Youth Football Head Impact Exposure Differences Based On Four Data Cleaning Methods

Youth tackle football athletes commonly experience repetitive head impacts (RHIs). Instrumented mouthguards (iMG) are capable of quantifying frequency and magnitude of RHIs. Using iMGs requires cleaning recorded data to confirm impacts. Data cleaning techniques include filtering, timestamping, and video verification. Video verification is currently regarded as best practice; however, several different cleaning methods are utilized in the literature. PURPOSE: To determine differences in number and rate per athlete exposure (IR) of RHIs among four data cleaning methods. METHODS: Participants (N=18, age=9±1years, height=139.7±10.2 cm, mass=36.2±8.6 kg) wore iMGs for six tackle football games. Impact rate ratio (IRR) and intraclass correlations (ICCs) along with 95% confidence intervals (CIs) were calculated to compare four different data cleaning methods. IRRs represent a ratio of compared methods(Table 1). IRRs with a 95% CI not containing 1.00 were considered significant. Method 1: Timestamp cleaning removed impacts occurring before or after gameplay and during breaktimes (e.g., halftime). Method 2: Video-verified impacts completed by a trained researcher. Method 3: iMG algorithm-determined real impacts. Method 4: iMG algorithm-determined real impacts only within timestamp bounds. RESULTS: RHI data collected resulted in 62 athlete exposures across six games. Data cleaning methods resulted in different numbers of RHIs and IRR (Table 1). ICCs indicated moderate agreement between methods 2 and 3 (ICC=0.514,p=0.012) and methods 2 and 4 (ICC=0.616,p=0.003), and excellent agreement between methods 3 and 4 (ICC=0.975,p<0.001). CONCLUSION: Method 1 produced the highest IR, 22 times more impacts than method 2, which utilized video verification. Overall, our results suggest that the method of data cleaning can affect IR and certain methods may result in an overestimation of RHIs. Therefore, video verification should still be considered best practice.R01HD110401