Background The clinical relationship between accumulated head impacts and chronic cognitive dysfunction is under increasing scrutiny. There is therefore a need to track an athlete's accumulation of head impacts.

Objective We sought to develop and test an inexpensive wireless device that could be widely used for monitoring blows to the head in football.

Design Prospective, descriptive.

Setting Stanford University scrimmages.

Subjects 16 males from 6 different positions.

Methods During four spring practice sessions, players wore mouthguards instrumented with three accelerometers and gyroscopes for measuring head motion. When the acceleration magnitude exceeded 15G (after 110 Hz low-pass filter), an event was recorded. Simultaneously, full-speed scrimmage sessions were video recorded from two perpendicular angles and later analysed for ‘helmet-helmet contacts.’ Plays were only considered if the subject's helmet was struck by the opposing player's helmet at the initial point of contact, or if the subject's helmet sustained “no-contact” at all during the play. Mouthguard events were then synchronised and evaluated during the time period of each play to determine agreement with video data.

Results Video analysis yielded 153 helmet-helmet contacts and 402 no-contacts. Mouthguard sensitivity for detecting observed helmet contact at the 15G threshold was 61% (95% CI 0.54 to 0.69) and the specificity was 91% (95% CI 0.88 to 0.93). Overall accuracy was 83%.

Conclusion Although the 15G threshold is based on published helmet telemetry clinical studies, 39% of observed helmet-helmet contacts were below this threshold. Further research is required to understand if lower acceleration contacts are clinically relevant, and if they can be accurately detected.

Acknowledgements We acknowledge the Stanford athletic training, videography, and football staff for their cooperation as well as X2 Impact Inc. for donation of mouthguards.