The internet is widely used for health information and support, often by vulnerable people. Internet-based research raises both familiar and new ethical problems for researchers and ethics committees. While guidelines for internet-based research are available, it is unclear to what extent ethics committees use these. Experience of gaining research ethics approval for a UK study (SharpTalk), involving internet-based discussion groups with young people who self-harm and health professionals is described. During ethical review, unsurprisingly, concerns were raised about the vulnerability of potential participants. These were dominated by the issue of anonymity, which also affected participant safety and consent. These ethical problems are discussed, and our solutions, which included: participant usernames specific to the study, a closed website, private messaging facilities, a direct contact email to researchers, information about forum rules displayed on the website, a ‘report’ button for participants, links to online support, and a discussion room for forum moderators. This experience with SharpTalk suggests that an approach to ethics, which recognises the relational aspects of research with vulnerable people, is particularly useful for internet-based health research. The solutions presented here can act as guidance for researchers developing proposals and for ethics committees reviewing them.
- discussion groups
- ethics committees/consultation
- information technology
- research ethics
- suicide/assisted suicide
- technology/risk assessment
Statistics from Altmetric.com
- discussion groups
- ethics committees/consultation
- information technology
- research ethics
- suicide/assisted suicide
- technology/risk assessment
Ethics for health research
Many countries have their own strict systems of regulation and ethical governance for health research. These are too numerous to list here, are country specific, and are widely and easily available to researchers. Ethical codes for health and social care research have much in common, and relate to informed consent, anonymity, privacy, confidentiality, harm minimisation and risks to research participants.1 Research with vulnerable people requires particular consideration, and relates to emotional issues and risks.2 Lakeman and FitzGerald3 report findings from a survey of ethics committees whose members were asked about the risks, benefits and ethical problems associated with suicide research. Concerns included access to vulnerable people, potential harm to participants or researchers, researcher competency, maintaining confidentiality, providing support to participants and responding sensitively to family needs. While issues such as access and confidentiality, for example, may be relevant to all participants in health research, risks may be compounded if an individual is emotionally vulnerable or immature or is at risk of coercion. Similarly, the subject, context or medium of the research itself may increase vulnerability. For example, when the focus is on suicidal thoughts, or self-harm or participants are anonymous throughout a study.
Internet research ethics
In 2009 an estimated 25% of the world's population had access to the internet.4 The Office for National Statistics in the UK reports that in 2009, 96% of people aged 16–24 in the UK had used the internet, with 86% using it every day or almost every day.5 Increasingly, online discussion groups function as ‘communities’ where people come together for support, using one of a number of forms of ‘social software’. Internet-based research has also increased in mental health—for example, the use of the internet for treatment6 and comparing online with face-to-face interventions in mental health and well-being.7 Flicker et al,8 discuss ‘ethical predicaments’ in research with internet communities: enrolling participants, protection from risks and links between public and private data. However, it remains unclear how guidance is used by researchers or ethics committees.
Concerns about internet-based research can relate to different aspects of research: the ‘virtual’ position of participants in ‘cyberspace’; the ethical implications of undertaking different types of research online (such as covert research (lurking) when researchers observe interactions in internet-based public forums); the relationship of the researcher to participants, and the public/private status of internet data.9–12
Professional and other organisations variously provide guidance for health researchers undertaking internet-based health research. Those provided by the British Psychological Society and the Association of Internet Researchers (AoIR) (see Box 1 below) are often duplicated or referenced within professional and academic guidance and there is a growing awareness of the specifics of internet research in health contexts. Tensions exist, however, between what Simon C and Mosavel M13 distinguish as procedural ethics (formal regulations and review processes) and ethics in practice (nuanced, everyday ethical issues, which may not be addressed in procedural ethics). While most researchers would agree that participants should be well informed about and give their consent to taking part in research, different settings and participant groups require tailored approaches and raise unique issues. Kidd and Finlayson discuss ethics as the ‘uncharted water’ of research, which affects researchers—for example, emotional intensity in the early stages of research or being a practitioner researcher.14 Researchers need to be aware of the uniqueness of ethical issues in different research contexts as well as any impacts on themselves. A sample of available ethics guidance relevant to internet-based health research is provided in Box 1, including some provided by university departments.
Guidance on ethics for internet-based research
AoIR (Association of Internet Researchers) ethics working committee provides a range of guidance for research via the internet.
Ethical decision-making and internet research: recommendations from the AoIR ethics working committee.
