Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Who benefits and how? Public expectations of public benefits from data-intensive health research:

Who benefits and how? Public expectations of public benefits from data-intensive health research: The digitization of society and academic research endeavours have led to an explosion of interest in the potential uses of population data in research. Alongside this, increasing attention is focussing on the conditions necessary for maintaining a social license for research practices. Previous research has pointed to the importance of demonstrating ‘‘public benefits’’ from research for maintaining public support, yet there has been very little consideration of what the term ‘‘public benefits’’ means or what public expectations of ‘‘public benefits’’ are. In order to address this pressing issue a series of deliberative workshops with members of the public were held across Scotland in May and June 2017. The workshops aimed to engage a cross-section of the Scottish population in in-depth discussions of the ways that the public – or publics – might benefit from data-intensive health research. The findings reported here discuss workshop participants’ understandings and expectations of health research; who they considered to be ‘‘the public’’ that should benefit from health research and; in what ways they felt ‘‘the public’’ should benefit. Workshop participants’ preference was clearly for the widest possible public benefit to be felt by all, but they also acknowledged the value in research aiming to primarily benefit vulnerable groups within society. A key focus of discussions was the extent to which workshop participants were confident that potential public benefits would be realised. A crucial consideration then is the extent to which mechan- isms and political support are in place to realise and maximise the public benefits of data-intensive health research. Keywords Public benefits, data science, health informatics, public engagement programme had significant impacts and resulted in the Background programme being put on hold. Such controversies have The digitization of society and academic research drawn attention to the importance of engaging with endeavours have led to an explosion of interest in the members of the public to ensure that data is used in potential uses of population data in research (McGrail ways which align with public values and interests to et al., 2018); this is particularly true in relation to health ensure that public concerns are adequately addressed. research (e.g., Aitken et al., in press; Wellcome Trust, This is essential for developing and maintaining a 2015). However, recent years have also brought a number of public controversies, particularly regarding proposed uses of health data (e.g., Carter et al., 2015; Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, UK Garrety et al., 2014). Two high-profile examples from the UK are the failed introduction of the care.data Corresponding author: scheme to link hospital and GP records (Ramesh, Mhairi Aitken, Usher Institute of Population Health Sciences and 2014) and Google Deep Minds’ involvement in process- Informatics, University of Edinburgh, Old Medical School, Teviot Place, ing health data at an NHS Trust in London (Hodson, Edinburgh EH8 9AG, UK. 2016). In the case of care.data, public opposition to the Email: mhairi.aitken@ed.ac.uk Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http:// www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society social licence for current and future practices (Carter and social well-being’’’ (Scott et al., 2018: 21). Given et al., 2015). the importance of ‘‘public benefits’’ for public accept- The growing interest in potential uses of population ability of data uses, and its central role in determining data, and the increasing recognition of the importance which researchers get access to data for what purposes, of ensuring a social licence for these uses, has resulted it is essential that a better understanding of what the in considerable interest in understanding public atti- term means is generated in a way that is grounded in tudes and views on these topics (e.g., Aitken et al., public views. This paper aims to provide a first step in 2016; Davidson et al., 2013; Wellcome Trust, 2016). developing such an understanding. It draws on the Accordingly there is a fast-growing body of inter- results of three deliberative workshops with a broad national literature reporting findings of studies examin- range of publics, organised as part of a programme of ing public attitudes towards data and health-related public engagement research around data linkage and data science practices (Aitken et al., 2016). A consistent sharing for health research. The workshops focussed theme to emerge through this literature is the import- on the meanings and expectations publics held and ance of demonstrating ‘‘public benefits’’ from research formed regarding such research use and the paper using population data (e.g., Davidson et al., 2013). maintains that such formulations should inform Much of the qualitative research undertaken has dis- research policy and governance in health-related data cussed perceived ‘‘trade-offs’’ between risks to individ- science. uals (e.g., loss of privacy) and broader public benefits of research. In a systematic review of literature on public Deliberative methods and mini-publics attitudes towards data linkage for health research it was found that ‘‘assurances that research would bring about Deliberative forms of public engagement have been public benefits – or at least that it had the potential to used to examine and address public interests in a wide bring about such benefits – were widely reported to be range of policy areas including: health (Mitton et al., fundamental for ensuring public support or accept- 2009); e-health (King et al., 2010); nanotechnology ance’’ (Aitken et al., 2016). While previous research (Pidgeon and Rodgers-Hayden, 2007); natural resource has revealed different sets of expectations around management (Halseth and Booth, 2003); transport ‘‘public benefits’’ when research is conducted by planning (Bickerstaff and Walker, 2001); waste man- public sector or commercial organisations (Davidson agement (Petts, 2005); and environmental policy et al., 2013), on the whole, this literature has not exam- (Owens et al., 2004). These approaches to public ined how members of the public conceptualise ‘‘public engagement are seen as means of increasing civic benefits’’ or what their expectations are of the ways in engagement and ensuring that plans or policies are rele- which research might bring ‘‘public benefits’’ (Scott vant to affected publics/communities (Coleman and et al., 2018). Gotze, 2010). ‘‘Public benefits’’ are emphasised in research and Deliberative public engagement methods represent governance processes, providing justification for the fora for developing public-spirited reasoning to reach use of health-related and other data generated in a collective agreements (Parkinson, 2004: 379). In the range of ways and promises of beneficial outputs and ideal of public deliberation ‘‘no force except that of outcomes. Many governance mechanisms which con- the better argument is exercised’’ (Habermas, 1975: trol access to population health data require demon- 108). Deliberation should engage with both reasons strable public benefit as an element of the approvals and emotions to enable diverse forms of expression processes before permission is granted for researchers and argumentation and bring about mutual under- to use ‘‘Big Data’’ for health and social research. For standing (Escobar, 2011; Morrell, 2010). ‘‘For the indi- example, projects in Scotland that require access to viduals involved, this process requires an open mind, a population health data have to apply to the Public spirit of reciprocity, and acceptance of the validity of Benefit and Privacy Panel (PBPP) and are required to others’ arguments’’ (Niemeyer, 2011: 105). demonstrate that ‘‘any impact on individual privacy is A well-established approach to deliberative public clearly outweighed by the public benefit resulting from engagement is that of ‘‘mini-publics’’. A mini-public is the processing [of the data]’’ (PBPP, no date: 3). Yet, a deliberative forum bringing together a group of ran- despite being of operational significance, the concept of domly selected members of the public in order to reflect ‘‘public benefit’’ remains under-theorised and little con- the range of demographic and attitudinal characteris- sideration is given to what this means in practice. As tics from the broader population – e.g., age, gender, Scott et al. (2018) have observed ‘‘public benefits’’ are income, opinion (Roberts and Escobar, 2015). As ‘‘variously glossed over as impacts that are ‘in the such, it is intended that the mini-public represents a public interest’, ‘for the greater good’, or ‘leading to microcosm of ‘‘the public’’ (Roberts and Escobar, the improvement of health, education or economic 2015). Engagements with a mini-public involve a mix Aitken et al. 3 of information provision and deliberation (Roberts and of the population and are less likely to be pre-condi- Escobar, 2015). Deliberations typically take place in tioned by having previously taken part in research. small groups and are facilitated by neutral facilitators. Face-to-face recruitment also promotes participation Through this process participants may revise or compared to written recruitment (Tausch and strengthen their initial views both in response to infor- Menold, 2016). Recruitment can be undertaken in the mation and evidence presented as well as from exchan- immediate run up to the events, so drop outs are less ging ideas with other participants. As such it is likely; and, participants have already met people important that the mini-public includes a range of per- involved in the study and participated in an initial on- spectives and viewpoints. street survey, so are more likely to agree to further, Through a deliberative process mini-publics often more intensive participation (Sixsmith et al., 2003). aim to produce a statement, recommendation or Potential participants were given a letter containing a report to inform policy-making processes regarding brief description of what would be involved in the the issue under consideration. Such outputs set out pos- event; the date, timing and location; and a map to the itions reached through consideration of diverse per- venue and were contacted by phone two days before spectives and as such reflect civic-minded positions the event was due to take place in order to confirm rather than private interests. their participation and to enable any drop outs to be replaced. On the day of the event, participants were provided Methods with further information about the study and the ways In order to engage mini publics in characterising public that data would be recorded and used and the potential benefit, a series of three deliberative workshops with outcomes of the research. Participants then completed a members of the public were held across Scotland in consent form. At the end of the workshop participants May and June 2017. The workshops engaged a cross- received £80 in cash as a thank you for their participa- section of the Scottish population in in-depth discus- tion. This amount was chosen to reflect current best sions of the ways that the public – or publics – might practice and a reasonable estimate of the necessary benefit from data-intensive health research. amount of money to ensure participation in a full day Three locations were chosen for these workshops to event taking account of the need to make the event reflect different demographic characteristics. The first accessible to everyone including those with childcare workshop was held in Perth, a small city in a largely requirements or living outside of the city. rural region of central Scotland. The second workshop The workshops were day events beginning at 10 am was held in Glasgow, Scotland’s largest city. The third and ending at 4 pm. The workshops were run in such workshop was held in Aberdeen, a city located in the a way as to ensure a relaxed and informal atmosphere North East of Scotland. This geographic spread to encourage open and fluid conversations. The day enabled the workshops to engage with and reflect the was structured around a mix of facilitated small group interests and perspectives of the Scottish public living in and plenary discussions, brief presentations and videos very different contexts and localities. to stimulate discussions and deliberation. While infor- Following the mini-publics approach, workshop mation provision was important for stimulating participants were recruited using quota sampling to discussions, presentations were kept brief and reflect key demographic characteristics of the local throughout the day the emphasis was on deliberation populations in each location. Quotas were set on and active involvement of participants. Topic guides gender, age, professional background and level of were developed for each of the group discussions but trust in institutions (based on an attitudinal question these were used flexibly to allow conversations to in the sampling questionnaire). We aimed to have 20– follow emerging topics of interest and to enable par- 25 participants in each workshop. In Perth there were ticipants to shape the agenda of the workshops. 