Objectives This study explored factors that play a role in psychological adaptation and recovery of young people with sarcoma.
Design Qualitative study.
Setting National Health Service hospitals in the UK.
Methods Using purposive sampling, participants were recruited for semistructured interviews over the telephone or face to face in order to answer questions about how cancer impacted various domains of their life. Data were analysed using a framework approach.
Results Thirty participants, aged 15–39 years with primary sarcoma diagnosis provided in-depth accounts of their experience. Emerging themes from the interviews were grouped into two overarching themes that relate to one’s adaptation to illness: individual level and environmental level. The qualitative nature of our study sheds light on meaningful connections between various factors and their role in one’s psychological adaptation to sarcoma. We devised a visual matrix to illustrate how risk and protective factors in adaptation vary between and within individuals.
Conclusions This study demonstrates that young people with sarcoma report an array of both positive and negative factors related to their illness experience. The route to recovery is a multifactorial process and a one-size-fits-all approach to psychosocial care proves inadequate. We propose that moving beyond the latent constructs of resilience and psychopathology towards a dynamic model of psychological adaptation and recovery in this population can result in optimisation of care. We offer some recommendations for professionals working with young people with sarcoma in clinic and research.
- qualitative research
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
Strengths and limitations of this study
Framework analysis allows a structured and organised approach for theme-based and case-based analyses, which produces a visual matrix of themes by participants.
The qualitative nature of our study sheds light on meaningful connections between various factors and their role in recovery and adaptation to life with and after sarcoma.
Participants in this study are a heterogeneous group and were self-selected, as such, the factors identified may be limited and not generalisable to other adolescents and young adults (AYA) populations.
Psychopathology and resilience in AYA with sarcoma may result from vastly different pathways, therefore, understanding one’s unique trajectory may help us in offering a more personalised approach to psychosocial care.
Sarcomas are a rare form of cancer but nevertheless represent an important health issue and often afflict young people.1 2 Adolescents and young adults (AYAs) present a particularly vulnerable group of patients due to their developmental period marked by physical, social and emotional maturation,3 all of which may be disrupted by a cancer diagnosis.4 5 Consequently, AYAs often experience different challenges and needs compared with younger or older patients facing cancer.6 7
Over the past two decades the improvements in treatments for sarcomas resulted in higher survival rates, however, these remain low compared with other tumours, hovering around 50%–60%.8 9 Patients with sarcoma are often subjected to an array of treatments; radiation and surgical procedures are the first line of treatment for soft tissue sarcoma,10 while osteosarcomas respond best to a multimodal approach including multiagent chemotherapy.11 Due to these treatments many individuals do not reach full recovery; up to 50% report lifelong disability and significantly compromised health-related quality of life.12 Qualitative work with adults shows that sarcoma diagnosis and treatment can be burdensome and psychologically taxing,13 however, a recent review of the literature revealed that the impact of sarcoma diagnosis and treatment on psychosocial well-being is not yet well understood.14
Psychological adaptation and recovery in cancer are complex phenomena with biological, individual and societal factors contributing varying degrees of risk for poor outcomes. Although distress is a normal response to life-threatening illness, a subset of young patients report psychopathology on reaching remission.15–17 The causes of this may be many and multifaceted; treatment and medications may directly interfere with the endocrinology,18 disrupt education or career plans,19 20 compromise fertility,21 22 or lead to a lifelong disability.6 On the other hand, studies have found that despite the challenges, many AYAs with cancer become resilient or experience post-traumatic growth.23–25
In the recent years, national as well as international initiatives have recognised the importance of psychosocial impact of cancer diagnosis in youth and are calling for research initiatives that investigate psychopathology and promote psychological well-being in young patients.26 27 However, the field lacks a common definition of resilience28 and still relies on latent models of assessing mental health (ie, having depression leads to fatigue and/or loss of interest).29
We propose a different way of understanding psychological adaptation and recovery in cancer and conceptualise the two as interlinked and dynamic processes that encompass individuals’ mental, physical and social function along the illness trajectory and into survivorship. Resilience and mental health are malleable constructs contained within, brought about by complex interactions between different factors. While successful adaptation and recovery (ie, maximised resilience factors and minimised risk factors for psychopathology beyond physical functioning) is a shared goal in increasingly patient-centred healthcare systems, we need to account for the fact that the configuration of these constructs will vary between people and may result from different pathways.
