No Performance Changes Following a Concussion Amongst National Hockey League Players

Abstract

Athletic performance following sport-related concussion (SRC) is an emerging area of research. Lingering neurological deficits in postural control and neurocognition may impair athletic performance. Previous reports have found no difference in performance after SRC among professional hockey players as measured by routine statistics (e.g. goals, assists), yet advanced metrics may identify subtler differences in performances have not been studied. PURPOSE: To examine changes in hockey player performance using advanced metrics upon returning from a SRC as compared to non-injury time loss players. METHODS: Between 2008 and 2015, 90 National Hockey League (NHL) players with documented concussions and 51 NHL players who missed games due to non-injury related causes were identified. Players with intervening missed times and confounding orthopaedic injuries were excluded. Advanced metrics were retrieved from WAR-on-ICE blog and available statistics were compared to NHL.com to ensure validity. A total of 51 metrics were identified and performance was compared between pre- and post-concussion over 5 games, 10 games, or the season (all games prior versus all games post-concussion). Dependent variables were compared with a 2 (Group) x 2 (Time) repeated measures ANOVA and a conservative alpha value of 0.01 was set due to lack of variable independence. RESULTS: There were no significant interactions between groups for the 5 game analysis. Significant interactions were noted only for Time on Ice - Percentage (percentage of time a player was on the ice) for 10 games (Concussion: +0.59%, Control: -0.97%; P<0.001, d=0.007) and full season (Concussion: 0.34% change, Control: +0.96%, P=0.04; d=0.055). Several variables including high danger scoring chances and PDO (on-ice save and shooting percentages) were below 0.05, but effect sizes were less than 0.1; suggesting the differences were not meaningful. CONCLUSIONS: These results suggest there are no meaningful performance differences in NHL players upon return from SRC. The only significant finding was an inconsistent change in time on ice following injury, but the effect sizes were very small (Cohen’s d<0.06). These findings are consistent with previous studies which found little impaired performance in professional hockey or basketball players upon return from SRC.