RESUMEN
OBJECTIVE: To describe the frequency of clinical signs and the accuracy of video diagnosis of sport-related concussion. DESIGN: An observational cross-sectional study. SETTING AND PARTICIPANTS: Videos from a database of all suspected sport-related concussion in TOP14 matches (French professional male rugby Championship) were used, from 2012 to 2015 seasons. The videos were analyzed by 4 observers, blinded to the concussion diagnosis, after a training phase, and an inter-rater reliability analysis. INTERVENTIONS: N/A. MAIN OUTCOME MEASURES: Video analysis for clinical signs and presumed diagnosis of sport-related concussion. The observer's diagnosis was compared with the team physicians' diagnosis. RESULTS: Four hundred seventeen videos were analyzed with 142 concussed players. Sport-related concussions happened mostly from an impact to the head (seen by the observers in 98.4%), against an opponent (73.2%), and during a tackle (74.6%). Video signs were observable in 98.6% for suspected loss of consciousness, tonic posturing (95%), ataxia (77.2%), dazed look (35.7%), abnormal behavior (18.2%), and seizures (96.2%). Ataxia was seen in 77.7% of concussed players, suspected loss of consciousness in 61.4%, dazed look in 63.2%, abnormal behavior in 55%, and tonic posturing in 7.1%. The observers diagnosed 79.8% of concussions. CONCLUSIONS: We described the frequency of video clinical signs of sport-related concussion, with a good accuracy of the blinded observers for the diagnosis. This emphasizes the importance of pitch-side video analysis as an extra tool for sport-related concussion diagnosis.