Article Text
PDF
Abstract
Objectives Government guidance to manage COVID-19 was challenged by low levels of health and digital literacy and lack of information in different languages. ‘Covid Confidence’ sessions (CC-sessions) were evaluated to assess their effectiveness in counteracting misinformation and provide an alternative source of information about the pandemic.
Design We worked with community anchor organisations to co-ordinate online CC-sessions serving three economically deprived, ethnically mixed, neighbourhoods. We conducted a qualitative, participatory process evaluation, in tandem with the CC-sessions to explore whether a popular opinion leader/local champion model of health promotion could mobilise pandemic responses. Group discussions were supplemented by final interviews to assess changes in community capacity to mobilise.
Setting Sheffield, England, September 2020 to November 2021.
Participants Community leaders, workers and volunteers representing a variety of local organisations resulted in 314 attendances at CC-sessions. A group of local health experts helped organisations make sense of government information.
Results CC-sessions fostered cross-organisational relationships, which enabled rapid community responses. Community champions successfully adapted information to different groups. Listening, identifying individual concerns and providing practical support enabled people to make informed decisions on managing exposure and getting vaccinated. Some people were unable to comply with self-isolation due to overcrowded housing and the need to work. Communities drew on existing resources and networks.
Conclusions CC-sessions promoted stronger links between community organisations which reduced mistrust of government information. In future, government efforts to manage pandemics should partner with communities to codesign and implement prevention and control measures.
- COVID-19
- Health Education
- Health Equity
- INFECTIOUS DISEASES
- PUBLIC HEALTH
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
STRENGTHS AND LIMITATIONS OF THIS STUDY
CC-Session agendas were based on local concerns, thereby ensuring good engagement by stakeholders.
Data collected during sessions on effective mobilisation was supplemented by interviews with key informants.
The relationships between engagement and vaccine uptake could not be established as it was not possible to link interactions with individual data.
Background
While the Government has generated a vast amount of information and guidance about the COVID-19 pandemic, people say that they struggle to keep up with it. The WHO announced in 2020 that the COVID-19 pandemic had triggered an infodemic, that is ‘an overabundance of information—some accurate and some not—that makes it hard for people to find trustworthy sources and reliable guidance when they need it’.1 Although rumours and misinformation are spread in all disease outbreaks, information now “goes faster and further, like the viruses that travel with people and go faster and further. So it is a new challenge, and the challenge is the [timing] because you need to be faster if you want to fill the void…What is at stake during an outbreak is making sure people will do the right thing to control the disease or to mitigate its impact. So it is not only information to make sure people are informed, it is also making sure people are informed to act appropriately”.2 When creating information strategies, governments are dealing with a wicked and complex problem, because COVID-19 is (a) a new phenomenon, where the virus, as well as the science needed to tackle it, is rapidly evolving and (b) treatments and policies for treatment are contested. Further, the Government strategies for spreading accurate information have met with varying success because willingness to accept the facts varies according to the beliefs of any particular group, and their attitudes and trust towards government information in general. The success of strategies is not only dependent on local attitudes and beliefs but also on the characteristics of the messenger, for example, the person or group that is delivering the information.
The success of using what are called local champions has been extensively documented since the HIV/AIDS pandemic began in the 1980s.3
Local champions are people, who are willing to promote local awareness and action via informal conversations with family, friends, neighbours, street outreach working with local organisations and virtual outreach, using social media channels. They are well placed to explore the barriers to acting on COVID-19 information and can serve as bridges to community organisations that can help to remove social and economic barriers to following guidance. This approach to making sense of guidance and promoting action rests on the identification of local people who are trusted by others in their community. Several key elements are needed to successfully move people from looking at information to taking action to protect health. These include enlisting locally known popular opinion leaders, using people with knowledge of the area/group to champion the initiative, providing training and ongoing support to ensure that local people have the confidence to spread the word, encouraging locally trained people to use their own local knowledge to ensure that messages are tailored to different concerns and groups and feedback to help those who are having the conversations see the impact.4 The relevance of local champions has been recognised in relation to COVID-19.5 We used these principles to organise ‘Covid Confidence’ sessions, with the aim of supporting people in economically deprived neighbourhoods to act appropriately in terms of managing risk.
