Concussion Risk Between Individual Football Players: Survival Analysis of Recurrent Events and Non-events

Area of Science:

• Biomechanics and Sports Medicine
• Statistical Modeling in Sports Injury Research

Background:

• Head impact exposure and concussion susceptibility vary significantly among football players.
• Previous analyses of head impact data often lack subject-specific risk characterization.
• Existing methods struggle to model recurrent events like concussions and non-concussive impacts simultaneously.

Purpose of the Study:

• To develop and present a novel survival analysis method for characterizing individualized concussion risk in football.
• To address limitations in previous analyses by modeling both concussions and non-concussive impacts.
• To account for variations in head impact data sampling across individual players.

Main Methods:

• Utilized a new survival analysis technique designed for real-world head impact data.
• Employed the Consistent Threshold method to generate subject-specific concussion risk curves.
• Developed an approach to select an optimal cumulative distribution function by minimizing injury prediction error.

Main Results:

• The proposed method appropriately models recurrent events (concussions) and recurrent non-events (head impacts without concussion).
• The analysis effectively accounts for uneven head impact sampling between players.
• Demonstrated that minor differences in distribution fit can lead to substantial predictive errors in injury risk.

Conclusions:

• The developed subject-level analysis provides a robust method for assessing concussion risk in football.
• This approach can be adapted to create population-specific risk curves for various sports.
• Findings can inform the development of targeted interventions to decrease concussion incidence in athletes.