Concussion knowledge, attitudes and reporting intention amongst UK university student-athletes: Implications for institutions, coaches and future research

This article reports on a survey of concussion knowledge (CK), attitudes (CA) and reporting intention (CRI) amongst a multisport cohort of UK university student-athletes. A cross-sectional study design was used, with 217 student-athletes completing the Rosenbaum Concussion Knowledge and Attitudes Survey – Student Version. The findings show that CK is broadly aligned with previous study cohorts including research in North America, and that sport type had no effect on CK, CA and CRI. Whilst gender did not affect CA or CRI, males returned higher CK scores than females. The respondents’ concussion history did not affect CA. Paradoxically, respondents with a history of concussion had significantly greater CK scores yet were more likely to report an intention to continue playing with SRC symptoms. There are three key implications of these findings. The comparison of findings between males and females highlights the need for a more nuanced conceptualisation of safety and risk in relation to SRC. Second, the evidence suggests a strong need for UK universities to introduce bespoke education programmes and policies formalising the healthcare support and return to learn of students post-SRC. Third, awareness of the complex interplay between SRC knowledge, experience and behaviour will enable coaches across all sports to exercise caution when managing SRC amongst students. The article concludes by arguing that there is a need for further qualitative research to understand how concussion is experienced by UK university students and provided for and accommodated within UK universities.


Introduction
:p.838:p.845Consequently, the provision of medically validated information is understood as an important part of the public health response to SRC, as an antidote to both the frequently inaccurate representation in the media 2 and the well-documented tendency of athletes to 'under-report' concussion injuries. 3,4This has stimulated a wide range of studies documenting SRC knowledge, attitudes and reporting intentions, spanning populations from clinicians to youth sport participants.Indeed, these properties lead SRC to be unique amongst sport-related injuries in having such a strong focus on education and awareness as a strategy for injury prevention.Studies are most frequently directed at single sport cohorts, and the vast majority have been undertaken within North America.
Much research has focused on college/university students and has shown a significant incidence of concussion. 5,6,7,8,9For instance, Chinn and Porter 10 and Kroshus et al. 4 respectively recorded that 38% and 41% of college student-athletes reported having previously been concussed.In line with epidemiological studies, 11 the reported incidence of SRC amongst college/university students varies across sports.Whilst studies have also shown gender differences in concussion knowledge (CK) amongst both college students 8,12 and sports participants more widely, 13,14 no discernible pattern is evident. 9,15ore generally, studies have demonstrated that concussion history is not associated with changes in attitude, 16 or reporting behaviour, 4 and may even have a negative influence on concussion attitudes (CA). 17here have been relatively few UK studies to date, but most have raised concerns about the 'limited' 18 and 'insufficient' 19 SRC knowledge within this population.Other UK studies have highlighted how higher levels of knowledge do not necessarily equate to more risk averse attitudes, reporting intentions, or predicted future behaviour. 20,21Both Hutchinson et al. 20 and Travis et al. 16 were generally positive about the overall levels of SRC knowledge in their cohorts, but also found that 65.6% and 45.3% respectively would continue playing after sustaining a concussion.The work of Register-Mihalik et al. 17 suggests that the phenomenon whereby enhanced knowledge does not equate with safer attitudes is a more universal response to SRC.
In the UK context, the students' knowledge of and attitudes towards SRC resonate with current debates about the scope of universities' duty of care.Whilst a UK university has no legal responsibility to provide healthcare, it must 'act reasonably to protect health, safety and welfare of its students'. 22Increasingly, these institutions recognise the moral duty, for instance, to counteract harms such as sexual violence and cyberbullying, 23 and provide bespoke mental health services for students.Additionally, Towl 24 has argued for the related commercial benefits in terms of marketing, reputation and ultimately recruitment.The Higher Education Policy Institute is currently campaigning for a statutory duty of care to be introduced for students in higher education (HE). 25hilst sport has not yet featured in these debates, there is a strong case that it should do.A large proportion of the student population participate in organised sport, with over 100,000 students participating in the British Universities and Colleges Sport (BUCS) sanctioned events in the 2021/2022 season and even more participating in lower level, intra-mural sport. 26Given the absence of any legal responsibility, healthcare provision is inevitably ad hoc and heterogeneous across universities.Some universities, including the one featured in the study, have appointed medical leads for concussion and established referral networks for cases that cannot be locally managed.It should be noted, however, that this institution is world-renowned for sport-related subjects and therefore likely has some of the best provision in the sector.More generally, UK universities typically do not provide students with access to SRC-trained medical personnel, emergency action plans and/or neurocognitive/neuropsychological testing.The formal BUCS position is that students should follow the sport-specific guidance issued by individual governing bodies and issues no pronouncement on provision of SRC healthcare.For example, the institution of the second author has no referral network in place resulting in SRCs either being self-managed by the student-athlete or reliant upon the primary care provided by the national healthcare service.
Therefore, there is little to suggest that student-athletes in UK HE are treated in accordance with international standards of SRC management. 1 By way of contrast, in the USA, the NCAA Concussion Safety Protocol Checklist contains 49 components across eight areas: pre-season education; pre-participation assessment; recognition and diagnosis; post-concussion management; return to play (RTP); return to learn (RTL); reducing exposure to head trauma; and administrative obligations. 27Within the University setting, the appropriate management of SRC assumes greater significance due to research evidencing the negative impact of concussion on academic performance, 28 and the embedding of RTL protocols in best practice concussion management. 1 With the concerns noted above, and in light of the paucity of information in UK HE settings, the aim of this study was to assess the SRC knowledge, attitudes and reporting intentions of university students in the UK.Additionally, we sought to examine perceptions of the SRC healthcare support available to students, and students' readiness to respond when SRCs occur.The nature of the research design enabled us to examine the differences across sports, the differences between male and female athletes, and the influence of a personal concussion history.