Methods
CC-Sessions were coproduced and hosted by community organisations, who expressed interest in working with the Sheffield Community Contract Tracers (SCCT) to mobilise responses to the pandemic. SCCT is a voluntary group of retired health professionals, comprised of nine former public health specialists, directors of Public Health and local general practitioners (GPs) with experience in infection control, communicable disease control, epidemiology, health promotion, primary care, participatory evaluation and community organisation. The group originally came together to pilot the effectiveness of community-based contact tracing and subsequently expanded their role to disseminating COVID-19 information, with the aim of promoting understanding to community workers and volunteers. Many of these experts/professionals had long-established links with local voluntary and statutory organisations. The CC-sessions dealt directly with issues of misinformation. They were provided in conjunction with online SCCT information sessions, which delivered up to date information. Topics were based on community-identified concerns. The information sessions aimed to provide people with key facts about COVID-19 exposure, transmission and protective behaviour, as well as the COVID-19 vaccines. The CC-sessions drew on local knowledge and expertise, using discussion to show how champions can support people to deal with issues arising during the pandemic. A participatory process evaluation was conducted, where participants observed and reflected on the utility of the sessions. Qualitative key informant interviews were conducted with a subset of participants at the end of the project.
The programme theory for Covid Confidence is based on well-known models for using popular opinion leaders and providing peer support to manage health.3 6 We began with a set of assumptions (table 1).
Preliminary logic model for COVID-19 Confidence
CC-session agendas were created by eliciting community concerns, via conversations with anchor organisation leads, and allowing these concerns to dictate the topics for discussion. This put communities in control over the nature and direction of support. Knowledge exchange and mobilisation were promoted, using a participatory action learning and research framework to explore challenges in supporting local people during the pandemic, figure out possible solutions, observe what happened and then reflect on what had worked and why.7 This process facilitates collective learning, collaboration and networking to promote social change. When complicated information was presented about COVID-19, participants were given time to engage in a process of collective sensemaking, to question implications, to explore how people might struggle to carry out the guidance and what the guidance would mean for people in different circumstances.8
Three different neighbourhoods expressed interest in developing COVID-19 Champions. Fifteen sessions were facilitated between September 2020 and November 2021 across three Sheffield neighbourhoods. These neighbourhoods have areas of high local deprivation, a mix of ethnic groups and 18%–22% of the community do not have English as their main language. Further, 18%–27% live in overcrowded housing, which is far more than the national average of 8.7%. 48 organisations took part, with a total of 198 participants (including repeat attenders). Participants drew on their local knowledge to consider how information could be used in their particular setting. They ‘road tested’ the information during conversations with people in challenging life circumstances and fed back in subsequent CC-sessions on how information needed to be tailored for local groups and local issues. Participants used concerns raised by the public to structure sessions and were involved in reflecting on the utility of the sessions throughout. This information was included in the process evaluation.
The participatory process evaluation’ was conducted in tandem with CC-sessions, generating data from participants’ reflections on the challenges and outcomes of taking action and the utility of the discussions. Participants agreed that CC-sessions would be recorded and notes would be taken during sessions, for the purposes of tracking progress with the action learning cycle. Notes were fed back to participants in subsequent CC-sessions to reflect on whether and how they enabled champions to have conversations with people and modify information for local groups. Over time, we hoped that repeated participation in the CC-sessions might lead to shared learning and networks of support, building community capacity to address the pandemic. Capacity building happens when community groups become more able to define, assess, analyse and act on health (or any other) things that their local members are concerned about (see table 2). Key indicators of capacity, developed via prior research,9 10 were used to review the CC-session notes and interviews, in order to assess how the process related to capacity building.
Indicators of community capacity building and mobilisation
The CC-session notes were organised by three researchers (JH, PRA and FA) using the above indicators as themes and members checked via 9 key informant interviews with 12 participants. These were conducted at the end of the project, to assess changes in community capacity to mobilise. The topic guide, developed from findings, asked general questions as well as exploring communication issues that arose about Council co-ordination of the information bus and the many positive comments that were made about a local coproduced film (box 1).
