Body-Checking Rules and Childhood Injuries in Ice Hockey

METHODS

Data from the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP, Health Canada) were used. CHIRPP is an injury-surveillance program that collects data on circumstances and outcomes of injury at selected pediatric emergency departments (mainly large teaching hospitals). The validity of CHIRPP data has been reported previously.^10,11 All participating hospitals in Ontario and Quebec were included except the Children's Hospital of Eastern Ontario in Ottawa because it serves children from Eastern Ontario and Western Quebec.

All boys between 10 and 15 years of age who presented to participating emergency departments with a hockey-related injury from September 1995 to the end of August 2002 were included. The patient (parents or child) is asked to complete information about the injury on the front of the CHIRPP form, including where the child was at the time of the injury, what the child was doing, and what went wrong to cause the injury. The treating physician completes clinical information on the back of the CHIRPP form.

Injuries were characterized as checking or nonchecking injuries based on the written description in the “what went wrong?” field. The methodology used was based on that developed by McFaull.¹² To be included as a checking injury, the written description had to include the words “checked,” “body checked,” “hit into boards,” or “hit by player” or “mis en échec,” “heurté,” “plaqué” (the common French terms for checking). All other injuries were classified as nonchecking injuries. Concussions and fractures were defined on the basis of the coding provided by the treating physician in the emergency department.

Because guidelines associated with body checking are related to the hockey level, boys were classified on the basis of their age at the time of the injury. Although the age groupings within the Canadian Hockey Association⁶ have now changed, the reported groupings of atom (10/11 years), peewee (12/13 years), and bantam (14/15 years) remained consistent throughout our study period. In 2002/2003 there were ∼89000 boys (10–15 years) registered in hockey in Ontario (population: ∼12 million) and 42000 boys (10–15 yeas) registered in Quebec (population: ∼7.5 million).

Checking was defined as “allowed” for (1) all 12- and 13-year-old players in Ontario (peewee level) and (2) 10- and 11-year-old players (atom level) in Ontario playing between 1998/1999 and 2001/2002 in all parts of Ontario except Kingston, which did not participate in the pilot project. Players were assumed to play hockey in the area in which the emergency-department visit occurred.

All 14/15-year-old players (bantam level) in both provinces were allowed to check, so the number of checking injuries and concussions are compared between provinces and analyzed separately.

The primary outcome was the proportion of injuries attributable to checking. In other words, the number of nonchecking injuries that was presented to a hospital served as an indicator of exposure to hockey, because we had no information on the number of players by hospital catchment area. Previous methodologic studies of CHIRPP data suggest that using age-specific proportions of injuries rather than counts is associated with improved validity of CHIRPP data.¹³

Statistical analyses were conducted by using the SAS 8.0 system (SAS Institute, Inc, Cary, NC). The odds ratios (ORs) for sustaining checking injuries (compared with nonchecking injuries) were calculated by using Mantel-Haenszel χ² statistics. For players in the younger age groups, the odds of suffering a checking injury, concussion, or fracture were compared between jurisdictions in which checking was allowed to areas in which checking was not allowed. For older players (14- and 15-year-olds), the odds of sustaining a checking injury, concussion, or fracture were compared between Ontario and Quebec. Players in Ontario had between 2 and 4 years of experience with allowed checking, whereas those playing in Quebec were experiencing permitted checking for the first time.

This study was approved by the Hospital for Sick Children's Research Ethics Board in March 2003.

RESULTS

There were 4736 hockey injuries reported to CHIRPP from participating hospitals in Ontario and Quebec from 1995 to 2002. Where checking was allowed, there were 2540 injuries. Table 1 lists the number and percentage of children with checking injuries, fractures, and concussions at the younger age-group levels in areas in which checking was allowed and areas in which checking was not allowed. The proportion of injuries attributable to checking was higher in areas in which checking was allowed. For example, for 10- and 11-year-olds, 49% of the injuries were checking-related where checking was allowed compared with 41% where checking was not allowed. The difference was more pronounced among 12- and 13-year-olds, with 48% of the injuries being checking-related where checking was allowed compared with 27% where checking was not allowed. The resulting OR for sustaining a checking injury (compared with a nonchecking injury) in areas in which checking was allowed was 1.83 (95% confidence interval [CI]: 1.58–2.11). Slightly higher proportions of concussions (OR: 1.42; 95% CI: 0.98–2.05) and significantly more fractures (OR: 1.25; 95% CI: 1.06–1.47) were also observed in jurisdictions in which checking was allowed.

Table 2 depicts the results for older (ages 14 and 15) players. At this level, checking is allowed in both provinces, so we compared the proportion of checking injuries, fractures, and concussions between Ontario and Quebec. Older players in Ontario were significantly more likely to have received a checking injury (OR: 1.90; 95% CI: 1.36–2.66). There was no significant difference in the proportion of concussions in this age group (OR: 1.6; 95% CI: 0.68–3.81) or in the proportion of fractures (OR: 0.82; 95% CI: 0.58–1.17).