Authors: Charles Ess and the AoIR ethics working committee
Approved by AoIR, 27 November 2002 http://aoir.org/reports/ethics.pdf
This guidance provides useful information on internet context and culture, internet users, consent online, legal aspects and risks and benefits of online research. Case studies are provided, as are useful US/European comparisons on law and expectations.
The British Psychological Society. (2007) Report of the working party on conducting research on the internet. Guidelines for ethical practice in psychological research online. http://www.admin.ox.ac.uk/media/global/wwwadminoxacuk/localsites/curec/documents/internetresearch.pdf
This guidance suggests that two dimensions guide ethical considerations: identifiability (identifiable/anonymous) and level of observation (explicit consent/invisibly observed). This document provides guidance on nine key topics including: verifying identity; public/private space; informed consent; levels of control; withdrawal; debriefing; deception; monitoring; protection of participants and researchers; data protection.
University of Leicester. Exploring online research methods. http://www.restore.ac.uk/orm/site/home.htm
Lancaster University. Social science research ethics. http://www.lancs.ac.uk/researchethics Contains a section on web-based research.
University of Bedfordshire. Ethical guidelines for the online researcher. http://www.beds.ac.uk/research/iasr/ethics/onlineresearch
Ethical guidance for health research also concerns researchers working with vulnerable people. For example, in the UK, the National Institute of Health Research provides guidance as part of its research governance provision and other countries provide their own. Professional and charitable organisations also provide guidance. All highlight issues such as a vulnerable person's understanding of what is being asked of them, consent and, for children, parental roles and methods of data collection. Box 2—provides a sample of available resources.
Examples of ethical guidance for working with vulnerable people
NSPCC. Ethical issues in research with children (resources) http://www.nspcc.org.uk/Inform/research/reading_lists/ethical_issues_in_research_with_children_wda55732.html
Links to a useful bibliography containing guidance on, and examples of, research with children, covering consent, research methods and children as co-researchers.
General Medical Council. Consent to research: research involving vulnerable adults http://www.gmc-uk.org/guidance/ethical_guidance/6471.asp
British Society of Criminology. Code of Ethics http://www.britsoccrim.org/codeofethics.htm
This guidance outlines researchers' responsibilities to colleagues and research participants and includes internet research, drawing attention to variations in legal and cultural jurisdictions.
NIHR Research Governance. HR good practice resource pack. The research passport: vetting and barring scheme guidance http://www.nihr.ac.uk/files/Research%20Passport%20Current/Research_Passport_and_the__Vetting_and_Barring_Scheme_Guidance.pdf
This guidance is part of the National Institute of Health research ethics guidance and sets out good research practice, drawing together other relevant policies relating to working with children and vulnerable people.
Ethical research with vulnerable people
As a team we were aware of the vulnerability of the young self-harmers whom we were hoping to recruit into our online study and also that the conditions of the research context (online, anonymous participation, no face-to-face intervention) would include risks for both participants and researchers. We therefore thought carefully about the ethical issues inherent in our study in order to provide as safe an environment as possible for the participants and take account of the impact of this kind of research on the researchers.
One useful concept was that of ‘ethics as process’,15 16 which views ethics as relationally dynamic and revisited frequently over the course of a study. This approach draws attention to the complexities of qualitative research approaches and sustaining relationships with participants over time, yet is equally applicable to internet-based research, especially with vulnerable people. Emphasis is on good intention, respect and awareness of the emotional, social and physical experience of the participant through research participation. This approach prioritises the researcher/participant relationship and participant representation and rights to withdraw and contribute as they wish, balanced against the aims of the researcher or study. In this approach participants can be encouraged to contribute to decision-making and discuss findings, in keeping with human rights and dignity.
Fisher's17 more recent discussion of a ‘justice-as-care’ approach to research ethics also provides a useful concept for researchers working with vulnerable people online. Fisher conceptualises vulnerability as a relational construct, which in a research context includes ‘susceptibility to harm’, a position, not solely reliant on societies' view of harm (physical, psychological, or social characteristics), but also acknowledging participants' dependency upon the actions of researchers within specific research contexts. Building on Goodin,18 Fisher suggests that in relational ethics, an obligation to protect the vulnerable should also acknowledge a potential ‘context of dependency’ and not just address the perceptions of the ‘agent’, in that the vulnerable participant is susceptible to risks from participation in research, and the researcher therefore has a role in alleviating risks, constituting a moral, rather than a ‘voluntary’ position.