Interview topic guide
SCCT role
What are your thoughts on the role that SCCT played in sharing information about COVID-19?
Did it generate trust? Reduce hesitancy? Form new connections and/or networks?
What has changed—for the long term, rather than temporarily?
Have local knowledge and experiences been valued and recognised by public health and the City Council?
Has the process enabled communities to have a voice in managing COVID-19?
Covid Confidence cascade effect
Some people came to the first sessions and went on to run smaller CC-sessions in local areas, feeding back information from the larger sessions. We would like to understand the differences in your process—what happened? How did your local sessions develop?
How was local knowledge used through Covid Confidence?
Were there any new ways of doing things? Will these approaches last/be taken forward or were they COVID-19 specific?
Were you able to provide advice and support to people who were concerned about symptoms?
Were you able to facilitate sessions that met local needs, for example, local language support?
Community bus
What do you think about the Council’s community bus?
Is it something that you have helped to support in your area?
Are you getting enough notice to mobilise local workers to show up and support the bus?
How do you think the bus helps to communicate?
Does it help your communities? Have you heard any local feedback about it?
Seldom Heard Communities film
Were you involved with the Seldom Heard Communities film?
People have mentioned the film as having been key in the process of managing the pandemic. What do you think?
What impacts do you think the film has had?
SCCT, Sheffield Community Contract Tracers.
We selected key informants from the larger group of participants. They were given a participant information sheet and asked if they had any questions prior to the interview. After checking understanding, verbal consent was taken.
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.
Findings
Findings are presented in terms of the indicators for community capacity building (see table 2).
Stakeholder participation records showed community leaders, workers and volunteers representing community groups, local people, health services, university and council departments attended 1 or more of 15 sessions (table 3).
Overview of groups represented
Four CC-sessions were hosted by two of the neighbourhoods. People from these neighbourhoods also attended the 11 CC-sessions hosted by the third neighbourhood, totalling 314 attendances. From this group of repeat attendees, a core group representing a diverse range of local organisations developed over time. The group became a forum for doing problem assessment, identifying shared problems, considering solutions and considering whether they had capacity and resources to take action (table 4).
Combating misinformation: possible solutions and actions taken
Participants exercised local leadership, identifying local organisations and informal groups that were willing to take action. This included health champions working to spread accurate information and supporting people to consider how they could reduce their risks. Organisations that had previously received little formal recognition found the process empowering and affirming, “The BAME community themselves got organised – realised that they were actually very active within their own communities” (OB02). These actions fostered stronger links between people who had previously little experience of working together. For example, volunteers from a number of organisations promoted City Council information about the vaccine programme, subsequently coming together to assist general practices with vaccine uptake. The links became empowering organisational structures, which were instrumental in helping communities come together to address COVID-19 issues. SCCT sessions “put us in touch with all sorts of people who are a mine of information…and certainly the workshops enabled us to meet people who were really helpful” (OD02). A number of people attended all of them, which created stronger links between people and organisations which “were able to share information from one group to another” (OD02). “They used sessions to share how to run COVID-19 Q&A sessions, how to build Covid Confidence, and how to adapt information and support to specific groups” (OS03).
Communities mobilised resources by recruiting volunteers fluent in Urdu, Punjabi, Arabic, Hindi and Bulgarian to produce videos. “Locals identified social media platforms that were commonly used to promote messages” (OS01). Members of Sheffield City Council (SCC) Public Health and Communications teams began to attend sessions, working with local people to coproduce more accessible written information. Over time, sessions became a conduit for disseminating vaccine clinic schedules to volunteers and COVID-19 Champions. When community leads noted that trusted local people were instrumental in promoting use of drop-in clinics, SCC arranged times and locations in tandem with them. People noted that “the partnership approach really helped with getting messages out” (OS01).
The CC-sessions enabled communities to establish a more equitable relationship with academics and Public Health experts. SCCT used their previous working relationships to engage current experts in seminars. These crosscutting relationships meant that they were able to facilitate dialogues between expert speakers and local people, with the express aim of making sense of information. Experts provided updated information on government guidance, explained local and national statistics and outlined the development of the vaccine strategy.
