Objectives Modern sport safeguarding strategies include published global rights declarations that enshrine athletes’ entitlements at the policy level. It is unclear how these documents translate to athletes’ lived experiences. The study aimed to determine athletes’ knowledge, attitudes and beliefs about their human rights in sports settings.
Setting Web-based survey.
Participants 1159 athletes from 70 countries completed a validated web-based survey. Over half of participants (60.1%) were between 18 and 29 years, currently competing (67.1%), not members of players’ unions (54.6%), elite (60.0%) and participating in individual (55.8%) non-contact (75.6%) Olympic (77.9%) sports. Gender distribution was equal.
Primary and secondary outcome measures Participant demographics (eg, gender, age) and athletes’ knowledge, attitudes and beliefs about their human rights in sports settings.
Results Most (78.5%) were unaware of any athletes’ rights declarations. Gender influenced participants’ confidence in acting on their rights in sport significantly. Males were more likely to accept pressure from coaches and teammates than females, but age affected how likely males were to accept this pressure. Paralympic athletes were less likely to agree that violence is acceptable in sports, compared with Olympic. Player union membership increased confidence in freely expressing one’s opinion in sports settings. Athletes’ rights-related awareness, knowledge and beliefs were disconnected.
Conclusions Awareness raising is not enough to prevent human rights violations in sports. The cultural climate of the entire ecosystem must be targeted, using systems-level strategies to shift stakeholders’ biases, beliefs and behaviours. This approach takes the onus of addressing abuse off athletes’ shoulders and places accountability on sports organisations.
- Sexual harassment
- Human rights
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
All forms of harassment and abuse in sports are a breach of human rights. Integrating human rights with sport is one important step in building sports environments that enable athletes to flourish.
The International Olympic Committee and World Players Association have developed and published athletes’ rights declarations.
Despite these documents, egregious cases of abuse still occur in sports, and athletes’ knowledge, attitudes and beliefs about their human rights in sports remain unclear. Athletes’ inability to access and confidently act on their human rights may perpetuate existing power imbalances between athletes and sports institutions, which sets the stage for unchecked unethical practices.
WHAT THIS STUDY ADDS
Athletes’ knowledge about their rights in sport is disconnected from their beliefs about the same.
Most athletes are unaware of the human rights declarations global sports authorities have published and promoted.
Gender is the strongest individual characteristic influencing athletes’ confidence in applying their rights in sports contexts. Having a disability, being subjectively judged, participating in a contact sport and being a member of a player’s union also significantly influenced athletes’ agency in accessing their rights while training and competing.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Safe and responsible sports environments must include tailored strategies for girls, women and similarly vulnerable groups who participate in sports.
Global sports organisations’ current communication strategies may be ineffective in reaching athletes.
Since beliefs are a stronger driver of behaviour than knowledge, safeguarding programs that emphasize athlete education, knowledge and awareness are unlikely to be effective without true culture change and organisational accountability.
The connection between sport and human rights
In its ideal form, sport is one of the most powerful global human rights promoters.1 Centred on tenets of mutual understanding and fair play, sport ‘done well’ can facilitate democracy building and humanism on small and large scales. As the Olympic Charter aspirationally asserts, ‘the practice of sport is a human right. Every individual must be able to practise sport, without discrimination of any kind … with respect for universal fundamental ethical principles … and the preservation of human dignity’.1 Ideally, all individuals have the right to feel safe and be treated with respect as they participate in sport. As summarised by the World Players Association (WPA), athletes stand at the intersection of sport and human rights.2–4
