players, though exposure hours were not taken into account in ICER calculations, limiting the comparability of results.25
While other studies have examined the impact of body checking on the rate of ED visits,9 ,26 ,27 injury-related costs were not assessed. Following a policy change which lowered the introduction of body checking by 1 year, Hagel et al26 found a twofold greater risk of injury presenting to EDs among 11-year-old players exposed to body checking. In contrast, Kukaswadia et al9 found no differences in the rate of injuries presenting to EDs. Following a similar policy change, Harris et al27 found that injury rates for fractures and injuries to the head and neck were lower (though not statistically significant).
The projection of our results onto the population of Alberta Pee Wee hockey players competing in the 2011–2012 hockey season was performed under the assumption that there are no costs associated with implementing a no body checking policy, and that injury rates in Alberta would be the same as those observed in Quebec if body checking was not allowed. We also assumed that injury rates and resource use patterns of players in rural Alberta did not differ from players in city leagues. Although the style of play and healthcare utilisation patterns may differ between rural and city leagues, the direction of these differences is unknown.
To ensure that a direct cost comparison could be made between Alberta and Quebec, unit costs and fee schedules obtained from Alberta were applied to healthcare resources utilised by players in both Alberta and Quebec. While direct costs of healthcare may differ in other provinces, the relative differences found in this study would remain consistent and are therefore generalisable to other provinces that allow body checking in Pee Wee ice hockey.
Limitations to our study related to dropout, missing exposure-hours and injury rates were previously addressed and do not suggest any systematic bias.7 While access to study therapists and physicians, standardised follow-up and return-to-play guidelines may have influenced healthcare utilisation practices, the direction of this influence is unknown and would not be expected to differ by province. We were unable to control for potential differences in healthcare utilisation practices between provinces, population differences or risk factors for injury (eg, year of play, previous injury or concussion, player size, position, level of play) in our analyses.
As described in the methods, we used conservative assumptions surrounding healthcare resource use for the few injuries that were missing details. In addition, we did not assess costs related to productivity loss for players and parents or guardians due to reduced working capacity since they did not fall within the scope of the perspective defined for our analysis. As these costs account for approximately 46% of the total costs associated with injuries in Canada,16 our results substantially underestimate the full economic impact of body checking in youth ice hockey. When extended across multiple provinces over many years, the costs arising from this voluntary excess exposure to injury place a considerable, unnecessary burden on the healthcare system.
Concern has been raised about the serious, long-term health consequences of these injuries, such as concussions.28–30 However, owing to the time horizon of our study (limited to one season of play), we did not capture the potential long-term impact of these injuries or the related implications for healthcare utilisation and quality of life. Given that the greatest differences in injury types between the two cohorts in our study were for concussions and fractures,7 our results may underestimate the magnitude of the difference in costs.
Our study demonstrates that delaying the age at which body checking is introduced is cost-saving. Going forward, recent policy changes in Canada and the USA provide a unique opportunity to prospectively examine the future injury and cost burden of body checking in youth ice hockey. Using an economic perspective to place a value on body checking may further inform other countries that continue to allow body checking in youth ice hockey.
What are the new findings?
Our results demonstrate that direct healthcare costs incurred by injured Pee Wee ice hockey players in Alberta, where policy allowed body checking, were over 2.5-fold greater than costs incurred by injured players in Quebec, where body checking was not allowed.
Our study provides evidence that policy that does not allow body checking in Pee Wee hockey is cost-saving (ie, fewer injuries and lower direct healthcare costs) compared to policy that allows body checking.
By examining the impact of injuries associated with body checking in youth ice hockey from an economic perspective, our results place a monetary value on the recently implemented body checking policies in Canada and the USA.
How might it impact on clinical practice in the near future?
Our results provide additional evidence to inform ice hockey governing bodies responsible for developing body checking policies in youth hockey.
Evaluating the cost–effectiveness of injury prevention strategies provides a unique, objective perspective that can inform policymakers responsible for the development of health policies.
The societal costs associated with excess injuries that occur through voluntary participation in youth sports require consideration when formulating health policies.
We thank the Pee Wee players, parents, coaches, team designates, physiotherapists, athletic therapists, student therapists, study coordinators and study physicians for their assistance in completing this research. This research would not have been possible without the support of Hockey Calgary, Hockey Edmonton, Hockey Quebec, Hockey Alberta and Hockey Canada.
Contributors SL, DAM, GC, NAK, JK and CAE contributed to the planning, conduct of analysis and reporting of the results. DAM is responsible for the overall content of the economic analysis. WHM contributed to the planning and reporting of the results. CAE is responsible for the overall content of this research as the guarantor.
Funding This study was funded by the Canadian Institutes of Health Research and Max Bell Foundation.
Competing interests DAM is supported by the Canada Research Chairs Program. CAE is supported by a Population Health Investigator Award from Alberta Innovates Health Solutions and a Professorship in Pediatric Rehabilitation in the Faculty of Medicine, University of Calgary (Alberta Children's Hospital Foundation).
Ethics approval University of Calgary, University of Alberta, McGill University, Université de Montréal and Laval University.
Provenance and peer review Not commissioned; externally peer reviewed.