Most of the written information is not clear – it really helped our community, ourselves as part of the community, as well as professionals. It’s good to get explanations, rather than just information. (OD01)
There was a widely shared perception that “SCC Public Health staff are very good at what they’re doing, but they’re not necessarily there when you need them”. Thus, the information sessions hosted by SCCT were seen as
Vital for our work – good to know they were there every 2 weeks. Though we were all community leaders and community activists and community people, we are not medical people. The questions we were taking to SCCT were the questions we were hearing from the vulnerable people in the community. I was getting really good answers from SCCT so that was a big asset (OB01).
Countering misinformation was a long-term challenge because of the constantly shifting information about COVID-19. Responsive and timely exchange led to what one participant called a “waterfall effect – they got information and were able to pass that back to others more locally in their area” (OS01). SCCT became an “anchor point - it meant that we were able to keep countering the misinformation that they were getting on a regular basis” (OD01; OD02).
Although the strategies to improve access to relevant information were effective, champions found that trust was a major issue.
As much as we shared the data, the statistics and what the Government were saying, we weren’t getting anywhere. There was no trust in the communities because we were working with the Government. (OSO1)
Community organisations decided that the only solution was to increase cross-sector stakeholder collaboration, which included local leaders.
The only way to get the information out to the communities was to work with the local Imams and local GPs. We worked with particular GPs who people from the community really look up to….One of these GPs was also very active on social media and the community organisations used his videos a lot. Some of these GPs also came to COVID-19 Confidence.
Champions said that the fact that these leaders were known and trusted was a tipping point in COVID-19 engagement:
Champions said that the fact that these leaders were known and trusted was a tipping point in COVID-19 engagement:
[So we said] ‘your GP and your Imams - two people you really look up to - would they be lying? Would they be putting you at risk?’ And we explained that they could be putting themselves more at risk by not getting tested or having the vaccine” (OSO1).
Organisations collaborated with SCC to promote test centres, which “really built a trusting relationship with SCC and Public Health – they learnt from us and we learnt from them” (OS01). The process increased local credibility, because it “meant we knew up to date information… people then trust you because you’ve known what’s going on” (OD02).
Dealing with mistrust and vaccine hesitancy led stakeholders to ask why relationships with government agencies were so poor. A docuseries, coproduced by SCCT and Black, Asian and minority ethnic (BAME) leads, engaged seldom heard communities in describing their situations (https://www.communitycontacttracers.com/shc/).
The collaboration triggered some changes in programme management. Community organisations - such as the BAME Group - were asked to join statutory sector steering groups and committees. As a result, voluntary sector organisations gained greater awareness of “what was happening on the ground” (OS01). They also became partners in the sense of comanaging COVID-19 issues. “SCC didn’t just take a top-down approach – felt they were asked how to best achieve things in the communities” (OS02). Community hubs and anchor organisations informed a number of statutory sector initiatives including COVID-19 Community Response Grants, the COVID-19 information bus, the vaccine van and developing a cadre of COVID-19 Champions. The ‘COVID-19 Community Response Grants’ have meant “Closer working relationship than there might normally be, especially with some organisations for example, around comms – continued focus on getting specific comms out to particular communities” (CO2). Being valued for their knowledge meant that organisations—particularly smaller ones that felt previously missed—believed that they were in a “definitely different relationship now, [where] people in the Council have now recognised the importance of the voluntary sector” (OS02; OS01).
Alongside description of capacity building, people also identified the underlying elements that contributed to effective community mobilisation. Drawing on local knowledge was essential to counteracting misinformation and ensuring that health promotion was seen to be relevant and appropriate. CC-session participants agreed that ‘trust was such a massive issue’ and it was important to start by involving people who were already locally known and trusted. For example, Roma Slovak and Bulgarian groups were all approached by finding people who had existing relationships with them, to act as a bridge for the COVID-19 volunteers. “In some neighbourhoods, a cadre of active volunteers already existed, usually coordinated by the local community anchor organisation. COVID-19 volunteers were in most cases community members, working in other capacities, and local volunteers already knew what people were saying/thinking about the vaccines” (OD01). When volunteers were new to an area, community leaders and people known to the various groups acted as gatekeepers, linking volunteers with the areas which needed support. This gave volunteers credibility. They noted that being endorsed by local people was crucial in gaining trust in an area—“Without that, we wouldn’t have got anywhere” (OB02).