As the majority of patients with sarcoma experience lifelong disability30 and the number of AYA sarcoma survivors keeps increasing, it is important to develop a novel model of understanding psychological adaptation and recovery in sarcoma and describe the factors that are important to individuals. This will allow us to better address the psychological needs of young patients with sarcoma.
Thus, this study had three aims: (1) To describe factors that impact psychological adaptation and recovery in sarcoma in AYA and demonstrate variability between individuals, (2) To contribute evidence towards a dynamic understanding of psychological adaptation and recovery in AYA with sarcoma, and (3) To provide recommendations for clinical care and research.
To best address these aims, we relied on qualitative methods that allow for an in-depth exploration of patient-reported outcomes, their personal circumstances and response styles.31
Materials and methods
Participants and recruitment
Individuals who responded to the study advertisement online (via social media) or in the recruiting UK hospitals were invited to participate. Eligible participants were those with any diagnosis of sarcoma, aged over 13 years (current age) and able to communicate in English. Patients less than 16 years of age gave their assent after consent had been received from their legal guardians; the interviews proceeded following receipt of consent. Purposive sampling32 was guided by a sampling framework (see online supplemental material for details) to recruit participants across the cancer trajectory, with experience of multiple treatment modalities and aimed at approximately representing a distribution of patients across demographic (eg, gender, age) and illness (eg, type of sarcoma) characteristics. The final sample size was determined when data saturation was reached, and no new themes emerged from the individual semistructured interviews.
The analysis presented in this paper represents a subgroup of the total sample (n=121) who were aged 15–39 years, corresponding to the National Institute of Health definition of AYAs, which has been identified as a unique subgroup of cancer patients and survivors.33
Design and procedure
This study is a part of a larger study of health-related quality of life in patients with sarcoma across the UK, aiming to develop a patient-reported outcome measure specific for patients with sarcoma. Data were collected in this phase from February 2017 to August 2017.
Two modes of interviews were offered, in person, or over the phone. In-person interviews were held at home and at the National Health Service (NHS) premises and phone interviews were conducted with individuals far from the study site. People aged under 16 years were interviewed face to face and had their parents present at their request.
The interview guide was based on the existing theoretical models of health-related quality of life34 35 and refined with an advisory group that included patient advocates, sarcoma experts, members of the sarcoma unit, as well as the research team. This guide, previously published elsewhere,36 was not prescriptive and was purposefully flexible to enable the researcher to explore new and emerging experiences.37 The topic guide was shared with participants prior to the interview, so they knew the topics covered; this allowed participants to have more control over the process. The participants were aware of the study aims, and the role of the researchers. No other information about the research team was shared with the participants.
Researchers with graduate-level training in qualitative methods were responsible for conducting and moderating the semistructured individual interviews. Individual interviews lasted on average 45 min, were audio recorded and transcribed verbatim. Names, locations or other identifiable information were omitted from the transcripts for confidentiality. This manuscript followed the COnsolidated criteria for REporting Qualitative research (COREQ) guidelines38 for reporting of qualitative research (online supplemental material).
This study is based on in-depth data analysis of 30 transcripts of the semistructured interviews. The transcripts were analysed using a framework approach.39 Framework analysis allows a structured and organised approach for theme-based and case-based analyses, which produces a visual matrix of themes by participants. The approach is appropriate for semistructured interviews that discuss similar topics, and results in sensible data reduction, which allows for multidisciplinary and collaborative approaches to analysis.40
Researchers (UK and AM) read the transcripts repeatedly to familiarise themselves with the data, and then independently coded them. The first phase of the analysis followed a deductive approach by devising codes and key concepts, which were mapped onto the first conceptual framework of psychological adaptation and recovery developed by UK (PhD candidate). The conceptualisation of some themes (ie, function, activity, participation) was informed by the WHO International Classification of Functioning, Disability and Health framework, which describes recovery as a process of improvement in the psychological, physical and social domain.41 The second phase of the analyses relied on an inductive approach, adding themes from the initial coding in order to complement the framework and consolidate the findings.42 This resulted in a framework (figure 1) on which data were mapped. Themes were then compared within and between individuals with the help of visual matrices (figures 2 and 3). The themes were described systematically, and the findings discussed among the team and embedded in the broader literature.
