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

Learn More →

The social licence for research: why care.data ran into trouble

The social licence for research: why care.data ran into trouble Current controversy The social licence for research: why care.data ran into trouble 1 2 1 Pam Carter, Graeme T Laurie, Mary Dixon-Woods Health Sciences, University of ABSTRACT Although care.data has numerous aims (box 1), Leicester, Leicester, UK In this article we draw on the concept of a social licence we focus specifically on its research purposes. We Department of Law, University to explain public concern at the introduction of care. begin by offering some brief background on the use of Edinburgh, Edinburgh, UK data, a recent English initiative designed to extract data and regulation of routine medical data before intro- Correspondence to from primary care medical records for commissioning and ducing the concept of a social licence. Professor Mary Dixon-Woods, other purposes, including research. The concept of a Health Sciences, University of social licence describes how the expectations of society Leicester, 2nd Floor, Adrian THE USE AND REGULATION OF MEDICAL regarding some activities may go beyond compliance Building, University of RECORDS FOR RESEARCH Leicester, University Road, with the requirements of formal regulation; those who Researchers have long relied on access to personal Leicester LE1 7RH, UK; do not fulfil the conditions for the social licence (even if medical information routinely collected during the md11@le.ac.uk formally compliant) may experience ongoing challenge course of patient care in order to conduct studies, and contestation. Previous work suggests that people’s Received 17 July 2014 including clinical trials and epidemiological cooperation with specific research studies depends on Revised 24 November 2014 research. However, the repurposing of routinely Accepted 30 December 2014 their perceptions that their participation is voluntary and collected data for research is not without risk to Published Online First is governed by values of reciprocity, non-exploitation and 8 relevant values, and measures such as anonymisa- 23 January 2015 service of the public good. When these conditions are tion (even when possible) do not solve all ethical, not seen to obtain, threats to the social licence for legal and technical problems; people may, for research may emerge. We propose that care.data failed example, have religious or moral objections to par- to adequately secure a social licence because of: 5 ticular studies or concerns about stigma and (i) defects in the warrants of trust provided for care.data, breaches of privacy. (ii) the implied rupture in the traditional role, Accordingly, researchers’ access to, and use of, expectations and duties of general practitioners, and personal data has traditionally been subject both to (iii) uncertainty about the status of care.data as a public tight controls and to efforts to promote public and good. The concept of a social licence may be useful in patient confidence in those controls. The Data explaining the specifics of care.data, and also in Protection Act 1998, which implements the reinforcing the more general lesson for policy-makers European Data Protection Directive, has always that legal authority does not necessarily command social recognised ‘medical purposes’ as a legal basis for legitimacy. processing personal data from medical records, subject to a proportionality test. While it contains a ‘research exemption’, it does not absolve data con- trollers completely of their obligations under the BACKGROUND law. Data protection has been bolstered by the Encouraging more and better health-related research requirement, introduced in 1999, that each NHS is now firmly established as a policy goal in organisation have a designated senior individual 1–3 England. A crucial feature of the current policy (known as a Caldicott Guardian) with responsibility drive is an emphasis on the use of the distinctive both for the confidentiality of patient information infrastructure of the personal medical records of —which is protected by the common law—and for individuals registered with the National Health enabling appropriate information sharing. These Service (NHS). Researchers have long argued that, legal measures also operate alongside guidance and contrary to the public interest, these records have codes of practice issued and updated regularly by remained under-exploited as a research resource. the professional regulators, by the NHS Research Yet the recent launch of care.data—an initiative to Governance Framework, by a system of research Open Access extract data from NHS primary care medical ethics committees, and by a variety of other Scan to access more records in England unless patients have purposefully mechanisms. free content opted out, in part to facilitate research—has proved Nonetheless, access to and secondary use of deeply controversial. Within weeks of the start of a medical information for research purposes has con- public information campaign run by NHS England, tinued to be a fraught and contested activity, prone public and professional concern had become so to concern over its ethical standing. The tensions overwhelming that in February 2014 the implemen- were discussed extensively in a 2006 report of the 5 4 tation of the scheme was suspended for 6 months. Academy of Medical Sciences (AMS), in a docu- Our aim in this article is not to assess the rights and ment that can be seen as an attempt at advocacy on wrongs of care.data—others have already offered the part of a clinical research community that saw such critiques —nor to offer an in-depth analysis of itself as beleaguered and misunderstood. The To cite: Carter P, Laurie GT, the legal background, but rather to show how the report offered multiple examples of the benefits of Dixon-Woods M. J Med concept of a social licence can help to explain the research that uses routinely collected personal data, Ethics 2015;41:404–409. challenges faced by care.data. including its possibilities as a means of answering 404 Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 Current controversy grounds that the risks involved were low in comparison with the likely benefits and in comparison with other forms of Box 1 Aims of care.data research (such as clinical trials). This claim was in keeping with the influential (although not necessarily uncontested) view that NHS England has described the care.data service as: ‘…anew, over-emphasis on individual consent fails to balance a broader modern data system for the NHS in England. Known as care. range of ethical principles and the argument that people have a data, its purpose will be to provide timely, accurate information 14 15 responsibility to participate in research. The tenor of the to citizens, clinicians and commissioners about the treatments arguments was that the proper solution to the challenges sur- and care provided by the NHS’ (http://www.england.nhs.uk/ rounding use of routine medical data for research purposes wp-content/uploads/2013/05/ces-tech-spec-gp-extract.pdf ). might then be one of adjusting the regulatory environment to The aims of the care.data programme are sixfold: accommodate what the public would support. 1. To support patients’ choice 2. To advance customer services A SOCIAL LICENCE FOR RESEARCH INVOLVING MEDICAL 3. To promote greater transparency RECORDS? 