It was also important that volunteers had some experiences in common with the local people they were reaching out to. This could be a shared language, a common culture, being a member of a vulnerable group and having COVID-19 or caring for a family member who had COVID-19. For example, a bilingual medical student whose father had gone through COVID-19 was trusted almost immediately even though he was not locally known. Being in a respected role counteracted misinformation. For example, a video of an imam and an Islamic pharmacist at the local mosque included the message given from the Quran saying you should save your own life first. “This really gave them confidence to get the jab” (OD01).
Local workers with established relationships were also able to influence people. A worker who was well known in her local community made a video of herself talking in Arabic about having her vaccine. “These pre-existing bridges of trust enabled outreach to a range of ethnic groups in different neighbourhoods” (OD01; OSO1).
Finally, the importance of repeated contacts was mentioned frequently. One off, written information is not effective on its own. Outreach often involves having more than one conversation with people, leaving them time to consider and come back with concerns and questions (Sharrow COVID-19 Confidence meeting Feb, June 2021). Conversations need to start by listening to people’s concerns before giving information. Further, information giving needs to be embedded and opportunistic. Opportunities to share information need to be found in the course of engaging people on what matters to them first, building a relationship and introducing COVID-19 information when appropriate. For example, women’s well-being sessions were used to “find out what is most important issue for them first – build that relationship and then get onto vaccines when they are ready. It can take a lot of conversations too – not a quick win” (CC-Session meeting 11.2022).
Findings were compared with our preliminary logic model for developing Covid Confidence (table 1) and used to develop a final model showing how the approach enabled people to communicate information and support to individuals and local groups (box 2). The effectiveness of this approach is supported by previous research using locally known and respected people to raise awareness of health risk during epidemics.3
The Covid Confidence approach
Providing information
Receiving information by trusted experts.
Being encouraged to ask questions
Helped workers and volunteers understand COVID-19 guidance.
Developing approaches to communication
Promoting discussion that focuses on
Identifying the challenges of explaining COVID-19.
Considering the concerns and needs of people in different groups and situations.
Sharing approaches to giving information to local people.
Helped people to develop approaches for successfully communicating the information to different groups.
Establishing trust
Using people who were known in the neighbourhood or had something in common.
Using familiar vocabulary and preferred language.
Listening to concerns before ‘telling’ people what to do.
Repeated contact and conversations.
Built trust which inclined people to consider following recommendations.
Reflecting on the process
Co-ordinating sessions where people were encouraged in.
Feeding back on successful ways of communicating information validated champions, recognising that their local knowledge combined with communication skills is effective.
In summary, the participants reported that the CC-sessions increased their access to up-to-date, accurate information, and the sense-making process meant that they were able to explain information to other people. They were able to coproduce translated material which was useful in local discussions, and coproduced videos were found to be useful (table 3). People who used the Covid Confidence approach (box 2) reported that they were able to influence people, in terms of their stated intentions to self-isolate and to get vaccinated. The surrounding context, however, contained a number of challenges that made it difficult for people to actually follow guidance. Workers reported that people were reluctant to get tested, because they needed to continue to work. People were unable to comply with self-isolation and distancing guidelines, due to overcrowded and poorly ventilated housing. People were challenged to feed themselves and their families if they were self-isolating, because their incomes were affected. The challenges of managing COVID-19, therefore, reflect the challenges of health inequalities.
Discussion
The process evaluation had several strengths and weaknesses. Participants set the agenda for each session, this increased attendance and ensured that local issues were addressed. Locally known and trusted Public Health and medical people facilitated discussion between local people and pandemic experts. This provided data on the challenges of contextualising of information and using local expertise and knowledge to generate appropriate engagement strategies. The data collected during sessions was supplemented by a final set of interviews with local workers and leaders who clarified and confirmed the findings. Unfortunately, quantitative relationships between engagement and vaccine uptake could not be established, as it was not possible to link interactions to decisions to get vaccinated.
