Return to Sport Following Adolescent Concussion: Epidemiologic Findings From a High School Population

Materials and Methods

After receiving institutional review board approval, the clinical records of 357 consecutive patients (14 to 18 years old) who presented to a single health care system between September 2013 and December 2016 with concussions were retrospectively reviewed. Four independent reviewers collected all data and performed extensive chart reviews. Patients with diagnosed neurological disorders were excluded.

Patient demographic and historical data, including age, sex, sport of concussion, and history of previous concussion, were collected from patient charts. Concussion-related variables such as loss of consciousness, amnesia, mechanism of injury, timing of diagnosis, timing of removal, and RTS were also obtained via retrospective chart review. History of concussion was used to denote a self-reported history of remote concussion by the patient. Previous concussion was used to denote a documented concussion that occurred during the study period. Patients were permitted to RTS after completing the 6-step pathway recommended by the Centers for Disease Control and Prevention, with each step requiring a minimum of 24 hours and the athlete to remain asymptomatic. The steps are as follows: no activity until asymptomatic at rest, light aerobic exercises, sport-specific exercise, noncontact training drills, full-contact practice, and RTS.⁸ Patient ImPACT scores were obtained from a separate database of ImPACT assessments collected by neuropsychologists at the authors' institution who perform and interpret Im-PACT tests on athletes.

The ImPACT metrics of verbal memory, visual memory, motor speed, reaction time, impulse control, and cognitive efficiency index and symptom scores were recorded at each round of testing, corresponding to each concussion in the study period. Athletes at the authors' institution also had baseline ImPACT testing performed prior to the season. For athletes not obtaining ImPACT testing prior to the season, a normative comparative data set was created based on baseline ImPACT testing of athletes with similar demographics who had no history of concussion. Data from the concussed athlete were then compared with either normative data matching the patient's demographics or baseline scores, for those applicable.

Verbal memory, visual memory, motor speed, and reaction time were expressed as a percentage of the normative data and/or baseline scores. Impulse control is expressed as the sum of errors committed during different phases of the test and has traditionally been used as a surrogate for test validity.⁹ Cognitive efficiency is expressed as an integer. It represents a measurement of percentage correct and reaction time in seconds on the symbol match test. A higher score indicates that the athlete performed well in both the speed and accuracy components of the symbol match test, and thus was more cognitively efficient.¹⁰ The symptom score is an integer representing the sum of an athlete's self-reported constitutional symptoms (eg, headache, dizziness, vomiting, and sensitivity to noise) data.⁹ A higher score reflects a higher symptom total.

Results

A total of 357 patients with concussions presented to the authors' institution between September 2013 and December 2016. No patients were excluded from this study. The average age of injury for all athletes was 15.5 years (range, 14–18 years), and 61.9% of patients were male. Football was the most common sport of injury (27.7%), followed by hockey (21.8%) and soccer (17.0%). Of those concussed, 6.7% reported loss of consciousness and 14.3% reported amnesia. A history of concussion was reported in 33.0% of patients. Overall, 99.1% of patients were able to RTS (Table 1). When examining the 3 most common sports of injury, 30.6% of football players, 48.1% of hockey players, and 33.3% of soccer players reported a historic previous concussion. Among concussions sustained during the study time period, hockey players had the highest rate of loss of consciousness and amnesia (10.4% and 15.6%, respectively).

Table 1: Demographics of All Athletes

On average, high school athletes required 30.4±23.3 days of recovery prior to RTS. There was a high incidence of history of concussion (33.1%) in high school athletes, and 32 athletes experienced a recurrent concussion. There was an increased time to RTS in athletes with a history of concussion (33.5±24.9 days vs 29.2±22.6 days, P=.12) or who sustained recurrent concussions (35.3±38.3 days, P=.77), although this was not statistically significant. Mean ImPACT verbal memory and visual memory scores increased between baseline ImPACT scores and Im-PACT scores after 1 and 2 concussion(s) for those athletes who sustained multiple concussions during the study period. Visual memory showed a significant increase in patients with a history of concussion as compared with those without a concussion (56.1%±25.5% vs 35.7%±25.9%, P=.015). Mean visual motor speed and reaction time decreased with each recurrent concussion (Figure 1).

Figure 1: Changes in Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) scoring with consecutive concussions (A). Changes in impulse control, symptom score, and cognitive efficiency index with each recurrent concussion (B).

A multivariable linear model controlling for possible covariates demonstrated that athletes with a history of concussion (P=.005), females (P<.001), and patients diagnosed in the clinic (P=.046) had a significantly longer RTS time (Table 2).