Both these concepts draw attention to the need for guidance that goes beyond what Fisher terms the ‘agent’s' perspective and encourages researchers to take account of both their relationship with the ‘researched’ person and the potential role of participants in establishing and maintaining ethical research processes.
So, while guidance is available for researchers on a case by case basis, nevertheless, examples of lessons learnt during internet health research involving vulnerable people can be of benefit to other researchers working in these kinds of sensitive areas, where participants may be anonymous and researcher/participant relationships may be founded on more egalitarian norms than traditional models. We discuss our experience of addressing these sorts of ethical issues in our internet research with Young People who Self-Harm (YPSH), and offer our strategies as exemplars for researchers and ethics committee members.
The SharpTalk study
Many young people use the internet for support and health information, yet social networking sites have raised concerns about privacy and security of information for both patients and professionals.19 20 Health professionals need to improve their skills in working online and may be reluctant to undertake online discussions or support.21 Within this context, we aimed to explore the potential of online communities to foster engagement and shared learning between NHS professionals, healthcare students and young people who self harm (YPSH). The study (known as SharpTalk) was funded by the National Institute for Health Research (NIHR) UK RISC programme in 2008, and an ethics approval application was prepared and submitted in December 2008 within the NHS ethical governance system.
Based on UK ethical guidance we outlined our plan to build an experimental online forum (SharpTalk) to bring together YPSH and health professionals, record their (typed) discussions and observe their behaviour. We planned to recruit anonymously young people aged 16–25 from existing online self-harm forums, final year and postgraduate healthcare students through course tutors, and recently qualified professionals through advertisements on professional websites. There would be three discrete discussion groups of young people, students and professionals in different proportions. Data would be collected on recruitment, retention and levels of participation online. Topics (identified by researchers and participants) would be discussed in the groups. Discussions would be observed, recorded, archived, and analysed using discourse analysis. At the close of the forum, participants would be asked to complete an online questionnaire about their involvement in SharpTalk. The full study and findings are reported elsewhere (C Owens, unpublished data, 2010).22 We also provide a brief summary of the study as it happened following ethical approval (see Box 3).
The SharpTalk study following ethical approval
An online experimental self-harm support forum (SharpTalk) was set up to observe and assess engagement and shared learning between young people who self-harm, health professionals and healthcare students. Seventy-seven (young people who self-harm) were recruited anonymously through existing online self-harm websites, and 18 NHS professionals and healthcare students were recruited through health professional websites and university tutors. Participants chose a unique email address and username to register and participate in the forum. Participants logged in to start threads, read posts and respond to others. Interaction was asynchronous, and six moderators were online in rotation to monitor posts, especially during busy periods in the late evening. The site, like other support sites, had ground rules about appropriate online behaviour and links to online support. Threads and posts remained visible in the forum for participants to look at if they wished.
Participants could choose to log in and post whenever they wished during the 14 weeks that the forum was open (15 June to 20 September 2009). Three types of ‘rooms’ were available to participants: a debate/discussion room for topics relating to self-harm; a crisis/support room where participants shared their problems, asked for, and gave advice and support; and a random room where participants chatted about life and subjects not relating to self-harm. Participants ‘knew’ each other online by their usernames. Researchers and moderators used their real names on the site and participants knew when they were online. Participants were aware, and had consented to, their interactions being observed and their posts analysed. Findings from discourse analysis of posts in the crisis room are reported elsewhere (S Sharkey, et al, unpublished data, 2011),23 suggesting a range of advice giving within a protective milieu. The forum was briefly re-opened 6 months later to discuss emerging findings with participants, who contributed to the interpretation.
As part of our ethics approval application we described our plan to take consent online and asserted that participants would be anonymous, known only by a chosen username within the forum groups. SharpTalk would be a closed site which would provide privacy for participants. We included details of our risk protocol and information about our expert ethical advisory panel. We also informed the ethics committee that the site would be moderated by one of the researchers, and a group of clinicians would be on call throughout the study to provide advice to researchers.
We included copies of rules and links to be displayed on the website: ‘netiquette’ (eg, no abusive or offensive posts, no advertising); rules specific to self-harm (eg, not sharing graphic details of self-harm methods); guidance on labelling posts as ‘triggering’ (ie, likely to make someone feel like self-harming); suggestions of ‘alternative things to do if you feel like self-harming’ and web links to support sites (eg, National Self-harm Network).
