Comparing post-concussive neurocognitive test data to normative data presents risks for under-classifying "above average" athletes.

We compared classification accuracy of post-concussion test data against baseline and normative data, accounting for baseline level of performance. Athletes (N = 250) completed baseline and post-concussion ImPACT assessments, within 7 days of concussion (verified by sports medicine professionals and self-reported symptoms). Athletes were classified as "below average," "average," or "above average" at baseline. Change from baseline was calculated using reliable change indices (RCIs) and regression-based measures (RBz), and comparison to normative data was achieved using z-scores. Normative comparisons identified fewer symptomatic, concussed athletes than RCIs and RBz. Both RCIs and RBz consistently identified "impairment" at 1 and 1.5 SD, regardless of baseline level, whereas normative comparisons identified 46-48% fewer athletes performing "above average" at baseline using a cut-off of 1 SD and 36-38% fewer using a cut-off of 1.5 SD. The use of normative comparisons may differentially classify concussed, symptomatic athletes who are outside the "average" range at baseline.