Prince Zeid Ra’ad Al Hussein’s and Rachel Davis’ 2020 Recommendations for an International Olympic Committee (IOC) Human Rights Strategy, commissioned by the IOC to ensure better alignment with United Nations’ (UN) human rights standards, has further concretised the formal link between sport and universal human freedoms.5 6 Antecedent to this work, in 1998 and again in 2004, former Secretary of the Committee on the Rights of the Child in the UN Office of the High Commissioner for Human Rights Paolo David articulated how protecting the human rights of athletes is critical for the elimination of discrimination and violence in sports and for the development of the sport itself.7 8 David was the first to suggest that the a priori integration of human rights into sports systems could improve their quality and their athletes’ well-being.9 This repositioned an ethic of care and human development as the true cornerstones and goals of sport10 11 and identified the ethical risks of disconnecting athletes’ fitness and performance goals from their fundamental human entitlements.7 9
Athlete exploitation is a human rights matter
Recently, against the backdrop of a groundswell of social justice activism and civil unrest outside sport,12 13 athletes’ individual and collective agency has been amplified in the public domain and, in some cases, formalised. As individual high-profile athletes from diverse backgrounds are publicly expressing their beliefs about myriad flaws in society and sport,14–16 an increasing number of athlete-centred organisations are coalescing to bolster the impact. For example, using a long-established international human rights framework, the WPA and IOC have enshrined athletes’ rights in sport at the policy level through their respective 2017 and 2018 athletes’ rights declarations.2 5 17 Despite this momentum, disturbing cases of athlete silencing and abuse continue to emerge, bringing the grim underbelly of sport—and in some cases, the disempowerment of athletes amidst overpowering social turmoil—to light.18–20 Thus, there is still a disconnect between athletes’ human rights, as expressed by the UN, WPA and IOC and endorsed by athletes, and athletes’ lived experience of psychological, physical, sexual, financial and political harms19 21–23—which are in some cases fatal.22 24 25
All forms of interpersonal violence, including harassment, bullying, hazing, disability stigma, neglect, gendered and racialised discrimination and physical, psychological and sexual abuse, constitute human rights violations.26 Long-standing systemic athlete abuse exists across the continuum of sports, from hyperfeminine (eg, synchronised swimming, gymnastics) to hypermasculine (eg, ice hockey, American football) sports.27 There is a heightened risk of abuse for elite athletes and children who are athletes,21 26 28 29 with further compounding where considerations of gender, ethnicity, disability and sexuality are included.30–34 It is clear that abuse in sports is prevalent, generally tolerated and underexamined relative to its scope and impact.18 It is also clear that though athletes are central to the sport, they may also be the most disempowered members of sports environments.
Prioritising human rights is a sport safeguarding strategy
Human rights are universal, inalienable, indivisible and inter-related.5 All rights have equal status and are necessary to protect human dignity. If athletes’ day-to-day experience of their rights is disconnected from the official policies that allege to protect athletes’ rights, the vulnerability of athletes in the sport is exacerbated. Our research aimed to determine athletes’ knowledge, attitudes and beliefs about their human rights in sport and to understand the correlation between their knowledge on the one hand and their attitudes and beliefs on the other.
A two-part web-based survey (table 1) was developed and validated.35 The knowledge construct consisted of five yes/no statements drawn directly from the WPA and IOC athletes’ rights declarations to determine athletes’ concrete knowledge of five rights. The attitudes/beliefs construct assessed the degree to which athletes agreed/disagreed with eight associated plain language statements via a seven-point Likert scale to test how athletes express and/or experience each right while they are training and competing in their sport. The full survey was distributed via Qualtrics software (Qualtrics, Seattle, Washington, USA). Data collection began in February 2020 and concluded in September 2020. All responses were anonymous.
A non-probability technique, snowball sampling, was used.36 A link to the survey was emailed to representatives from the international sports federation’s medical committees, athlete commissions and administrative staff within the study authors’ professional networks. Each representative was asked to forward the survey to competing or retired athletes within their institution. Additionally, the survey was posted on various sports medicine and sports science social media forums, where viewers were asked to complete the survey if they were athletes or forward the survey to athletes.
