Ice hockey is a rapidly growing sport across 76 countries1. It is widely recognized as one of the most physically demanding sports worldwide, with high injury rates. Risks associated with ice hockey include high-speed collisions, sudden changes in direction, and contact with skate blades, sticks, and frozen pucks2-4. In the 2010 Olympic Winter Games, ice hockey players had one of the highest injury rates, with 18% of players affected5.
Previous studies analyzing the International Ice Hockey Federation Adult World Championship tournament for several years found injury rates of 14.2 per 1,000 player-games and 6.4 per 1,000 player-games, respectively6,7. These high injury rates can have debilitating consequences for the player and the team, including loss of playing time, risk of injury recurrence, reduced performance, and deterioration of quality of life8,9.
Several types of injuries can happen while playing hockey10. However, the most common injuries in ice hockey are acute injuries such as lacerations, contusions, sprains, concussions, and fractures. The most prevalent among the aforementioned injuries are fractures7,11. The risk of injury is affected by a number of variables such as player position, game versus practice exposure, and the level of participation12,13. Therefore, precise information on the prevalence, incidence, characteristics, localization, and mechanisms of injury is crucial in effectively preventing and managing injuries in ice hockey.
Despite the growing popularity of ice hockey in Korea after the 2018 Pyeongchang Winter Olympics14, there is a lack of research on the injuries associated with it. The study aimed to bridge this gap and identify the epidemiology of elite adult ice hockey injuries in Korea, including injury characteristics, locations, and mechanisms, using a self-reported questionnaire. A comprehensive understanding of ice hockey injuries is essential for effective injury prevention, treatment, and a safe return of players to sports.
The retrospective study utilized a self-reported questionnaire to collect the injury records of ice hockey players from three semiprofessional teams in Korea. The Institutional Review Board of Korea University Anam Hospital (Seoul, Republic of Korea) approved the study protocol (No. 2022AN0466). Written informed consent was obtained from all the participants.
During the period of 2018 and 2019, 68 players from three semiprofessional ice hockey teams in Korea voluntarily consented and were thus included in the study. The injuries were retrospectively investigated, and each player completed a questionnaire following a brief explanation of the survey.
Inclusion criteria were as follows: (1) written informed consent from the players; (2) orthopedic injuries; (3) moderate-to-severe injuries resulting in time loss of ≥7 days; (4) injuries that occurred either during ice hockey competition or training; (5) injuries that occurred post-high school; (6) injuries evaluated by a physician and diagnosed in clinics; and (7) agreement of the head coach to ensure player participation and to collect injury information. The exclusion criteria were as follows: (1) foreign players; (2) non-orthopedic injuries, such as a concussion or facial laceration; and (3) minor injuries resulting in a time loss of <7 days.
A self-report questionnaire was developed by the research team to collect data on injuries resulting from ice hockey competitions and training (Fig. 1). The questionnaire was partially adapted from the National Collegiate Athletic Association Injury Surveillance System and other previously reported injury surveillance systems4,6,7,9,15. It included demographic and injury characteristics such as age, ice hockey career, position, stick side, number and location of injuries, cause of injury, circumstances in which the injury occurred, treatment, and the return process. Moreover, to reduce recall bias, the questionnaire was designed to provide a clear and specific definition of injury, with prompts to assist participants in recalling events and injuries16. Prior to administering the questionnaire, the researchers provided a detailed briefing to the players who were participating. This briefing included the objectives of the study and essential elements of the questionnaire. In the study, an injury was defined as one that (1) occurred due to participation in either ice hockey training or competition and (2) resulted in performance limitations, leading to an inability to maintain the usual training load for more than 7 days. In cases where a player had multiple injuries, the injury was considered independent if it affected a different body part or if the player incurred the same injury in the same location after achieving complete recovery from the previous one. To ensure the reliability of the results obtained from the questionnaire, all the ice hockey players included in the study completed the questionnaire twice within a 3-month interval17. In case of any differences between the results of the two questionnaire, the researchers, players, trainers, and coaches gathered and discussed until an accurate result was drawn, and any disagreements were resolved by consensus. Before statistical analysis, all data were reviewed for accuracy and completeness by the researchers, and any ambiguous or missing information was explained by the player and/or trainer.
Fig. 1. Injury survey questionnaire form.
Continuous variables, such as age and length of career, were presented as mean±standard deviation, while categorical variables, including stick side, player position, injury experiences, circumstances, mechanisms, locations, injuries that required surgeries, and decision-makers for returning to sports were reported as numbers (percentages). Subgroup analyses were performed using the Pearson chi-square test to identify the injury characteristics in two groups stratified by the decision-maker for returning to sports (“By medical staff” vs. “By players themselves or non-medical staff”). Additionally, logistic regression was conducted to examine the effects of each injury characteristic, including players’ position, circumstances, mechanisms, anatomic regions of injury, fracture status, and injuries that required surgery, the decision of the medical staff to allow players to return to sports.
