In recent years, attention has been drawn to the study of sport-related concussion and the evaluation of the neuropsychological measures used in diagnosing and assessing progress over the course of the injury. Most often, baseline assessments are administered to athletes at the beginning of their respective sports season to provide a reference for comparison in the event that the athlete sustains a head injury.1,2 Baseline measurements allow the examiner to determine whether the results of the individual's assessment following a traumatic event are different when compared to his or her own previous tests instead of comparisons to others. Additionally, concussion symptoms are highly individualized in every athlete.3–5 Individual differences may persist in performance on several tests, in addition to their recovery pace and time.3
Most collegiate and professional sports associations implement baseline testing annually.5 Recommendations for clinical practice continue to support the use of neuropsychological assessment after the incidence of concussion as part of an athlete's overall treatment and concussion management.4,6,7 Commonly, the athlete is returned to play only after post-trauma follow-up testing approximates the baseline outcomes.4
Scattered through the concussion management literature is discussion of administering “rebaseline” testing some time after the athlete has been returned to play. Ambiguous guidelines for rebaseline testing are noted in the National Athletic Trainers' Association Position Statement, which states, “A new baseline examination should be completed annually for adolescent athletes, those with a recent concussion, and when feasible, all athletes.”4 This rebaseline test serves as an updated benchmark for the remainder of the sports season and is conducted following the return of the athlete to practice and competition. Some authors and institutions suggest this take place in the same sport season,8 some recommend annual baseline testing,4,9 and some do not specify the timing.10 For example, the HeadMinder Concussion Resolution Index professional manual recommended that it is “crucial” to establish a new baseline after resolution of player symptoms; however, specifics regarding time interval are unspecified.11 There are no recommendations regarding rebaseline testing provided in the Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) clinical manual.10 Lovell12 noted that ImPACT test–retest reliability declines with the passage of time between serial assessments, suggesting that periodic new baseline assessments represent good practice.
There has only been one study that assessed rebaseline testing. Lynall et al8 investigated the utility of baseline and rebaseline assessments in athletes after concussion in 34 National Collegiate Athletic Association (NCAA) Division I athletes participating in multiple sports. Athletes were administered the CNS Vital Signs computerized neurocognitive test at initial baseline, post-trauma evaluation, final post-injury before return-to-play, and then again at rebaseline. The median number of post-injury evaluations was two and the median time in which the re-baseline tests were administered from the initial baseline was 397 and 169.5 days from post-injury to rebaseline. Researchers found insufficient clinical utility for the administration of rebaseline evaluations after return-to-play for athletes sustaining one concussion throughout their respective sports season.8 The need for rebaseline testing of athletes who have recently suffered a concussion has not been documented well in the literature and the necessity of rebaseline testing for athletes needs to be evaluated for its clinical utility.
The current study aimed to investigate the clinical utility of administration and evaluation of rebaseline tests following the return-to-play of athletes who had a concussion during the sports season. We aimed first to evaluate the effectiveness of athletes' latest follow-up assessment following concussion as their final rebaseline for the season. The second aim addressed whether re-baseline testing provided incremental information useful in future clinical evaluations. Third, we addressed whether rebaseline or follow-up testing does anything detrimental to the testing sample, such as introducing practice effects. Correct assessment of the evaluation will enable accurate interpretation of concussive symptoms using the appropriate baseline measure. Finally, we employed the ImPACT computerized neurocognitive test, which has been reported to be used by more than 90% of colleges and universities.13
Data from a total of 41 NCAA Division II student-athletes were used in this study. All 41 athletes sustained a concussion during either the 2015–2016 or 2016–2017 athletic season and represented 10 sports and 14 teams. Demographic characteristics of the sample are noted in Table 1.
Demographic Characteristics of the Sample
During the concussion education session at pre-season collection of baseline information, participants consented to allow their de-identified information to be used for research purposes. The study had been approved annually by the Florida Institute of Technology Institutional Review Board.
