Objective To document false-positive scores on the Sport Concussion Assessment Tool(SCAT3) in high school athletes.

Design Pre-Test Only.

Setting Six high schools in the Mid-West and Central Region of the USA.

Participants The sample was composed of a total of 489 high school athletes who completedall components of the SCAT3: 157 female (Mean age=15.6, SD=1.8 years) and 322 male(Mean age=15.9, SD=1.5 years) adolescent athletes.

Intervention The SCAT3 was administered to all athletes.

Outcome measures Cases were identified as “false-positive” by surpassing cutoff scores using 95% confidence intervals from the sample on SCAT3 components: Standardised Assessment ofConcussion (SAC: Mean 26.06±2.34, Cutoff<21), total severity of symptoms (Males: Mean:3.24±5.93, Cutoff>15; Females: Mean: 4.16±6.96, cutoff>18), tandem gait (Mean:15.55±4.52, Cutoff>25 sec), and total errors on the modified Balance Error Scoring System(mBESS: Mean: 3.42±2.67, Cutoff>9 errors).

Main results Overall, 16% of athletes (n=78) obtained scores falling beyond false-positivecutoffs on at least one sub-component of the SCAT, and only 2% (N=10) obtained two or morescores beyond cutoffs. False-positive rates on individual components were as follows: SAC:5.7% (n=28), mBESS: 5.5% (n=27), tandem gait: 3.9% (n=19), total symptom severity: 4.9% (n=24).

Conclusions While rates of false positives fell generally within 95% CIs for individual components of the SCAT, the overall SCAT3 false-positive rate was 16%. Comparisons between post-concussion and baseline SCAT3 data should be based on 2 or more components falling outside 95% CIs.

Competing interests Dr. Schatz serves on the ImPACT Scientific Advisory Board