4. To improve outcomes Based on the reasoning that the public would accept and 5. To increase accountability endorse a less restrictive approach to the use of medical records 6. To drive economic growth by making England the default for research purposes than the custodians of the system realised, location for world-class health services research. the AMS’s clarion call was to ‘be bold’. Although it did not use NHS, National Health Service. the term explicitly, the core of the AMS argument was that the social licence for the use of medical records for research was more permissive than the operation of the regulatory environ- primary research questions, as a prelude to clinical trials, as a ment allowed. means for disease surveillance, and as a resource for identifying The concept of a social licence is discussed in two distinct lit- potential participants in trials or other studies. But such activ- eratures. One important literature derives from the work of the ities were, the AMS report argued, continually thwarted by the sociologist Everett Hughes who describes the relationship complexities and frustrations of multiple legal, regulatory and between a profession and society in terms of the two concepts professional restrictions. of licence and mandate. For Hughes, licence is granted to The AMS report and others, while accepting the need for certain occupational groups to carry out particular activities; a regulation, argued that the regulatory environment for the use mandate is claimed by members of the group to define for itself of personal data for health research in England was, by 2006, proper conduct in relation to matters concerned with their not in the public interest: it was disproportionate relative to the work. A second, and mostly distinct, corpus of work on cor- risks and benefits involved and involved multiple overlapping porate social responsibility describes the concept of the ‘social layers, institutions and actors as well as contested and sometimes licence to operate’ as the expectations of society regarding the conflicting standards and expectations for research governance. conduct and activities of corporations that go beyond the While accepting the broad principle of the need for regulation, requirements of formal regulation. Thus, industries with sig- the report suggested that NHS organisations were overly cau- nificant environmental impact may find that operating within tious about legal risk in areas of interpretive uncertainty, and the law but outside the boundaries of social approval can result tended to default to a conservative approach. The way the rules in corporate damage—for example, by having a negative impact relating to data protection were understood, interpreted and on a company’s brand or provoking new and restrictive regula- applied by NHS organisations, research ethics committees and tion. The Australian mining industry is often offered as an others was a focus of particular criticism. The overall effect, the example of an industry required to earn a social licence, and report and others argued, was that NHS organisations were gen- maintain it, by behaving in a trustworthy and responsible way; if erally reluctant to share personal information without explicit it were otherwise, it would face ongoing challenge and costly consent from patients or anonymisation of data to remove them delays and interference in its activities. from the reach of data protection legislation: the dichotomous Some analogies have been drawn between a social licence for ‘consent or anonymise’ approach was the norm. the mining of minerals and for the process of data mining. Yet this was not the only possible approach, nor was it Regarding these and similar activities, what the social licence required by law. Under what was Section 60 of the Health and emphasises is the possible need for those (whether they are Social Care Act (2001) (later Section 251 of the NHS Act public or private bodies, or specific occupational groups) under- 2006), a Patient Information Advisory Group (PIAG), acting on taking activities likely to provoke public disquiet to go ‘beyond behalf of the Secretary of State, was allowed to approve process- compliance’ with legal requirements. We propose that a poorly ing of patient information without seeking patients’ consent for informed understanding of the social licence for secondary use research purposes. However, the AMS report criticised the of personal medical data, and a failure to recognise that legal PIAG’s approach, processes and membership, and its alleged authority might not be enough to secure the social licence, tendency to stress its role in protecting privacy or confidentiality seems to have been at the heart of the controversy underlying ‘without equal emphasis on the benefits derived from well- care.data. conducted research’. PIAG was argued to suffer from mission creep, applying a stricter standard of absolute and proven neces- Legal authority for care.data sity rather than assessing whether applications involved a pro- The Health and Social Care Act 2012 established the Health portionate and justifiable interference in privacy. and Social Care Information Centre (HSCIC) as an Executive Central to the AMS report’s argument for reform was that Non-Departmental Public Body empowered to obtain patient- PIAG appeared to adopt a more conservative approach than identifiable information from general practices. Practices them- patients and the public would favour. It claimed strong public selves were, under their NHS contract, obliged to transfer the approval and support for research using medical records, on the data to HSCIC unless patients explicitly opted out. Using a Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 405 Current controversy centrally managed General Practice Extraction Service (GPES) defined, consenting patients can be generalised to the broader based on commercially supplied software, care.data was conception of NHS citizenship implicit in the new policy direc- intended to gather data monthly, capturing most routine general tion is not clear. practice consultations. The data would then be made available What is clear is that individuals’ cooperation with specific research studies is usually secured through three principal in aggregate form to the HSCIC, with six aims (box 1) identi- fied for their use, including that of driving ‘economic growth by mechanisms: their expectations about how research is conducted making England the default location for world-class health ser- and regulated; their trust in the institutions and individuals who vices research’. recruit them; and their beliefs in the wholesomeness and public Applications to use the data are to be reviewed by the value of the research endeavour. More broadly, the public legit- HSCIC’s Data Access Advisory Group. If requests to use data imacy and acceptability of health research rests heavily on its are classified as sensitive and identifiable, researchers are subject status as a socially valuable enterprise conducted in the service to additional requirements, including that of undergoing scru- of the public good. tiny from the Confidentiality Advisory Group of the Health There are many reasons to doubt that care.data could reason- Research Authority. This group replaced the PIAG and an ably assume that the public would automatically confer upon it interim body (the Ethics and Confidentiality Committee operat- the same legitimacy and endorsement as that enjoyed by ing under the National Information Governance Board). It con- research where individual informed consent is sought and clear tinues the role of advising on data use for research when information about study aims is provided. For instance, the individual consent has not been obtained on the basis of mobility of electronic data and the practical difficulties of speci- so-called Section 251 approvals. fying in advance the research questions for which data might be Other changes relevant to the creation of care.data included used or the populations to be studied mean that care.data was in amendments in 2012 to the NHS Constitution to offer a many ways quite distinct from conventional research projects. ‘pledge’ to inform English NHS patients about research studies Three threats to the social licence faced by care.data are espe- in which they might be eligible to participate, and also, crucially, cially important in explaining the challenges it faced: (i) defects an expectation that patients would be willing to share their in the warrants of trust provided for care.data, (ii) the implied medical information for healthcare planning and for research rupture in the traditional role, expectations and duties