The ImPACT is a computerized battery of tests designed specifically to detect changes in neurocognitive functioning following sports-related concussion. In addition to collecting basic demographic and historical (eg, concussion and education history) and symptom information, the ImPACT tests neurocognitive functioning, with an emphasis on four composite scores: verbal memory, visual memory, motor and processing speed, and reaction time. In addition to composite scores, expected improvement over time is adjusted via a Reliable Change Index (RCI) score. The RCI score uses the standard error of difference to establish a confidence interval between the initial baseline and the post-trauma score to evaluate change and reduce error for interpretation.10 The measure has been deemed both sensitive and specific in the “assessment of neurocognitive and neurobehavioral sequelae of concussion.”14 As noted previously, the ImPACT is the most widely used computerized instrument in concussion management programs in North America.15
All participants were administered the ImPACT as part of the university's routine preparticipation medical evaluation. Twenty-four to 48 hours after sustaining a concussion, athletes were readministered the test. This evaluation, known as a “post-trauma,” included the addition of a clinical interview to assess the nature of the trauma and the symptoms following the incident. Physical and cognitive rest were prescribed acutely, with return to moderate aerobic exercise as soon as it was tolerated with no exacerbation of symptoms. Participation in team practices and contests was suspended. When the athletes reported to their athletic trainers that they were symptom free (or no more symptomatic than at baseline), they were readministered the ImPACT to determine whether recovery was sufficiently complete to allow return-to-play. If athletes were cleared, they were allowed to participate in their sport's practice again; however, if they were still deemed to be recovering, they were required to wait an additional amount of time before being retested. Approximately 2 weeks after their last follow-up evaluation that returned them to play, a new ImPACT baseline (rebaseline) was obtained to set a new standard for the remainder of the season. Because of logistical and individual peculiarities (eg, out of town trips and spring break), in actuality there was an average of 31 ± 17.33 days from return-to-play to establishment of the rebaseline measurement.
All analyses were conducted using IBM SPSS software (version 24.0; IBM Corporation). A series of repeated measures analyses of variances were conducted to compare the ImPACT composite scores of athletes across the four testing sessions: baseline, post-trauma, follow-up, and rebaseline. All pairwise comparisons between testing sessions were assessed using the Sidak correction to limit the risk of type I error. An alpha level of 0.05 was used for all inferential statistical analyses.
The composites verbal memory, visual memory, visual motor, and reaction time were compared. Because the impulse control composite score and the Cognitive Efficiency Index (CEI) that evaluates for speed and accuracy are not generally used in clinical decision making, they are not reported. Means and significant differences for composite scores across testing sessions are shown graphically in Figure 1 and described below. Scatterplots with boxes housing points within the 95% confidence interval and limit lines showing maximum and minimum scores are shown for the four ImPACT composites. Significant differences between testing times appear as lines above the boxes. As was expected, scores on all four ImPACT composite scores in the initial post-trauma evaluation were significantly lower than at baseline and the final follow-up evaluation. The comparison of primary interest was between the final follow-up and the rebaseline. As shown in Figure 1, none of those comparisons were statistically significant. The outcomes of the statistical comparisons for each of the four milestone times of measurement are described below.
Means and significant differences for composites across testing sessions. *Significant differences across testing at the P < .05 level. **Significant differences across testing at the P < .01 level. ***Significant differences across testing at the P < .001 level.
A significant difference was found across the testing sessions for verbal memory composites (F(3,120) = 18.08, P < .001). Composites were significantly lower during post-trauma testing than initial baseline (P = .004), follow-up (P < .001), and rebaseline (P < .001). Significant differences were also found between baseline and follow-up testing (P = .051), with higher scores on follow-up. No other significant differences were observed for verbal memory.
A significant difference was found across the testing sessions for visual memory (F(3,120) = 12.33, P < .001). Composites were significantly lower during post-trauma testing than initial baseline (P = .017), follow-up (P = .003), and rebaseline (P < .001). No other significant differences were observed for visual memory.
A significant difference was found across the testing sessions for visual motor composites (F(3,120) = 29.05, P < .001). Visual motor composites were significantly lower during post-trauma testing than initial baseline (P = .040), follow-up (P < .001), and rebaseline (P < .001). Significant differences were also noted between baseline and follow-up testing (P < .001) with follow-up scores being higher, and baseline and rebaseline testing (P < .001) with rebaseline scores being higher. No other significant differences were observed for visual motor.
A significant difference was found across the testing sessions for reaction time (F(3,120) = 9.57, P < .001). Composites were significantly higher and slower during post-trauma testing than follow-up (P = .008) and rebaseline (P = .012). Significant differences were also noted between initial baseline and follow-up (P = .001), with faster reaction times in follow-up testing. No other significant differences were observed for reaction time.
Our finding that rebaseline composite scores on the ImPACT did not differ from the scores on the final follow-up evaluation suggests that conducting a rebaseline evaluation in close proximity to an athlete's follow-up evaluation was unnecessary and unwarranted. In the only other study that looked directly at rebaseline testing, Lynall et al8 reported that three of the nine tests comprising the CNS Vital Signs test (visual memory, complex attention, and processing speed) differed significantly from final follow-up to rebaseline, which left equivocal whether the rebaseline testing added value. We saw no such ambiguity with the ImPACT composite scores and concluded that no new information of clinical value was provided by the rebaseline evaluation.