Correlation analysis tested the association between each of the knowledge statements and their associated attitudes/beliefs statements. Only associated statements were tested. Correlation coefficients of 0.4 and below represent negligible associations. The significance level of the correlation was tested through Spearman’s rho: as the data are non-paramedic, Spearman correlations were conducted, testing the dependence of one variable on the other, measured by Rho (r). Guidelines for strength of the association suggest strong correlation when r>0.6 and very strong correlations when r>0.8.37
Ordinal logistic regression models tested whether agreeing to each of the attitude/beliefs statements is a function of knowledge of associated rights and other athlete-related and sport-related characteristics. A total of eight ordinal logistic regressions were run, with each of the eight attitudes/beliefs statements as dependent variables. Independent variables for each regression were the associated knowledge statements (eg, K1 for regression with AB3 as the dependent variable, K5 for regression with AB7 as the dependent variable, as per table 2) and selected athlete-related and sport-related demographic variables deemed relevant based on the prior literature26 27 30 31 38 39: gender, age group, the highest level of competition, Olympic/Paralympic sport classification, competition status, union membership status, participation in a secondary sport, knowledge of the WPA/IOC athletes’ rights declarations, subjective judging, contact sport and sport structure (team vs individual). As there was a high correlation between age and gender, an interaction variable was computed between the two variables, with the reference group being women aged>30. A total of 15 independent variables in each regression were used.
Patient and public involvement
No patients were involved.
Athlete and sports characteristics
A total of 1159 responses were recorded. Cases were excluded if no questions were answered in any section (n=12), no knowledge or attitudes/beliefs questions were answered (n=410) or the participant indicated that they were not an athlete (eg, ‘rugby league referee’, ‘anything I enjoy and that keeps me fit’, ‘physiotherapist’ or ‘movement is medicine’) (n=4). Responses that recorded ‘I do not understand this statement’ for all statements in the knowledge construct were also excluded (n=3). As such, 730 athletes submitted sufficiently complete surveys for analysis (63%).
Over half of the participants (60.1%) were between 18 and 29 years, currently competing (67.1%) and were not members of a player’s union (54.6%). There was a roughly equal distribution among genders (49.3% men, 50.7% women). Overall, 23% competed in Paralympic sport, 77.9% in Olympic sport. Most (78.5%) were unaware of the WPA or IOC athletes’ rights declarations. Over half of respondents competed at international levels (60.0%), more than half played individual sports (55.8%) and three-quarters (75.6%) competed in non-contact sports. A third (34.3%) engaged in sports with subjective judging(As proposed by Balmer et al, 2003, we have been classified sports into those that involve objective judging/scoring for example, athletics, swimming and weightlifting and sports that include judging which is predominantly subjective such as boxing, gymnastic, synchronised swimming and judo and sports that involve subjective decisions including rugby and hockey).40 Almost 15% of respondents also played a secondary sport (competitive level not elicited). Most athletes (84.5%) reported participating in only one primary sport (table 3).
Athletes from 70 countries participated, with the largest percentage of citizenship in Europe and North America (44% and 24.8%, respectively). Continental distribution was broad but uneven. Where athletes selected more than one country of citizenship, only the first three countries were included. Where athletes indicated both a primary and secondary sport (14.6%), only the primary sport was included.
Overall, 49 sports were represented through the 730 athletes who submitted complete surveys, the top 3 of which were swimming (22.8% of respondents), football/soccer (9.7% of respondents) and basketball (non-disabled) (7.3% of respondents) (figure 1). Overall, 32 sports were not included in the Olympic or Paralympic Games’ programmes, and 19 sports were represented by fewer than 5 survey participants (Biathlon, Equestrian, Ice Hockey, Ski Jumping, Wheelchair Fencing, Shooting Sports, Triathlon, Modern Pentathlon, Rugby, American Football, Baseball, Beach Volleyball, Boccia, Boxing, Freestyle Mogul Skiing, Goalball, Luge, Skeleton, Wheelchair Tennis).
Athletes’ knowledge of their rights
Most respondents recognised each of the five rights statements as athlete rights (86.8%–95.9%) (figure 2). Athletes’ rights to the protection of their name, image and performance, and their right to freedom of expression were endorsed least frequently, with 12.5% and 7.8% of participants responding ‘no, this is not a right of athletes’, respectively. For most of the knowledge statements, very few athletes indicated they did not understand the statement, except for the right to access an effective remedy when rights are not respected and upheld, where 7.3% of participants responded, ‘I do not understand this statement’.