Ethical issues and solutions
Ethics approval for SharpTalk was not initially granted by the ethics committee, which fed back concerns about anonymity, safety and consent. Taking account of the internet-based context, researcher and participant roles and relationships, and the vulnerable nature of the research participants, we discuss key ethical issues and then provide summaries of our solutions for researchers and ethics committees to consider (see Boxes 4–6). In practice these solutions often relate to more than one ethical problem.
Internet research with vulnerable people. Our solutions: anonymity
Solutions from SharpTalk—an ‘ethics as process’ approach
Acknowledge the lived experiences and expectations of participants.
Balance safety considerations and risk of needed research not being carried out.
Display information on website so that participants interested in joining the study can assess:
levels of anonymity (including how data will be used and reported, such as different identifiers used with verbatim quotes);
consenting process (if appropriate, different levels of consent for different levels of participation);
study procedures and expectations (including, in SharpTalk, that no face-to-face intervention was offered).
Clearly display discussion forum ground rules.
Require that participants create a new email and a unique username to join the study.
Provide direct confidential contact with named researcher.
Clearly display online support links.
Provide a private messaging facility observable by researchers.
Ensure a secure closed site (spam and search protected).
Internet research with vulnerable people. our solutions: safety
Solutions from SharpTalk—an ‘ethics as process’ approach
Provide appropriate moderator cover and training to ensure monitoring at peak times and ‘checking’ at other times in 24 h period (in SharpTalk six moderators in team, clinical advisor team, rota coordinator, rota ensured at least one moderator online at a time with back-up).
Consider a moderator ‘hand-over’ room.
Risk management protocol for moderators and researchers.
Report button for participants to report online abuse/concerns to moderators.
Private messaging facility.
Display and refer participants to forum rules about acceptable behaviour.
Provide a support page, including a distraction page, web links to other sources of support (online and face-to-face).
Provide direct email contact with named researcher.
Clearly display information about duration of the forum, right to withdraw and debriefing.
Consider participant advisory role as a means to stay engaged beyond the research.
Internet research with vulnerable people. Our solutions: verification and consent-taking
Solutions from SharpTalk—an ‘ethics as process’ approach
Consider targeted recruitment and required registration questionnaire for participants entering study.
Provide clear website links and information to enable participants to read information and understand the study, and what is being asked, to underpin informed decision about participation.
Phased consent-taking to give time and increase understanding to underpin informed consent (In SharpTalk consent was taken for registration, participation and use of quotes).
Provide opportunities for participants to discuss principles and boundaries of ethics particular to the study procedures, use as part of study decision trail.
Anonymity is desirable to maintain privacy, but in internet research concerns may relate to confidentiality of stored data,12 participants being identified via website breaches, or web searches using verbatim quotes.24 The ethics committee expressed concern about the research team's plan to recruit anonymously and the consequent inability of the team to intervene to help a participant in the event of a crisis. We considered the impact on recruitment of alternative strategies, such as, face-to-face in schools and asking participants for names and addresses. We rejected these alternatives as discouraging to potential participants,12 while still not guaranteeing our ability to intervene face to face. We considered the context of the YPSH, who frequently uses online discussion groups, and who expects anonymity.9 The National Self-Harm Network (an online support forum—http://www.nshn.co.uk/), for example, emphasises confidentiality as a condition of membership, and bans posts on its support forums that might compromise a member's anonymity.
In response to the ethics committee we reiterated the importance of anonymity within online culture and added information on this for them to consider. We emphasised the protective nature of anonymity and that participants would create a username and new email account, and different identifiers would be used in reported findings. Participants would have online access to information about internet privacy and would be asked to read this as part of the consent process. A closed site meant that participants would not be subject to unwanted external internet sites or spam emails. There would be online links to other websites providing support.
During the study, the importance of anonymity for online engagement with YPSH was borne out. Some participants discussed (posted about) different levels of anonymity, linked to perceptions about trust and type of website, as this extract from one of the discussion threads in SharpTalk shows:
Thread: Disclosing Identity
User 059 (usernames, date and time of post omitted to protect identity)
As a general rule I don't disclose my identity publicly…I have five (active) email addresses, one for uni, one for work, then three levels for people I trust…It's the people who try to find out who we are that scare me, if they try to find out my details without asking me then they are deleted/blocked because I can't trust them, after all, if their intentions were honest wouldn't they be content to ask?
This was mirrored in participant feedback:
With a topic like this I feel that it is important for anonymity as it has been concidered a tabo subject and if it was not anon then I would be more reluctant to post
(original grammar and typographical errors are retained in extracts).
