A Guide to the Interpretation and Reliability of Annual Collegiate ImPACT Baselines by Sex

To create a sex-specific clinicians guide to evaluating significant immediate post-concussion assessment and cognitive testing (ImPACT) score changes using minimal detectable change (MDC) ranges, to evaluate the sex-specific test-retest reliability using intraclass correlation coefficients (ICCs), and to compare male and female performance across all composite and subtest scores.

Data from 390 National Collegiate Athletic Association Division I athletes (n = 390; male = 106, female = 284) across four athletic seasons (2020-2021 through 2024-2025) who had completed ImPACT remotely for two consecutive seasons were used to calculate the MDC and ICC at the 95% confidence intervals (CIs) by sex, as well as to compare sex-based performance for all ImPACT scores (subtest and composite).

The visual memory composite score, as well as the subtests of design memory correct distractors (immediate), design memory hits (delay), XO total correct memory, XO total correct (interference), XO average correct reaction time (RT; interference), symbol match total correct RT (visible), and color match total correct, exhibited significant differences between males and females at the first and/or second remote time point. ICCs for females demonstrated poor reliability on 14 of the 33 subtest scores (ICC = 0.06-0.49) and on the visual memory composite score (ICC = 0.49). For males, 11 of the 33 subtests had poor reliability (ICC = 0.21-0.48). Many of the subtests present poor reliability for both genders, particularly as it relates to the three letters task.

This study found sex differences on ImPACT performance, suggesting that sex-specific MDC values are needed to evaluate significant changes in ImPACT scores. Additionally, this study found several ImPACT subtest scores to be unreliable, suggesting only moderate to good scores be used in clinical evaluation.