of purposes. The relevant section—‘Respect, consent and confiden- general practitioners (GPs), and (iii) uncertainty about the status tiality’ (p. 8)—explains that patients have the right to request of care.data as a public good. that their information should not be shared but also that such requests may not be upheld if the public interest is found to out- WARRANTS OF TRUST weigh the individual case. Ensuring that people are aware of how data from medical Thus was the administrative infrastructure for care.data records has been or might be used, and the protections that are created—but not, it seems, the social licence. The amendments in place, might be assumed to be critical to ensuring their confi- to the Health and Social Care Act 2012 and to the NHS dence that such use is legitimate and well-governed. A system- Constitution seem to have been based on a re-imagining and atic review of public awareness of, and views on consent to, the re-responsibilisation of patients as active citizens. There has secondary use of medical records for health has identified a gen- been an assumption that the implicit social contract underpin- eralised lack of awareness and understanding. Yet level of ning the NHS would mean most patients would be happy to ‘awareness’ itself may not be a reliable guide to people’s faith in support sharing of their personal data in the interests of the six a system. Individuals’ willingness to cooperate with health aims of the care.data programme, including economic goals to research often depends on powerful heuristics rather than be realised through research. detailed understanding of research governance procedures and Yet the extent to which confidence in such assumptions was standards. Even when they agree to take part in specific well-founded was not clear. Previous social science research sug- studies, it is not the detail of consent forms and information gests a nuanced and delicate understanding of societal support sheets that matters to participants so much as an overall faith in for, and cooperation with, health research is needed. It has the legitimacy of the endeavour and a reassurance of protection 22 23 27 28 found that the public’s support and tolerance for research, and from risk. Important, too, is a belief that agreeing to its associated risks, often depends far more on an often fragile participate in research will not make individuals vulnerable to set of cues about the safety and social good of research partici- risk of harm, exploitation or charges of gullibility. This is not pation, and on institutional and professional credentials, than it to imply that consent forms and information sheets are of no does on the formal architecture of research regulation, or on value or relevance; on the contrary, they act as ‘symbolic rational assessment of the detail of information sheets or other tokens’, as vitally important warrants of the trustworthiness of 22 23 22 documents aimed at gaining ‘informed consent’. the process. Further, most of what is known about patients’ support for Thus, a lack of detailed awareness might have mattered little research is based on quite particular examples of research par- in the roll-out of care.data had the right warrants of trust been ticipation—often those where patients already have an interest in place. But the approach adopted did not appear to provide in a medical condition and where they are asked for quite spe- them. The implied consent model underlying care.data did cific consent to a project or programme. Where patients specific- allow for patients to opt out, but the process for ensuring that ally consent to a particular study or programme of research, they were enabled to do so was precarious. The practical they may (depending on the study) expect the possibility of delivery of the requirement to ensure patients were aware of some personal benefit, as well as being able to experience sec- care.data and their right to opt out proved problematic. The ondary benefits of their consent, such as the ‘warm glow’ asso- NHS England information leaflet about the initiative was ciated with altruistic behaviour that benefits an identifiable addressed to households rather than individuals and might have community (eg, of those with a specific disease). The extent appeared to be unsolicited junk mail. to which the findings of this body of research about participa- For those seeking to improve their understanding, the leaflet tion in specific, relatively well-bounded studies or cohorts by itself did not appear to be fully informative. Entitled ‘Better 406 Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 Current controversy Information Means Better Care’, it made no explicit reference to opportunity to address novel research questions but also gen- care.data other than to direct those seeking further information erates new ethical dilemmas in relation to re-identification of 14 41–43 to an NHS website. It did not include an opt-out form. The bene- individuals or losses of data and other intrusions asso- fits of information sharing were explained in the patient informa- ciated with risks to anonymity and privacy. Again, the extent to tion leaflet, but the risks were not. HSCIC’sown which the social licence for these potentially profound altera- acknowledgement that ‘a small risk of re-identification due to tions to the nature of the doctor–patient relationship was small numbers/rare diseases’ was not conveyed. Questions secured for care.data is in question. about who would access data, for which purposes and on which terms, were also left unclear, despite evidence showing that the STATUS OF CARE.DATA AS A PUBLIC GOOD public wanted more information about how governments and Values of reciprocity and fairness underpin the legitimacy of companies collect, share and use data. Additionally, it was not health research: if people participate voluntarily, they expect clear that busy GPs would have the time or sufficient information that their contribution will be used to improve the care of 23 24 to engage in complex discussions with patients wishing to opt others, and that their good faith will not be exploited. out, nor was any extra resource or support provided for the task, Much depends, therefore, on the extent to which uses of per- even for vulnerable groups such as care home residents. The con- sonal data are seen as serving the public interest and conducted cerns of GPs regarding losing control of patient data during this by those with a public interest orientation. Yet neither of these process were also not addressed. Further, few opportunities were features of the social licence seems to have been fully addressed 33–35 available to the public to influence the direction of policy. for care.data. Overall, in contrast to the cues usually available to research The ‘public regime of justification’ that was provided for candidates, the warrants of trust provided by care.data appeared care.data stressed its benefits at a national rather than an individ- inadequate and were likely to have undermined the social ual level. Of the six aims of care.data (box 1), the first five are licence rather than strengthen it. Lack of detail in the informa- rooted in the use of data to improve quality and delivery of care tion sheets was unlikely to have been the direct cause of the and the governance of healthcare, with the recommendations of failure to secure the social licence, but it does provide visible Sir Robert Francis arising from his damning report on Mid evidence of how the operationalisation of the programme was Staffordshire NHS Trust invoked to supply part of the motiv- founded in poor understanding of what was needed for the ation. However, there are indications that members of the public and professionals to have trust and confidence in it. public expect to see benefits of such data use made explicit. Inadequate consultation and engagement meant a failure to take The sixth aim of care.data refers to research, but in a way that into account ‘the right things’ and thus secure the social licence. links