The on- and off-pitch assessment of sports-related concussion: A systematic review of diagnostic test accuracy studies

Background: The diagnostic accuracy of sports concussion assessment tools for on-pitch decision-making is not clear. The objective of this systematic review is to identify and synthesise the evidence on the diagnostic accuracy of any item/tool used to assess suspected concussion at the time of injury with a key focus on football (soccer).

Methods: Studies were eligible for inclusion if they were original data-based studies reporting diagnostic accuracy data for same-day concussion assessments of athletes aged ≥13 years. Studies were excluded if they did not provide data enabling calculation of diagnostic accuracy outcomes. Risk of bias and applicability was assessed using the QUADAS-2 tool. Information sources included MEDLINE, Embase, CINAHL, SPORTDiscus, CENTRAL, PsycINFO and Web of Science plus manual reference screening. Results are presented in a narrative synthesis with studies grouped by assessment item domain, with further stratification by sport, assessment setting, timing, and participant characteristics.

Results: Sixty-five studies were included in the review. Fifty-five (85%) studies were judged to have high risk of bias, 9 (14%) had unclear risk of bias and one (1.5%) study had low risk of bias. Index tests were generally well-reported, but critical information on how the clinical diagnosis of concussion was confirmed was either lacking or absent in 37 (57%) studies. Two studies focused exclusively on same-day concussion assessment in football players. Only one study (rugby union) reported data from the on-pitch assessment of concussion. Cognitive assessment items were the most frequently reported assessment domain. Symptom scales and the King-Devick test consistently reported high sensitivity and specificity. Balance assessments demonstrated low sensitivity across six studies.

Conclusion: Concussion assessment protocols vary substantially in diagnostic accuracy. Only two football specific studies were identified and one on-pitch assessment study. No studies were identified which evaluated the diagnostic accuracy of on-pitch concussion assessment items in football. This highlights the urgent need to validate a streamlined, concussion assessment tool for use in on-pitch football environments.