25 participants: 15 women and 10 men; in Glasgow 21 Discussions were facilitated from a neutral perspective participants: 11 women and 10 men; and in Aberdeen and participants were encouraged to speak openly and 23 participants: 12 women and 11 men. In all groups freely. The presentations were delivered by members people were aged between their early twenties and late of the research team who also facilitated the discus- seventies and included a mix of those unemployed, sions. The research team consisted of social scientists employed part time and in full-time employment. interested in understanding social and ethical dimen- Participants were recruited in person on the street in sions of data-intensive health research, these interests each of the three cities. This method was chosen since were explained at the outset of the workshop and it it was considered more robust than recruiting from was made clear that the research team did not intend other pre-existing ‘‘panels’’, because participants are to persuade or convince participants of the merits or less self-selecting and so tend to be more representative otherwise of proposed uses of population data but 4 Big Data & Society rather were interested in understanding public views to reflect on the discussions over the course of the around these. day. It was reiterated that we were not aiming to At the beginning of the workshops participants were reach a consensus around these issues but rather asked to agree a set of ground rules for the day. wanted to capture and reflect everyone’s views. Participants suggested potential ground rules which During the workshops all small group discussions were then agreed by the whole group. Across each of were recorded and detailed notes were taken of the plen- the workshops the ground rules created consistently ary discussions. Following the workshops the record- emphasised the importance of allowing everyone a ings of the discussions were typed up by the research chance to speak and respecting differences of opinion. team. A narrative account of each of the workshop It was stressed that we were not necessarily aiming to sessions was developed through assimilating each of reach a consensus about the topics that would be dis- the researchers’ notes from the events along with the cussed but rather we were interested in hearing the transcripts from the recordings. Notes from each ses- range of views people might hold about these. sion were collated and key themes and topics were iden- Following this the day began with small group dis- tified, as well as the identification of nuanced or cussions on the topic of health research very broadly, in divergent perspectives. In reporting the workshop dis- which participants discussed their understandings and cussions in the following sections of this paper, we draw expectations of health research. This was followed by a on transcripts of the small group discussions (which presentation introducing the topic of data-intensive allow verbatim quotes) and detailed notes of the plen- health research and giving an overview of the ways ary discussions (which do not allow verbatim quotes). that data is currently used in health research and the The following sections of this paper will set out the governance systems in place. Small group discussions findings from the workshop discussions and key con- then focussed on data-intensive health research. A siderations which emerged as central for meeting public second presentation set out the importance of ‘‘public expectations concerning public benefits from data- benefits’’ as a driver for data-intensive health research. intensive health research. This was a brief presentation which described the importance given to ‘‘public benefits’’ for determining applications for data access and also previous research Findings identifying ‘‘public benefits’’ as crucial for public Expectations of health research acceptability of data use. The presentation noted the difficulty in clearly defining the concept of ‘‘public In the first small group discussion of the day, workshop benefits’’ and led into facilitated small group discus- participants were asked to describe their expectations of sions around what this term means and what people’s health research and what they understood the term expectations of public benefits would be. ‘‘health research’’ to mean. Participants typically A series of videos were then shown of data scientists responded by describing health research as being discussing research projects which they had either aimed at improving health or health services, and/or recently completed or were planning to undertake. leading to new treatments. Across each of the work- Videos were used instead of face-to-face presentations shops it was evident that health research was expected in part as a result of practical challenges associated with to be aimed at a clear purpose of improving health securing participation of data scientists at each of the outcomes or health services. three workshops and also because this virtual presence Workshop participants described a variety of types encouraged participants to be frank and honest in their of health research, demonstrating awareness of diverse responses. The three data scientists gave an overview of forms that such research can take. The examples given their research, why they considered it important and were offered spontaneously by workshop participants. what the anticipated public benefits of the research While some participants described only research under- would be. Following the videos, workshop participants taken in clinical settings others gave varied examples discussed the extent to which they considered each to including: surveys; research around lifestyles; research have public benefits. in laboratories; and audit of health services. These par- In the final session of the workshops participants ticipants noted that health research can use a variety of were asked to come up with statements to answer two sources of information including donated organs or key questions: (1) Who is the public that should benefit tissue samples, genetic or clinical information as well from data-intensive health research and; (2) In what as information from medical records or surveys. While way should they benefit? Participants were given no some participants did not initially think of this diversity further direction on how they should answer these of health research, once mentioned by others these questions and were not provided with examples or broad approaches and methods were acknowledged as options to choose from, but rather were encouraged comprising health research. Aitken et al. 5 When asked who they would expect to be conduct- that academic research is not necessarily independent of ing health research, in many groups workshop partici- any financial interests: pants initially discussed health research conducted by charities and it was clear that this was often the first M2: It would depend where the university was getting type of research that came to mind. Participants often the funding from. There’s a chance the university could stated that this was due to having an awareness of this be getting their funding from a pharmaceutical research as a result of fundraising campaigns, TV company. adverts or leaflets (for example from Cancer Research F2: Again, that’s trust, isn’t it? You’re trusting those UK). However, a broader range of research organisa- researchers to be ethical with their findings, because tions were also discussed and workshop participants you do get the picture that the money issue is always listed many potential researchers including those breathing down their neck within the NHS, universities, charities or private com- M3: I would think, or I would hope, that universities panies. In most groups the involvement of private com- and that were more neutral, that they don’t have any- panies in health research was a contentious subject. thing to win or lose out of it. They’re looking at the Private companies were described as having a different bigger picture. (Glasgow, Group 3) set of objectives from academic or public sector organ- isations, and in some discussions commercial interests In discussing the different types of researchers who may were conceptualised as being in opposition to public be involved in health research participants acknowl- benefits. However, participants also noted the blurri- edged that there can be a role for commercial organisa- ness of lines between sectors and acknowledged that tions but concerns were raised about funding and the university research can also lead to commercial out- interests that go along with that. Funding of research puts. For example, the following discussion reflected was an area of interest in discussions across the work- one group’s shifting position as they considered the shops. It was acknowledged that research could be sup- potential involvement of private companies in health ported by either public or private funding. Some research: participants expressed mistrust of private research fun- ders, for example a large number of participants con- F2: the right people would be people in the NHS, in the tended that private companies would suppress research medical profession, because they’re not in it for the results if it enabled them to increase their profits. In money. In my point of view, I don’t want people to particular, a number of participants stated that they profit from it, from the data believed that a ‘‘cure for cancer’’ may have already M1: no been found but that pharmaceutical companies are M3: but if they could cure someone holding on to it to make the most money: F2: but if they could cure someone and then charge them massively F1: People always say there’s no cure for the common M1: no financial gain cold – so what chance do we have of finding cures for F2: I don’t want financial gain, yeah cancer if there’s no research into it MA: what if a university research group collaborated M1: some people say they have got a cure for cancer with a pharmaceutical company? but they’re holding it back to sell to the highest bidder F2: Sometimes I would imagine that has to happen F1: Conspiracy theories M3: for new drugs M1: There’s truth in it MA: in that situation there may be some profit that F2: Everybody thinks that, the higher authorities are all goes with it in it for what they can make out of it F1: yeah, there has to be some profit M2: are they withholding a cure? M3: without profit there won’t be any progress F2: unfortunately not everyone has a heart, they don’t F1: you need the companies involved – it’s a necessary care enough. (Perth, Group 1) evil F2: so long as there is good with it M2: I believe that there is research that has been done F1: more good than bad that has proven to actually help benefit certain illness F2: it’s about the balance and whatnot but nothing will be done about it because M3: It’s never going to be 100 per cent the pharmaceutical companies are making far too much F2: some people are only going to be in it for the money money from selling other products to help combat these M1: that’s just the way it is. (Perth, Group 1) things. M4: I think also the problem is that you have huge com- Similarly when asked about their expectations of uni- panies that are making billions and billions of pounds and versity researchers, participants typically acknowledged that obviously promotes distrust because they’re making 6 Big Data & Society so much money. But at the same time, you need that level NHS and everything. Everybody makes money from it of funding to be able to produce the advances. An exam- so, why would they? They don’t care about the little ple would be the Ebola crisis last year, they did ten years’ person. (Aberdeen, Group 3) worth of research in nine months because the WHO just piled billions into it, and they need that funding level, and This related to wider discussions around the ways in they got a cure out of it. But then at the same time, which benefits of health research are realised and a because they’re so profitable, it’s not necessarily ethical. widely held perception that currently the benefits are (Glasgow, Group 3) not realised equitably across society and that different groups or people in different locations across the UK There was concern that research conducted or funded experience health services differently as well as experi- by private companies would be directed at maximising encing different health outcomes. profits rather than maximising public benefits. For Throughout these discussions a recurring theme was example, one participant in Aberdeen stated: that the potential benefits of health research were not always or consistently realised. A range of factors were The money in research is directed at money-making noted as limiting the realisation of public benefits from things. Like, weight loss gets a lot of money put into health research, these included commercial interests, it, more than malaria, which affects people a lot more political priorities and limited public funding. than being overweight, which is generally, or can be, a cosmetic thing. It’s not down to illness. Not always – Who is the public that should benefit from sometimes it is – but, malaria, if they found a cure for data-intensive health research? that, billions of people, that would be their life improved. So ... where the money is they’ll research. When posed the question ‘‘who is the public?’’, work- (F1, Aberdeen, Group 3) shop participants typically treated this as a straightfor- ward and obvious question. Across all of the Furthermore, there was also discussion of the potential workshops participants almost always initially for governments to be involved in suppressing discov- responded by stating that: ‘‘the public is everyone’’. eries of new cures if they are too expensive or not However, these responses quickly became more aligned with government priorities: nuanced and complex when discussions moved to think about what this meant in practice and, in particu- M4: I often wonder if they want to find a cure for it. All lar about the ways in which ‘‘the public’’ could benefit the pharmaceutical companies they’re making millions from data-intensive health research. Two key consider- out of it. If they find a cure for it, what are they going ations emerged as important in conceptualising ‘‘the to do now? public’’. These were scale and need. M1: You could say the same about the government – Firstly, there were discussions around scale with does the government really want to find cures? Because, workshop participants expressing a range of views on at the end of the day, the longer people live, the more it whether the public should be conceptualised as every- costs to give them everything: pensions, health service, one globally, nationally or locally. and what have you. So, do they want us to be living to 100 years old? Probably not. M3: It’s us – all of us. Everyone in the world. F2: You lot are a load of doom and gloom! F4: It’s society. As in, what stands below the govern- F3: Five minutes in and we’re into the conspiracy the- ment. The individual who goes to work every day, ories! [laughter] doesn’t put a suit on, doesn’t sit and make a decision M4: They’ve been researching for maybe twenty years, on what our lives are going to be doing. (Perth, you’d think somebody would’ve found a breakthrough Group 3) M2: With all the money that’s been ploughed into it. M4: Do they really want to find a cure for it? F2: Presumably that’s just us, Joe Bloggs. F3: Why would they waste money investing in trying to F3: Everyone find it in the first place? F2: We’re the public, are we not? M1: Cos they’ve got to look like they’re trying to find it EC: Are we? [...] When you think about ‘‘the public’’, F3: No, they are trying to ... I’ve never