Factors Associated With Concussion Nondisclosure in Collegiate Student-Athletes

Research Design and Participants

This was a cross-sectional study of NCAA collegiate student-athletes from 4 Division I and 2 Division II universities in Michigan, Ohio, Pennsylvania, and South Carolina. Data were collected from fall 2017 to spring 2018. All collegiate student-athletes included in the study were between 18 and 24 years old and participated in an NCAA-sanctioned sport at their respective institution at the time of study participation. A total of 1125 collegiate student-athletes initiated the study survey. The survey response rate could not be calculated for the current study and is acknowledged as a limitation. Before survey administration, athletes at each participating university were instructed that they must be 18 years or older. To ensure that only athletes 18 or older took the survey, the first question asked, “Are you 18 years or older?” Nine (0.8%) collegiate student-athletes were excluded, as they were younger than 18 years of age at the time of the study and were instructed to stop the survey. To reduce response biases, collegiate student-athletes were excluded if they sustained a concussion within the previous 3 months (n = 27; 2.4%), reported impairment from a concussion sustained longer than 3 months before or were receiving treatment for protracted impairment due to concussion at the time of recruitment (n = 16; 1.4%), or had missing or incomplete data (n = 332; 29.5%). After applying the exclusion criteria, we analyzed data for 741 participants.

Measures

We used a 10- to 15-minute survey that asked about personal and sport demographics, diagnosed concussion and nondisclosed concussion history, concussion knowledge, and level of agreement regarding pressures to play after a head impact experienced during collegiate sport participation. The demographics section addressed age, sex, academic year, level (eg, Division I, II, III), and sport. For the diagnosed concussion history, participants were asked to report the number of concussions they sustained that were diagnosed by a physician or certified athletic trainer. In addition, nondisclosed concussion history was obtained by asking participants to provide the number of possible concussions they had sustained that they did not report to their coach, certified athletic trainer, parent, or teammate. For the purposes of the current study, we transformed diagnosed concussion and nondisclosed concussion history into dichotomous variables (yes or no). Participant concussion knowledge was assessed using an adapted version of a previously published instrument developed by Register-Mihalik et al.²⁰ Participants were asked to recognize the signs and symptoms of concussion and answer general concussion-knowledge questions for a total knowledge score (maximum = 49). Higher scores indicated greater knowledge of concussion. An additional 7 concussion signs and symptoms (sensitivity to light and noise, balance problems, fogginess, feeling more emotional, fatigue or low energy, and feeling slowed down) were added to the original instrument to be consistent with current concussion-identification and -management practices. The adapted concussion-knowledge assessment used in the current study had acceptable internal consistency with a Cronbach α of 0.77. Finally, participants were asked to indicate their level of agreement with the following statement: “As a college athlete, I have felt pressure from _______ to return to play after sustaining a hit to the head after playing sports.”⁶ Pressure from external sources (ie, coaches, teammates, fans, parents or family members) was rated on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). We transformed the pressure item into a dichotomous variable, with participants who reported slightly agree, agree, or strongly agree categorized as experiencing pressure from external sources and athletes who reported anything else categorized as not experiencing pressure from external sources. The final survey was assessed for face validity by 3 content experts and pilot tested with 10 collegiate undergraduate students who did not participate in sport.

Procedures

Before recruitment and data collection, we obtained university institutional review board exemption approval. We sent head athletic trainers at each participating institution an informative email detailing the study. Certified athletic trainers at each participating institution served as liaisons between the researchers and sports teams to schedule and coordinate data-collection sessions. Data collection occurred in athletic training rooms or designated classrooms, at team practices or meetings, before or after practice, or as participants received treatment for an unrelated musculoskeletal injury. Participants were first provided with an informed consent form to review, and consent was indicated when they voluntarily completed the study survey. The survey was administered to participants via iPad (Apple Inc, Cupertino, CA) or paper and pencil depending upon the resources available at each data-collection site. Survey responses collected via iPad were recorded on the Qualtrics survey platform (Provo, UT). Paper-and-pencil responses were later entered into Qualtrics by a trained researcher. Data were imported into and analyzed using SPSS (version 24; IBM Corp, Armonk, NY).

Data Analysis

Demographic information including age, sex, academic year, division, and sport was described using means, standard deviations (SDs), frequencies, and percentages. Sport was transformed into a binary variable of high (basketball, football, field hockey, ice hockey, lacrosse, soccer, softball, volleyball, wrestling) and low (baseball, bowling, cross-country, golf, rowing, swimming and diving, tennis, track and field, water polo) concussion-risk sports, based on epidemiologic findings from Zuckerman et al.¹ Age, sex (female or male), concussion-risk type (high or low risk), diagnosed concussion history (yes or no), concussion knowledge, and experienced pressure (yes or no) from coaches, teammates, fans, and family or parents were included in analyses as potential factors associated with concussion nondisclosure history (yes or no). Before data analysis, we determined that the assumptions of linearity of the logit and multicollinearity were met. Univariable logistic regression analyses were conducted on each factor to determine the odds of not disclosing a concussion. Significant univariable factors were then used in a multivariable logistic regression analysis. Significance was set a priori at P ≤ .05.