A further purpose of the current study addressed whether rebaseline measurements might have contributed to practice effects. Such an effect could not be studied directly because no evaluation was conducted following the rebaseline session. However, we noted significantly improved scores from original baseline to follow-up evaluation (and rebaseline) in all domains except visual memory. Although the ImPACT employs multiple forms of the subtests across the testing conditions (baseline, post-trauma, follow-up, and rebaseline), repeated test administration might be expected to enhance understanding of procedural sets so that practice effects will emerge. In making sense of their own results, Lynall et al8 reported no improvement in visual memory over time, although other measures did show positive change. They concluded that practice effects might be occurring except in the visual memory domain where athletes could be confusing the shapes seen in the earlier administrations, thus inducing a proactive interference with respect to the later discriminations. Although no improvement in visual memory occurred in the current study from original baseline through rebaseline, a suggestion of proactive interference would not explain differences in outcomes for visual and verbal memory (because both shapes and words may differ across administrations).
The issue of practice effects also is affected by the latency between evaluations. In the current study, an average of 1 month occurred between final follow-up and rebaseline. This contrasted with the Lynall et al study, where the average interval was 397 days. Clearly, if practice effects eventuated, they would have been more likely in the current study than in Lynall et al's study.
During follow-up evaluations, collegiate athletes may be strongly motivated to perform their best to return-to-play for many reasons, including retaining their position on the team and preserving their scholarship.16 Although motivation may be at play more for the follow-up evaluations than baseline evaluations, this theory is weakened for rebaseline evaluations because the athletes have already been allowed to return-to-play. Their performance on the testing does not have a bearing on their participation in their respective sport; thus, one would expect their performance to return to the original baseline measurement. Although it can be posited that because the athletes have already returned to play, the rebaseline would not generate as much motivation to perform as the follow-up, there are no data to suggest the athletes believe that poor or poorer performance on rebaseline testing would be ignored. The current study suggests there is reasonable evidence to conclude that the follow-up scores will be higher than the original baseline score. The current findings complement Lynall et al's8 suggestion that the improvement in test performance may well reflect a practice effect. Although the current results indicate that all testing after the initial post-trauma assessment was higher than both the original baseline and post-trauma evaluation, the mechanism for this difference is unclear. In addition to practice effects, motivation and effort play a key role in follow-up and rebaseline assessments in anticipation of return-to-play.
In comparison with the Lynall et al8 study's use of the CNS Vital Signs test, the current study used the ImPACT, which is by far the most widely used computerized neurocognitive test in college and university concussion management programs.13 The current study included athletes who sustained more than one concussion to the pool of participants, addressing concerns noted by Lynall et al8 about multiple concussions. Also, the current study provided a shorter window of time from follow-up to rebaseline evaluation. The Lynall et al study reported a range of 37 to 333 days for rebaseline testing, with a median of 169.5 days.8 The current study condensed this timeframe to a range of 6 to 97 days, with a mean of 31.02 days from the final follow-up assessment to rebaseline testing.
Although the current study's sample size was larger than the solitary study in the literature that addressed rebaseline testing,8 the size still was smaller than desirable. As such, it would add confidence to generalizability to have a larger sample. Furthermore, by splitting the sample into smaller groups to evaluate for specific variables, such as sex, sport, and history of concussion, the smaller sample size produced a limitation in the significance of an outcome. Differences were noted in the time between baseline and follow-up assessments, as well as follow-up and rebaseline. However, because the results were consistent regarding changes in composites across testing intervals, this may not be a major limitation. One aim was to see whether rebaseline assessment did anything detrimental to the concussion protocol for athletes. It was found to be difficult to determine due to varying factors, such as practice effects and motivation of the athlete. Future research could explicitly assess for these variables. Although the specific cause may be unclear as to why scores were higher in follow-up and rebaseline testing than in the original baseline test, attribution of practice effects and/or motivational effects remains consistent with conclusions from previous studies.
Implications for Clinical Practice
The use of rebaseline assessments is suggested throughout the literature after an athlete has returned to play; however, specifics regarding time interval and course are unspecified. Some concussion management programs have supported the use of these evaluations following the return of the athlete to practice and competition. The current study provides a strong evidence base for the elimination of the rebaseline procedure, at least in the college milieu.
The current study suggests that rebaseline testing with the ImPACT following return-to-play in athletes within close proximity to follow-up is both unnecessary and unwarranted. No significant differences were found between follow-up and rebaseline testing on any composite scores measured on the ImPACT. Additionally, there is reasonable evidence that the athletes' follow-up testing score will be higher than their original baseline score, likely due to practice effects and/or motivational changes. Further, baseline testing comes with several burdens (ie, time constraints, administration, and cost), especially for settings with limited resources. Due to these limitations associated with the administration of rebaseline assessments, the recommendations of the current study ease these burdens. The study unambiguously supports the use of athletes' final follow-up evaluation as their rebaseline measurement for the remainder of the sports season in the event of an additional concussion.