even thought who’s in your mind? that they weren’t trying to. They’re doing all the F2: Myself research and everything, they’re trying to find cures M1: Working class people. [...] F5: The people that the research affects. M2: They make money by giving drugs. The pharma- M2: Or benefits. cies make money by getting the prescriptions from the F3: I just think everyone. Aitken et al. 7 EC: In Scotland? participant stated that the public should benefit ‘‘in F3: Just everyone. every way they are entitled to’’ and another stated EC: In the world? that the public should benefit in ‘‘the most beneficial F1: In the whole wide world. way possible’’. These responses highlight workshop F2: It depends what the research is about. If it’s merely participants’ recognition of the wide range of public based on Scotland, it would be the people of Scotland, I benefits that could be possible. Many participants would think. People of voting age. (Perth, Group 2) were clearly avoiding giving answers which might limit the ways that benefits were pursued and demon- There was generally a preference that ‘‘the public’’ strated a clear preference that benefits should be max- should be considered as inclusively as possible, meaning imised across society in whatever form they take. that there was a preference for research to benefit the Given that workshop participants had previously widest possible public (i.e., benefit the maximum described the purpose of health research as being pri- number of people). One workshop participant even marily about leading to improvements in health and/or stated that we should think of the public as ‘‘everyone health services, it is unsurprising that many responses in the universe’’ – this was stated only partially as a joke related to this theme. Many participants suggested that and demonstrated a strong preference to be as inclusive finding cures for diseases and making new drugs avail- as possible when thinking about ‘‘the public’’. able was a clear way in which members of the public However, the second key consideration – need – led should benefit. In particular, cancer, dementia and to different sets of preferences. Workshop participants mental health were often flagged as pressing conditions acknowledged that the findings and outcomes of about which more research was needed. More broadly research can have greater or lesser potential impact than just cures and increased medicalization, consider- for different groups within the public. For example, it able time in the discussions focused upon improving was acknowledged that health research could – and in lives; with a focus on health improvement, better qual- many instances should – have the greatest benefits for ity of life and enhanced lifestyle, with people living people with major health issues or groups affected by longer and healthier lives and lives that are less stress- particular conditions related to the subject of the ful. Linked to better quality of life and outcomes, par- research. This led to recognition of the value of ticipants suggested that future generations should be health research being targeted at vulnerable groups thought about so they do not face similar health and who would potentially benefit the most. In this sense lifestyle burdens, with better understanding and imple- public benefits were conceptualised as being benefits to mentation of preventative measures. Participants also particular smaller groups within the public rather than stated that benefits of research should be measurable, benefits to the wider public as a whole. However, it was through better quality care and services. Improved allo- also noted that while the immediate benefits of research cation of resources was also a way in which participants might be targeted at those considered to be in greatest thought the public could benefit. need, in the long-term this would be of benefit to every- Public benefits were also conceptualised in less direct one in society, suggesting that the two positions were ways. For example, participants considered that there far from incompatible. can be ‘‘knock-on effects’’ of small numbers of people benefiting from health research. Improving health and/ M1: I think it’s everybody because even if you’re not or quality of life of vulnerable groups was expected to aware of it or affected by it, that doesn’t mean it’s not lead to wider benefits for society. Therefore, participants of relevance to you. thought that there should be proactiveness to address F3: Just because you’re not associated with it at the particular needs of vulnerable communities to ensure time it doesn’t mean it won’t impact you later on in that communities, and society as a whole, can lead your life. (Perth, Group 3) better quality lives. This was also described as requiring empowerment of individuals and communities. The public were also described as benefitting from research through the creation and dissemination of new In what ways should members of the public knowledge. Workshop participants suggested that benefit? increasing scientific knowledge is in and of itself a bene- fit and that members of the public could benefit from When asked in what ways the public should benefit from data-intensive health research, a very broad set greater engagement with the scientific community. This of responses was given, demonstrating wide-ranging led to discussions of the need for science to do more to conceptualisations of ‘‘public benefits’’. Responses indi- publicise research results, and, in particular, to engage cated that workshop participants considered that with the public. However, it was frequently stressed public benefits should be as broad as possible. One that it is not sufficient for research to create new 8 Big Data & Society knowledge or understanding, rather it is considered diverse and varied forms that public benefits might essential that policy makers, governments and/or the take, as well as the diverse forms that research itself health service act on research findings in order to realise might take. However, there were some clear patterns the potential public benefits. regarding which projects were regarded most favour- ably. Voting preferences generally reflected the close- M5: The point is, it’s about what the outcome of the ness or relevance of research subject matter to research is, because if it says people coming from individuals’ lives. For example, older participants poorer backgrounds or deprived areas [are more at tended to vote for Project A which related to admis- risk of health conditions] .. . is anyone actually going sion to care homes. At all three of the workshops to spend any money to change it? Or is just a case of– Project C (which related to interception of illegal F2: ‘‘This is the research, we’ve done it, that’s it’’. drugs) consistently received the fewest votes yet was M4: The findings can be there but nothing’s actually also always the project which generated the greatest done. (Aberdeen, Group 3) discussion. The key discussion points which emerged unprompted in group deliberations related to: who These discussions around the ways in which the would benefit from the project; which project would public might benefit from data-intensive health have the greatest impact; and to what extent work- research revealed openness to the possible forms shop participants expected that the potential benefits public benefits could take and an unwillingness to would be realised. set limits or constraints on what public benefits In discussing their reasons for voting for particular might be. However, there was considerably more projects, participants considered which project would interest in the likelihood of the benefits being realised. be likely to have the greatest impact. This was discussed Reflecting earlier discussions around perceived bar- both in terms of which projects would benefit the most riers to realising benefits of health research, partici- people and which would have the biggest benefit for pants across all workshops were less concerned with those in greatest need – the two key ways in which what the benefits would be compared with assurances these ‘‘mini publics’’ characterised public benefit. that any benefits would be realised. While noting that Project A focussing on care homes was seen to be knowledge itself brought benefits, the participants very relevant in the context of an aging population wanted reassurance that further public benefits with care homes expected to be of increasing relevance could ensue through applying knowledge to improve in years to come. Additionally this project was con- lives. sidered important since it related to benefitting vulner- able people. Responses to three research projects M2: [Care] is putting a huge drain on resources at the The preceding sections have focussed on workshop moment. We’re living in an older society. participants’ discussions of public benefits from data- F3: People are living a lot longer. These are people who intensive health research at a general level. The work- have paid into the pot for so many years and then shops moved on to encourage participants to consider they’re just getting left... need to put more resources the ways in which particular research projects might into it. (Perth, Group 3) lead to public benefits. Workshop participants were shown videos of three Similarly, Project B which focussed on air pollution was data scientists who presented their research and the considered to potentially benefit the whole of society as ways in which they considered it would lead to public well as the environment and non-human life. Since this benefits. The three research projects presented were project was focussing on impacts on babies, many chosen to reflect very different subject matter. Project workshop participants regarded it as potentially A related to admissions to care homes; Project B having the biggest public benefits as it was starting related to impacts of air pollution on babies; and ‘‘right at the beginning of life’’. This was a point that Project C related to evaluating impacts of one particu- was made in all workshops. lar illegal drug interception policy. As a catalyst for discussions, after viewing the F3: For me, if [Project B] was just air pollution, without videos workshop participants were asked to vote on the pregnancy component, I probably would have gone which project they considered to have the greatest for care homes [Project A]. But because his research is potential public benefits. There was widespread agree- mixed with the pregnancy, as a mum, I want to give my ment that each of the three research projects had – at children the best start in life. least potential – public benefits. This reflects the find- F2: Anything to do with pregnancy is important – it’s ing discussed above that participants recognised the the start in life. (Aberdeen, Group 3) Aitken et al. 9 Participants noted that having a poor start in life can While the subject matter and aims of the project were affect the whole of the rest of life. widely considered to be of high importance, workshop participants were not confident that this project would F3: I voted [for Project B], we all know the risks about be able to make a significant difference. This highlights smoking in pregnancy and now its became unacceptable the importance of demonstrating mechanisms being in that women should be smoking because the harm it does place to enable impact. to the baby and now we know that pollution is likely to Drug use is a more controversial subject compared be just as bad, no one seems to be bothered about it and with care homes or air pollution so it is not we know the effects it can have on babies, more pollution surprising that Project C generated the most discussion. smaller babies then it would force them to do something Given that workshop participants typically acknowl- about it. That would prevent poorer health in childhood edged that their preferences for particular projects and all the things that go along with being premature were influenced by the extent to which they were per- and all the things that go along with it. ceived as having relevance to their own lives, as well as F4: I voted for [Project B] too and I did that because it society more generally, the potential stigma of drug use is the start of life, a baby being born, because it affects and what they saw as its social complexity, might have your whole life. I actually liked all of them. I worked in had some influence in shaping responses. It is note- a dementia unit and I know how it affects everyone, the worthy that one workshop participant who spoke whole family, but the baby one was trying to do some- about the ways in which his own life and that of his thing at the start of life. (Perth, Group 3) friends and family had been affected by drug addiction made an emotional and passionate plea to fellow par- Conversely, workshop participants were more sceptical ticipants to give Project C greater consideration. of the extent to which Project C (focussing on drugs The deliberations regarding the three projects illus- enforcement) would have a big impact. Some partici- trate the findings reported in preceding sections of this pants described this as ‘‘fighting a losing battle’’ and paper. The voting exercise was intended as catalyst for said that ‘‘drugs are going to be used no matter what’’. discussions rather than a priority setting exercise and Across the workshops drug use was recognised to be an workshop participants typically stressed that they area of major, but intractable, concern with significant appreciated the potential benefits of all three projects. implications for society. Drugs were described as a big Not voting for a project did not mean that a participant problem to be tackled, but also a complex social issue. did not believe it could bring public benefits. As such Much of the discussion about Project C related to the participants recognised and acknowledged the very dif- extent to which this research would actually have an ferent types of public benefits which could come from impact in addressing problems of drug use: it was the three projects. However, the discussions consist- described as reaching ‘‘just the tip of the iceberg’’. ently focussed on key considerations (which were not This meant that people were less inclined to vote for prompted by the facilitators): which project would be this project in comparison to the others, although as the likely to have the biggest impact, and whether partici- dialogue below illustrates, people had different kinds of pants had confidence that potential public benefits reservations. would be realised. These