Box 4 describes the strategies we used in dealing with the need for anonymity for the young participants, while attempting to minimise the risks associated with online anonymity.
Safety online is especially important for vulnerable young people discussing sensitive topics.8 Discussion in SharpTalk was within a National Health Service (UK) context and, although small and closed, involved participants sharing problems relating to self-harm and supporting each other. SharpTalk therefore mirrored interactions in larger ‘public’ self-harm support forums, where face-to-face intervention is not expected (face-to-face intervention was not part of the SharpTalk protocol), but advice and emotional support may be given by group members and moderators, especially during crisis. As researchers we believed it important to undertake research with YPSH online, in order to improve understanding of their lived experience and world view and help improve health service provider dealings with vulnerable and often marginalised groups of young people. Although face-to-face intervention was not part of the study, researchers need to consider what happens after their internet study. The ethics committee was worried that discussion in SharpTalk might promote self-harm and that professionals might also disclose self-harm. Additional issues included not being able to help vulnerable people face-to-face, the risks of promoting self-harm and the need to support young people after the study. The committee was particularly concerned about the adequacy of the moderation arrangements and the need for moderator training to ensure safety in the study.
Box 5 describes the strategies we used to optimise participant safety in SharpTalk, including: targeted recruitment (experienced and confident internet users)2; a registration questionnaire; clearly displayed study information and forum ground rules; a risk protocol (available on request from lead author), and a moderation system (team of trained moderators, a planned rota, a moderator ‘hand-over’ room, clinical advisor back up); external ethics advisors; links to online support; private messaging facility, and direct confidential contact with a researcher. Participants also had an alert button to report breaches of forum ground rules.
During the study, participants presented as confident internet users, reinforced the forum rules and helped shape the online space. In our ongoing research programme with YPSH, we ensured that the participants had clear information about the expected duration of the forum and were also able to provide a longer-term role as advisors to the programme to any of the young people who were interested. Within crisis and discussion room interactions, and in private messaging, participants supported each other and there was no incitement to self-harm. Moderation on SharpTalk was viewed positively, especially the supportive interaction, along with the small, closed nature of the site. During participant feedback there was a consensus that moderators were needed on self-harm forums, and a strong consensus that moderators ought to get involved in providing support.
Thread: Your Views on Moderation
I think the mods on here are better then any of the ones on other sites I have used, but perhaps that is because this is on a much smaller scale then the others.
Mods on here, get involved and offer support that maybe users can't always do because they are still goin through it themselves, they also provide imspiration that we can get through these difficult times, and you know they will always listen
Verifying identity and consent online
Linked to anonymity and safety, in internet research, researcher and participants are not brought together physically, so verification of identity is difficult.25 This does not mean that online research is more ‘misrepresentative’ than face to face.10 Verification ‘safeguards’, such as credit card checks, IP addresses, emails or hard copies of consent forms are not always practicable or failsafe.19 Some authors have suggested that it may be easier for participants to deceive researchers online.12 Others suggest that fraudulence is rare.26 While deception may happen, this may be no more likely than in face-to-face research. Nevertheless, the ethics committee had some concerns that participants might not provide accurate information. We were interested in how the participants in SharpTalk presented themselves and constructed their identity online, rather than the ‘truthfulness’ of what they were saying.
Our approach (box 6) aimed to enhance trust and commitment, and increase understanding to support informed consent. Detailed study information was displayed on the website during recruitment. However, extra measures may be no guarantee of understanding (online or face to face), and, in online information to participants, a balance is needed to avoid excessive wording, which may also exacerbate the tendency to agree without reading.27 Phased consent allows participants time to consider and reflect on different aspects of participation. We used this approach in SharpTalk, taking initial consent to use registration data, and subsequent consent for participation and analysis of discussion data. We also sought consent to use direct (anonymised) quotes. Recruitment of YPSH into SharpTalk was brisk (Within the first 24 h, 26 young people had registered and following the 2-week recruitment window, 77 young people were admitted into the study). The young participants appeared confident and committed posters, readily discussing health research and problems relating to self-harm.
Debriefing and withdrawal
Information about the right to withdraw needs to be clearly stated in internet research and debriefing can include a summary of findings either sent individually or via a link to the study website. The ethics committee suggested these issues needed greater emphasis on SharpTalk. We added an ‘I want to withdraw from the study’ page onto the website, emphasising participants' right to withdraw at any time without consequence, along with a debriefing page. We explained to the participants that withdrawing data which had been collected from interactions in an online forum would affect the meaning of the remaining data. Interestingly, by tracking which website pages on SharpTalk the young participants visited, we knew that no-one had used the ‘I want to withdraw’ facility, yet during the study participants did withdraw, by simply not posting any further. This may reflect the apparent ease of withdrawal (and expectation of such) of participants online.