The first author (UK) kept a reflexive journal to explore the concepts and repetitive themes. In order to account for participants’ and researchers’ cultural context and understanding of the interviews, meanings from data were produced jointly within the team discussions. Due to the nature of this analysis the exact inter-rater reliability was not determined.43 Interpretations on both sides enabled confirmation or rejection of researchers’ perceptions of data.44 The goal of this work was not to address the questions of prevalence and extent, therefore quasi-statistical phrases such as ‘some’, ‘many’ or ‘several’ are used to not make the findings quantifiable and consequently misleading.45 Please see figure 4 for the detailed steps of the analysis.
Patient and public involvement
Patients and advocates for sarcoma were involved in the design of the overarching research study on health-related quality of life in sarcoma and included as co-applicants on the grant. They were invited to help with the recruitment of participants and dissemination strategy, as well as in the analysis, interpretation of the results and, when appropriate, acknowledged in the publication.
A total of 175 patients expressed interest in participating in the study, of whom 121 provided valid consent and were interviewed or joined a focus group. This study includes the 30 young people who were aged 15–39 years who participated in interviews. Descriptive demographic and medical information of our sample can be found in table 1.
Factors that play a role in psychological adaptation and recovery in sarcoma
Our analyses revealed 7 themes and 26 subthemes that can be viewed as factors, which were grouped into two overarching themes: individual-level factors and environmental-level factors. When describing these factors, individuals’ accounts were rated as either describing a given factor in a negative, positive or ambivalent light. These factors form the basis for our conceptualisation of psychological adaptation and recovery (see figure 1).
Illustrative quotes from participants are provided within the text and fully anonymised for confidentiality reasons. The remaining quotes are included in table 2, which is organised to follow the themes’ framework.
Individual-level factors encompass those that were individually bound and sometimes independent from external circumstances. These were not necessarily mutually exclusive with environmental-level factors, and sometimes overlapped on the continuum between individual and societal. Figure 2 demonstrates the discussed themes by all participating individuals.
Function refers to an individual’s physical or bodily functioning including psychological well-being and mental health.41 Treatment side effects were endorsed entirely as negative, impacting participants’ quality of life, and though they varied in magnitude and number, participants commonly reported trouble eating, feeling nauseous, and experiencing fatigue or pain. Side effects limited individuals’ participation in activities, however, they tended to subside after the treatment ended. For some participants symptoms persisted over time. Among the late effects, issues and worries related to fertility were the most commonly endorsed and discussed equally between genders.
Mental health concerns and acute episodes of psychopathology were reported by 17 participants. Two participants reported that their mental health was directly influenced by their physical function and resolved with physical recovery (see ‘Other’ in figure 2). Several participants reported thoughts about their mortality, which provoked anxiety and negative emotions, and the majority emphasised that the mental health problems and psychopathology started after the phase of active treatment.
I think the main thing that I struggled with was the bit afterwards when some people were ‘You’re alright now. It’s finished, you’re okay.’ I found that hard to talk to some people about, because they saw that actually the physical side of it is over, but actually the mental side isn’t.
Twelve participants also endorsed feelings of guilt, which tended to persist beyond active treatment. Young people expressed feeling guilty in relation to failing to fulfil their roles as partners or parents, unable to carry out their daily tasks or having to rely on their parents for financial or logistical support. Some participants experienced guilt in their relationship to other cancer peers who had a poorer prognosis, although others expressed feeling ‘lucky’.
Body image concern was also prevalent; 15 participants reported negative feelings associated with it, and 3 individuals talked about their body image improving with illness due to weight loss or improved self-care and healthier lifestyle brought on by their illness (see table 2).
Activity is the ability to carry out a task or an action.41 The most commonly described factor was loss of independence by 17 participants, which often resulted from physical limitations associated with treatment procedures. Among those who reported feeling dependent, two individuals did not perceive it as problematic because of the temporary nature of dependence and adjustments made to cope. One individual reported remaining relatively independent, which resulted in a sense of accomplishment, and resultantly positive attitude.