All statistical analyses were conducted using Python software (Python Foundation; https://www.python.org). Two-sided p-values were reported, and statistical significance was set at a p-value less than 0.05.
This study included a total of 68 players, with a mean age of 27.00±3.58 years and a mean career length of 17.51±4.19 years (Table 1). Among the players, 21 (30.9%) used the left stick. The positions of the 68 players were as follows: 38 forwards (55.9%), 22 defensemen (32.4%), and eight goalies (11.8%). Moreover, 10 players (14.7%) had never experienced any moderate-to-severe orthopedic injuries that might have resulted in a time loss of more than 7 days, while 17 players (25.0%) had experienced three or more moderate-to-severe orthopedic injuries.
Table 1 . Baseline characteristics included 68 semiprofessional players
| Characteristic | Data |
|---|---|
| No. of players | 68 |
| Age (yr) | 27.00±3.58 |
| Lengthof career (yr) | 17.51±4.19 |
| Left stick side | 21 (30.9) |
| Position | |
| Forward | 38 (55.9) |
| Defensemen | 22 (32.4) |
| Goalie | 8 (11.8) |
| Injuryexperiences (time) | |
| None | 10 (14.7) |
| 1 | 25 (36.8) |
| 2 | 16 (23.5) |
| ≥3 | 17 (25.0) |
Values are presented as number only, mean±standard deviation, or number (%).
Among the players, 58 had sustained a total of 120 injuries (Table 2). Of these, 93 injuries (77.5%) occurred during games, whereas 27 (22.5%) during training. The most frequent injury mechanism was player-to-player contact resulting in 66 injuries (55.0%), followed by noncontact mechanisms with 23 injuries (19.2%), player-to-puck or stick contact with 16 injuries (13.3%), and player-to-board contact with 15 injuries (12.5%). The most commonly affected anatomic regions were the shoulder in 34 patients (28.3%) and the knee in 24 patients (20.0%), followed by the wrist and hand, foot and ankle, pelvis, lower back, upper arm, and elbow (Fig. 2). Of 120 cases, 16 injuries required surgery, with seven shoulder injuries (43.8%), four involving the forearm, wrist, and hand (25.0%), two affecting the knee (12.5%), two affecting the upper arm and elbow (12.5%), and one affecting the lower leg, foot, and ankle (6.3%). Medical staff made the decision regarding the players’ return to sports in 53 cases (44.2%), whereas players and nonmedical staff made that decision in 67 cases (55.8%).
Table 2 . Characteristics of 120 injuries in 58 players with injury experience
| Characteristic | Data |
|---|---|
| Circumstances of injury | |
| Game | 93 (77.5) |
| Practice | 27 (22.5) |
| Mechanisms of injury | |
| Player-to-player contact | 66 (55.0) |
| Player-to-board contact | 15 (12.5) |
| Player-to-puck or stick contact | 16 (13.3) |
| Noncontact mechanism | 23 (19.2) |
| Injuries by anatomic region | |
| Shoulder | 34 (28.3) |
| Upper arm and elbow | 7 (5.8) |
| Forearm, wrist, and hand | 19 (15.8) |
| Lower back | 9 (7.5) |
| Pelvis, groin, and thigh | 11 (9.2) |
| Knee | 24 (20.0) |
| Lower leg, foot, and ankle | 16 (13.3) |
| Injuries that required surgery (n=16) | |
| Shoulder | 7 (43.8) |
| Upper arm and elbow | 2 (12.5) |
| Forearm, wrist, and hand | 4 (25.0) |
| Knee | 2 (12.5) |
| Lower leg, foot, and ankle | 1 (6.3) |
| Return-to-sports decision | |
| By medical staff | 53 (44.2) |
| By himself or nonmedical staff | 67 (55.8) |
Values are presented as number (%).
Fig. 2. Pictogram of injuries by anatomic region.
The subgroup analysis based on the decision-maker for returning to sports showed no significant differences in players’ position (χ2=1.9, degree of freedom [df]=2, p=0.388), circumstances (χ2= 3.0, df=1, p=0.084), mechanisms (χ2=4.1, df=3, p=0.255), and anatomic region of injury (χ2=7.2, df=6, p=0.303) as shown in Table 3. However, for injuries that required surgery, the proportion of cases in which the medical staff made the decision to return to sports was significantly higher as compared to the cases where the decision was made by the players themselves or nonmedical staff (χ2=18.5, df=1, p<0.001).