Athletes’ attitudes and beliefs about their rights
The majority of athletes (81.1%–96.7%) strongly agreed, agreed or somewhat agreed with attitudes/beliefs statements pertaining to freedom of expression, personal identity protection, access to the appropriate remedy and privacy (statements 1, 2, 7, 8) (figure 3). Overall, 18.8% recorded neutral to strongly disagree responses to ‘(When I am training and competing in my sport) if I experience behaviour that I deem inappropriate, I can seek assistance without fear of consequences or retaliation’, 14.1% recorded neutral to strongly disagree responses to the statement ‘…it is always ok for me to freely express my opinion’ and 3.3% recorded neutral to strongly disagree responses to the statement ‘…if someone has access to my personal information, I should know how it is being used’. Nearly one in ten (9.6%) recorded ‘neutral’ to ‘strongly agree’ responses to the statement ‘…it is sometimes ok for coaches to use any kind of violence toward me (eg, intimidation, assault or coercion)’, while conversely 7.5% recorded neutral to strongly disagree responses to the same statement made about teammates. Overall, 52.7% of athletes agreed to varying degrees or were neutral about the statement, ‘…it is sometimes ok for coaches to pressure me in any way’, while 48.5% agreed to varying degrees or were neutral about the same statement made about teammates and others.
Correlation between athletes’ knowledge of and attitudes/beliefs about their rights
The correlation between athletes’ knowledge of their rights and associated attitudes/beliefs was negligible across all paired statements. The highest correlation coefficient among all pairs of interest was 0.21. Five of the nine pairs reached statistical significance (Spearman’s correlation, p<0.05, table 4).
Individual and sports characteristics associated with attitudes/beliefs about athletes’ rights
Table 2 summarises the results of the eight ordered logistic regressions, presenting only statistically significant results (p<0.05). The standard interpretation of the ordered logit coefficient is that for a one-unit increase in the predictor, the response variable level is expected to change by its respective regression coefficient in the ordered log-odds scale. In contrast, the other variables in the model are held constant (Stata, UCLA nd.). Two of the eight regression analyses did not return overall statistical significance (prob>χ2>0.05). Only the predictor variables with statistically significant effects are presented and discussed below.
Knowledge of rights
Only three of the eight theorised relationships showed knowledge of the right had a significant impact on attitudes/beliefs about the right:
Athletes who knew that every athlete is entitled to equality of opportunity in the pursuit of sport (K1) were less likely to agree that it is sometimes ok for teammates and others to use any kind of violence towards them in sport (AB4) (−0.73).
Athletes who knew that every athlete has the right to freedom of opinion and expression (K2) were more likely to agree that it is ok for them to freely express their opinion in sport (AB1) (1.26).
Athletes who knew that every athlete must be able to access an effective remedy when their rights are not respected and upheld (K5) were more likely to agree that they could seek assistance without fear of consequences if they experience behaviour they deem inappropriate in sport (ab7) (0.55).
Gender was the leading determinant of attitudes/beliefs related to rights. Male athletes were more likely than female athletes to accept pressure from coaches (0.99) and teammates (0.83). Compared with younger male athletes, however, male athletes older than 30 years old were less likely (−0.70) to accept pressure from their coach.
Male athletes were also more likely than female athletes to accept violence from coaches (0.66) and teammates (0.76). On the other hand, female athletes were less likely to agree that if they experience behaviour they deem inappropriate, they can freely seek assistance without fear of consequences or retaliation in their sports setting (0.43 log odds, male).
Paralympic athletes were less likely to agree that it is ok for coaches (−1.17) and teammates (−0.69) to use any kind of violence in sport. There were no differences, however, between Paralympic and Olympic athletes in their likelihood to accept pressure from coaches or teammates.
Subjective judging and contact status
Athletes in sports with subjective judging components were less likely than athletes in sports without subjective judging to accept pressure from teammates (−0.45). Athletes in contact sports were likelier to agree that the use of violence by coaches and teammates is ok (0.48 and 0.55, respectively).
Athletes are more likely to agree that it is ok to freely express their opinion in sports if they are a member of a player’s union (0.45).