Participation involves active engagement in one’s life and daily activities such as education or work, parenting, engagement in one’s care, and romantic relationships and intimacy.41 Disruption in education or work was reported by 29 participants. Sixteen reported a negative impact and felt like they fell behind and had to change their educational and vocational plans, which was ‘hard to accept’. Nine participants felt ambivalent about their vocational disruption, most often explaining that it was a ‘small price to pay’, or the disruption was temporary. Four participants, all of them young adults, endorsed disruption as a positive thing, allowing them to rethink their priorities and pursue more fulfilling endeavours (see table 2).
Financial strain resulting from work disruption was discussed by 24 participants, and mostly felt by those with limited support from their family or partners. Young adults in particular reported hidden costs of sarcoma, which included travelling for treatment, parking or family lodging at a hotel. A few participants reported limited impact because their employers continued to provide them with a salary, or they were financially secure.
Of 30 participants, 7 were parents. Five of them reported that their parenting role was compromised by illness, and they expressed worry about their children. One individual reported that being a parent was his most important role and endorsed it as a positive factor, explaining that it provided him with a sense of purpose during treatment and recovery. Twenty-seven participants also spoke about romantic relationships and intimacy, which were commonly impacted by the diagnosis and treatment. Several talked about romantic relationships in a negative way and suggested that their relationships fell apart, describing illness and associated distress as the main reasons. Eight individuals felt ambivalent about the impact of the sarcoma on their relationship, while six reported that their illness strengthened the appreciation of their romantic partner and made the relationship stronger as they learnt ways to remain intimate despite the limitations imposed by illness.
A widely endorsed positive factor was individuals’ participation in one’s own care. Twenty-four young people reported being proactive about their care and sought ways to improve their health. Shared decision making and patient engagement were seen as empowering.
Lived experience describes other factors that are pertinent in one’s participation and recovery. A common theme discussed by 28 participants was the need for information. Young people were not always satisfied with the amount of information they were provided. In a few cases the information felt overwhelming, and several reported inadequate information at the time of diagnosis (see table 2).
Twelve participants spoke about illness triggers or ‘reminders’, which had a negative impact on all but one person. These triggers, ranging from bodily sensations to visits to the hospital, reminded participants of illness and caused feelings of distress and anxiety. One participant reported the illness reminder—a large scar—as a positive sign, giving her a sense of pride and strength.
Another aspect that proved important in one’s understanding and adaptation was prior experience. Prior experience defines one’s experience with cancer, be it their own or familial, or significant prior experience with the medical care (eg, major surgery, unrelated to cancer). Six participants discussed prior experience, two endorsed it as negative because they have lost a person close to them to cancer. They worried about their own prognosis and felt like they could not openly talk about it with their family because of the negative connotations. On the other hand, three participants reported their prior experience as a positive factor that allowed them to better manage their expectations, navigate the medical system and advocate for themselves.
Finally, nine participants talked about the added burden resulting from external factors not related to their illness, for example, breaking a limb and family bereavement. This burden was an additional stressor and was exclusively negative.
Environmental-level factors refer to aspects of individuals’ social and environmental context within which they found themselves during their illness experience. They range from more intimate circles of friends and family to one’s medical team. Figure 4 demonstrates the discussed themes by all participants.
Family and social circles
Twenty participants endorsed family as a positive factor. Many reported their relationship within the family improving and resulting in a more cohesive environment due to support and greater appreciation of one another. Some individuals could only draw partial support from their family, and for a few, family was not a source of support but rather an impediment (see table 3).
Twelve participants spoke about peers and close friends in a positive light. Close friends provided additional support for the individual, which was different from familial support and provided them with a sense of normalcy. However, several participants reported that they also lost friends and felt abandoned due to illness.
Participants often spoke about their medical teams and healthcare professionals. The initial impact of diagnosis often stemmed from the time they learnt their diagnosis. Learning the diagnosis was a negative experience for 15 individuals and almost entirely resulted in feelings of shock and distress. Seven participants learnt the news over the phone, which did not allow them to ask questions. Some felt that the bad news was attenuated by the fact that they were not alone at the time of diagnosis, and some found great relief in getting an answer for their previously unexplained symptoms.