DISCUSSION

Increased injuries attributable to checking were observed among younger players when checking was allowed. Injuries related to checking were more severe, with a statistically significant increase in the proportion of fractures and an increase in the proportion of concussions that approached statistical significance. Older players in Ontario, with earlier checking experience, had greater odds of sustaining checking injuries than their Quebec counterparts and slightly greater odds of receiving a concussion; therefore, it is unlikely that there was a protective effect of learning to check earlier.

The primary limitation of this study is that it is not population-based, so the calculation of rates is not possible. In addition, the participating hospitals are mostly pediatric teaching hospitals, which do not capture all hockey injuries. The results may not be generalizable to other health care settings. However, previous studies regarding the validity of CHIRPP data suggest that using age-specific proportions of injuries generally produces valid results for comparative studies.¹³

The injury data used in this study were gathered within the context of the health care system. All of the players in our study were treated at a hospital emergency department. The advantage of CHIRPP data is that the diagnoses and injury mechanisms are recorded very reliably, and the advantage of emergency-department data is that the injuries included are of sufficient severity to be worth preventing. Furthermore, routinely collected health care data minimize the possibility for bias resulting from parents' self-report in the context of a study of hockey-related injuries. Parents knowingly participating in a study of changes to hockey rules may over- or underreport the role of body checking in their child's injury; this concern would not apply to our study. Finally, this study compares injuries between 2 provinces with similar hockey organizations, aside from differences in their body-checking rules.

Checking was defined as “allowed” among 10- and 11-year-old players in Ontario subsequent to the rule change that allowed checking at the elite (rep) level. However, within the injury data, it was not possible to distinguish between players who played at the elite level from other players. We hypothesize that the relatively small difference in checking injuries seen in the youngest age group reflects this methodologic limitation.

Checking injuries were noted in situations in which checking was not permitted. In some cases, the injury may have been a result of contact with another player that was not intended to be a check. Hockey Canada defines body contact as a “defense tactic designed to legally block or impede the progress of an offensive puck carrier…. The defensive player may not hit the offensive player in an effort to initiate contact. There must be no action where the puck carrier is pushed, hit, or shoved into the boards.”⁶ Some of the injuries classified as checking injuries may have been the result of body contact rather than body checking. Some other checking injuries may have occurred in the context of informal play in which no rules were in place. However, we believe that the difference is unlikely to vary systematically between the 2 provinces and therefore are unlikely to bias the results presented here.

Injuries were classified as checking only when checking was mentioned explicitly as the cause in the written description. This approach may underestimate the proportion of contact injuries, which have been reported to account for up to 86% of hockey injuries.² By including injuries explicitly recorded as involving checking, we feel that we minimized the possibility of misclassification bias even if we may have underestimated the total number of checking injuries.

Several studies have highlighted the need for caution when allowing young children to body check in organized hockey.^1,8 The results of our study support their conclusions. Body checking among younger players was associated with an increased risk of injury, a risk that was not reduced among older players. Ice hockey has been associated with a high incidence of head injury.¹⁴

Current recommendations have been published by the American Academy of Pediatrics and the Canadian Academy of Sports Medicine. These recommendations are based on reviews of the literature and expert panels. The American Academy of Pediatrics guideline suggests that “[b]ody checking not be allowed in youth hockey for children 15 years old or younger.”⁷ Similar recommendations from the Canadian Academy of Sports Medicine propose that “Bantam hockey (age 14–15) is a more appropriate age at which to begin teaching the techniques, but in a graduated fashion (i.e. hip check and blocking only, no contact near boards). Full body checking can begin at the Midget (16–17) level when less variability between player's size exists.”¹⁵ Current practice in Quebec falls within these guidelines. The Canadian Hockey Association's current policy is to allow body checking at the peewee level (ages 12 and 13), and they endorsed the decision to allow body checking at the atom (ages 10 and 11) rep level in Ontario. Body checking at these levels is inconsistent with recommendations of both the American Academy of Pediatrics and the Canadian Academy of Sports Medicine. It is possible that training programs for players and coaches may mitigate the adverse effects of checking. Evaluation of such programs will continue to inform the debate about body checking.

What this study adds is evidence, from a hospital perspective, that ice hockey injuries resulting from body checking are more common when checking is permitted for younger players (ages 10–13). Severe injuries such as concussions and fractures are also more common when checking is allowed for younger players. In addition, a protective effect of learning to check earlier was not observed.

CONCLUSIONS

Policies that permit younger players to body check were accompanied by an increase in body-checking injuries, particularly fractures and a slightly increased risk of concussion. On the basis of our results, we suggest that children should play hockey only in noncontact leagues until at least the age of 14.