Study design
A cross-sectional survey was used to assess CK, attitudes and reporting intention in a university student-athlete population.This approach was chosen due to the exploratory nature of the research as well as our desire to provide an observational overview to inform future research and policy development (discussed below). 29A notable feature of this field of study is that no standardised measurement tool exists, and researchers frequently develop bespoke instruments for the assessment of SRC knowledge.Whilst many of those have established face validity, our initial review of literature showed that two instruments vie in terms of frequency of use, either in their original or modified form; namely, the Rosenbaum Concussion Knowledge and Attitudes Survey -Student Version (RoCKAS-ST), 30 and the tool developed by Register-Mihalik et al. 3 The RoCKAS-ST has the advantage of being the only tool with psychometric properties, 31 and whilst the reliability of its attitudes (though not knowledge) measure has been questioned, 6 its validation and wider use 32 made it the most appropriate survey tool for this study.

Participants and procedure
After institutional ethical approval was obtained, the survey was advertised to all students studying at Loughborough University.The university consistently ranks in the top 10 in national league tables and is internationally renowned for sport-related subjects, providing an ideal institution for this multisport study of UK university student-athletes CK, attitudes and reporting intentions.The survey was distributed online (the responses were received from 8 December 2017 to 29 March 2018) and was publicised on social media by the student elected head of the Athletic Union and by the university's sports services department.Self-identified involvement in university sport was the only inclusion criterion.

