Objective To investigate differences between baseline self-report of concussion symptoms with two commonly used concussion assessment tools.

Design Retrospective, cross-sectional.

Setting University of Alberta.

Subjects One hundred and ninety-four varsity athletes (117 males; 77 females).

Intervention Prior to the competitive season, subjects completed baseline testing, including self-report of concussion symptoms, with the Subjective Concussion Assessment Tool (SCAT2) and the on-line computerised Immediate Post-Concussion Assessment of Cognition (ImPACT) test, under supervision of an athletic therapist. Testing was counterbalanced to control for serial position effect.

Outcome Measures The 18 shared self-report concussion symptoms were compared, and a general linear model used to identify possible interactions between gender and group type (SCAT2 vs ImPACT). Finally correlations were calculated.

Results There was an interaction between gender and group type for 1 symptom (trouble falling asleep). Surprisingly, pairwise comparisons identified differences (p≤ 0.05) on 10/18 paired symptoms, with correlations ranging from 0.08 to 0.58.

Conclusions Subjects respond differently when self-reporting concussion symptoms, depending on type of instrument used, and mode of administration. Despite test items being similar in content, there are significant differences in symptom reports in the same subject between SCAT2 and ImPACT tests. Use of similar data collection methods for baseline testing, assessment post-injury and prior to return-to-play is indicated, to increase both the diagnostic and prognostic utility of such tests. This further emphasises current recommendations that these tests not be used in isolation to make therapeutic decisions regarding athletes with concussion; and underscores the importance of serial clinical evaluations by a suitably qualified physician.

Competing interests None. All authors have signed the disclosure form.