Our findings align with previous research on the elements needed to effectively communicate risk to people during epidemics3 and recent community engagement research which found that community leaders, volunteers and multilingual approaches targeted to specific groups can effectively disseminate COVID-19 information and expand access to testing.11 12 Recent research calls for authentic community outreach13 but as of yet there are few studies confirming effectiveness.14 There is evidence from a recent systematic review that community engagement can prevent and control disease during an epidemic, when local leaders, community organisations and networks, key stakeholders and local people communicate social and behavioural risk and get logistical support from health sectors.15 Stories from our Covid champions indicated that support promoted vaccine uptake, but we decided not to ask individuals to share personal details allowing us to link them with vaccination decisions, because of the widespread lack of trust. Community-based contact tracing can also promote vaccination,16 but in Sheffield this local initiative was not supported by the UK Public Health system. In our project, local neighbourhoods used CCC- sessions to foster collective relationships among organisations, which in turn codeveloped communication strategies. We found that communities had to embark on a process of ‘recontextualising’ government information by finding locally appropriate ways of explaining risk.17 Linkages with health systems came late in the process, translation of written information, information sharing about the timing of the community bus and co-ordination of clinic schedules finally occurred in months 9–14. Our study echoed review findings regarding inadequate community resources and weak health support infrastructures.15 Our communities were able to identify issues quickly, and also keep track of emerging needs, using existing voluntary sector knowledge to draw on local social networks. This ability to rapidly and flexibly mobilise is identified as crucial in other studies18 which note that community mobilisation at early stages can compensate for slower restructuring at statutory level. Other studies also note that given the evidence of effectiveness, communities need to be engaged from the beginning in codeveloping and co-implementing public health strategies, sustaining a two-way dialogue to consistently provide transparent and accurate information.19 20
Last, but not least, barriers in the broader context, which have been a problem in past epidemics, continue to limit abilities to connect people to local organisations that can work with them to prevent COVID-19 transmission, and remove barriers to self-isolation if infected.
Conclusions
In participatory research, impact is defined as the changes that occur during the process of collaborative enquiry and reflection, this includes changes in interactions between individuals and organisations as well as across systems. Our evaluation found that Covid Confidence increased interaction across stakeholders, improved ability to assess local problems and generate solutions and fostered stronger links that increased community capacity. Trusted outside agents—SCCT members—who were able to enlist topic experts increased confidence of local people in communicating information, counteracting misinformation and supporting people to make appropriate decisions. Local people began to trust champions as a result. Providing a forum where organisations can identify problems and possible solutions, adapt information and share effective approaches was an important element in developing cross-organisational relationships and supportive networks. Early support can enable communication across local groups to take action. The mobilisation can, however, be challenging to integrate with statutory sector management in the early days of a pandemic. Fewer changes were seen at the level of government systems. Mobilisation, was challenging to integrate with statutory sector management in the early days of the pandemic. Despite having capacity to manage at community level, however, the broader context of social determinants of health diluted effectiveness of mobilisation. In the case of COVID-19, individual ability to follow guidance was undermined by the need to continue work, often in front line jobs, and overcrowded housing.
In conclusion, community mobilisation can be an instrumental component in public health pandemic management. However, communities need to be actively involved in codesign and implementation of public health strategies. The strategies need to be underpinned by government recognition of underlying inequalities that make it difficult to follow COVID-19 guidance. As the urgency of the pandemic wanes, the resources needed by community organisations to continue to manage the long-term effects of COVID-19 need to be retrospectively assessed and coherent funding strategies put in place that support them in continuing to address the underlying issues of health inequality that have been highlighted during the pandemic.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Sheffield Hallam University (ER30632144). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
The authors would like to thank the organisations and local people who participated in Covid Confidence and cocreated data via their reflective practice; the members of SCCT who reviewed drafts of the evaluation report; and key informants who participated throughout and member checked the findings. The authors would also like to thank BMJ Open reviewers for their useful comments.
Footnotes
Contributors All four authors contributed equally to this paper. JH designed the evaluation and led on data analysis with contribution from PRA and FA; JH, FA and PRA collected data; and PRE contributed to the analysis and writing. JH is responsible for the overall content and acts as the guarantor of the study.
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 but members of the public were involved at all stages
Provenance and peer review Not commissioned; externally peer reviewed.