A disconnect among knowledge, beliefs and awareness
Most athletes had some knowledge of their rights but did not consistently believe they could act on them while training and competing. Additionally, nearly all respondents were unaware of the rights declarations global sports authorities publish and promote. These data suggest though many safeguarding programmes emphasise education, knowledge and awareness raising—this approach may lack efficacy in the absence of culture change41 42 and organisational accountability.43 Beliefs are a strong driver of behaviour44 but in sports, athletes’ beliefs appear disconnected from their knowledge. This may be for a good reason. We must recognise that changing beliefs is inextricable from organisational accountability: athletes who do not feel safe and supported by their organisations will not report safeguarding issues no matter how high their knowledge is. Changing beliefs alongside greater organisational accountability, while potentially more challenging than transmitting knowledge alone, has greater potential to influence safeguarding. The influence of player’s unions on athletes’ perception of personal agency in accessing rights and/or organisational accountability also needs to be explored.
Gender, violence and power
Gender was the strongest individual characteristic influencing beliefs about violence and confidence in applying rights in sport. It is well-documented that girls and women are at greater risk of interpersonal violence in sports. It is no surprise that women in this study have a stronger belief than men that violence is inappropriate in sports.26 42 Though men are more likely than women to accept violence and pressure from their coaches and teammates, women in this study are less likely to believe that they can freely seek assistance without fear of consequences or retaliation when inappropriate or harmful behaviour is experienced. This resonates with what we know about the consequences of reporting harassment and abuse45; these data are indicative of societal norms and men’s and women’s experiences of violence more broadly.
For boys and men, violence and pressure are tied up in hegemonic masculinity—social norms around what it means to ‘be a man’. Sport remains a culture where, by and large, idealised ‘masculine’ norms are continually reified and reinforced. This includes the many covert and overt ways in which boys and men are taught to accept pressure and violence as ways of ‘being a man’ often through acts of consensus and coercion that enforce compliance46: participating in roughhousing a new team member, ignoring pain after an injury, or experiencing the deep stigma of sexual violence against boys/men. In this way, while boys and men might internally reject pressure and violence and understand rights violations in this regard, societal norms demand they accept and even play a role in upholding the dominant form of masculinity. Our research reflects this conformity. This has implications for safeguarding men and boys in any sport that has yet to be widely acknowledged and considered. There is particular relevance in male team sports and environments where bullying and hazing are normalised parts of belonging rituals, such as men’s rugby, Australian rules football, baseball, American football, ice hockey, lacrosse and others.
Gender-based violence is pervasive for girls and women and is a ‘serious violation of human rights and a life-threatening health and protection issue’ rooted in gender inequality.47 Existing research has shown that in Europe, prevalence rates of gender-based violence in sports are between 1% and 64% (due to different methodologies and definitions).48 While knowledge of sexual harassment and abuse of girls in sports has increased, and to some extent, social stigma has decreased, our research shows that it is important to understand broader sociocultural reasons why girls and women feel less confident to freely report violations of their personal rights, including their reasons for and experiences of consequences and retaliation for doing so.45 It is also unclear which athletes feel more or less comfortable speaking up on behalf of others when they observe—rather than directly experience—rights violations in sport.
Gender-related study results may reflect the ways society places blame on female assault victims/survivors and socialises girls and women into fearing the consequences of holding power to account.49–52 It must be recognised that men are over-represented in dominant positions of power in elite sports contexts.53 Time and again, there are stories of girls and women not being believed and perpetrators being protected by their communities, law enforcement and society. The fear of societal rejection, including from sport itself, is a narrative all girls and women are socialised into. Understanding this context brings new implications for safeguarding, including emphasising the protection of more vulnerable athletes and leveraging unique elements of sport (ie, emphasis on contextual and systemic approaches to injury prevention) to mitigate harm against women and girls.42 54
One important finding was that ‘violence’ emerged as a distinct and complicated concept among respondents. On the heels of the UN World Report on Violence Against Children, UNICEF supported research to protect children from violence in global sports,28 55 culminating in an invited 2010 report that put the issue of violence against children in sports firmly on the world map.55 In the 2017 and 2018 global athletes’ rights policies, ‘violence’ is conflated with other, less severe concepts such as ‘harassment’, ‘intimidation’ and ‘pressure’. Study results point to the need to separate ‘violence’ from related but conceptually distinct nomenclature as the UN and others do. For example, an equal proportion of athletes reported that ‘pressure’ was ok from both coaches and teammates, but that ‘violence’ was ok from coaches only but not from teammates. These findings, as well as gender differences in acceptance of violence versus pressure, underline violence as a separate and unequally weighted idea at the athlete level.