emerged as key themes throughout discussions at all three of the workshops. M1: I didn’t vote for it, not because I didn’t think there would be any benefit but we have been talking about Discussion health data and I just found it much harder to directly correlate, obviously we are talking about drugs and not Each of the three workshops resulted in open and wide- necessarily everyone taking these drugs will be taking ranging discussions leading to nuanced and complex them to a degree which significantly affect their health, positions regarding the ways that data-intensive you might get the people who take them now and health research might bring public benefits. The work- again. I just thought it would be very difficult to state shop discussions demonstrated that ‘‘public benefits’’ a clear public benefit ... were conceptualised in a number of different ways F1: with that one it’s hard to see how you could do and participants had varied expectations of what this something directly with the findings, I mean maybe fur- implied for actual research practice. Nonetheless, there ther down the line, but straight away, it would be dif- was a clear consensus that public benefits should be to ficult to get a clear public benefit from. individuals, specific groups and to society more widely. M5: I can see the clear public benefit from that but it It is interesting to note that no one spoke of societal looks like a hard fight, it’s just gonna take a long time, I benefits in terms of economic benefit, although this is just felt the other one, for me, was more beneficial. often portrayed as a form of public or societal benefit (Perth, Group 3) by governments and funding agencies. For example, 10 Big Data & Society The European Commission Strategy Centre recently maintaining relationships between research and called data ‘‘the lifeblood of the global economy’’ policy – and more broadly science and society. There (European Commission, 2017), whereas benefits to the was consensus that there needs to be commitment and economy through new data-driven industries were willingness by all stakeholders involved in the research never mentioned by workshop participants. Where eco- process to implement findings and maximise public nomic dimensions were raised, these related to potential benefits. Workshop participants were much less con- impacts on individuals or groups within society. For cerned with what form public benefits would take, com- example, when discussing care homes and the pressures pared with seeking assurances that potential benefits of an aging population, workshop participants were would be realised. Addressing this area of concern typically concerned with the economic impact on indi- may be vital for securing public support and establish- viduals through a pressured care system and how that ing a social license for future health-related data science affected health and wellbeing. In considering how best and demands transparency regarding how impact can to generate and sustain a social license for health- be achieved. related data science, funders and researchers may need to more clearly articulate how economic benefits Study limitations would result in wider social impact. Workshop participants had a broad and inclusive The research was undertaken following a ‘‘mini- approach to conceptualising both the public and publics’’ approach which aimed to engage with a public benefit. They did not perceive a conflict between cross-section of the Scottish population. Participants desiring wide benefit for society as a whole and promot- were sampled to represent demographic characteristics ing the needs of particular groups. Indeed, they recog- broadly representative of local populations in each of nised the role of indirect public benefits – those the locations. Nevertheless, the total number of partici- accruing to society when the pressing needs of some pants in the study was 69 and as such is not sufficient to are met. It was generally acknowledged that targeting claim that the sample is representative of the wider benefits for those in need was appropriate and neces- Scottish population. Deliberative public engagement sary. This resonates with a recent study which exam- methods are effective for enabling in-depth examination ined public sector professionals’ expectations of public of participants’ viewpoints and rationales but are time benefits and found that when evaluating the relative and resource-intensive which constrained the scope of value of different public benefits some focussed on the this study. The aim of this qualitative, deliberative number of people benefitting, others focussed on the research was not to produce a set of findings which level of need of those benefitting and, for others it could be generalised or be considered representative. was the potential for long-term impacts (Scott et al., Rather it intended to explore the ways in which mem- 2018). bers of the public would engage with this important Workshop participants were unwilling to narrowly subject and examine the range of views that would be define or constrain public benefits and preferred to keep expressed. In that context, it is noteworthy that the this definition open recognising the very many forms findings produced were broadly consistent both within public benefits could take. They were more concerned and between each of the workshops. with the likelihood that benefits would be realised – The responses reported throughout this paper were that research would make a difference – a theme con- generated through discussion, with few prompts from sistent across all workshops. The discussions high- the facilitators. The research team did not set out to lighted the need for action in response to research examine public views on any particular potential types findings, suggesting that ensuring impact itself was an of benefits but rather to explore how members of the important component of achieving public benefit. They public would conceptualise public benefits and what recognised the wider processes involved in this, beyond their expectations are. Future research might usefully the research community itself to include policy makers engage the public in discussions of particular potential and practice communities. Across all the workshops outcomes and research aims. there was widespread agreement that, currently, the benefits of health research are impeded by commercial Conclusions or political interests. There was considerable scepticism of the extent to which necessary mechanisms and pol- This research has demonstrated the value of engaging itical support are in place to enable research to lead to with members of the public to address conceptually, public benefits. This indicates that workshop partici- socially and ethically challenging issues related to cur- pants recognised that the pathways between research rent and future health-related data science practices. and impact are not straightforward or inevitable and The social licence for the research use of health and reflects widespread awareness of the importance of other ‘‘Big Data’’ needs to be continuously negotiated Aitken et al. 11 Bickerstaff K and Walker G (2001) Participatory local gov- as public expectations, preferences and views shift in ernance and transport planning. Environment and Planning response to changing contexts and experiences. The A 33: 431–451. boundaries of acceptability will need to be renegotiated Carter P, Laurie GT and Dixon-Woods M (2015) The social on an ongoing basis to ensure that actual practices align licence for research: Why care.data ran into trouble. with public values and concerns. This may be of par- Journal of Medical Ethics 41: 404–409. ticular importance when considering economic benefits, Coleman S and Gotze J (2010) Bowling together: Online particularly but not only with respect to commercial public engagement in policy deliberation. Available at: involvement in research (Davidson et al., 2013; www.hansardsociety.org.uk (accessed 21 November 2018). Wellcome Trust, 2016). Davidson S, McLean C, Treanor S, et al. (2013) Public The nuanced discussions at each of these workshops acceptability of data sharing between the public, private have demonstrated the value of bringing together and third sectors for research purposes. Scottish diverse groups to discuss conceptually challenging Government Social Research. Available at: https:// topics enabling considered and informed opinions to www2.gov.scot/resource/0043/00435458.pdf. be developed. As noted by Scott et al. (2018: 32) Escobar O (2011) Public Dialogue and Deliberation: A Communication Perspective for Public Engagement ‘‘there is a need for further direct engagement with Practitioners. Edinburgh: Edinburgh Beltane – UK the public to better understand where their boundaries Beacons for Public Engagement. Available at: www.belta- of acceptability lie within the context of data sharing’’. nenetwork.org/resources/beltane-publications (accessed 21 This study reaffirms much of the literature concerning November 2018). deliberative public engagement methods, in demon- European Commission (2017) Enter the data economy: EU strating the value of these methods for informing and policies for a thriving data ecosystem. (21): 16. Available examining public attitudes around complex subject at: https://ec.europa.eu/epsc/publications/strategic-notes/ matter. enter-data-economy_en. Garrety K, McLoughlin I, Wilson R, et al. (2014) National electronic health records and the digital disruption of Declaration of conflicting interests moral orders. Social Science & Medicine 101: 70–77. The author(s) declared no potential conflicts of interest with Habermas J (1975) Legitimation Crisis. Cambridge: Beacon respect to the research, authorship, and/or publication of this Press. article. Halseth G and Booth A (2003) ‘‘What works well; What needs improvement’’: Lessons in public consultation from British Columbia’s resource planning processes. Funding Local Environment 8(4): 437–455. This research was funded by The Farr Institute of Health Hodson H (2016) Revealed: Google AI has access to huge Informatics Research. The UK HIRN is supported by a 10- haul of NHS patient data. New Scientist Available at: funder consortium: Arthritis Research UK, the British Heart www.newscientist.com/article/2086454-revealed-google-ai- Foundation, Cancer Research UK, the Economic and Social has-access-to-huge-haul-of-nhs-patient-data/ (accessed 20 Research Council, the Engineering and Physical Sciences March 2018). Research Council, the Medical Research Council, the King G, Heaney DJ, Boddy D, et al. (2010) Exploring public National Institute of Health Research, the National perspectives on E-health: Findings from two citizen juries. Institute for Social Care and Health Research (Welsh Health Expectations 14(4): 351–360. Assembly Government), the Chief Scientist Office (Scottish McGrail K, et al. (2018) A position statement on population Government Health Directorates), the Wellcome Trust, data science. International Journal for Population Data (MRC grant no: MR/M501633/2). Science 3(1): 4. Mitton C, Smith N, Peacock S, et al. (2009) Public participa- Ethical approval tion in health care priority setting: A scoping review. Health Policy 91(3): 219–228. This research was granted ethical approval by the Usher Morrell ME (2010) Empathy and Democracy: Feeling, Institute Ethics Committee. Thinking, and Deliberation. University Park: Pennsylvania State University Press. References Niemeyer S (2011) The emancipatory effect of deliberation: Empirical lessons from mini-publics. Politics & Society Aitken M, Jorre JDS, Pagliari C, et al. (2016) Public 39(1): 103–140. responses to the sharing and linkage of health data for Owens S, Rayner T and Bina O (2004) New agendas for research purposes: A systematic review and thematic syn- appraisal: Reflections on theory, practice, and research. thesis of qualitative studies. BMC Medical Ethics 17(1): 73. Environment and Planning A 36: 1943–1959. Aitken M, Tully MP, Porteous C, et al. (in press) Consensus Parkinson J (2004) Why deliberate? The encounter between Statement on Public Involvement and Engagement with Data-Intensive Health Research. International Journal of deliberation and new public managers. Public Population Data Science. Administration 82: 377–395. 12 Big Data & Society Petts J (2005) Enhancing environmental equity through deci- wind farms in Scotland. Available at: www.climatex- sion-making: Learning from waste management. Local change.org.uk/media/1438/citizens_juries_-_full_report. Environment 10(4): 397–409. pdf (accessed 21 November 2018). Pidgeon N and Rogers-Hayden T (2007) Opening up nano- Scott K, Burall S, Perrin H, et al. (2018) Data for public technology dialogue with the publics: Risk communication benefit: Balancing the risks and benefits of data sharing. or ‘upstream engagement’? Health, Risk & Society 9(2): Available at: http://understandingpatientdata.org.uk/ 191–210. news/data-public-benefit (accessed 21 November 2018). Public Benefit and Privacy Panel (PBPP) (no date) Guiding Sixsmith J, Boneham M and Goldring JE (2003) Accessing principles and policy for decision making. Available at: the community: Gaining insider perspective from the out- www.informationgovernance.scot.nhs.uk/pbpphsc/wp- side. Qualitative Health Research 13(4): 578–589. content/uploads/sites/2/2015/04/Guiding-Principles.docx Tausch AP and Menold N (2016) Methodological aspects of (accessed 8 June 2018). focus groups in health research. Global Qualitative Nursing Ramesh R (2014) NHS patient data to be made available Research 3: 1–12. for sale to drug and insurance firms. The Guardian, Wellcome Trust (2015) Enabling data linkage to maximise the 19 January. Available at: https://www.theguardian.com/ value of public health research data: Full report. Available society/2014/jan/19/nhs-patient-data-available-compa- at: www.wellcome.ac.uk (accessed 21 November 2018). nies-buy. Wellcome Trust (2016) The one-way mirror: Public attitudes Roberts J and Escobar O (2015) Involving communities in to commercial access to health data. Available at: www. deliberation: A study of 3 citizens’ juries on onshore wellcome.ac.uk (accessed 21 November 2018). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Big Data & Society SAGE

Who benefits and how? Public expectations of public benefits from data-intensive health research:

Loading next page...
 
/lp/sage/who-benefits-and-how-public-expectations-of-public-benefits-from-data-XqtuXkw0kq

References (30)

Publisher
SAGE
Copyright
Copyright © 2022 by SAGE Publications Ltd, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses.
