Reliability of the Sport Concussion Assessment Tool 5 baseline testing: A 2-week test-retest study

To examine short-term test-retest reliability of the Sport Concussion Assessment Tool 5 (SCAT5).

Longitudinal study.

Sixty-two professional male ice hockey players (mean age=22.5, SD=3.2) completed a preseason baseline test twice over a two-week interval. Half of the players were tested by the same assessor on both testing sessions. Spearman's correlations (rs) were used to determine linear agreements, and Wilcoxon signed rank tests (sig r) were used to determine mean differences, between testing sessions.

Symptoms had high test-retest reliability (Score: rs=0.85, p<0.001, sig r p<0.001; Severity: rs=0.84, p<0.001, sig r p<0.001). The reliability coefficients for the SAC (rs=0.58, p<0.001, sig r: p=0.412), and mBESS (rs=0.40, p=0.001, sig r: p=0.607) were considerably lower than symptoms. More than half (52%) of the athletes reported at least one baseline symptom (Md=1, M=2.2, SD=3.3). The most commonly reported symptoms were fatigue or low energy and neck pain. The broad ranges of SAC total scores (range=28-45, Md=35, M=35.4, SD=4.2) and SAC test-retest change scores (range -7 to +11) were mostly due to variability on the memory performance, tested using 10-item word lists. The number of mBESS single leg stance errors (Md=1, Md 3.2, SD=4.0) was greater than Tandem stance errors (Md=0, Md=2.0, SD=3.6).

The two-week test-retest reliability of the SCAT5 baseline scores varied from moderate to high. However, there was considerable individual variability on the SAC and mBESS scores and most players have notable short-term fluctuation on performance even if uninjured. Recommendations for interpreting change on the SCAT5 are provided.