Objective To establish test-retest reliability of the Sport Concussion Assessment Tool 2 (SCAT2) in healthy elementary school children and young adults.

Design Repeated-measures design.

Setting Elementary Schools in the Montreal area (children) and McGill University Athletics Department (adults).

Participants Twenty-two children (14 females and eight males; mean age 10.3±0.5 years) and 73 young adults (42 female, 31 males; mean age 22.6±1.8 years) from University students participated in this study.

Intervention None.

Main Outcome Measurements Participants were tested using the Sport Concussion Assessment Tool-2 (SCAT2) on two occasions, 1-week apart (6.4±2.7 days). The SCAT2 measures several domains of functioning such as delayed memory, balance and post-concussion symptoms. Test-retest reliability was assessed for the total SCAT2 score as well as for the SCAT2 components separately.

Results Intraclass Correlation Coefficients for the total SCAT2 score were only fair for both groups of participants (ICC=0.367 for young adults and 0.446 for children). The ICCs were also below 0.5 for individual components of the SCAT2 in young adults. In children, individual components yielded better reliability than the total score (Balance ICC=0.725; SAC ICC=0.523; symptom severity ICC=0.488).

Conclusions The results indicate that in view of the low levels of test-retest reliability achieved for the total SCAT2 score, professionals should be cautious when using change in score over time for clinical decisions. Recommending the use of the SCAT2 as a standard of care to diagnose a concussion or to ascertain recovery from one may be premature at this time.

Acknowledgments Participants and their families, The Priory School, Gardenview Elementary School for allowing testing within their institutions, Fonds de recherche du Québec-Santé, School of Physical and Occupational Therapy McGill University.

Competing interests None.