ISSN
2053-9517
eISSN
2053-9517
DOI
10.1177/2053951718816724
Publisher site
See Article on Publisher Site

Abstract

The digitization of society and academic research endeavours have led to an explosion of interest in the potential uses of population data in research. Alongside this, increasing attention is focussing on the conditions necessary for maintaining a social license for research practices. Previous research has pointed to the importance of demonstrating ‘‘public benefits’’ from research for maintaining public support, yet there has been very little consideration of what the term ‘‘public benefits’’ means or what public expectations of ‘‘public benefits’’ are. In order to address this pressing issue a series of deliberative workshops with members of the public were held across Scotland in May and June 2017. The workshops aimed to engage a cross-section of the Scottish population in in-depth discussions of the ways that the public – or publics – might benefit from data-intensive health research. The findings reported here discuss workshop participants’ understandings and expectations of health research; who they considered to be ‘‘the public’’ that should benefit from health research and; in what ways they felt ‘‘the public’’ should benefit. Workshop participants’ preference was clearly for the widest possible public benefit to be felt by all, but they also acknowledged the value in research aiming to primarily benefit vulnerable groups within society. A key focus of discussions was the extent to which workshop participants were confident that potential public benefits would be realised. A crucial consideration then is the extent to which mechan- isms and political support are in place to realise and maximise the public benefits of data-intensive health research. Keywords Public benefits, data science, health informatics, public engagement programme had significant impacts and resulted in the Background programme being put on hold. Such controversies have The digitization of society and academic research drawn attention to the importance of engaging with endeavours have led to an explosion of interest in the members of the public to ensure that data is used in potential uses of population data in research (McGrail ways which align with public values and interests to et al., 2018); this is particularly true in relation to health ensure that public concerns are adequately addressed. research (e.g., Aitken et al., in press; Wellcome Trust, This is essential for developing and maintaining a 2015). However, recent years have also brought a number of public controversies, particularly regarding proposed uses of health data (e.g., Carter et al., 2015; Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, UK Garrety et al., 2014). Two high-profile examples from the UK are the failed introduction of the care.data Corresponding author: scheme to link hospital and GP records (Ramesh, Mhairi Aitken, Usher Institute of Population Health Sciences and 2014) and Google Deep Minds’ involvement in process- Informatics, University of Edinburgh, Old Medical School, Teviot Place, ing health data at an NHS Trust in London (Hodson, Edinburgh EH8 9AG, UK. 2016). In the case of care.data, public opposition to the Email: mhairi.aitken@ed.ac.uk Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http:// www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society social licence for current and future practices (Carter and social well-being’’’ (Scott et al., 2018: 21). Given et al., 2015). the importance of ‘‘public benefits’’ for public accept- The growing interest in potential uses of population ability of data uses, and its central role in determining data, and the increasing recognition of the importance which researchers get access to data for what purposes, of ensuring a social licence for these uses, has resulted it is essential that a better understanding of what the in considerable interest in understanding public atti- term means is generated in a way that is grounded in tudes and views on these topics (e.g., Aitken et al., public views. This paper aims to provide a first step in 2016; Davidson et al., 2013; Wellcome Trust, 2016). developing such an understanding. It draws on the Accordingly there is a fast-growing body of inter- results of three deliberative workshops with a broad national literature reporting findings of studies examin- range of publics, organised as part of a programme of ing public attitudes towards data and health-related public engagement research around data linkage and data science practices (Aitken et al., 2016). A consistent sharing for health research. The workshops focussed theme to emerge through this literature is the import- on the meanings and expectations publics held and ance of demonstrating ‘‘public benefits’’ from research formed regarding such research use and the paper using population data (e.g., Davidson et al., 2013). maintains that such formulations should inform Much of the qualitative research undertaken has dis- research policy and governance in health-related data cussed perceived ‘‘trade-offs’’ between risks to individ- science. uals (e.g., loss of privacy) and broader public benefits of research. In a systematic review of literature on public Deliberative methods and mini-publics attitudes towards data linkage for health research it was found that ‘‘assurances that research would bring about Deliberative forms of public engagement have been public benefits – or at least that it had the potential to used to examine and address public interests in a wide bring about such benefits – were widely reported to be range of policy areas including: health (Mitton et al., fundamental for ensuring public support or accept- 2009); e-health (King et al., 2010); nanotechnology ance’’ (Aitken et al., 2016). While previous research (Pidgeon and Rodgers-Hayden, 2007); natural resource has revealed different sets of expectations around management (Halseth and Booth, 2003); transport ‘‘public benefits’’ when research is conducted by planning (Bickerstaff and Walker, 2001); waste man- public sector or commercial organisations (Davidson agement (Petts, 2005); and environmental policy et al., 2013), on the whole, this literature has not exam- (Owens et al., 2004). These approaches to public ined how members of the public conceptualise ‘‘public engagement are seen as means of increasing civic benefits’’ or what their expectations are of the ways in engagement and ensuring that plans or policies are rele- which research might bring ‘‘public benefits’’ (Scott vant to affected publics/communities (Coleman and et al., 2018). Gotze, 2010). ‘‘Public benefits’’ are emphasised in research and Deliberative public engagement methods represent governance processes, providing justification for the fora for developing public-spirited reasoning to reach use of health-related and other data generated in a collective agreements (Parkinson, 2004: 379). In the range of ways and promises of beneficial outputs and ideal of public deliberation ‘‘no force except that of outcomes. Many governance mechanisms which con- the better argument is exercised’’ (Habermas, 1975: trol access to population health data require demon- 108). Deliberation should engage with both reasons strable public benefit as an element of the approvals and emotions to enable diverse forms of expression processes before permission is granted for researchers and argumentation and bring about mutual under- to use ‘‘Big Data’’ for health and social research. For standing (Escobar, 2011; Morrell, 2010). ‘‘For the indi- example, projects in Scotland that require access to viduals involved, this process requires an open mind, a population health data have to apply to the Public spirit of reciprocity, and acceptance of the validity of Benefit and Privacy Panel (PBPP) and are required to others’ arguments’’ (Niemeyer, 2011: 105). demonstrate that ‘‘any impact on individual privacy is A well-established approach to deliberative public clearly outweighed by the public benefit resulting from engagement is that of ‘‘mini-publics’’. A mini-public is the processing [of the data]’’ (PBPP, no date: 3). Yet, a deliberative forum bringing together a group of ran- despite being of operational significance, the concept of domly selected members of the public in order to reflect ‘‘public benefit’’ remains under-theorised and little con- the range of demographic and attitudinal characteris- sideration is given to what this means in practice. As tics from the broader population – e.g., age, gender, Scott et al. (2018) have observed ‘‘public benefits’’ are income, opinion (Roberts and Escobar, 2015). As ‘‘variously glossed over as impacts that are ‘in the such, it is intended that the mini-public represents a public interest’, ‘for the greater good’, or ‘leading to microcosm of ‘‘the public’’ (Roberts and Escobar, the improvement of health, education or economic 2015). Engagements with a mini-public involve a mix Aitken et al. 3 of information provision and deliberation (Roberts and of the population and are less likely to be pre-condi- Escobar, 2015). Deliberations typically take place in tioned by having previously taken part in research. small groups and are facilitated by neutral facilitators. Face-to-face recruitment also promotes participation Through this process participants may revise or compared to written recruitment (Tausch and strengthen their initial views both in response to infor- Menold, 2016). Recruitment can be undertaken in the mation and evidence presented as well as from exchan- immediate run up to the events, so drop outs are less ging ideas with other participants. As such it is likely; and, participants have already met people important that the mini-public includes a range of per- involved in the study and participated in an initial on- spectives and viewpoints. street survey, so are more likely to agree to further, Through a deliberative process mini-publics often more intensive participation (Sixsmith et al., 2003). aim to produce a statement, recommendation or Potential participants were given a letter containing a report to inform policy-making processes regarding brief description of what would be involved in the the issue under consideration. Such outputs set out pos- event; the date, timing and location; and a map to the itions reached through consideration of diverse per- venue and were contacted by phone two days before spectives and as such reflect civic-minded positions the event was due to take place in order to confirm rather than private interests. their participation and to enable any drop outs to be replaced. On the day of the event, participants were provided Methods with further information about the study and the ways In order to engage mini publics in characterising public that data would be recorded and used and the potential benefit, a series of three deliberative workshops with outcomes of the research. Participants then completed a members of the public were held across Scotland in consent form. At the end of the workshop participants May and June 2017. The workshops engaged a cross- received £80 in cash as a thank you for their participa- section of the Scottish population in in-depth discus- tion. This amount was chosen to reflect current best sions of the ways that the public – or publics – might practice and a reasonable estimate of the necessary benefit from data-intensive health research. amount of money to ensure participation in a full day Three locations were chosen for these workshops to event taking account of the need to make the event reflect different demographic characteristics. The first accessible to everyone including those with childcare workshop was held in Perth, a small city in a largely requirements or living outside of the city. rural region of central Scotland. The second workshop The workshops were day events beginning at 10 am was held in Glasgow, Scotland’s largest city. The third and ending at 4 pm. The workshops were run in such workshop was held in Aberdeen, a city located in the a way as to ensure a relaxed and informal atmosphere North East of Scotland. This geographic spread to encourage open and fluid conversations. The day enabled the workshops to engage with and reflect the was structured around a mix of facilitated small group interests and perspectives of the Scottish public living in and plenary discussions, brief presentations and videos very different contexts and localities. to stimulate discussions and deliberation. While infor- Following the mini-publics approach, workshop mation provision was important for stimulating participants were recruited using quota sampling to discussions, presentations were kept brief and reflect key demographic characteristics of the local throughout the day the emphasis was on deliberation populations in each location. Quotas were set on and active involvement of participants. Topic guides gender, age, professional background and level of were developed for each of the group discussions but trust in institutions (based on an attitudinal question these were used flexibly to allow conversations to in the sampling questionnaire). We aimed to have 20– follow emerging topics of interest and to enable par- 25 participants in each workshop. In Perth there were ticipants to shape the agenda of the workshops. 