it explicitly to economic growth. The multiplicity of aims towards which care.data is directed, RUPTURE IN TRADITIONAL EXPECTATIONS and the linking of care.data and research to the so-called ‘health The legitimacy of health research draws heavily and crucially on and wealth’ agenda, may be an unfamiliar rationale for citizens 7 45 people’s trust in organisational and professional credentials, used to a public service model of healthcare. Patients may mis- including the trust placed in professionals who have a duty of con- trust commercial interests, especially where these might be per- 36 26 32 35 46–48 fidentiality and relationships developed through continuity of ceived as profiteering or resulting in excessive profit care and empathy. GPs are constrained both by the common law or where patients have concerns about the extent to which risks duty of confidentiality and statutory requirements under the Data and benefits are evenly distributed and whether their contribu- Protection Act 1998 and Human Rights Act 1998 to respect tions to research will be reciprocated by a contribution to the patient confidences and to process personal data or disclose confi- public good. The persistent problem of non-publication of study dential information only within clearly defined parameters. results is just one element of how non-reciprocation may mani- However, in neither legal regime is the protection of privacy abso- fest. Other concerns may be linked to questions of who will be lute. A 2013 review of the system of information governance in able to access care.data for which purposes with which risks, how health and social care services clarified, for example, that legally the credentials of bona fide researchers can be established, and patients do not own their data. Notwithstanding, what is legally what mandate commercial organisations will have to use data permissible and what is socially acceptable do not necessarily coin- that originated from private consultations between patients and cide. For some, care.data represents a significant and unwelcome their GPs. alteration to traditional understandings of the private and confi- The complexity is amplified by the institutional fragmentation dential nature of the relationship between GP and patient. of the NHS in England, which is no longer a single unitary Some accounts argue that routine monthly extraction of elec- public sector entity. Transfer of information ‘within the NHS’ in tronic individual level patient data from GP records represents a practice means transfer between entities with distinct legal step-change in terms of the level of intrusion into patients’ status. This takes place in a complex framework of statutory private lives, to the extent that it has been suggested that it provisions and legal contracts, and it increasingly occurs with erodes the fundamental right to respect for private life offered those who are private contractors of services to the NHS or by the Human Rights Act. Concerns have been expressed that may operate outside the NHS entirely. Concerns about insur- patients may not wish intimate details of their medical records ance companies and others having access to personal data to be accessible to those outside of their ‘circle of care’, and suggest that considerable further dialogue between the public, might decline to trust and confide in their healthcare providers affected stakeholders such as GPs, and policy makers is required if they know that their information will be shared. The law before a social licence can be said to be in place for care.data. has long recognised the importance of protecting this trust for both the private and public interests that it serves (see, for CONCLUSIONS example, the cases of X v Y (1988) 2 ALL ER 648 and also The concept of a social licence offers insight into why the Campbell v Mirror Group Newspapers (2004) UKHL 22). implementation of care.data, an initiative to extract data from These concerns are linked in part to the advances in techno- routine general practice consultations for use in research and logical capability associated with ‘big data’ which offer the other activities, has met with concern and controversy. Although Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 407 Current controversy Informatics Research (MRC Grant No: MR/K007017/1, MR/K006525/1, MR/ the infrastructure was in place, the activities were perfectly K006584/1 and MR/K006665/1). Professor Mary Dixon-Woods holds a Senior lawful, and a case had been made for the possible benefits that Investigator Award from the Wellcome Trust (WT097899) that has supported this might be generated, the experience of care.data starkly exposes work. an enduring truism about the limits of law: legal authority does Competing interests None. not necessarily command social legitimacy. A parliamentary Provenance and peer review Not commissioned; externally peer reviewed. majority may allow legislation to be passed, but that does not Open Access This is an Open Access article distributed in accordance with the equate to a societal seal of approval or to securing the trust and terms of the Creative Commons Attribution (CC BY 4.0) license, which permits confidence of patients, citizens, healthcare professionals and others to distribute, remix, adapt and build upon this work, for commercial use, researchers. Securing a social licence may require something provided the original work is properly cited. See: http://creativecommons.org/ other than a legal mandate. licenses/by/4.0/ Although England has remained firmly on the legislative path, different solutions to the same challenges relating to the use of REFERENCES personal medical information have been found north of the 1 Department of Health. Best research for best health: a new national health research border. Rather than legislative change, actors operating within strategy. The NHS contribution to health research in England. London: Department of Health, 2006. the existing legal frameworks have worked together under the 2 Liddell A, Ayling M, Reid G. Innovation, health and wealth accelerating adoption Scottish Health Informatics Programme (SHIP) to develop and diffusion in the NHS. Department of Health, 2011. mechanisms of principled proportionate governance—building 3 HM Treasury and Department for Business Innovation and Skills. The plan for on public engagement exercises—to deliver responsive, risk- growth. Crown Copyright. 2011. based approaches to data linkage for health research. Consent, 4 The Academy of Medical Sciences. Personal data for public good: using health information in medical research. London: The Academy of Medical Sciences, 2006. anonymisation and authorisation are all available as governance 5 Taylor M. Information governance as a force for good? Lessons to be learnt from tools to be deployed, as appropriate, in any given data linkage care.data. Script Ed 2014;11(1):1–8. proposal after an assessment of the risks and tolerances 6 Pollock AM, Roderick P. Trust in the time of markets: protecting patient information. involved. Lancet 2014;383(9928):1523–4. 7 Dixon-Woods M, Ashcroft RE. Regulation and the social licence for medical As the Scottish example shows, this does not mean that research. Med Health Care Philos 2008;11:381–91. unanimous social consensus is required for all developments. 8 Clark S, Weale A. Information governance in health. Nuffield Trust, 2011. Rather, a social licence for research will require, as a minimum, 9 Caldicott F. Information: to share or not to share? The Information Governance that certain conditions of social engagement have been Review. Crown Copyright. London: Williams Lea for the Department of Health, respected. Genuine dialogic engagement that might result in a 2013. 10 Taylor MJ. Health research, data protection, and the public interest in notification. broad licence must be distinguished from more narrowly Med Law Rev 2011;19(2):267–303. focused public relations exercises that seek to ‘capture’ the 11 Department of Health. Research governance framework for health and social care. public, that is, to persuade the public of the legitimacy of deci- 2nd edn. https://www.gov.uk/government/uploads/system/uploads/attachment_data/ 33 35 sions already taken by experts, and from simple ‘awareness- file/139565/dh_4122427.pdf (accessed 24 Oct 2014). 