Beyond gendered understandings of violence, our study may also reveal a ‘taken for grantedness’ around violence that might have to do with the athletes’ lack of power in sport and the normalisation of athlete abuse generally. For example, findings show that violence is often perceived as part of the fabric of contact sports. Athletes may not understand that violence normally refers to non-consensual injurious acts.56 Given this context, we must explore how and why athletes of different genders and sports settings understand and experience violence.
Disability and violence
The UN Convention on the Rights of Persons with Disabilities (2007) guarantees that each child is born free and equal in dignity and due merits.57 For sport, this convention was part of a shift to seeing sportspersons with disabilities as athletes with equal claim to full participation in the world of sport. In the current study, disability status significantly influenced athletes’ rights experiences: Paralympians were more likely than Olympians to agree to pressure from teammates but less likely to agree to violence from coaches or teammates.
It is well-known that athletes with disabilities are at greater risk of interpersonal violence in sports. In the study, Para athletes had a stronger belief than non-disabled athletes that violence is inappropriate in sport.26 30 31 58 This finding may reflect athletes with disabilities coming to sport with increased a priori exposure to subtle and overt forms of violence and harm through a range of life experiences outside sport. This relative sensitisation may influence tolerance of violence. This finding may alternatively reflect Para athletes’ struggle to garner the full respect of society, coaches and non-disabled peers.59 This patronising form of disability stigma may change the dynamics of Para training environments, potentially removing or minimising extreme, harsh or demanding behaviours of any description, including violent behaviours. Differences between Para and non-disabled sports settings need to be examined, as do the cultural features that influence the perception of acceptable behaviours.
Study strengths and limitations
Study strengths include using a large data set, a valid survey codeveloped with athletes, sports experts and academics, and two coding systems for correlation analyses as well as ordered logistic regression to explain how individual and sports characteristics, as well as knowledge of rights, influence athletes’ attitudes and beliefs.35 Limitations of the study include respondent clustering on two continents. Limitations around demographic data are also acknowledged, including a lack of analysis around ethnicity, sexuality and gender as a spectrum. These considerations are important for future research.
Once one is labelled an athlete, ‘it is frequently the case that their identity as (human beings) first is lost and their rights as (human beings) are eroded’,60 paving the way for a range of rights violations, including: excessive intensive physical training; psychological, physical, sexual abuse and neglect; violence on and off the field of play; doping; economic exploitation; displacement; trafficking and sale (eg, athlete contracts); transfers and reduction of freedom of association; limits to the right to education; and limits to civil rights and freedoms of athletes.7 60 This is the fundamental disconnect: sport can become more dangerous when human rights are not integrated with sport. This study reveals an additional disconnect between athletes’ rights-related knowledge, beliefs, awareness and experiences. As a result, any genuine effort to prevent human rights violations in sports must target the cultural climate of the entire sports ecosystem, not just ‘knowledge’. A shift in all stakeholders’ belief-driven behaviours is required. This approach takes the onus of identifying, processing and addressing rights violations off athletes’ shoulders and places accountability on sports organisations.
Ultimately, athletes should not have to choose between sport and human rights, and those delivering sport should not insist that this choice be made. Athlete-centred safeguarding systems should have the capacity to go beyond policies and uphold the moral and legal obligations sports entities have towards their membership.
Data availability statement
Data are available upon reasonable request.
Patient consent for publication
This study involves human participants. This study was deemed exempt by the Yale Institutional Review Board on 26 June 2019 (Protocol ID 2000025862). Participants gave informed consent to participate in the study before taking part.
The authors thank Catherine Stratton MPH for helping to format the paper and responding to editorial feedback.
Twitter @YetsaTuakli, @margo.mountjoy, @shereebekker
Correction notice The article has been corrected since it was published online. The co-author Natalie's name has been corrected to Natalie R Galea.
Contributors NRG and YAT-W conceived the work, which was further developed with SB. All authors contributed to the acquisition, analysis and interpretation of data, manuscript drafting, manuscript revising and final approval of the version to be published. All authors agree to be accountable for all aspects of the work. YAT-W is acting as guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.