25 participants: 15 women and 10 men; in Glasgow 21 Discussions were facilitated from a neutral perspective participants: 11 women and 10 men; and in Aberdeen and participants were encouraged to speak openly and 23 participants: 12 women and 11 men. In all groups freely. The presentations were delivered by members people were aged between their early twenties and late of the research team who also facilitated the discus- seventies and included a mix of those unemployed, sions. The research team consisted of social scientists employed part time and in full-time employment. interested in understanding social and ethical dimen- Participants were recruited in person on the street in sions of data-intensive health research, these interests each of the three cities. This method was chosen since were explained at the outset of the workshop and it it was considered more robust than recruiting from was made clear that the research team did not intend other pre-existing ‘‘panels’’, because participants are to persuade or convince participants of the merits or less self-selecting and so tend to be more representative otherwise of proposed uses of population data but 4 Big Data & Society rather were interested in understanding public views to reflect on the discussions over the course of the around these. day. It was reiterated that we were not aiming to At the beginning of the workshops participants were reach a consensus around these issues but rather asked to agree a set of ground rules for the day. wanted to capture and reflect everyone’s views. Participants suggested potential ground rules which During the workshops all small group discussions were then agreed by the whole group. Across each of were recorded and detailed notes were taken of the plen- the workshops the ground rules created consistently ary discussions. Following the workshops the record- emphasised the importance of allowing everyone a ings of the discussions were typed up by the research chance to speak and respecting differences of opinion. team. A narrative account of each of the workshop It was stressed that we were not necessarily aiming to sessions was developed through assimilating each of reach a consensus about the topics that would be dis- the researchers’ notes from the events along with the cussed but rather we were interested in hearing the transcripts from the recordings. Notes from each ses- range of views people might hold about these. sion were collated and key themes and topics were iden- Following this the day began with small group dis- tified, as well as the identification of nuanced or cussions on the topic of health research very broadly, in divergent perspectives. In reporting the workshop dis- which participants discussed their understandings and cussions in the following sections of this paper, we draw expectations of health research. This was followed by a on transcripts of the small group discussions (which presentation introducing the topic of data-intensive allow verbatim quotes) and detailed notes of the plen- health research and giving an overview of the ways ary discussions (which do not allow verbatim quotes). that data is currently used in health research and the The following sections of this paper will set out the governance systems in place. Small group discussions findings from the workshop discussions and key con- then focussed on data-intensive health research. A siderations which emerged as central for meeting public second presentation set out the importance of ‘‘public expectations concerning public benefits from data- benefits’’ as a driver for data-intensive health research. intensive health research. This was a brief presentation which described the importance given to ‘‘public benefits’’ for determining applications for data access and also previous research Findings identifying ‘‘public benefits’’ as crucial for public Expectations of health research acceptability of data use. The presentation noted the difficulty in clearly defining the concept of ‘‘public In the first small group discussion of the day, workshop benefits’’ and led into facilitated small group discus- participants were asked to describe their expectations of sions around what this term means and what people’s health research and what they understood the term expectations of public benefits would be. ‘‘health research’’ to mean. Participants typically A series of videos were then shown of data scientists responded by describing health research as being discussing research projects which they had either aimed at improving health or health services, and/or recently completed or were planning to undertake. leading to new treatments. Across each of the work- Videos were used instead of face-to-face presentations shops it was evident that health research was expected in part as a result of practical challenges associated with to be aimed at a clear purpose of improving health securing participation of data scientists at each of the outcomes or health services. three workshops and also because this virtual presence Workshop participants described a variety of types encouraged participants to be frank and honest in their of health research, demonstrating awareness of diverse responses. The three data scientists gave an overview of forms that such research can take. The examples given their research, why they considered it important and were offered spontaneously by workshop participants. what the anticipated public benefits of the research While some participants described only research under- would be. Following the videos, workshop participants taken in clinical settings others gave varied examples discussed the extent to which they considered each to including: surveys; research around lifestyles; research have public benefits. in laboratories; and audit of health services. These par- In the final session of the workshops participants ticipants noted that health research can use a variety of were asked to come up with statements to answer two sources of information including donated organs or key questions: (1) Who is the public that should benefit tissue samples, genetic or clinical information as well from data-intensive health research and; (2) In what as information from medical records or surveys. While way should they benefit? Participants were given no some participants did not initially think of this diversity further direction on how they should answer these of health research, once mentioned by others these questions and were not provided with examples or broad approaches and methods were acknowledged as options to choose from, but rather were encouraged comprising health research. Aitken et al. 5 When asked who they would expect to be conduct- that academic research is not necessarily independent of ing health research, in many groups workshop partici- any financial interests: pants initially discussed health research conducted by charities and it was clear that this was often the first M2: It would depend where the university was getting type of research that came to mind. Participants often the funding from. There’s a chance the university could stated that this was due to having an awareness of this be getting their funding from a pharmaceutical research as a result of fundraising campaigns, TV company. adverts or leaflets (for example from Cancer Research F2: Again, that’s trust, isn’t it? You’re trusting those UK). However, a broader range of research organisa- researchers to be ethical with their findings, because tions were also discussed and workshop participants you do get the picture that the money issue is always listed many potential researchers including those breathing down their neck within the NHS, universities, charities or private com- M3: I would think, or I would hope, that universities panies. In most groups the involvement of private com- and that were more neutral, that they don’t have any- panies in health research was a contentious subject. thing to win or lose out of it. They’re looking at the Private companies were described as having a different bigger picture. (Glasgow, Group 3) set of objectives from academic or public sector organ- isations, and in some discussions commercial interests In discussing the different types of researchers who may were conceptualised as being in opposition to public be involved in health research participants acknowl- benefits. However, participants also noted the blurri- edged that there can be a role for commercial organisa- ness of lines between sectors and acknowledged that tions but concerns were raised about funding and the university research can also lead to commercial out- interests that go along with that. Funding of research puts. For example, the following discussion reflected was an area of interest in discussions across the work- one group’s shifting position as they considered the shops. It was acknowledged that research could be sup- potential involvement of private companies in health ported by either public or private funding. Some research: participants expressed mistrust of private research fun- ders, for example a large number of participants con- F2: the right people would be people in the NHS, in the tended that private companies would suppress research medical profession, because they’re not in it for the results if it enabled them to increase their profits. In money. In my point of view, I don’t want people to particular, a number of participants stated that they profit from it, from the data believed that a ‘‘cure for cancer’’ may have already M1: no been found but that pharmaceutical companies are M3: but if they could cure someone holding on to it to make the most money: F2: but if they could cure someone and then charge them massively F1: People always say there’s no cure for the common M1: no financial gain cold – so what chance do we have of finding cures for F2: I don’t want financial gain, yeah cancer if there’s no research into it MA: what if a university research group collaborated M1: some people say they have got a cure for cancer with a pharmaceutical company? but they’re holding it back to sell to the highest bidder F2: Sometimes I would imagine that has to happen F1: Conspiracy theories M3: for new drugs M1: There’s truth in it MA: in that situation there may be some profit that F2: Everybody thinks that, the higher authorities are all goes with it in it for what they can make out of it F1: yeah, there has to be some profit M2: are they withholding a cure? M3: without profit there won’t be any progress F2: unfortunately not everyone has a heart, they don’t F1: you need the companies involved – it’s a necessary care enough. (Perth, Group 1) evil F2: so long as there is good with it M2: I believe that there is research that has been done F1: more good than bad that has proven to actually help benefit certain illness F2: it’s about the balance and whatnot but nothing will be done about it because M3: It’s never going to be 100 per cent the pharmaceutical companies are making far too much F2: some people are only going to be in it for the money money from selling other products to help combat these M1: that’s just the way it is. (Perth, Group 1) things. M4: I think also the problem is that you have huge com- Similarly when asked about their expectations of uni- panies that are making billions and billions of pounds and versity researchers, participants typically acknowledged that obviously promotes distrust because they’re making 6 Big Data & Society so much money. But at the same time, you need that level NHS and everything. Everybody makes money from it of funding to be able to produce the advances. An exam- so, why would they? They don’t care about the little ple would be the Ebola crisis last year, they did ten years’ person. (Aberdeen, Group 3) worth of research in nine months because the WHO just piled billions into it, and they need that funding level, and This related to wider discussions around the ways in they got a cure out of it. But then at the same time, which benefits of health research are realised and a because they’re so profitable, it’s not necessarily ethical. widely held perception that currently the benefits are (Glasgow, Group 3) not realised equitably across society and that different groups or people in different locations across the UK There was concern that research conducted or funded experience health services differently as well as experi- by private companies would be directed at maximising encing different health outcomes. profits rather than maximising public benefits. For Throughout these discussions a recurring theme was example, one participant in Aberdeen stated: that the potential benefits of health research were not always or consistently realised. A range of factors were The money in research is directed at money-making noted as limiting the realisation of public benefits from things. Like, weight loss gets a lot of money put into health research, these included commercial interests, it, more than malaria, which affects people a lot more political priorities and limited public funding. than being overweight, which is generally, or can be, a cosmetic thing. It’s not down to illness. Not always – Who is the public that should benefit from sometimes it is – but, malaria, if they found a cure for data-intensive health research? that, billions of people, that would be their life improved. So ... where the money is they’ll research. When posed the question ‘‘who is the public?’’, work- (F1, Aberdeen, Group 3) shop participants typically treated this as a straightfor- ward and obvious question. Across all of the Furthermore, there was also discussion of the potential workshops participants almost always initially for governments to be involved in suppressing discov- responded by stating that: ‘‘the public is everyone’’. eries of new cures if they are too expensive or not However, these responses quickly became more aligned with government priorities: nuanced and complex when discussions moved to think about what this meant in practice and, in particu- M4: I often wonder if they want to find a cure for it. All lar about the ways in which ‘‘the public’’ could benefit the pharmaceutical companies they’re making millions from data-intensive health research. Two key consider- out of it. If they find a cure for it, what are they going ations emerged as important in conceptualising ‘‘the to do now? public’’. These were scale and need. M1: You could say the same about the government – Firstly, there were discussions around scale with does the government really want to find cures? Because, workshop participants expressing a range of views on at the end of the day, the longer people live, the more it whether the public should be conceptualised as every- costs to give them everything: pensions, health service, one globally, nationally or locally. and what have you. So, do they want us to be living to 100 years old? Probably not. M3: It’s us – all of us. Everyone in the world. F2: You lot are a load of doom and gloom! F4: It’s society. As in, what stands below the govern- F3: Five minutes in and we’re into the conspiracy the- ment. The individual who goes to work every day, ories! [laughter] doesn’t put a suit on, doesn’t sit and make a decision M4: They’ve been researching for maybe twenty years, on what our lives are going to be doing. (Perth, you’d think somebody would’ve found a breakthrough Group 3) M2: With all the money that’s been ploughed into it. M4: Do they really want to find a cure for it? F2: Presumably that’s just us, Joe Bloggs. F3: Why would they waste money investing in trying to F3: Everyone find it in the first place? F2: We’re the public, are we not? M1: Cos they’ve got to look like they’re trying to find it EC: Are we? [...] When you think about ‘‘the public’’, F3: No, they are trying to ... I’ve never even thought who’s in your mind? that they weren’t trying to. They’re doing all the F2: Myself research and everything, they’re trying to find cures M1: Working class people. [...] F5: The people that the research affects. M2: They make money by giving drugs. The pharma- M2: Or benefits. cies make money by getting the prescriptions from the F3: I just think everyone. Aitken et al. 7 EC: In Scotland? participant stated that the public should benefit ‘‘in F3: Just everyone. every way they are entitled to’’ and another stated EC: In the world? that the public should benefit in ‘‘the most beneficial F1: In the whole wide world. way possible’’. These responses highlight workshop F2: It depends what the research is about. If it’s merely participants’ recognition of the wide range of public based on Scotland, it would be the people of Scotland, I benefits that could be possible. Many participants would think. People of voting age. (Perth, Group 2) were clearly avoiding giving answers which might limit the ways that benefits were pursued and demon- There was generally a preference that ‘‘the public’’ strated a clear preference that benefits should be max- should be considered as inclusively as possible, meaning imised across society in whatever form they take. that there was a preference for research to benefit the Given that workshop participants had previously widest possible public (i.e., benefit the maximum described the purpose of health research as being pri- number of people). One workshop participant even marily about leading to improvements in health and/or stated that we should think of the public as ‘‘everyone health services, it is unsurprising that many responses in the universe’’ – this was stated only partially as a joke related to this theme. Many participants suggested that and demonstrated a strong preference to be as inclusive finding cures for diseases and making new drugs avail- as possible when thinking about ‘‘the public’’. able was a clear way in which members of the public However, the second key consideration – need – led should benefit. In particular, cancer, dementia and to different sets of preferences. Workshop participants mental health were often flagged as pressing conditions acknowledged that the findings and outcomes of about which more research was needed. More broadly research can have greater or lesser potential impact than just cures and increased medicalization, consider- for different groups within the public. For example, it able time in the discussions focused upon improving was acknowledged that health research could – and in lives; with a focus on health improvement, better qual- many instances should – have the greatest benefits for ity of life and enhanced lifestyle, with people living people with major health issues or groups affected by longer and healthier lives and lives that are less stress- particular conditions related to the subject of the ful. Linked to better quality of life and outcomes, par- research. This led to recognition of the value of ticipants suggested that future generations should be health research being targeted at vulnerable groups thought about so they do not face similar health and who would potentially benefit the most. In this sense lifestyle burdens, with better understanding and imple- public benefits were conceptualised as being benefits to mentation of preventative measures. Participants also particular smaller groups within the public rather than stated that benefits of research should be measurable, benefits to the wider public as a whole. However, it was through better quality care and services. Improved allo- also noted that while the immediate benefits of research cation of resources was also a way in which participants might be targeted at those considered to be in greatest thought the public could benefit. need, in the long-term this would be of benefit to every- Public benefits were also conceptualised in less direct one in society, suggesting that the two positions were ways. For example, participants considered that there far from incompatible. can be ‘‘knock-on effects’’ of small numbers of people benefiting from health research. Improving health and/ M1: I think it’s everybody because even if you’re not or quality of life of vulnerable groups was expected to aware of it or affected by it, that doesn’t mean it’s not lead to wider benefits for society. Therefore, participants of relevance to you. thought that there should be proactiveness to address F3: Just because you’re not associated with it at the particular needs of vulnerable communities to ensure time it doesn’t mean it won’t impact you later on in that communities, and society as a whole, can lead your life. (Perth, Group 3) better quality lives. This was also described as requiring empowerment of individuals and communities. The public were also described as benefitting from research through the creation and dissemination of new In what ways should members of the public knowledge. Workshop participants suggested that benefit? increasing scientific knowledge is in and of itself a bene- fit and that members of the public could benefit from When asked in what ways the public should benefit from data-intensive health research, a very broad set greater engagement with the scientific community. This of responses was given, demonstrating wide-ranging led to discussions of the need for science to do more to conceptualisations of ‘‘public benefits’’. Responses indi- publicise research results, and, in particular, to engage cated that workshop participants considered that with the public. However, it was frequently stressed public benefits should be as broad as possible. One that it is not sufficient for research to create new 8 Big Data & Society knowledge or understanding, rather it is considered diverse and varied forms that public benefits might essential that policy makers, governments and/or the take, as well as the diverse forms that research itself health service act on research findings in order to realise might take. However, there were some clear patterns the potential public benefits. regarding which projects were regarded most favour- ably. Voting preferences generally reflected the close- M5: The point is, it’s about what the outcome of the ness or relevance of research subject matter to research is, because if it says people coming from individuals’ lives. For example, older participants poorer backgrounds or deprived areas [are more at tended to vote for Project A which related to admis- risk of health conditions] .. . is anyone actually going sion to care homes. At all three of the workshops to spend any money to change it? Or is just a case of– Project C (which related to interception of illegal F2: ‘‘This is the research, we’ve done it, that’s it’’. drugs) consistently received the fewest votes yet was M4: The findings can be there but nothing’s actually also always the project which generated the greatest done. (Aberdeen, Group 3) discussion. The key discussion points which emerged unprompted in group deliberations related to: who These discussions around the ways in which the would benefit from the project; which project would public might benefit from data-intensive health have the greatest impact; and to what extent work- research revealed openness to the possible forms shop participants expected that the potential benefits public benefits could take and an unwillingness to would be realised. set limits or constraints on what public benefits In discussing their reasons for voting for particular might be. However, there was considerably more projects, participants considered which project would interest in the likelihood of the benefits being realised. be likely to have the greatest impact. This was discussed Reflecting earlier discussions around perceived bar- both in terms of which projects would benefit the most riers to realising benefits of health research, partici- people and which would have the biggest benefit for pants across all workshops were less concerned with those in greatest need – the two key ways in which what the benefits would be compared with assurances these ‘‘mini publics’’ characterised public benefit. that any benefits would be realised. While noting that Project A focussing on care homes was seen to be knowledge itself brought benefits, the participants very relevant in the context of an aging population wanted reassurance that further public benefits with care homes expected to be of increasing relevance could ensue through applying knowledge to improve in years to come. Additionally this project was con- lives. sidered important since it related to benefitting vulner- able people. Responses to three research projects M2: [Care] is putting a huge drain on resources at the The preceding sections have focussed on workshop moment. We’re living in an older society. participants’ discussions of public benefits from data- F3: People are living a lot longer. These are people who intensive health research at a general level. The work- have paid into the pot for so many years and then shops moved on to encourage participants to consider they’re just getting left... need to put more resources the ways in which particular research projects might into it. (Perth, Group 3) lead to public benefits. Workshop participants were shown videos of three Similarly, Project B which focussed on air pollution was data scientists who presented their research and the considered to potentially benefit the whole of society as ways in which they considered it would lead to public well as the environment and non-human life. Since this benefits. The three research projects presented were project was focussing on impacts on babies, many chosen to reflect very different subject matter. Project workshop participants regarded it as potentially A related to admissions to care homes; Project B having the biggest public benefits as it was starting related to impacts of air pollution on babies; and ‘‘right at the beginning of life’’. This was a point that Project C related to evaluating impacts of one particu- was made in all workshops. lar illegal drug interception policy. As a catalyst for discussions, after viewing the F3: For me, if [Project B] was just air pollution, without videos workshop participants were asked to vote on the pregnancy component, I probably would have gone which project they considered to have the greatest for care homes [Project A]. But because his research is potential public benefits. There was widespread agree- mixed with the pregnancy, as a mum, I want to give my ment that each of the three research projects had – at children the best start in life. least potential – public benefits. This reflects the find- F2: Anything to do with pregnancy is important – it’s ing discussed above that participants recognised the the start in life. (Aberdeen, Group 3) Aitken et al. 9 Participants noted that having a poor start in life can While the subject matter and aims of the project were affect the whole of the rest of life. widely considered to be of high importance, workshop participants were not confident that this project would F3: I voted [for Project B], we all know the risks about be able to make a significant difference. This highlights smoking in pregnancy and now its became unacceptable the importance of demonstrating mechanisms being in that women should be smoking because the harm it does place to enable impact. to the baby and now we know that pollution is likely to Drug use is a more controversial subject compared be just as bad, no one seems to be bothered about it and with care homes or air pollution so it is not we know the effects it can have on babies, more pollution surprising that Project C generated the most discussion. smaller babies then it would force them to do something Given that workshop participants typically acknowl- about it. That would prevent poorer health in childhood edged that their preferences for particular projects and all the things that go along with being premature were influenced by the extent to which they were per- and all the things that go along with it. ceived as having relevance to their own lives, as well as F4: I voted for [Project B] too and I did that because it society more generally, the potential stigma of drug use is the start of life, a baby being born, because it affects and what they saw as its social complexity, might have your whole life. I actually liked all of them. I worked in had some influence in shaping responses. It is note- a dementia unit and I know how it affects everyone, the worthy that one workshop participant who spoke whole family, but the baby one was trying to do some- about the ways in which his own life and that of his thing at the start of life. (Perth, Group 3) friends and family had been affected by drug addiction made an emotional and passionate plea to fellow par- Conversely, workshop participants were more sceptical ticipants to give Project C greater consideration. of the extent to which Project C (focussing on drugs The deliberations regarding the three projects illus- enforcement) would have a big impact. Some partici- trate the findings reported in preceding sections of this pants described this as ‘‘fighting a losing battle’’ and paper. The voting exercise was intended as catalyst for said that ‘‘drugs are going to be used no matter what’’. discussions rather than a priority setting exercise and Across the workshops drug use was recognised to be an workshop participants typically stressed that they area of major, but intractable, concern with significant appreciated the potential benefits of all three projects. implications for society. Drugs were described as a big Not voting for a project did not mean that a participant problem to be tackled, but also a complex social issue. did not believe it could bring public benefits. As such Much of the discussion about Project C related to the participants recognised and acknowledged the very dif- extent to which this research would actually have an ferent types of public benefits which could come from impact in addressing problems of drug use: it was the three projects. However, the discussions consist- described as reaching ‘‘just the tip of the iceberg’’. ently focussed on key considerations (which were not This meant that people were less inclined to vote for prompted by the facilitators): which project would be this project in comparison to the others, although as the likely to have the biggest impact, and whether