12 Doyal L, Tobias JS. Informed consent in medical research. London: BMJ Books, raising’ information exercises. While legitimate disagreement is inevitable, if a social licence is to be maintained, both the final 13 Walley T. Using personal health information in medical research. BMJ 2006;332 result and process used to achieve that result must be one which (7534):130–1. reasonable citizens can at least recognise as defensible on the 14 McHale J. Reforming the regulation of health research in England and Wales: new grounds that it reflects common social values and goals. challenges: new pitfalls. J Med Law Ethics 2013;1(1):23–42. 15 Harris J. Scientific research is a moral duty. J Med Ethics 2005;31(4):242–8. Trust and confidence in research governance or the ‘social 16 The Academy of Medical Sciences. A new pathway for the regulation and licence for research’ depends upon ideas about the public good governance of health research. The Academy of Medical Sciences, 2011. http:// that are not straightforwardly synonymous with the aim of www.acmedsci.ac.uk/p99puid209.html increasing the UK’s gross domestic product. What patients care 17 Hughes EC. Men and their work. Glencoe, Illinois: Free Press, 1958. 18 Gunningham N, Robert A, Thornton D. Social license and environmental protection: about as patients cannot be equated with what patients care why businesses go beyond compliance. Law Soc Inq 2004;29(2):307–41. about as citizens who are part of a much wider social endeav- 19 Allen G. Public affairs and mining in Australia–a rescued license to operate. J Public our. If care.data is to succeed, patients need to have the confi- Aff 2011;11(4):382–6. dence that their medical records will be held securely, 20 Rooney D, Leach J, Ashworth P. Doing the social in social license. Soc Epistemol anonymised appropriately, and that secondary use of this per- 2014;28(3–4):209–18. 21 Department of Health, London. NHS Constitution—the NHS belongs to us all. sonal data is in the public interest: the conditions of the social licence need to be respected in ways that go beyond compliance 22 Armstrong N, Dixon-Woods M, Thomas A, et al. Do informed consent documents laid down in a legal framework. Necessary—and hopefully suffi- for cancer trials do what they should? A study of manifest and latent functions. cient—conditions for social licence include: (i) reciprocity, Sociol Health Illn 2012;34(8):1230–45. 23 Dixon-Woods M, Ashcroft RE, Jackson CJ, et al. Beyond “misunderstanding”: which must begin with sound two-way communication, (ii) non- written information and decisions about taking part in a genetic epidemiology exploitation, which must exclude the spectre of disempower- study. Soc Sci Med 2007;65(11):2212–22. ment, and (iii) service of the public good, which need not 24 Dixon-Woods M, Tarrant C. Why do people cooperate with medical research? exclude a wealth agenda so long as there is confidence that Findings from three studies. Soc Sci Med 2009;68(12):2215–22. research governance and information governance systems can 25 Dixon-Woods M, Wilson D, Jackson C, et al. Human tissue and ‘the public’: the case of childhood cancer tumour banking. BioSocieties 2008;3(1):57–80. hold researchers, and others with custodial responsibility for 26 Hill EM, Turner EL, Martin RM, et al. “Let’s get the best quality research we can”: medical information, to account. public awareness and acceptance of consent to use existing data in health research: a systematic review and qualitative study. BMC Med Res Methodol 2013;13(1):1–10. Twitter Follow Mary DixonWoods at @MaryDixonWoods 27 Kenyon S, Dixon-Woods M, Jackson CJ, et al. Participating in a trial in a critical situation: a qualitative study in pregnancy. Qual Saf Health Care 2006;15 Contributors PC conceived of the project and drafted the article. MD-W and GL (2):98–101. reviewed several drafts, suggested new literature and commented on and revised the 28 Dixon-Woods M, Tarrant C, Jackson CJ, et al. Providing the results of research to structure and intellectual content. participants: a mixed-method study of the benefits and challenges of a consultative Funding Graeme Laurie holds a Wellcome Trust Senior Investigator Award approach. Clin Trials 2011;8(3):330–41. (WT103360MA) and is Co-Investigator on the 10-funder Farr Institute of Health 29 Callaway E. UK push to open up patients’ data. Nature 2013;502:283. 408 Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 Current controversy 30 NHS England. Better information means better care, 2014. 41 Brown I, Brown L, Korff D. Using NHS patient data for research without consent. 31 Health and Social Care Information Centre. General Practice Extraction Service Law Innov Technol 2010;2(2):219–58. (GPES) customer requirement summary NIC-178106-MLSWX. Health and Social 42 Thomas R, Walport M. Data sharing review. Copyright Richard Thomas and Mark Care Information Centre, 2012. Walport, 2008. 32 Sciencewise. Big data: public views on the collection, sharing and use of personal 43 Ohm P. Broken promises of privacy: responding to the surprising failure of data by government and companies. Sciencewise, 2014. http://www. anonymization. UCLA Law Rev 2010;57:1701. sciencewise-erc.org.uk/cms/big-data 44 Francis R. Report of the mid Staffordshire NHS Foundation Trust public inquiry. 33 Raman S, Mohr A. A social licence for science: capturing the public or London: Stationery Office, 2013. http://cdn.midstaffspublicinquiry.com/sites/default/ co-constructing research? Soc Epistemol 2014;28(3–4):258–76. files/report/Executive%20summary.pdf 34 Garrety K, McLoughlin I, Wilson R, et al. National electronic health records and the 45 Macfarlane A, Pollock A. Opting out of care.data is not the answer. 2014. http:// digital disruption of moral orders. Soc Sci Med 2013;101(0):70–7. www.statslife.org.uk/opinion/1210-opting-out-of-care-data-is-not-the-answer 35 Davidson S, McLean C, Treanor S, et al. Public acceptability of data sharing 46 Haddow G, Laurie G, Cunningham-Burley S, et al. Tackling community concerns between the public, private and third sectors for research purposes. Scottish about commercialisation and genetic research: a modest interdisciplinary proposal. Government Social Research, 2013. http://www.scotland.gov.uk/socialresearch Soc Sci Med 2007;64:272–82. 36 Wellcome Trust. Towards consensus for best practice: use of patient records from 47 Ipsos MORI. Public Dialogue Workshops Report prepared for the Health Research general practice for research. London, 2009. Authority. Ipsos MORI, 2013. 37 Stone MA, Redsell SA, Ling JT, et al. Sharing patient data: competing demands of 48 Hunn A. Survey of the general public: attitudes towards health research. Health privacy, trust and research in primary care. Br J Gen Pract 2005;55(519):783–9. Research Authority, 2013. 38 Perera G, Holbrook A, Thabane L, et al. Views on health information sharing and 49 Chalmers I, Glasziou P, Godlee F. All trials must be registered and the results privacy from primary care practices using electronic medical records. Int J Med Inf published. BMJ 2013;346:f105. 2011;80(2):94–101. 50 McCartney M. Care.data: why are Scotland and Wales doing it differently? BMJ 39 medConfidential. What’s the story? 2013. http://medconfidential.org/ 2014;348:g1702. whats-the-story/ (accessed 7 Nov 2013). 51 Sethi N, Laurie GT. Delivering proportionate governance in the era of 40 Murtagh M, Thorisson G, Wallace S, et al. Navigating the perfect [data] storm. eHealth: making linkage and privacy work together. Med Law Int 2013;13 Norsk Epidemiol 2012;21(2):203–9. (2–3):168–204. Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 409 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Medical Ethics British Medical Journal

The social licence for research: why care.data ran into trouble

Loading next page...