Table 2: Multivariable Linear Model and Multivariable Poisson Model: Effect of Demographic Variables on Time to Return and Number of Visits for All Athletes

Discussion

Risk factors for and outcomes of concussions in the youth athlete remain an important topic. This study found a high prevalence of high school athletes with a history of concussion, and these athletes require approximately 1 month of recovery prior to RTS. Females, athletes with a previous history of concussion, and those diagnosed in-clinic rather than in-game were found to have increased RTS times. In analyzing ImPACT scores, the authors found that as athletes sustained multiple concussions, scores related to memory increased, whereas scores related to motor speed and reaction time decreased.

Previous studies have evaluated RTS timing after a concussion. In 2000, Guskiewicz et al¹¹ performed an epidemiological study on high school and college football players who sustained concussions. Their study demonstrated that an alarming 30.8% of concussed players returned to competition on the same day, and 20% of athletes never left the game. They attributed these findings to a historic disregard of RTS guidelines by clinicians. A recent study by Putukian et al¹² suggested that high school athletes take approximately 30 days to RTS after a concussion, as compared with 3 to 5 days in professional athletes. This finding was likely due to a more contemporary approach to young athletes, emphasizing more careful and conservative treatment.^12–14 The current study had similar findings, as high school athletes required approximately 30 days of recovery before RTS after a concussion. Additionally, the current study found a significant increase in RTS timing in those with a previous history of concussion. These findings suggest that there is greater compliance to contemporary RTS guidelines; however, there is no standardized RTS time because presentation and severity of concussions vary greatly.¹³ Furthermore, the results suggest that prior concussions may have lingering effects and that incomplete recovery may cause additive cortical and subcortical pathogenesis in a dose-response manner.^15–18

Researchers have evaluated the effects of concussion on ImPACT scores. Tsushima et al¹⁹ evaluated recurrent concussions in high school athletes and found that they did not have significant changes in their ImPACT scores of verbal memory, visual memory, reaction time, processing speed, or symptom score with recurrent concussions. Conversely, Asken et al²⁰ investigated the effect of ImPACT scoring on RTS in collegiate athletes and found that 60% of their patients exhibited a reliable decline in 1 or more ImPACT composite scores after a single concussion. However, their study found a paradoxical increase in visual memory after only 1 concussion. Similarly, the current study found that patients with a history of concussion showed a statistically significant 20% increase in visual memory after 1 concussion as compared with patients without a history of concussion. The current authors also found that patients who went on to have subsequent concussions demonstrated an increase in memory ImPACT testing. These results suggest that as patients sustain multiple concussions and undergo repeated ImPACT testing, they gain familiarity and comfort with the test and are able to perform the test's memory tasks more accurately and expeditiously, thus increasing their memory scores. When evaluating other ImPACT scores more representative of cognitive function, the current study found a reliable decline in average visual motor speed and reaction time. These ImPACT domains are not as susceptible to learning and thus show a dose-related decrease in score, reflecting previous findings that subsequent concussions have a dose-response decrease in cognitive functions among athletes.¹⁸

The analysis of athlete characteristics on outcomes following concussions has led to various findings.^21–24 Berz et al²⁵ found that females have a postconcussion symptom score nearly twice as high as those of males (30.9 vs 15.8, P<.05), and females may take longer to become symptom free after a sport-related concussion. Broshek et al²⁴ and Covassin et al²² demonstrated that females have decreased simple and complex reaction times and processing speed as well as persistently lower visual memory composite scores after concussion. The current study found similar results, with females demonstrating significantly longer RTS times compared with males. These findings have been hypothesized to be due to worse vestibular and oculomotor impairment after concussion in females. This impairment may manifest as symptomatic headache and dizziness, delaying RTS.²¹ Finally, the current study found that athletes diagnosed in-clinic had a longer RTS as compared with those diagnosed in-game. These findings suggest that athletes diagnosed in-clinic had delayed diagnosis and ultimately delayed initiation of treatment as compared with those diagnosed in-game. Previous studies have found that athletes who began cognitive and physical rest immediately after injury (ie, in-game diagnosis) were more likely to recover within 30 days compared with patients who had delayed cognitive and physical rest for 1 to 7 days after their injury (the time it takes for an athlete to obtain a clinic appointment).²⁶ These findings underscore the importance of vigilance during the game and prompt diagnosis of concussions by team physicians.

Conclusion

This study found a high prevalence of previous concussions in high school athletes. On average, players required 1 month of recovery before RTS. However, female athletes, players with a previous history of concussion, and those with delayed diagnosis required increased time to RTS. Impact scores for memory (visual and verbal) were found to increase with recurrent concussions, whereas visual motor speed and reaction time decreased.