Athlete concussion history recall is underestimated: a validation study of self-reported concussion history among current professional rugby union players

Abstract

ABSTRACT Objective: To assess the concurrent validity and test re-test reliability of the Michigan Traumatic Brain Injury Identification Method (MTBIIM). The psychometric properties of this concussion index were investigated by comparing the agreement between player self-reported diagnosed concussions and medical record diagnosed concussions among professional rugby union players. Study Design: Cross-sectional study i) validation and ii) test re-test reliability. Methods: The MTBIIM was administered via a structured interview to obtain the number and nature of player self-reported concussion histories from players, while contracted to the host club. Self-reported concussion history information was compared to medically recorded data captured between 2008 and 2017. A mixed-effects logistic regression model explored predictors of player self-report accuracy. Results: Data from 62 players (25.39[4.36] years) included 99 unique rugby related concussions. Medically documented concussions (n = 92) per player (1.48 [1.96]) were 30% more than the mean number of self-reported diagnosed (n = 63) concussions per player (1.02 [1.21] events). Overall, self-reported diagnosed concussions and medical record diagnosed concussion histories had a ‘fair’ level of agreement (k=0.274; SE [0.076]), p=.001). Self-reported lifetime concussion history was signicantly negatively correlated with recall of concussions. Conclusions: Initial concurrent validity of the MTBIIM was found to be fair with the average athlete under-reporting the number of clinically diagnosed concussions.