 
/lp/british-medical-journal/the-social-licence-for-research-why-care-data-ran-into-trouble-Wml1cv69ya

References (60)

Publisher
British Medical Journal
Copyright
Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://group.bmj.com/group/rights-licensing/permissions
ISSN
0306-6800
eISSN
1473-4257
DOI
10.1136/medethics-2014-102374
Publisher site
See Article on Publisher Site

Abstract

Current controversy The social licence for research: why care.data ran into trouble 1 2 1 Pam Carter, Graeme T Laurie, Mary Dixon-Woods Health Sciences, University of ABSTRACT Although care.data has numerous aims (box 1), Leicester, Leicester, UK In this article we draw on the concept of a social licence we focus specifically on its research purposes. We Department of Law, University to explain public concern at the introduction of care. begin by offering some brief background on the use of Edinburgh, Edinburgh, UK data, a recent English initiative designed to extract data and regulation of routine medical data before intro- Correspondence to from primary care medical records for commissioning and ducing the concept of a social licence. Professor Mary Dixon-Woods, other purposes, including research. The concept of a Health Sciences, University of social licence describes how the expectations of society Leicester, 2nd Floor, Adrian THE USE AND REGULATION OF MEDICAL regarding some activities may go beyond compliance Building, University of RECORDS FOR RESEARCH Leicester, University Road, with the requirements of formal regulation; those who Researchers have long relied on access to personal Leicester LE1 7RH, UK; do not fulfil the conditions for the social licence (even if medical information routinely collected during the md11@le.ac.uk formally compliant) may experience ongoing challenge course of patient care in order to conduct studies, and contestation. Previous work suggests that people’s Received 17 July 2014 including clinical trials and epidemiological cooperation with specific research studies depends on Revised 24 November 2014 research. However, the repurposing of routinely Accepted 30 December 2014 their perceptions that their participation is voluntary and collected data for research is not without risk to Published Online First is governed by values of reciprocity, non-exploitation and 8 relevant values, and measures such as anonymisa- 23 January 2015 service of the public good. When these conditions are tion (even when possible) do not solve all ethical, not seen to obtain, threats to the social licence for legal and technical problems; people may, for research may emerge. We propose that care.data failed example, have religious or moral objections to par- to adequately secure a social licence because of: 5 ticular studies or concerns about stigma and (i) defects in the warrants of trust provided for care.data, breaches of privacy. (ii) the implied rupture in the traditional role, Accordingly, researchers’ access to, and use of, expectations and duties of general practitioners, and personal data has traditionally been subject both to (iii) uncertainty about the status of care.data as a public tight controls and to efforts to promote public and good. The concept of a social licence may be useful in patient confidence in those controls. The Data explaining the specifics of care.data, and also in Protection Act 1998, which implements the reinforcing the more general lesson for policy-makers European Data Protection Directive, has always that legal authority does not necessarily command social recognised ‘medical purposes’ as a legal basis for legitimacy. processing personal data from medical records, subject to a proportionality test. While it contains a ‘research exemption’, it does not absolve data con- trollers completely of their obligations under the BACKGROUND law. Data protection has been bolstered by the Encouraging more and better health-related research requirement, introduced in 1999, that each NHS is now firmly established as a policy goal in organisation have a designated senior individual 1–3 England. A crucial feature of the current policy (known as a Caldicott Guardian) with responsibility drive is an emphasis on the use of the distinctive both for the confidentiality of patient information infrastructure of the personal medical records of —which is protected by the common law—and for individuals registered with the National Health enabling appropriate information sharing. These Service (NHS). Researchers have long argued that, legal measures also operate alongside guidance and contrary to the public interest, these records have codes of practice issued and updated regularly by remained under-exploited as a research resource. the professional regulators, by the NHS Research Yet the recent launch of care.data—an initiative to Governance Framework, by a system of research Open Access extract data from NHS primary care medical ethics committees, and by a variety of other Scan to access more records in England unless patients have purposefully mechanisms. free content opted out, in part to facilitate research—has proved Nonetheless, access to and secondary use of deeply controversial. Within weeks of the start of a medical information for research purposes has con- public information campaign run by NHS England, tinued to be a fraught and contested activity, prone public and professional concern had become so to concern over its ethical standing. The tensions overwhelming that in February 2014 the implemen- were discussed extensively in a 2006 report of the 5 4 tation of the scheme was suspended for 6 months. Academy of Medical Sciences (AMS), in a docu- Our aim in this article is not to assess the rights and ment that can be seen as an attempt at advocacy on wrongs of care.data—others have already offered the part of a clinical research community that saw such critiques —nor to offer an in-depth analysis of itself as beleaguered and misunderstood. The To cite: Carter P, Laurie GT, the legal background, but rather to show how the report offered multiple examples of the benefits of Dixon-Woods M. J Med concept of a social licence can help to explain the research that uses routinely collected personal data, Ethics 2015;41:404–409. challenges faced by care.data. including its possibilities as a means of answering 404 Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 Current controversy grounds that the risks involved were low in comparison with the likely benefits and in comparison with other forms of Box 1 Aims of care.data research (such as clinical trials). This claim was in keeping with the influential (although not necessarily uncontested) view that NHS England has described the care.data service as: ‘…anew, over-emphasis on individual consent fails to balance a broader modern data system for the NHS in England. Known as care. range of ethical principles and the argument that people have a data, its purpose will be to provide timely, accurate information 14 15 responsibility to participate in research. The tenor of the to citizens, clinicians and commissioners about the treatments arguments was that the proper solution to the challenges sur- and care provided by the NHS’ (http://www.england.nhs.uk/ rounding use of routine medical data for research purposes wp-content/uploads/2013/05/ces-tech-spec-gp-extract.pdf ). might then be one of adjusting the regulatory environment to The aims of the care.data programme are sixfold: accommodate what the public would support. 1. To support patients’ choice 2. To advance customer services A SOCIAL LICENCE FOR RESEARCH INVOLVING MEDICAL 3. To promote greater transparency RECORDS? 4. To improve outcomes Based on the reasoning that the public would accept and 5. To increase accountability endorse a less restrictive approach to the use of medical records 6. To drive economic growth by making England the default for research purposes than the custodians of the system realised, location for world-class health services research. the AMS’s clarion call was to ‘be bold’. Although it did not use NHS, National Health Service. the term explicitly, the core of the AMS argument was that the social licence for the use of medical records for research was more permissive than the operation of the regulatory environ- primary research questions, as a prelude to clinical trials, as a ment allowed. means for disease surveillance, and as a resource for identifying The concept of a social licence is discussed in two distinct lit- potential participants in trials or other studies. But such activ- eratures. One important literature derives from the work of the ities were, the AMS report argued, continually thwarted by the sociologist Everett Hughes who describes the relationship complexities and frustrations of multiple legal, regulatory and between a profession and society in terms of the two concepts professional restrictions. of licence and mandate. For Hughes, licence is granted to The AMS report and others, while accepting the need for certain occupational groups to carry out particular activities; a regulation, argued that the regulatory environment for the use mandate is claimed by members of the group to define for itself of personal data for health research in England was, by 2006, proper conduct in relation to matters concerned with their not in the public interest: it was disproportionate relative to the work. A second, and mostly distinct, corpus of work on cor- risks and benefits involved and involved multiple overlapping porate social responsibility describes the concept of the ‘social layers, institutions and actors as well as contested and sometimes licence to operate’ as the expectations of society regarding the