Table 3 . Subgroup analysis of injury characteristics based on the decision-making for returning to sports
| Variable | Decision-maker for returning to sports | p-value | Pearson χ2 (df) | |
|---|---|---|---|---|
| Medical staff (n=53) | Players themselves or nonmedical staff (n=67) | |||
| Positions of players | 0.388 | 1.89 (2) | ||
| Forward | 29 (54.7) | 38 (56.7) | ||
| Defensemen | 16 (30.2) | 24 (35.8) | ||
| Goalie | 8 (15.1) | 5 (7.5) | ||
| Circumstances of injury | 0.084 | 2.99 (1) | ||
| Game | 45 (84.9) | 48 (71.6) | ||
| Practice | 8 (15.1) | 19 (28.4) | ||
| Mechanisms of injury | 0.255 | 4.06 (3) | ||
| Player-to-player contact | 32 (60.4) | 34 (50.8) | ||
| Player-to-board contact | 8 (15.1) | 7 (10.5) | ||
| Player-to-puck or stick contact | 7 (13.) | 9 (13.4) | ||
| Noncontact mechanism | 6 (11.3) | 17 (25.4) | ||
| Injuries by anatomic region | 0.303 | 7.20 (6) | ||
| Shoulder | 19 (35.9) | 15 (22.4) | ||
| Upper arm and elbow | 4 (7.5) | 3 (4.5) | ||
| Forearm, wrist, and hand | 8 (15.1) | 11 (16.4) | ||
| Lower back | 2 (3.8) | 7 (10.5) | ||
| Pelvis, groin, and thigh | 2 (3.8) | 9 (13.4) | ||
| Knee | 11 (20.8) | 13 (19.4) | ||
| Lower leg, foot, and ankle | 7 (13.2) | 9 (13.4) | ||
| Injuries that required surgery | <0.001* | 18.41 (1) | ||
| No | 38 (71.7) | 66 (98.5) | ||
| Yes | 15 (28.3) | 1 (1.5) | ||
Values are presented as number (%).
df: degree of freedom.
*Statistically significant.
The decision-making process by the medical staff for returning to sports was significantly associated with the player’s position, with goalies having a higher odds ratio (OR) of 7.869 (95% confidence interval [CI], 1.345–46.032; p=0.022) than forwards (Table 4). In addition, injuries that required surgery significantly affected the decision-making process for returning to sports by medical staff as compared to those that did not require surgery, with an OR of 30.081 (95% CI, 2.752–328.836; p=0.005).
Table 4 . Injury characteristics affecting the decision-making process for returning to sports by medical staff
| Variables | OR (95% CI) | p-value |
|---|---|---|
| Players’ position | ||
| Forward | Reference | |
| Defensemen | 1.326 (0.490−3.585) | 0.579 |
| Goalie | 7.869 (1.345−46.032) | 0.022* |
| Circumstances of injury | ||
| Game | Reference | |
| Practice | 0.372 (0.098−1.412) | 0.146 |
| Mechanisms of injury | ||
| Player-to-player contact | Reference | |
| Player-to-board contact | 3.212 (0.792−13.027) | 0.102 |
| Player-to-puck or stick contact | 1.481 (0.272−8.072) | 0.650 |
| Noncontact mechanism | 0.691 (0.156−3.063) | 0.626 |
| Injuries by anatomic region | ||
| Shoulder | Reference | |
| Upper arm and elbow | 0.776 (0.083−7.262) | 0.824 |
| Forearm, wrist, and hand | 0.361 (0.057−2.279) | 0.278 |
| Lower back | 0.369 (0.034−3.968) | 0.411 |
| Pelvis, groin, and thigh | 0.687 (0.086−5.504) | 0.724 |
| Knee | 1.149 (0.345−3.828) | 0.821 |
| Lower leg, foot, and ankle | 0.528 (0.114−2.439) | 0.413 |
| Injuries that required surgery | ||
| No | Reference | |
| Yes | 30.081 (2.752−328.836) | 0.005* |
OR: odds ratio, CI: confidence interval.
*Statistically significant.
To the best of our knowledge, this is the first study investigating the characteristics of orthopedic injuries and decision-making regarding return to sports among elite ice hockey players in South Korea, utilizing a self-reported questionnaire. Our study is significant for two major reasons. First, the prevalence of severe orthopedic injuries was substantial, with 58 of 68 ice hockey players (85.3%) reporting at least one occurrence, underscoring the high incidence of injuries among elite ice hockey athletes throughout their careers. Second, our findings indicate that decisions regarding return to sports after severe orthopedic injuries were more frequently made by the players themselves or non-medical staff than by medical professionals.