partici- dialogue below illustrates, people had different kinds of pants had confidence that potential public benefits reservations. would be realised. These emerged as key themes throughout discussions at all three of the workshops. M1: I didn’t vote for it, not because I didn’t think there would be any benefit but we have been talking about Discussion health data and I just found it much harder to directly correlate, obviously we are talking about drugs and not Each of the three workshops resulted in open and wide- necessarily everyone taking these drugs will be taking ranging discussions leading to nuanced and complex them to a degree which significantly affect their health, positions regarding the ways that data-intensive you might get the people who take them now and health research might bring public benefits. The work- again. I just thought it would be very difficult to state shop discussions demonstrated that ‘‘public benefits’’ a clear public benefit ... were conceptualised in a number of different ways F1: with that one it’s hard to see how you could do and participants had varied expectations of what this something directly with the findings, I mean maybe fur- implied for actual research practice. Nonetheless, there ther down the line, but straight away, it would be dif- was a clear consensus that public benefits should be to ficult to get a clear public benefit from. individuals, specific groups and to society more widely. M5: I can see the clear public benefit from that but it It is interesting to note that no one spoke of societal looks like a hard fight, it’s just gonna take a long time, I benefits in terms of economic benefit, although this is just felt the other one, for me, was more beneficial. often portrayed as a form of public or societal benefit (Perth, Group 3) by governments and funding agencies. For example, 10 Big Data & Society The European Commission Strategy Centre recently maintaining relationships between research and called data ‘‘the lifeblood of the global economy’’ policy – and more broadly science and society. There (European Commission, 2017), whereas benefits to the was consensus that there needs to be commitment and economy through new data-driven industries were willingness by all stakeholders involved in the research never mentioned by workshop participants. Where eco- process to implement findings and maximise public nomic dimensions were raised, these related to potential benefits. Workshop participants were much less con- impacts on individuals or groups within society. For cerned with what form public benefits would take, com- example, when discussing care homes and the pressures pared with seeking assurances that potential benefits of an aging population, workshop participants were would be realised. Addressing this area of concern typically concerned with the economic impact on indi- may be vital for securing public support and establish- viduals through a pressured care system and how that ing a social license for future health-related data science affected health and wellbeing. In considering how best and demands transparency regarding how impact can to generate and sustain a social license for health- be achieved. related data science, funders and researchers may need to more clearly articulate how economic benefits Study limitations would result in wider social impact. Workshop participants had a broad and inclusive The research was undertaken following a ‘‘mini- approach to conceptualising both the public and publics’’ approach which aimed to engage with a public benefit. They did not perceive a conflict between cross-section of the Scottish population. Participants desiring wide benefit for society as a whole and promot- were sampled to represent demographic characteristics ing the needs of particular groups. Indeed, they recog- broadly representative of local populations in each of nised the role of indirect public benefits – those the locations. Nevertheless, the total number of partici- accruing to society when the pressing needs of some pants in the study was 69 and as such is not sufficient to are met. It was generally acknowledged that targeting claim that the sample is representative of the wider benefits for those in need was appropriate and neces- Scottish population. Deliberative public engagement sary. This resonates with a recent study which exam- methods are effective for enabling in-depth examination ined public sector professionals’ expectations of public of participants’ viewpoints and rationales but are time benefits and found that when evaluating the relative and resource-intensive which constrained the scope of value of different public benefits some focussed on the this study. The aim of this qualitative, deliberative number of people benefitting, others focussed on the research was not to produce a set of findings which level of need of those benefitting and, for others it could be generalised or be considered representative. was the potential for long-term impacts (Scott et al., Rather it intended to explore the ways in which mem- 2018). bers of the public would engage with this important Workshop participants were unwilling to narrowly subject and examine the range of views that would be define or constrain public benefits and preferred to keep expressed. In that context, it is noteworthy that the this definition open recognising the very many forms findings produced were broadly consistent both within public benefits could take. They were more concerned and between each of the workshops. with the likelihood that benefits would be realised – The responses reported throughout this paper were that research would make a difference – a theme con- generated through discussion, with few prompts from sistent across all workshops. The discussions high- the facilitators. The research team did not set out to lighted the need for action in response to research examine public views on any particular potential types findings, suggesting that ensuring impact itself was an of benefits but rather to explore how members of the important component of achieving public benefit. They public would conceptualise public benefits and what recognised the wider processes involved in this, beyond their expectations are. Future research might usefully the research community itself to include policy makers engage the public in discussions of particular potential and practice communities. Across all the workshops outcomes and research aims. there was widespread agreement that, currently, the benefits of health research are impeded by commercial Conclusions or political interests. There was considerable scepticism of the extent to which necessary mechanisms and pol- This research has demonstrated the value of engaging itical support are in place to enable research to lead to with members of the public to address conceptually, public benefits. This indicates that workshop partici- socially and ethically challenging issues related to cur- pants recognised that the pathways between research rent and future health-related data science practices. and impact are not straightforward or inevitable and The social licence for the research use of health and reflects widespread awareness of the importance of other ‘‘Big Data’’ needs to be continuously negotiated Aitken et al. 11 Bickerstaff K and Walker G (2001) Participatory local gov- as public expectations, preferences and views shift in ernance and transport planning. Environment and Planning response to changing contexts and experiences. The A 33: 431–451. boundaries of acceptability will need to be renegotiated Carter P, Laurie GT and Dixon-Woods M (2015) The social on an ongoing basis to ensure that actual practices align licence for research: Why care.data ran into trouble. with public values and concerns. This may be of par- Journal of Medical Ethics 41: 404–409. ticular importance when considering economic benefits, Coleman S and Gotze J (2010) Bowling together: Online particularly but not only with respect to commercial public engagement in policy deliberation. Available at: involvement in research (Davidson et al., 2013; www.hansardsociety.org.uk (accessed 21 November 2018). Wellcome Trust, 2016). Davidson S, McLean C, Treanor S, et al. (2013) Public The nuanced discussions at each of these workshops acceptability of data sharing between the public, private have demonstrated the value of bringing together and third sectors for research purposes. Scottish diverse groups to discuss conceptually challenging Government Social Research. Available at: https:// topics enabling considered and informed opinions to www2.gov.scot/resource/0043/00435458.pdf. be developed. As noted by Scott et al. (2018: 32) Escobar O (2011) Public Dialogue and Deliberation: A Communication Perspective for Public Engagement ‘‘there is a need for further direct engagement with Practitioners. Edinburgh: Edinburgh Beltane – UK the public to better understand where their boundaries Beacons for Public Engagement. Available at: www.belta- of acceptability lie within the context of data sharing’’. nenetwork.org/resources/beltane-publications (accessed 21 This study reaffirms much of the literature concerning November 2018). deliberative public engagement methods, in demon- European Commission (2017) Enter the data economy: EU strating the value of these methods for informing and policies for a thriving data ecosystem. (21): 16. Available examining public attitudes around complex subject at: https://ec.europa.eu/epsc/publications/strategic-notes/ matter. enter-data-economy_en. Garrety K, McLoughlin I, Wilson R, et al. (2014) National electronic health records and the digital disruption of Declaration of conflicting interests moral orders. Social Science & Medicine 101: 70–77. The author(s) declared no potential conflicts of interest with Habermas J (1975) Legitimation Crisis. Cambridge: Beacon respect to the research, authorship, and/or publication of this Press. article. Halseth G and Booth A (2003) ‘‘What works well; What needs improvement’’: Lessons in public consultation from British Columbia’s resource planning processes. Funding Local Environment 8(4): 437–455. This research was funded by The Farr Institute of Health Hodson H (2016) Revealed: Google AI has access to huge Informatics Research. The UK HIRN is supported by a 10- haul of NHS patient data. New Scientist Available at: funder consortium: Arthritis Research UK, the British Heart www.newscientist.com/article/2086454-revealed-google-ai- Foundation, Cancer Research UK, the Economic and Social has-access-to-huge-haul-of-nhs-patient-data/ (accessed 20 Research Council, the Engineering and Physical Sciences March 2018). Research Council, the Medical Research Council, the King G, Heaney DJ, Boddy D, et al. (2010) Exploring public National Institute of Health Research, the National perspectives on E-health: Findings from two citizen juries. Institute for Social Care and Health Research (Welsh Health Expectations 14(4): 351–360. Assembly Government), the Chief Scientist Office (Scottish McGrail K, et al. (2018) A position statement on population Government Health Directorates), the Wellcome Trust, data science. International Journal for Population Data (MRC grant no: MR/M501633/2). Science 3(1): 4. Mitton C, Smith N, Peacock S, et al. (2009) Public participa- Ethical approval tion in health care priority setting: A scoping review. Health Policy 91(3): 219–228. This research was granted ethical approval by the Usher Morrell ME (2010) Empathy and Democracy: Feeling, Institute Ethics Committee. Thinking, and Deliberation. University Park: Pennsylvania State University Press. References Niemeyer S (2011) The emancipatory effect of deliberation: Empirical lessons from mini-publics. Politics & Society Aitken M, Jorre JDS, Pagliari C, et al. (2016) Public 39(1): 103–140. responses to the sharing and linkage of health data for Owens S, Rayner T and Bina O (2004) New agendas for research purposes: A systematic review and thematic syn- appraisal: Reflections on theory, practice, and research. thesis of qualitative studies. BMC Medical Ethics 17(1): 73. Environment and Planning A 36: 1943–1959. Aitken M, Tully MP, Porteous C, et al. (in press) Consensus Parkinson J (2004) Why deliberate? The encounter between Statement on Public Involvement and Engagement with Data-Intensive Health Research. International Journal of deliberation and new public managers. Public Population Data Science. Administration 82: 377–395. 12 Big Data & Society Petts J (2005) Enhancing environmental equity through deci- wind farms in Scotland. Available at: www.climatex- sion-making: Learning from waste management. Local change.org.uk/media/1438/citizens_juries_-_full_report. Environment 10(4): 397–409. pdf (accessed 21 November 2018). Pidgeon N and Rogers-Hayden T (2007) Opening up nano- Scott K, Burall S, Perrin H, et al. (2018) Data for public technology dialogue with the publics: Risk communication benefit: Balancing the risks and benefits of data sharing. or ‘upstream engagement’? Health, Risk & Society 9(2): Available at: http://understandingpatientdata.org.uk/ 191–210. news/data-public-benefit (accessed 21 November 2018). Public Benefit and Privacy Panel (PBPP) (no date) Guiding Sixsmith J, Boneham M and Goldring JE (2003) Accessing principles and policy for decision making. Available at: the community: Gaining insider perspective from the out- www.informationgovernance.scot.nhs.uk/pbpphsc/wp- side. Qualitative Health Research 13(4): 578–589. content/uploads/sites/2/2015/04/Guiding-Principles.docx Tausch AP and Menold N (2016) Methodological aspects of (accessed 8 June 2018). focus groups in health research. Global Qualitative Nursing Ramesh R (2014) NHS patient data to be made available Research 3: 1–12. for sale to drug and insurance firms. The Guardian, Wellcome Trust (2015) Enabling data linkage to maximise the 19 January. Available at: https://www.theguardian.com/ value of public health research data: Full report. Available society/2014/jan/19/nhs-patient-data-available-compa- at: www.wellcome.ac.uk (accessed 21 November 2018). nies-buy. Wellcome Trust (2016) The one-way mirror: Public attitudes Roberts J and Escobar O (2015) Involving communities in to commercial access to health data. Available at: www. deliberation: A study of 3 citizens’ juries on onshore wellcome.ac.uk (accessed 21 November 2018).

Journal

Big Data & SocietySAGE

Published: Dec 6, 2018

Keywords: Public benefits; data science; health informatics; public engagement

There are no references for this article.