Measures
The RoCKAS-ST self-report survey comprises of 55 items from which Concussion Knowledge Index (CKI) and Concussion Attitude Index (CAI) scores are derived.The CKI includes 25 items comprised of 17 true/false and eight symptom recognition questions with a possible score range of 0-25 points with higher scores suggesting greater CK.The CAI includes 15 items assessing levels of agreement with a range of scenarios (5-point Likert scale; 1 = strongly disagree and 5 = strongly agree) and a possible score range of 15-75 points.Higher scores suggest safer attitudes towards concussion.Concussion Reporting Intention (CRI) was measured using a single item from Section 3 of the RoCKAS-ST.The item states 'I would continue playing a sport while also having a headache that resulted from a minor concussion' and is scored on a 5-point Likert scale (Possible score 1-5).This item is reverse scored to derive the overall CAI score; thus, a higher score on the CRI measure suggests greater intention to report symptoms.This measurement has been used in previous research and is considered to be a face valid proxy for concussion symptom reporting. 4The RoCKAS-ST instrument has an internal validity index consisting of three true/false questions that are each scored with one point for correct answers (possible scores 0-3).
The RoCKAS-ST was supplemented with demographic questions to identify age, gender, ethnicity, year of study/ degree level, participation level (specifying Athletic Union club membership or intra-mural sport), concussion history (medically or self-diagnosed) and primary sport.Sports were then grouped into combat (e.g.martial arts), collision (e.g.rugby union), contact (e.g.soccer) and high and low risk non-contact sports (e.g.equestrianism and tennis, respectively).
Because we were also interested in the broader environment of SRC healthcare in universities, the participants were asked about how they would respond to a concussion (their own or a team mates').The participants were also asked to rate (on a 5-point Likert scale) both self-perceived confidence in their ability to help a teammate who had a suspected concussion, and the degree to which they felt that healthcare support for concussion injuries at the university was available.

Data analysis
Data were subject to validity checking prior to analysis.As per parameters described by Rosenbaum and Arnett, 30 a survey response was deemed valid if validity index score was 2 (or higher), and at least 90% of the CKI questions and 90% of the CAI questions had been completed.The participants that did not meet these criteria were deemed invalid and removed before analyses.
Valid data were analysed using SPSS (version 28, IBM corp., Armonk, NY, USA).Descriptive statistics were generated to present demographic characteristics and outcome measures.Mean ± standard deviation is presented for CKI (0-25), CAI (15-75) and CRI (1-5) scores, alongside the frequency and percentage of the responses to the CK and CA items.Data were deemed to be non-normally distributed via Kolmogorov-Smirnov tests, and thus, nonparametric tests were conducted to assess differences between gender, sport type and concussion history against the outcome measures CKI, CAI and CRI.Mann-Whitney U tests were used for assessing differences between two independent groups (gender and concussion history).For analysis of gender, two responses of 'Prefer not to say' were removed to allow for bilateral analysis (male vs female), making the sample 215 for gender.Kruskal-Wallis tests were used to assess differences between sport types.A Chi-square test of independence was performed to assess the influence of gender, sport type and concussion history on frequency of survey response rates to the 5-point Likert scale from which CRI scores were derived.Statistical significance for all tests was accepted at p < 0.05.

Results
A total of 231 student-athletes responded to the survey.All the participants passed the validity check; however, 14 respondents were excluded for not meeting the inclusion criteria of involvement in university sport or completing at least 90% of the survey items.Therefore, 217 responses were analysed (57.6% female, age = 19.8 ± 1.6 years).The demographic characteristics of the participants are presented in Table 1.

Concussion knowledge and attitudes
The mean CKI score for student-athletes was 19.3 ± 1.9 (Table 2).There was widespread recognition of accurate SRC statements, including acknowledgement that: concussion symptoms can last for several weeks (CKI 1.8, 97.7%), that second impact syndrome can occur (CKI 1.1, 92.6%), and that concussion can lead to emotional disruptions (CKI 1.16, 92.6%).The respondents demonstrated a good understanding of common falsehoods attributed to concussion, namely, that concussion: necessarily entails loss of consciousness (99.5% stated false to CKI 1.5), negatively impacts on intelligence (CKI 1.12, 97.7%), and does not impact upon sporting performance (CKI 2.3, 96.3%).The misunderstandings highlighted from the survey related to the ideas that: 'people who have had one concussion are more likely to have another concussion' (49.8% stating false to CKI 1.3), and 'after 10 days, symptoms of a concussion are usually completely gone' (41.5% stating false to CKI 1.13).Of the valid responses, only 12.4% recognised that 'an athlete who gets knocked out is experiencing a coma' (CKI 1.17).The mean response to the  The responses in bold signify correct answers.
16-item symptom recognition checklist (with eight correct symptoms) was 6.8 ± 1.2.The mean score for CAI amongst the student-athlete sample was 62.5 ± 6.3.Table 3 displays frequency of the responses to the CA items sections 3 and 4. The safest attitudes were expressed in relation to the views that coaches should use extreme caution when making RTP decisions (CAI 3.2, 97.7%).Most respondents disagreed that athletes had a responsibility to RTP whilst still experiencing symptoms (CAI 3.6, 93.1%), and that 'concussions are less important than other injuries' (CAI 3.5, 92.2%).