The high prevalence of injuries identified in this study aligns with previous research emphasizing the physically demanding nature of ice hockey and the associated risk of injuries1,18. In the 2010 Vancouver Olympics, ice hockey was among the sports with the highest injury rates, alongside bobsleigh, short-track speed skating, and alpine skiing5. Similarly, at the 2018 Pyeongchang Olympics, ice hockey was ranked ninth in terms of injury frequency, maintaining its status as a sport with a high incidence of injuries19. A notable finding of our study was the high proportion of injuries occurring during games (77.5%) compared to those resulting from the training sessions (22.5%). This observation is consistent with previous studies that reported a higher rate of injuries during games owing to increased intensity, physical contact, and competitive pressure12,13. The most common injury mechanism in our study was player-to-player contact (55.0%), followed by noncontact (19.2%). These results are consistent with those of previous studies, suggesting that physical contact is a major contributor to ice hockey injuries2,3. Interestingly, our study identified a higher proportion of noncontact injuries than earlier reports7,11. This difference could be attributed to variations in skill levels, competitiveness, and rule enforcement among different leagues and countries. Moreover, disparities in rink dimensions and ice quality can also influence gameplay and injury occurrence. Our results demonstrated that the most frequently injured anatomical regions were the shoulder (28.3%) and knee (20.0%), which is consistent with previous studies reporting a high prevalence of shoulder and knee injuries in ice hockey players10,11. The elevated rate of shoulder injuries may be due to frequent player-to-player or player-to-board contact, in addition to the vulnerability of the shoulder joint during collisions. Knee injuries are frequently linked to the distinct features of ice hockey, such as rapid changes in direction, abrupt stops, and high-speed movements, which exert considerable stress on the knee joint and its surrounding ligaments.
The decision-making process for returning to sports after an injury was found to be significantly influenced by the player’s position and whether the injury required surgery. Our study demonstrated that goalies had a 7.689 times higher likelihood of medical staff involvement in their return-to-play decisions than in those for the forwards. This finding might be attributed to the unique demands and biomechanics of the goaltender position, necessitating specialized treatment and rehabilitation for a safe return to the sport. Considering that ice hockey teams typically have only one or two goalies, a more cautious approach may be taken to ensure their safety and optimal performance when deciding on their return to sports. Injuries requiring surgery were more likely to involve medical staff in the return-to-play decision (OR, 30.081), which is an expected finding given the increased severity and complexity of surgical cases. It is reasonable to assume that, owing to their heightened severity, most cases requiring surgical intervention would necessitate medical staff involvement in the decision-making process for a safe return to sports.
A notable finding of this study is that return-to-sports decisions were made more frequently by the players themselves or nonmedical staff than by the medical staff. This observation may be attributed to the limited availability of dedicated medical staff for ice hockey teams in South Korea, which could have hindered in-depth discussions among players, medical staff, and nonmedical staff during the recovery process. The shortage of dedicated medical staff is a common issue in many Korean professional sports. Given the increasing popularity of ice hockey, it is crucial for the teams to actively employ and utilize medical staff. This would ensure more extensive involvement of medical professionals in the decision-making process for a safe and effective return of the players to sports.
This study has several limitations. Firstly, it relies on self- reported data, which may be influenced by recall bias. However, we attempted to mitigate this bias by providing clear definitions and prompts in the questionnaire. Secondly, the cross-sectional design of our study constrained our ability to draw causal inferences. Thirdly, the relatively small sample size might affect the generalizability of the findings. Fourthly, the questionnaire did not incorporate game exposure time or training hours for ice hockey players, thus restricting the analysis of injury incidence per unit of game exposure or training time. Fifthly, the data was not officially collected from the Korean Ice Hockey Association. For future systematic analysis of ice hockey orthopedic injuries in Korea, it is essential to establish a comprehensive system within the Korean Ice Hockey Association to collect injury occurrence data.
Our study offers crucial insights into the epidemiology of orthopedic injuries among elite ice hockey players in South Korea. The results emphasize the high prevalence of injuries and highlight the necessity of implementing effective injury prevention strategies that target the most commonly injured anatomical regions and mechanisms. Moreover, our findings revealed that nonmedical staff or players were more frequently involved in return-to-play decisions than were the medical staff. This stresses the importance of elite ice hockey players actively engaging in consultations with medical staff regarding injury recovery and return to competition to ensure a successful comeback.
No potential conflict of interest relevant to this article was reported.
The authors thank all the teams and physicians who participated in this study. Special thanks go to Gangho Jin (Director, JINT GYM) for his great efforts in data collection.
Conceptualization: DK, ICC. Data curation: DK, JJK, WKJ, JHL. Formal analysis, Visualization: DK, JHL. Methodology: JJK. Resources: WKJ. Supervision: ICC. Writing–original draft: DK. Writing–review & editing: ICC.