Nine participants spoke about their positive relationship with the medical team. Those who reported feelings of trust and being able to openly express their concerns reported greater levels of satisfaction with their care and less distress associated with treatment. Additionally, interim feedback on disease and treatment status was desired. Several individuals emphasised that receiving positive news during treatment was encouraging. Two participants reported feeling a negative impact due to the lack of positive news.
In addition to the medical team, the specialist nurses or in some cases AYA coordinators were endorsed as the most helpful resource by 19 participants. Specialist nurses were able to spend more time with participants and answer questions related to their illness and its impact.
My MacMillan nurse, she’s phenomenal. If I’m ever feeling a bit low or anything, when I was on chemo or anything, she was there. She was fantastic. Absolutely fantastic.
Fourteen participants spoke about psychotherapy and counselling. Some received it as part of their oncological care, while others sought it externally. For eight, counselling was a positive experience, teaching them adaptive coping mechanisms. However, four participants found counselling services inadequate in addressing their needs as sarcoma patients, which resulted in frustration. Two individuals explicitly asked for and wanted psychological support but were unable to obtain any (see figure 3, ‘Other’).
Healthcare system at large
Two-thirds of our participants reported a prolonged route to diagnosis. In all but one case this was a negative factor. Participants reported several months, in one case 2 years, before their symptoms were recognised and treated seriously. Several resorted to private care in order to speed up the diagnostic process. Two individuals reported an initial misdiagnosis.
Sixteen individuals also discussed difficulties in navigating the healthcare system. Some individuals had to travel long distances for their treatment, and two reported that their care was not age-appropriate. Managing the logistics of care was burdensome.
Participants were often encouraged to join cancer support groups, face-to-face or online, some of which were specific to sarcoma. Nine participants endorsed them as positive, but young adult men rarely discussed the topic or engaged with the services. Among cancer peers, participants felt supported, understood and less isolated due to their illness. However, half of the participants felt as though the groups were not tailored to their needs, found them too depressing due to generally poor prognosis, or hard to relate to due to different outlooks or age differences.
This qualitative study aimed to describe both positive and negative factors along the illness trajectory in young people with sarcoma. By relying on qualitative methodology, we demonstrated the extent of between-individual variability in mental, physical and social function along the illness trajectory, and contribute evidence towards a dynamic approach to understanding of psychological adaptation and recovery. Practical recommendations for clinical care and future research are discussed below.
Thirty AYA participants from across the UK provided in-depth subjective accounts about their experience with sarcoma. The present study contributes to the existing body of knowledge with a thorough investigation and description of commonly reported factors and demonstrates that when faced with cancer, individuals stem from vastly different backgrounds and life experience. Consequently, different individual-level or environmental-level factors will contribute to one’s psychological adaptation and recovery. As psychosocial well-being remains an unmet need and among the most important topics for AYAs with cancer,26 we have to develop a better understanding of how different subgroups of patients, though afflicted with the same diagnosis, respond differently based on their individual biopsychosocial profile.
We propose that successful psychological adaptation and recovery may not be considered as latent entities, but rather as dynamic systems composed of factors or symptoms. As such, the concept extends on the static models of psychopathology (ie, meeting the symptom threshold for a mental disorder) and resilience (ie, absence of psychopathology given adversity),46 and accounts for the fact that one might at once experience psychosocial growth, as well as symptoms of psychopathology.
Moving beyond the latent constructs of resilience and psychopathology in youth with sarcoma may be particularly important because unlike in the general population where mental health disorders are thought to occur via the activation of symptoms through adverse life events,47 48 in patients with chronic conditions (ie, cancer) the pathway to psychopathology may be different.49 50 The illness may be perceived as both an adverse life event, as well as a chronic health condition; we suggest that psychological adaptation and recovery in AYA with sarcoma will follow unique pathways and factors or symptom presentation. For example, our results imply that education and work are salient aspects of AYAs’ identities and often disrupted by illness. So, we might have two individuals experiencing anxiety (ie, not achieving full recovery) who are of the same age, gender, same type of sarcoma and with a comparable treatment plan. While one’s feelings of anxiety will revolve around the thoughts of mortality and follow-up consultations (follow-up consultation → thoughts of mortality → anxiety), another might experience anxiety due to occupational disruption and financial strain (work disruption → financial strain → anxiety). Similarly, we show that certain concepts such as body image, which most commonly carry negative connotations,51 52 can in fact act as resilience factors if appraised as positive (healthier lifestyle → improved body image after cancer → improved self-image). This further strengthens our claim that individual circumstances vary in important ways and should be accounted for in clinical practice. The high level of individual variability also suggests that the use of person-centred statistical approaches (eg, growth mixture models)53 would be informative in further study of this heterogeneous population.