Responses to concussion
CRI was assessed by the response to item 3.1 in the survey, 'I would continue playing a sport while also having a headache that resulted from a minor concussion'.Whilst overall respondents were neutral towards this statement (with a mean CRI score of 3.4 ± 1.2), 55.8% expressed the safest attitudes towards continuing to play sport whilst experiencing symptoms of a concussion, and 44.2% responded either 'neutral', 'agree' or 'strongly agree' to this item (Table 3).The responses in bold signify the safest response.Additionally, the respondents were asked to list the three most important things to do if they felt concussed or suspected that someone else had been concussed.The responses are reported in a word cloud (Figure 1).The respondents self-rated confidence in their ability to deal with a teammate experiencing concussion was obtained from the following survey item: 'On a scale of 1-5 (where 1 is not at all confident and 5 is extremely confident) how comfortable would you feel dealing with a teammate or fellow player who had concussion?'.The respondents were neutral with the mean response being 3.0 ± 1.2.
This neutrality was replicated in the respondents' evaluations of the healthcare support for concussion injuries available at the university.On a similar scale (1 = not well supported, 5 = extremely supported), the mean response was 3.2 ± 1.0.

Demographic differences
The student-athlete outcome measures (CKI, CAI and CRI) were analysed for differences by demographic variables.Table 4 displays mean CKI, CAI and CRI scores with standard deviation, stratified by sample characteristic.
Sport type.No significant differences for CKI or CAI were found between sport types (p = 0.752 and p = 0.227, respectively).The relationship between sport type and CRI was non-significant, X 2 (16, N = 217) = 11.5, p = 0.779.Sport type was not found to influence CK, CA, or CRI.
Gender.Statistical analysis found that males scored significantly higher on the CKI than females (M = 19.9 ± 1.9 vs F = 19 ± 1.9) (p = 0.01).The three greatest differences were found in relation to: potential duration of symptoms (49.6% of females compared to 71.1% of males correctly stating true to CKI 1.13), the belief that concussion only results from a direct hit to the head (66.4% of females compared to 84.4% of males correctly stating false to CKI 1.6), and the connection between loss of consciousness and coma (8.0% females compared to 18.9% males correctly stating true to CKI 1.17).No significant differences were found between males and females on CAI score (p = 0.662) or CRI score (X 2 (4, N = 215) = 6.7, p = 0.15).Concussion history.Amongst the participants, 27.0% reported having experienced a concussion in the previous 12 months, and 56.7% reported having experienced a concussion at some time in the past.Overall, a greater proportion of males (63.7%) than females (37.4%) reported that they had previously been concussed.Those who took part in collision sports were the most likely to have experienced concussion in the past (72.0%),and those who participated in low risk non-contact sports were the least likely (34.0%).Fewer than half of combat sport participants reported a previous concussion (47.6%).
No significant differences for CAI scores were found between those that reported experiencing a concussion in the last 12 months and those that did not (p = 0.395).There was, however, a significant difference in CKI scores, with those reporting a concussion having greater knowledge scores than the respondents not reporting a concussion (19.8 ± 1.9 vs 19.2 ± 1.9, respectively) (p = 0.023).Knowledge was most markedly enhanced in relation to the impact of concussion on memory (CKI 1.9), the potential longer-term impacts of concussion (CKI 2.1, CKI 2.2), and the physical damage to the brain sustained through concussion (CKI 1.11).The relationship between concussion history and CRI was significant, X 2 (4, N = 217) = 17.3, p = 0.002.Mean CRI scores for those that reported suffering a concussion in the last 12 months was 2.8 ± 1.2 compared to 3.6 ± 1.2 for those that had not suffered a concussion.Those that reported having experienced a concussion in the last 12 months were more likely to report continuing playing a sport whilst experiencing a headache because of a minor concussion (Figure 2(a)).
No significant differences for CAI scores were found between those that reported experiencing a concussion at any time in the past and those that did not (p = 0.811).There was a significant difference in CKI scores, with those reporting a concussion having greater knowledge scores than the respondents not reporting a concussion (19.8 ± 1.9 vs 18.7 ± 1.9, respectively) (p = 0.001).The relationship between concussion history and CRI was significant, X 2 (4, N = 217) = 12.9, p = 0.011.Mean CRI for the participants experiencing a concussion at any time in the past was 3.1 ± 1.2 versus 3.6 ± 1.2 for those that had not experienced a concussion.Those that reported experiencing a concussion at any time in the past were more likely to report continuing playing a sport whilst experiencing a headache because of a minor concussion (Figure 2(b)).

