Between-seasons test-retest reliability of clinically measured reaction time in national collegiate athletic association division I athletes.
Eckner JT, Kutcher JS, Richardson JK. J Athl Train. 2011;46(4):409-14.
Reaction time is commonly impaired after a concussion and is often assessed in computerized concussion assessment programs. Unfortunately, many sports medicine clinicians, particularly those working with younger athletes, may not have access to computerized concussion assessment programs. Therefore, Eckner et al developed a simple clinical measurement of reaction time; however, the test-retest reliability (consistency) of their assessment as well as computerized reaction time assessments have not been well evaluated between seasons. The purpose of this study was to measures the 1-year test-retest reliability of the clinical and a computerized measure of reaction time among 102 NCAA Division 1 athletes (78 American football, 10 women’s soccer, and 14 wrestling athletes). Clinical and computerized assessments of reaction time were performed during preparticipation physical examinations before two consecutive seasons. The clinical assessment of reaction time used a vertical shaft with a cylindrical weight at the bottom and the distance the shaft dropped between the time of release and hand closure could be measured to determine the reaction time (the article offers more details and figures). Each athlete performed 2 practice trials and 8 data collection trials with the examiner randomizing the time until release from 2 to 5 seconds (to prevent the athlete from anticipating the drop). The computerized assessment of reaction time was completed with a computerized neuropsychological test session (pressing the “k” as rapidly as possible when shown a particular image). Unfortunately, only 62 (61%) athletes had usable computerized assessment tests at baseline and follow-up (27 athletes had an invalid testing session and 13 missed a computerized assessment). The clinical measure of reaction time had higher test-retest reliability (between year 1 and year 2) than the computerized assessment but these were both statistically similar. Among participants with both clinical and computerized assessments the clinical assessments led to faster reaction times. Furthermore, the clinical assessment had faster reaction times at follow-up, suggestive of a learning effect by the athletes. The computerized test did not differ between years 1 and 2. The clinical assessments of reaction time did not differ between athletes with and without 2 valid computerized tests; reducing the risk that athletes with an invalid test perform tests differently.
This study may be valuable to sports medicine clinicians working in settings with limited budgets and the lack of access to computerized neuropsychological tests. The authors provide previously reported test-retest reliability for computer-based reaction time measures and the data in this study are consistent with other studies that evaluated test-retest over prolong periods of times (> 30 days). It will be important to follow this research team to see how this test performs when evaluating athletes with and without concussions. Athletes showed a learning effect with this test so it will be important to see if concussions lead to changes in scores beyond measurement error and learning effects (for example, if athletes improve 11 msec due to a learning effect then a concussion may need to impair reaction time by greater than 11 msec). It will also be interesting to see the impact of repeated exposures to the test and how this test performs among younger athletes that typically do not have access to computerized testing. The authors acknowledge that reaction time is only one of many signs indicating a possibly concussion and therefore should be assessed as part of a thorough evaluation. Furthermore, they note that this test may be an adjunct to computerized testing but that it should not replace computerized testing. Remember that this test was developed as a simple and low cost measure for athletes that do not have access to computerized concussion assessments. Have any of you tried this clinical measure of reaction time before? If so, what are your experiences?
Written by: Jeffrey Driban
Reviewed by: Stephen Thomas
Related Posts:
Eckner JT, Kutcher JS, & Richardson JK (2011). Between-seasons test-retest reliability of clinically measured reaction time in national collegiate athletic association division I athletes. Journal of Athletic Training, 46 (4), 409-14 PMID: 21944073
Having used this test in my every day practice, I'm interested to see where the research takes us. I think that this test, as well as some of the other standardized testing methods, could yield big gains for easy access, on the field concussion diagnosis.
Thanks for the comment Meghan. Do you find the test pretty easy to use with a large number of athletes?
I have never used this test before, but I could definitely see it being useful in my practice. I work with cross country and track athletes who are not given a computerized baseline due to cost and relatively low incidence of injury. However, some of my athletes receive concussions from other activities (pick-up basketball games, gym classes, military activities, etc.) and a test like this would be useful in helping to determine return to activity.
Abby: That sounds like a great idea. That's the type of situation that this test seems to be designed for. Good luck implementing it!
Thanks for this post! This particular device is definitely of interest due to its ability to assess such a key aspect of cognitive delay following sports concussion at a low price. Of interest to me was the fact that one third of the computerized baseline measures were invalid, many of them due to poor effort. At least we have some peace of mind following a valid computerized test that we have a true representation of the athlete’s abilities. My fear with the falling weight used in the study is that we don’t have an index of effort to make sure we are getting good baseline measures. I look forward to seeing more studies on this device, as I can definitely see great utility for it in the clinical setting.
The main question I have with this method is if the weight can change circumferences to accommodate for hand size. I work at a middle school/high school and I'm starting to think that larger hand sizes would result in a slightly faster result than smaller hands.
Overall, I think this is an effective method since I have a limited number of neurocognitive tests I can use this year, so I've used this method as a tool to see progress more frequently.
What I am really interested to find out is if this method is as valid regardless to hand size.
Thanks for posting this!
Meredith and Jay: Thanks for the comments!
Meredith: your concerns are well noted and it was nice to see the authors acknowledge these limitations as well. They definitely were not suggesting that this test was a new gold standard but rather as potential back up when other tests are not available. I'm also looking forward to see how the move forward with this test.
Jay: You raise an interesting point. If the key is whether the person's reaction time changes over time do you think the hand size is still a key issue (since the person is not being compared to another individual)? Let me know your thoughts.
Thanks for the comments!