- Johnston K, McCrory P, Mohtadi N, Meeuwisse W. Evidence based review of sport-related concussion: clinical science. Clin J Sport Med. 2001;11:150–159. doi:10.1097/00042752-200107000-00005 [CrossRef]
- Centers for Disease Control and Prevention. Traumatic brain injury and concussion. https://www.cdc.gov/traumaticbraininjury/. Accessed March 28, 2017.
- Iverson GL, Lovell MR, Collins MW. Interpreting change in ImPACT following sport concussion. Clin Neuropsychol. 2003;17:460–467. doi:10.1076/clin.17.4.460.27934 [CrossRef]
- Broglio SP, Cantu RC, Gioia GA, et al. National Athletic Trainers' Association position statement: management of sport concussion. J Athl Train. 2014;49:245–265. doi:10.4085/1062-6050-49.1.07 [CrossRef]
- Webbe FM, Zimmer A. History of neuropsychological study of sport-related concussion. Brain Inj. 2015;29:129–138. doi:10.3109/02699052.2014.937746 [CrossRef]
- Guskiewicz KM, Bruce SL, Cantu RC, et al. National Athletic Trainers' Association position statement: management of sport-related concussion. J Athl Train. 2004;39:280–297.
- McCrory P, Meeuwisse WH, Aubry M, et al. Consensus statement on concussion in sport–the 4th International Conference on Concussion in Sport held in Zurich, November 2012. Clin J Sport Med. 2013;23:89–117. doi:10.1097/JSM.0b013e31828b67cf [CrossRef]
- Lynall RC, Schmidt JD, Mihalik JP, Guskiewicz KM. The clinical utility of a concussion rebaseline protocol after concussion recovery. Clin J Sport Med. 2016;26:285–290. doi:10.1097/JSM.0000000000000260 [CrossRef]
- Asken BM, Bauer RM, Guskiewicz KM, et al. Immediate removal from activity after sport-related concussion is associated with shorter clinical recovery and less severe symptoms in collegiate student-athletes. Am J Sport Med. 2018;46:1465–1474. doi:10.1177/0363546518757984 [CrossRef]
- ImPACT Applications, Inc. ImPACT Administration and Interpretation Manual. Pittsburgh: ImPACT Applications, Inc.; 2016.
- HeadMinder, Inc. Concussion Resolution Index (CRI) Professional Manual. New York: HeadMinder, Inc. and PanMedix, Inc.; 2007.
- Lovell MR. The ImPACT neuropsychological test battery. In: Echemendia R, ed. Sports Neuropsychology: Assessment and Management of Traumatic Brain Injury. New York: The Guilford Press; 2006:193–215.
- Buckley TA, Burdette G, Kelly K. Concussion-management practice patterns of National Collegiate Athletic Association Division II and III athletic trainers: how the other half lives. J Athl Train. 2005;50:879–888. doi:10.4085/1062-6050-50.7.04 [CrossRef]
- Schatz P, Pardini JE, Lovell MR, Collins MW, Podell K. Sensitivity and specificity of the ImPACT Test Battery for concussion in athletes. Arch Clin Neuropsychol. 2006;21:91–99. doi:10.1016/j.acn.2005.08.001 [CrossRef]
- Bruce J, Echemendia R, Meeuwisse W, Comper P, Sisco A. 1 year test-retest reliability of ImPACT in professional ice hockey players. Clin Neuropsychol. 2013;28:14–25. doi:10.1080/13854046.2013.866272 [CrossRef]
- Bailey CR, Echemendia RJ, Arnett PA. The impact of motivation on neuropsychological performance in sports-related mild traumatic brain injury. J Int Neuropsychol Soc. 2006;12:475–484. doi:10.1017/S1355617706060619 [CrossRef]
Demographic Characteristics of the Sample
|Characteristic||N||Mean ± SD||Range|
|Age (y)||41||19.61 ± 1.52||18 to 23|
|Education||41||13.32 ± 1.15||12 to 16|
|Duration of recovery (days)||41||18.44 ± 20.69||4 to 114|
|History of prior concussion||41||0.73 ± 1.05||0 to 4|
|Days from post-trauma to follow-up||41||18.44 ± 20.69||4 to 114|
|Days from follow-up to rebaseline||41||31.02 ± 17.33||6 to 97|