Knowledge, attitudes and incidence
This study constitutes one of the largest UK surveys of sports participants' SRC knowledge and attitudes.To the extent that it is possible to make comparisons, the knowledge of concussion exhibited by this cohort is broadly in line with previous studies of this type, 10 but greater than in most previous UK surveys. 20,21he incidence of SRC reported in this survey exceeds previous UK studies, 18,20 and previous studies of North American college/university student cohorts. 4,10Whilst the incidence of SRC varied significantly across different types of sport, no statistically significant differences were found for the SRC knowledge and attitudes of different sports participants.

Gender
In contrast to studies that have previously reported either no significant differences in SRC knowledge according to gender, 9,31,33 or higher levels of knowledge amongst female university students, 8,12 this study found greater levels of knowledge amongst male respondents.The significant differences found in CKI scores in the current study should, however, be interpreted with caution due to the small differences reflected in the mean scores between groups.
Whilst this research aligns with studies which have shown no, or only slight, differences in the overall attitudes towards concussion expressed by males and females, 9,15,33 it also supports previous research identifying gender differences in attitudes to specific issues.For example, Hunt et al. found that females are more likely than males to report concussions, 15 and Sanderson et al. 14 and Jorgensen et al. 12 concluded that females are more likely than males to continue playing sport when concussed.Similarly, Hunt et al. 15 found that boys expressed a greater concern over the seriousness of concussions, and O'Connor and Molloy 33 noted that females reported worrying more about the potential long-term effects.Whilst others have questioned the reliability of the RoCKAS-ST attitudes measure, 6 the complexity of this relationship revealed across studies suggests that it is an oversimplification to consider attitudes as a binary choice between safety and risk. 34Rather, it may be more useful to consider attitudes as multidimensional and shaped by the interplay of demographic differences and environmental experience.
In line with this, a unique finding of this study is that gender-related attitudinal differences are largely related to a contrast between individual behaviour (with females expressing greater personal caution) and perceptions of the broader cultural norms that influence 'most' athletes' behaviour.This broadly aligns with Colon et al. 35 who reported differences in the number of concussions and duration of symptoms between genders and, perhaps most significantly, with Jorgensen et al. 12 who found that females reported safer attitudes towards concussion, but less safe behaviours than their male counterparts.Whilst it must be acknowledged that our current understanding of how gender mediates SRC knowledge, attitudes and behaviours is incomplete, this research provides a useful social and environmental perspective to recent debates about the apparent increased incidence and severity of concussion amongst females, 36 along the lines identified within the developing concussion research agenda. 37

Responses to concussion
This study aligns with research that has revealed counterintuitive responses to the experience of SRC. 13 Finding that for both recent and more historic concussions knowledge increased, attitudes remained the same, and reporting intentions declined, this study further demonstrates that concussion history is not necessarily associated with more precautionary attitudes or reporting behaviours. 4It is notable that whilst the CKI and CAI findings of this survey compare favourably with parallel studies, almost half of the respondents indicated that they might or would not withdraw from sport if experiencing a common symptom of concussion.Moreover, 14.3% did not believe that loss of consciousness should necessarily entail seeking emergency medical treatment, and nearly half did not believe that sustaining one concussion made them more susceptible to future concussions.Compounding the obvious concerns that follow from this, almost half did not have confidence in their ability to assist a teammate with concussion and, as Figure 1 illustrates, students' likely response to concussion is to defer to the expertise of others (medical staff and coaches).This shows the importance of ensuring that student-athletes are not left to manage concussions themselves.
The specific context of student sport may amplify the potential risks associated with student-athlete responses to SRC.Students are subject to distinct peer pressures that invariably involve the consumption of alcohol. 38,39UK student-athletes are often members of sporting teams or societies with strong social cohesion whereby the 'group norms' can be highly influential on individual behaviours.Graupensperger et al.'s 40 study on student-athletes in the USA reported that at the group-level, individuals in student sport teams with greater social cohesion held riskier attitudes towards playing through a concussion.
Here, the team-sport environment was hypothesised as facilitating pressures to put the 'team' before individual health, something that has been reported in other research on student-athletes. 41Any policy response should address these student-specific contexts and responses to SRC.

