doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.

Objectives: Determine, through video reviews, how often concussions occur in combat sport matches, how well non-medical personnel can be trained to recognize concussions and how often fights are judged to continue too long. Methods: This is a retrospective video analysis by an 8 person panel of 60 professional fights (30 boxing and 30 mixed martial arts). Through video review, medical and non-medical personnel recorded details about each probable concussion and determined if and when they would have stopped the fight compared to the official stoppage time. Results: A concussion was recorded in 47/60 fights. The fighter that sustained the first concussion ultimately lost 98% of the time. The physician and non-physician raters had 86% agreement regarding the number of concussions that occurred to each fighter per match. The mean number of concussions per minute of fight time was 0.08 (0.06 for boxers and 0.10 for MMA). When stratifying by outcome of the bout, the mean number of concussion per minute for the winner was 0.01 compared to the loser at 0.15 concussions per minute. The physician raters judged that 24 of the 60 fights (11 boxing [37%]; 13 MMA [43 %]) should have been stopped sooner than what occurred. Conclusion: Recognizing that the losing fighter almost always is concussed first and tends to sustain more concussions during the fight, along with the demonstration that non-physician personnel can be taught to recognize concussion, may guide policy changes that improve brain health in combat sports.