conflicting standards and expectations for research governance. conduct and activities of corporations that go beyond the While accepting the broad principle of the need for regulation, requirements of formal regulation. Thus, industries with sig- the report suggested that NHS organisations were overly cau- nificant environmental impact may find that operating within tious about legal risk in areas of interpretive uncertainty, and the law but outside the boundaries of social approval can result tended to default to a conservative approach. The way the rules in corporate damage—for example, by having a negative impact relating to data protection were understood, interpreted and on a company’s brand or provoking new and restrictive regula- applied by NHS organisations, research ethics committees and tion. The Australian mining industry is often offered as an others was a focus of particular criticism. The overall effect, the example of an industry required to earn a social licence, and report and others argued, was that NHS organisations were gen- maintain it, by behaving in a trustworthy and responsible way; if erally reluctant to share personal information without explicit it were otherwise, it would face ongoing challenge and costly consent from patients or anonymisation of data to remove them delays and interference in its activities. from the reach of data protection legislation: the dichotomous Some analogies have been drawn between a social licence for ‘consent or anonymise’ approach was the norm. the mining of minerals and for the process of data mining. Yet this was not the only possible approach, nor was it Regarding these and similar activities, what the social licence required by law. Under what was Section 60 of the Health and emphasises is the possible need for those (whether they are Social Care Act (2001) (later Section 251 of the NHS Act public or private bodies, or specific occupational groups) under- 2006), a Patient Information Advisory Group (PIAG), acting on taking activities likely to provoke public disquiet to go ‘beyond behalf of the Secretary of State, was allowed to approve process- compliance’ with legal requirements. We propose that a poorly ing of patient information without seeking patients’ consent for informed understanding of the social licence for secondary use research purposes. However, the AMS report criticised the of personal medical data, and a failure to recognise that legal PIAG’s approach, processes and membership, and its alleged authority might not be enough to secure the social licence, tendency to stress its role in protecting privacy or confidentiality seems to have been at the heart of the controversy underlying ‘without equal emphasis on the benefits derived from well- care.data. conducted research’. PIAG was argued to suffer from mission creep, applying a stricter standard of absolute and proven neces- Legal authority for care.data sity rather than assessing whether applications involved a pro- The Health and Social Care Act 2012 established the Health portionate and justifiable interference in privacy. and Social Care Information Centre (HSCIC) as an Executive Central to the AMS report’s argument for reform was that Non-Departmental Public Body empowered to obtain patient- PIAG appeared to adopt a more conservative approach than identifiable information from general practices. Practices them- patients and the public would favour. It claimed strong public selves were, under their NHS contract, obliged to transfer the approval and support for research using medical records, on the data to HSCIC unless patients explicitly opted out. Using a Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 405 Current controversy centrally managed General Practice Extraction Service (GPES) defined, consenting patients can be generalised to the broader based on commercially supplied software, care.data was conception of NHS citizenship implicit in the new policy direc- intended to gather data monthly, capturing most routine general tion is not clear. practice consultations. The data would then be made available What is clear is that individuals’ cooperation with specific research studies is usually secured through three principal in aggregate form to the HSCIC, with six aims (box 1) identi- fied for their use, including that of driving ‘economic growth by mechanisms: their expectations about how research is conducted making England the default location for world-class health ser- and regulated; their trust in the institutions and individuals who vices research’. recruit them; and their beliefs in the wholesomeness and public Applications to use the data are to be reviewed by the value of the research endeavour. More broadly, the public legit- HSCIC’s Data Access Advisory Group. If requests to use data imacy and acceptability of health research rests heavily on its are classified as sensitive and identifiable, researchers are subject status as a socially valuable enterprise conducted in the service to additional requirements, including that of undergoing scru- of the public good. tiny from the Confidentiality Advisory Group of the Health There are many reasons to doubt that care.data could reason- Research Authority. This group replaced the PIAG and an ably assume that the public would automatically confer upon it interim body (the Ethics and Confidentiality Committee operat- the same legitimacy and endorsement as that enjoyed by ing under the National Information Governance Board). It con- research where individual informed consent is sought and clear tinues the role of advising on data use for research when information about study aims is provided. For instance, the individual consent has not been obtained on the basis of mobility of electronic data and the practical difficulties of speci- so-called Section 251 approvals. fying in advance the research questions for which data might be Other changes relevant to the creation of care.data included used or the populations to be studied mean that care.data was in amendments in 2012 to the NHS Constitution to offer a many ways quite distinct from conventional research projects. ‘pledge’ to inform English NHS patients about research studies Three threats to the social licence faced by care.data are espe- in which they might be eligible to participate, and also, crucially, cially important in explaining the challenges it faced: (i) defects an expectation that patients would be willing to share their in the warrants of trust provided for care.data, (ii) the implied medical information for healthcare planning and for research rupture in the traditional role, expectations and duties of purposes. The relevant section—‘Respect, consent and confiden- general practitioners (GPs), and (iii) uncertainty about the status tiality’ (p. 8)—explains that patients have the right to request of care.data as a public good. that their information should not be shared but also that such requests may not be upheld if the public interest is found to out- WARRANTS OF TRUST weigh the individual case. Ensuring that people are aware of how data from medical Thus was the administrative infrastructure for care.data records has been or might be used, and the protections that are created—but not, it seems, the social licence. The amendments in place, might be assumed to be critical to ensuring their confi- to the Health and Social Care Act 2012 and to the NHS dence that such use is legitimate and well-governed. A system- Constitution seem to have been based on a re-imagining and atic review of public awareness of, and views on consent to, the re-responsibilisation of patients as active citizens. There has secondary use of medical records for health has identified a gen- been an assumption that the implicit social contract underpin- eralised lack of awareness and understanding. Yet level of ning the NHS would mean most patients would be happy to ‘awareness’ itself may not be a reliable guide to people’s faith in support sharing of their personal data in the interests of the six a system. Individuals’ willingness to cooperate with health aims of the care.data programme, including economic goals to research often depends on powerful heuristics rather than be realised through research. detailed understanding of research governance procedures and Yet the extent to which confidence in such assumptions was standards. Even when they agree to take part in specific well-founded was not clear. Previous social science research sug- studies, it is not the detail of consent forms and information gests a nuanced and delicate understanding of societal support sheets that matters to participants so much as an overall faith in for, and cooperation with, health research is needed. It has the legitimacy of the endeavour and a reassurance of protection 22 23 27 28 found that the public’s support and tolerance for research, and from risk. Important, too, is a belief that agreeing to its associated risks, often depends far more on an often fragile participate in research will not make individuals vulnerable to set of cues about the safety and social good of research partici- risk of harm, exploitation or charges of gullibility. This is not pation, and on institutional and professional credentials, than it to imply that consent forms and information sheets are of no does on the formal architecture of research regulation, or on value or relevance; on the contrary, they act as ‘symbolic rational assessment of the detail of information sheets or other tokens’, as vitally important warrants of the trustworthiness of 22 23 22 documents aimed at gaining ‘informed consent’. the process. Further, most of what is known about patients’ support for