Implications for UK universities
Of particular concern to UK universities should be: (a) the incidence of concussion reported here; (b) the rate of reported intention to continue playing after experiencing concussion symptoms; (c) student-athletes' self-reported lack of confidence in their ability to respond to their own or a teammate's concussion; and (d) student-athletes' lack of confidence in the existing SRC healthcare provision.By inference, student-athletes are currently not well supported and will need more direct guidance to align with the best practice protocols and precautionary principles for managing SRC.Where the responses to the incidence of SRC are largely student-directed, there is a particularly strong case for the provision of education and awareness programmes.Ideally, these would be designed to bridge the evident gap between knowledge and behaviour and be bespoke to the population.In particular, the design of educational programmes needs to incorporate the likely impact of the participants' concussion history, the influence of gender, and sport-specific norms and understandings of concussion.
Also, given that students' primary response to concussion would be to seek external help, there is also a strong case for more, and differently targeted, SRC healthcare support for this population.In the absence of a legal obligation, healthcare provision in student sport will remain variable across institutions.UK universities also typically provide greater healthcare provision for elite/representative teams than for intra-mural competitions, so standard guidelines for (sport-related) healthcare would ensure that provision in the sector better aligns with the medical ethical principle of justice.
Whilst concussion guidance documents routinely outline more precautionary concussion management for children compared to adults, 1 with gradations towards less caution for older subjects, university students are in a somewhat anomalous position.Indicatively, the preamble to the only UK protocol to directly address students -The Sport and Recreation Alliance's Concussion Guidelines for the Education Sectorexplicitly recognises this lacuna, stating that whilst the guidelines are 'aimed at school-aged children, [they] can also be applied to over 18 s in the absence of other advice' (emphasis added).The first UK national concussion guidelines provide a graduated return to activity outline that refers to accommodations made by employers and schools but not universities. 42Whilst levels of SRC knowledge may be generally increasing over time, 31 and more highly educated cohorts may have been shown to have better CK, 43 it appears the case that UK university student practice in relation to SRC is likely to diverge from public health and sport-specific guidance.The current paucity of concussion awareness and educational materials directly aimed at university students in the UK is a cause for concern and should be addressed.
Finally, whilst North American studies suggest that there may be challenges to implementing RTL protocols in HE, 29 there is at least evidence of their institutional acceptance. 30n light of the ongoing debates about UK universities' duty of care for students, 25 the continued absence of formalised provision appears unsustainable.Because many universities market sporting opportunities as part of the broader 'student experience' offered, demonstrating a duty of care and being candid in communicating the potential risks and the measures they make in mitigation is likely to have both ethical 22 and commercial benefits. 24Given the rates of SRC incidence reported in this study, and evidence of behaviours that potentially exacerbate health risks, there is a moral and commercial case for UK universities to follow the lead of the NCAA and formally introduce concussion safety protocols.