Sex and Concussion-Risk Sport Type

Male athletes and those in high-risk sports were more likely to not report a concussion than were females and those in low-risk sports. The majority of research supports these findings, with female athletes more likely to report their concussion to an authoritative figure than were male athletes.^5,21,22 Kerr et al⁵ indicated that male collegiate athletes were 2.88 times more likely to not disclose their concussion than were female collegiate athletes. However, other authors^9,23 described no sex differences in collegiate athletes regarding disclosing their concussions. Researchers reported that athletes participating in high concussion-risk sports such as cheerleading, football, lacrosse, and soccer were more likely to not disclose their concussion than were athletes participating in low risk sports.²⁴ In a study of retired NCAA athletes, 68% of football athletes were more likely to not disclose their concussions compared with 11% of females in low-risk or noncontact sports.⁵ In addition, almost two-thirds of football athletes were more likely to continue playing with a headache after a hit to the head compared with 35% of female athletes.²⁵ Thus, it is imperative that clinicians continue to educate and monitor male collegiate athletes and those in high concussion-risk sports for possible concussions, as they are more likely to not disclose a concussion to an authority figure. Also, behavior modification should be encouraged to reduce the risk of concussion in this population.

Diagnosed Concussion History

Concussion history was a significant factor associated with concussion nondisclosure. Specifically, collegiate student-athletes who had a history of at least 1 diagnosed concussion were 2.3 times more likely to not report a suspected concussive injury than those who had not sustained a concussion. These findings are similar to those recently reported in the literature on high school athletes,¹⁵ collegiate football players,¹⁴ Irish collegiate student-athletes,⁹ and US military cadets.¹⁶ Collectively, these results strongly suggest that prior personal experience with concussion is a dissuasive factor for injury disclosure. Those who sustained a concussion previously may choose to not report a future injury to avoid what they consider to be negative consequences of the management and recovery process (eg, removal from participation, loss of status on their athletic team, isolation, mental health concerns).¹⁵ Another explanation is that those who successfully recovered from a concussion may feel that they can self-manage their symptoms using strategies that worked for their previous injury. Moving forward, health care providers should take note of the modifying role a diagnosed concussion history has on injury disclosure. We can work to overcome this barrier by being transparent about the postinjury management process and actively seeking opportunities to engage previously concussed athletes in conversations about the importance of disclosure and the potential complications that could result from not reporting.

Concussion Knowledge

The relationship between concussion knowledge and concussion nondisclosure has been continuously investigated among collegiate and high school cohorts. The literature consistently demonstrated that concussion knowledge does not always equate to behavior modification or disclosure of a possible concussion. Among Irish collegiate-athletes, concussion knowledge did not affect nondisclosure.⁹ In our study, the odds of nondisclosure increased by 7.3% for each 1-point increase in concussion knowledge, independent of male sex, concussion history, and pressure from coaches. These results support the ongoing narrative that more knowledge does not, in fact, equate to better reporting of concussions, and they reveal an inverse relationship between concussion knowledge and disclosure. Moreover, our findings affirm those of Piana et al²⁶ that female collegiate ice-hockey athletes with higher concussion-knowledge scores were more likely than their peers to have sustained a concussion and not reported it. These outcomes continue to show that concussion knowledge remains an auxiliary item on the pyramid of influence²⁷ and that behavior modifications will likely continue to not change without changes in the structural foundation of values that shape reporting and nondisclosure.

Pressure from Coaches to Play After a Head Impact

In our sample, nearly 1 out of every 5 collegiate student-athletes indicated that he or she had felt pressured by a coach to play after a head impact. The percentage of participants who felt pressure from coaches to play after a head impact in this study (18.2%) was less than in previous research among Irish collegiate athletes (49.6%).⁹ Our observations were more similar to those of Kroshus et al,⁶ who noted that 13.68% of a US collegiate student-athlete sample experienced pressure from a coach during the previous season to continue playing after a head impact. Although our participants stated they felt pressure from other external stakeholders (ie, teammates, fans, and parents or family), pressure from coaches had the greatest negative influence on concussion nondisclosure. Earlier authors^20,21,28 determined that athletes were deterred from reporting concussive symptoms by concern about their coaches' reactions and how being removed from play could negatively affect their team. These results highlight the critical role of interpersonal relationships in regard to concussion reporting. The power of coaches to affect team concussion culture has been documented,²⁹ and evidence suggests that positive coach communication regarding concussion identification and reporting significantly increases athletes' intentions to disclose an injury.³⁰ Moving forward, coaches should be educated further about their role in shaping team norms for sport safety⁶ and encouraged to have open conversations about injury reporting with their athletes to normalize concussion disclosure.

Our study had a few limitations that need to be considered when interpreting the results. First, the survey response rate could not be calculated. Before the data-collection sessions, athletes were given the exclusion criteria and did not appear for the session if they were excluded or did not want to participate. We do not know how many student-athletes did this; therefore, the results should be interpreted with caution due to the potential for nonresponse bias. Second, although our dataset was still considered large, female athletes accounted for just over a quarter of the sample. Third, the data were collected from 2 states in the Midwest, 1 state in the mid-Atlantic, and 1 state in the Southeast. Thus, our findings cannot be generalized to all NCAA collegiate student-athletes. Finally, as with all survey research, recall bias may have influenced the results. Our athletes may not have remembered all of their concussions or disclosed a concussion when it was either a subconcussive injury or due to other factors (ie, dehydration). Future investigators should focus on factors that influence disclosure using socioeconomic models, specifically focusing on concussion interventions that target the cultural norms of “winning at all costs.” Further work should also address the age at which athletes start to not disclose their concussions, so that interventions can be implemented early in their sport participation. Finally, future authors should concentrate on interpersonal factors and concussion nondisclosure, focusing specifically on the coach-athlete relationship and ways to build trust and relationships between coaches and student-athletes.