A process as ethics approach
Our approach to ethics took account of participant perspectives and researcher relationships with participants. One example from the SharpTalk study serves to illustrate the benefits of a process approach to ethics. Two weeks into the life of the forum it became clear that one of the discussion groups had fewer people posting, with less opportunity for peer support, thereby increasing risk. The moderators, in light of few professionals posting, had taken on an increased support role, but this was no substitute for the peer support that participants were getting in other groups in SharpTalk. We discussed concerns with participants and the ethics committee, seeking a solution, which resulted in the three forum groups being reduced to two, increasing availability of peer support for all participants. This approach enabled the participants to have a collaborative and advisory role within the study, in keeping with their expectations about relationships and trust within the forum. The SharpTalk ethics application was resubmitted and approval was granted in February 2009. Box 3 outlines the SharpTalk study.
We have attempted, through examples from our own research, to show how a process approach to ethics can be useful to internet researchers running discussion groups involving vulnerable people (see Fisher and Ramcharan and Cutcliffe in our introduction) by prioritising the relationships within a study and acknowledging the unique participant perspective and context, and their potential role in helping address ethical issues in research. In our SharpTalk research we acknowledged the vulnerability of our participants, not positioning them as dependent in any negative sense, but recognising their expectations of an egalitarian context and how they would relate to us as researchers. They were able to discuss these issues with us, demonstrating their knowledge and expertise online and we sought to collaborate, identifying risks and subsequent risk-management strategies as the research progressed. Berry28 suggests a negotiated approach to consent and involvement, to avoid exploitation and alienation of internet research participants. By adopting a ‘process’ approach in SharpTalk, we were able to recognise the relationship between participants, moderators and researchers as part of a risk management strategy, which we recommend as important in internet discussion group research.
Gaining ethics approval for sensitive internet research can take time, and ours was not a unique experience,19 but we have nevertheless reflected on what might have optimised the ethics approval process for our study, and benefit other researchers and ethics committees. Internet researchers should not overestimate the scope of knowledge of ethics committees. It is important to anticipate information needs of committees when seeking approval, in order that committee members can fully understand the online context or risks, or indeed, the wishes and perspectives of young vulnerable participants. We suggest that NHS research ethics committees (RECs) undertake specific training to increase awareness of internet research with vulnerable groups. Some useful online information sites are available and these could be built upon for REC training (see Boxes 1 & 2). It is not reasonable to expect committee members to have knowledge and skills in every area of research and it might be useful if committees had a range of advisors, reflecting emergent research methods, such as research with vulnerable people in online discussion groups. Research teams might also be asked to recommend such advisors on a one-off basis, to help speed the approval process for both researchers and committees. Within a national system different RECs might have different expertise, such as specialising in online research, which would allow collaborative thinking between REC members and researchers about particularly novel or contentious areas.
SharpTalk ethics was dominated by anonymity, affecting other areas of ethical concern, such as safety and consent. Internet research requires attention to participant expectations and culture when balancing safety and the need for research, and anonymity and a collaborative approach was an expectation among SharpTalk participants. Our process approach to ethics enabled us to be responsive to participants and consult the ethics committee, which proved important when unexpected ethical issues emerged during the study.
We are grateful to the ethics committee who reviewed our ethics application and discussed ethical issues with us during SharpTalk. We thank all the participants in SharpTalk who shared their experiences and views with us. We also thank the following for their contributions to the study: Jayne Clarke, Peter Aitken (supported by the NIHR CLAHRC for the Southwest Peninsula), Fraser Reid, Matthew Gibbons, Priscilla Alderson, Nicola Madge and Mary Gilhooly.
Funding This paper presents independent research commissioned by the National Institute for Health Research (NIHR) under its Research for Innovation, Speculation and Creativity (RISC) Programme (grant reference number RC-RG-0407-10098). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. SS and CO were partly supported during the writing of this paper by the NIHR CLAHRC for the Southwest Peninsula.
Competing interests None.
Ethics approval This study was conducted with the approval of the Southampton NHS REC with, This study was conducted with NHS REC approval.
Provenance and peer review Not commissioned; externally peer reviewed.