Implications for coaches
The findings have a number of important implications for coaches in the university setting.First, coaches should not assume that well-educated cohorts of athletes, who are (or appear) relatively knowledgeable about the medical proscriptions for SRC, will necessarily conform to precautionary guidelines.Moreover, whilst athletes with a history of SRC may exhibit enhanced knowledge and indeed be increasingly confident of this knowledge, 18 they may also need to be more closely managed as they are less likely report symptoms to those who have the capacity to ensure that they protect themselves and leave the field of play.Third, this study shows that SRC is a concern for coaches across all sports, and not just those high-profile contact sports that dominate media coverage.Whilst sport type influences incidence of concussion, we see similar degrees of knowledge, types of attitude and reporting intentions across all sports.Moreover, the anomaly of a relatively low incidence of concussion in combat sports reported here suggests the potential for different definitions and/or symptom thresholds of concussion to predominate in different sporting subcultures.Finally, coaches within the UK university setting also have a role to play in evaluating the SRC healthcare provision available to student-athletes and lobbying for changes where appropriate.

Limitations and future research
The authors acknowledge the limitations of this study.As with all self-reported data, the results of this study should be interpreted with caution.Furthermore, the study presents a cross-sectional case study of one institution conducted 5 years ago and reservations are made about the generalisability to the wider UK student population.The rationale for using the RoCKAS-ST was outlined in the methods; however, the limitations of this instrument have been noted. 6Additionally, with the instrument now over a decade old, and given the developing field of SRC, the relevance of some items on the instrument should continue to be questioned. 44e also recognise that the additional test items we introduced were exploratory and may lack validity.Whilst we now know that student-athletes do not evaluate the SRC healthcare available to them highly, it remains to be established what they would perceive as more adequate healthcare, or how that would align with clinically defined priorities.On a broader, methodological level, our findings support critiques of such quantitative tools that provide only a 'snapshot' of what is a highly complex, multidimensional and dynamic phenomenon. 34Consequently, this data should be viewed as a point of departure for our understanding of experiences of SRC amongst university students.
Further research is therefore needed.Qualitative methodologies are required to build a deeper, and more nuanced, understanding of student-athlete experiences of SRC at various levels of competitiveness and formal organisation within universities.Qualitative research would also help us to better understand how SRC healthcare in universities could be enhanced to both meet the needs of students and ensure best practice in concussion management.Further research is also needed to explore whether there exists genderrelated attitudinal differences towards SRC, and how these might affect behaviours.Specifically, future research needs to identify which specific attitudes amongst which particular groups of people, constrain health-promoting behaviours, and how coaches might be enabled to respond.
Finally, research is needed to identify what universities are currently doing to support student-athletes in relation to SRC.Areas for focus include official policies, healthcare provision and student accommodations both within the classroom and in university halls of residence.This research would highlight any variance in service, identify elements of best practice, and constitute the basis for guidelines that can be introduced for all levels of sport across the HE sector.

Conclusions
Student-athletes at universities in the UK are a unique population to consider for SRC management.The study reported in this article establishes a baseline for improving understanding of the nexus between student knowledge, attitudes and reporting intentions towards concussion.The findings suggest that whilst enhancing SRC knowledge continues to remain important, greater emphasis should be placed on understanding how to translate knowledge into health-promoting behaviours amongst athletic populations, and student-athlete populations in particular.Coaches should be centrally placed in this process.In light of rising public health concerns about SRC, and the apparent lack of responses in the UK HE setting, it is incumbent on universities to adopt more formal and standardised concussion intervention programmes.

Figure 2 .
Figure 2. The responses to concussion reporting intention item (N = 217) by reported concussion in the last 12 months (a) and reported concussion at any time in the past (b).
a Suspected or formally diagnosed.

Table 2 .
Frequency of responses to concussion knowledge index (N = 217).
2.3Even though Player F is still experiencing the effects of the concussion, his performance will be the same as it would be had he not suffered a concussion 3.7% 96.3%

Table 3 .
Frequency of responses to concussion attitude index sections 3 and 4 (N = 217).Most athletes would feel that Athlete M should have returned to play during the first game of the season Most athletes would feel that Athlete H should tell his coach about the symptoms 4.7 I feel that the physiotherapist/sports scientist, rather than Athlete R, should make the decision about returning Athlete R to play 4.8 Most athletes would feel that the physiotherapist/sports scientist, rather than Athlete R, should make the decision about returning Athlete R to play 0.5% 9.8% 18% 57.1% 14.7% 4.9 I feel that Athlete H should tell his coach about the symptoms