Files
Abstract
The neurocognitive assessment has been regarded as the gold standard for concussionassessment. With the recent advances in technology this assessment has become computerized,and several programs now available specifically for sport concussion assessment. While theseprograms are based on the traditional pencil and paper assessments, the psychometric propertieshave not been established. More specifically, the reliability of the computer programs usingclinically relevant assessment intervals has not been performed. In addition, recent evidencesuggests that test performance during multiple administrations may be influenced by test takereffort. Therefore, the purpose of this project was to evaluate the test-retest reliability of threecommercially available computer-based concussion assessment programs while simultaneouslycontrolling for participant effort. One-hundred and eighteen (N=118) healthy, college agedstudents were recruited for this study. Each participant completed the Headminder CRI, theConcussion Sentinel, and the ImPACT concussion assessment tests at three time points: baseline,day 45, and day 50. Greens Memory and Concentration Test for effort was also administeredon each day. The data were reviewed and cleaned of invalid baseline tests or those with a poorunderstanding of the test administration. Seventy-three (n=73) participants were included fordata analysis. Intraclass correlation coefficients (ICC) were calculated for each output and eachsubtest score. Repeated measures analyses of variance was used to evaluate for changes in effortacross administration days. The ICC test-retest reliability results were lower than previouslyreported in the literature. All measures of effort were deemed high on all days of testing withsignificant increases on delayed recall (F1.81,130.54 = 6.464, p=.003), consistency (F2,144 = 5.800, p =.004), and free recall (F = 15.935, p < .000) variables. Results from this study indicate1.655, 119.191the test-retest reliabilities of three commercially available computer-based concussionassessment programs are not as high as previously reported. Differences in our results and thosepreviously reported are not likely attributed to effort by the participants. The differences may beaccredited to differing test-retest time intervals or the inability of the programs to consistentlymeasure neurocognitive functioning.