It would be erroneous to expect young people exposed to cancer to show uniform outcomes in facing their illness. Most of the existing literature in AYA cancer survivorship addressing psychological well-being relies on the conceptualisations of psychopathology and resilience from healthy populations, which have not been adapted to young people with cancer. Therefore, an understanding of psychological adaptation and recovery as dynamic systems is needed to take into account the interplay between many resilience and risk factors on an intraindividual, as well as interindividual level. Our study offers a first step in this conceptualisation and provides invaluable insights for future work in this field.
Limitations and future directions
Certain limitations are important to note in the present study. Our sample was heterogeneous, including individuals with all types of sarcoma, from a broad age range, and is limited to the UK, which might limit generalisability of the findings, particularly in the domain of the healthcare system. Additionally, participants were a self-selected group and mostly provided retrospective accounts, which may have been impacted by recall bias. Though asked about their illness and sarcoma history, this information was self-reported. Lastly, this study relied on an existing data set, which might explain why some concepts were not explored in-depth by all the participants, or in greater detail. The themes mapped onto the framework were not mutually exclusive and often spanned across domains. Despite its limitations, our study is the largest in-depth exploration of the impact of sarcoma on AYA, which adds important insights for future research and targeted therapeutic approaches for this vulnerable population of patients.
Conclusion and clinical implication
Sarcoma impacts young people on psychological, physical and social levels. Understanding the subjective experience and individual circumstances is at the centre of personalised approach to healthcare. Our findings suggest that young people with sarcoma require holistic guidance and care beyond active treatment. A one-size-fits-all approach is inadequate and healthcare professionals working with this population should actively engage with their patients from the time of diagnosis. Considering individual context will aid in identifying those in need of help in a specific domain (eg, return to work). While certain factors cannot be prevented (ie, infertility), promoting a cohesive atmosphere between the patients and medical professionals can result in better management of an individual and optimisation of service and resources utilisation. Furthermore, across the UK, nationally based services for AYAs mostly end at the age of 24 years.54 In our study we included individuals up to 39 years of age and demonstrated that young individuals who are ineligible for the NHS AYA services also present with needs that may not be adequately addressed in adult cancer units. We recommend that special attention is paid to young people outside the dedicated AYA services. A summary of recommendations for improvements of care and future research directions can be found in table 4.
The authors thank all the participants who took part in this study and all the Principal Investigators at the Trusts opened for recruitment: Fiona Cowie; Helen Hatcher; Amit Kumar; Dan Stark; Craig Gerrand; Robert Ashford; Claire McKenzie; Sarah Massey; Julie Woodford; Jonathan Gregory; Robin Young; Mariam Jafri; Paula Wilson; Robert Ashford; Nicola Keay and Paul Cool. The authors also thank all the professionals involved in patient recruitment.
Contributors RMT, MO, RW and CG were involved in developing the protocol; AM coordinated the running of the study and was responsible for data acquisition; UK, AM, LB, CG, RW and RMT contributed to the analysis; UK and AM drafted the manuscript. All authors critically revised and approved the final manuscript.
Funding This work was supported by Sarcoma UK (grant number SUK102.2016LG) and Bone Cancer Research Trust (grant number BCRT/49/16). RMT is a National Institute for Health Research (NIHR) Senior Nurse Research Leader; RMT acknowledges support from University College London Hospitals BRC. UK would also like to acknowledge the financial support from Economic and Social Research Council (ESRC) as well as Ad Futura (Slovenia).
Disclaimer The views are of the authors and do not necessarily reflect those of Sarcoma UK or Bone Cancer Research Trust, or the NIHR/NHS. None of the funders were involved in conducting this review or drafting this manuscript.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval The Study was approved by the London Stanmore National Health Service (NHS) Research Ethics Committee (reference: 16/LO/2152; IRAS reference: 217605).
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement No data are available.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
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.