Long-term test-retest reliability of baseline cognitive assessments using ImPACT

• Published in: The American journal of sports medicine Vol. 38; no. 1; p. 47
• Main Author: Schatz, Philip
• Format: Journal Article
• Language: English
• Published: United States 01.01.2010
• Subjects: Athletic Injuries - complications; Cognition Disorders - diagnosis; Cognition Disorders - etiology; Confidence Intervals; Humans; Neuropsychological Tests; Post-Concussion Syndrome - complications; Psychometrics; Reaction Time; Reproducibility of Results; Software; Statistics as Topic; Time Factors
• ISSN: 1552-3365
• DOI: 10.1177/0363546509343805

Computer-based assessment programs are commonly used to document baseline cognitive performance for comparison with postconcussion testing. There are currently no guidelines for how often baseline assessments should be updated, and no data documenting the test-retest stability of baseline measures over relevant time periods. To establish long-term test-retest reliability of baseline assessments using ImPACT, and to compare various statistical methods for establishing test-retest reliability. Case series; Level of evidence, 4. Participants were 95 collegiate varsity athletes completing baseline cognitive testing at 2 time periods, approximately 2 years apart. No participant sustained a concussion between assessments. All athletes completed the ImPACT test battery; dependent measures were the composite scores and total symptom scale score. Intraclass correlation coefficient estimates for visual memory (.65), processing speed (.74), and reaction time (.68) composite scores reflected stability over the 2-year period, with greater variability in verbal memory (.46) and symptom scale (.43) scores. Using reliable change indices and regression-based methods, only a small percentage of participants' scores showed "reliable" or "significant" change on the composite scores (0%-6%), or symptom scale scores (5%-10%). The current results suggest that college athletes' cognitive performance at baseline remains considerably stable over a 2-year period. These data help establish the effects of longer, clinically pragmatic testing intervals on test-retest reliability. The current results suggest that stretching the time between baseline assessments from 1 to 2 years may have little effect on the clinical management of concussions in collegiate athletes. These results should not be generalized to collegiate football players, who were not included in this sample. Youth athletes (high school and younger) should continue to receive annually updated baseline assessments until prospective study of the stability of baseline assessments for this younger age group can be completed.