Thus, a lack of detailed awareness might have mattered little research is based on quite particular examples of research par- in the roll-out of care.data had the right warrants of trust been ticipation—often those where patients already have an interest in place. But the approach adopted did not appear to provide in a medical condition and where they are asked for quite spe- them. The implied consent model underlying care.data did cific consent to a project or programme. Where patients specific- allow for patients to opt out, but the process for ensuring that ally consent to a particular study or programme of research, they were enabled to do so was precarious. The practical they may (depending on the study) expect the possibility of delivery of the requirement to ensure patients were aware of some personal benefit, as well as being able to experience sec- care.data and their right to opt out proved problematic. The ondary benefits of their consent, such as the ‘warm glow’ asso- NHS England information leaflet about the initiative was ciated with altruistic behaviour that benefits an identifiable addressed to households rather than individuals and might have community (eg, of those with a specific disease). The extent appeared to be unsolicited junk mail. to which the findings of this body of research about participa- For those seeking to improve their understanding, the leaflet tion in specific, relatively well-bounded studies or cohorts by itself did not appear to be fully informative. Entitled ‘Better 406 Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 Current controversy Information Means Better Care’, it made no explicit reference to opportunity to address novel research questions but also gen- care.data other than to direct those seeking further information erates new ethical dilemmas in relation to re-identification of 14 41–43 to an NHS website. It did not include an opt-out form. The bene- individuals or losses of data and other intrusions asso- fits of information sharing were explained in the patient informa- ciated with risks to anonymity and privacy. Again, the extent to tion leaflet, but the risks were not. HSCIC’sown which the social licence for these potentially profound altera- acknowledgement that ‘a small risk of re-identification due to tions to the nature of the doctor–patient relationship was small numbers/rare diseases’ was not conveyed. Questions secured for care.data is in question. about who would access data, for which purposes and on which terms, were also left unclear, despite evidence showing that the STATUS OF CARE.DATA AS A PUBLIC GOOD public wanted more information about how governments and Values of reciprocity and fairness underpin the legitimacy of companies collect, share and use data. Additionally, it was not health research: if people participate voluntarily, they expect clear that busy GPs would have the time or sufficient information that their contribution will be used to improve the care of 23 24 to engage in complex discussions with patients wishing to opt others, and that their good faith will not be exploited. out, nor was any extra resource or support provided for the task, Much depends, therefore, on the extent to which uses of per- even for vulnerable groups such as care home residents. The con- sonal data are seen as serving the public interest and conducted cerns of GPs regarding losing control of patient data during this by those with a public interest orientation. Yet neither of these process were also not addressed. Further, few opportunities were features of the social licence seems to have been fully addressed 33–35 available to the public to influence the direction of policy. for care.data. Overall, in contrast to the cues usually available to research The ‘public regime of justification’ that was provided for candidates, the warrants of trust provided by care.data appeared care.data stressed its benefits at a national rather than an individ- inadequate and were likely to have undermined the social ual level. Of the six aims of care.data (box 1), the first five are licence rather than strengthen it. Lack of detail in the informa- rooted in the use of data to improve quality and delivery of care tion sheets was unlikely to have been the direct cause of the and the governance of healthcare, with the recommendations of failure to secure the social licence, but it does provide visible Sir Robert Francis arising from his damning report on Mid evidence of how the operationalisation of the programme was Staffordshire NHS Trust invoked to supply part of the motiv- founded in poor understanding of what was needed for the ation. However, there are indications that members of the public and professionals to have trust and confidence in it. public expect to see benefits of such data use made explicit. Inadequate consultation and engagement meant a failure to take The sixth aim of care.data refers to research, but in a way that into account ‘the right things’ and thus secure the social licence. links it explicitly to economic growth. The multiplicity of aims towards which care.data is directed, RUPTURE IN TRADITIONAL EXPECTATIONS and the linking of care.data and research to the so-called ‘health The legitimacy of health research draws heavily and crucially on and wealth’ agenda, may be an unfamiliar rationale for citizens 7 45 people’s trust in organisational and professional credentials, used to a public service model of healthcare. Patients may mis- including the trust placed in professionals who have a duty of con- trust commercial interests, especially where these might be per- 36 26 32 35 46–48 fidentiality and relationships developed through continuity of ceived as profiteering or resulting in excessive profit care and empathy. GPs are constrained both by the common law or where patients have concerns about the extent to which risks duty of confidentiality and statutory requirements under the Data and benefits are evenly distributed and whether their contribu- Protection Act 1998 and Human Rights Act 1998 to respect tions to research will be reciprocated by a contribution to the patient confidences and to process personal data or disclose confi- public good. The persistent problem of non-publication of study dential information only within clearly defined parameters. results is just one element of how non-reciprocation may mani- However, in neither legal regime is the protection of privacy abso- fest. Other concerns may be linked to questions of who will be lute. A 2013 review of the system of information governance in able to access care.data for which purposes with which risks, how health and social care services clarified, for example, that legally the credentials of bona fide researchers can be established, and patients do not own their data. Notwithstanding, what is legally what mandate commercial organisations will have to use data permissible and what is socially acceptable do not necessarily coin- that originated from private consultations between patients and cide. For some, care.data represents a significant and unwelcome their GPs. alteration to traditional understandings of the private and confi- The complexity is amplified by the institutional fragmentation dential nature of the relationship between GP and patient. of the NHS in England, which is no longer a single unitary Some accounts argue that routine monthly extraction of elec- public sector entity. Transfer of information ‘within the NHS’ in tronic individual level patient data from GP records represents a practice means transfer between entities with distinct legal step-change in terms of the level of intrusion into patients’ status. This takes place in a complex framework of statutory private lives, to the extent that it has been suggested that it provisions and legal contracts, and it increasingly occurs with erodes the fundamental right to respect for private life offered those who are private contractors of services to the NHS or by the Human Rights Act. Concerns have been expressed that may operate outside the NHS entirely. Concerns about insur- patients may not wish intimate details of their medical records ance companies and others having access to personal data to be accessible to those outside of their ‘circle of care’, and suggest that considerable further dialogue between the public, might decline to trust and confide in their healthcare providers affected stakeholders such as GPs, and policy makers is required if they know that their information will be shared. The law before a social licence can be said to be in place for care.data. has long recognised the importance of protecting this trust for both the private and public interests that it serves (see, for CONCLUSIONS example, the cases of X v Y (1988) 2 ALL ER 648 and also The concept of a social licence offers insight into why the Campbell v Mirror Group Newspapers (2004) UKHL 22). implementation of care.data, an initiative to extract data from These concerns are linked in part to the advances in techno- routine general practice consultations for use in research and logical capability associated with ‘big data’ which offer the other activities, has met with concern and controversy. Although Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 407 Current controversy Informatics Research (MRC Grant No: MR/K007017/1, MR/K006525/1, MR/ the infrastructure was in place, the activities were perfectly K006584/1 and MR/K006665/1). Professor Mary Dixon-Woods holds a Senior lawful, and a case had been made for the possible benefits that Investigator Award from the Wellcome Trust (WT097899) that has supported this might be generated, the experience of care.data starkly exposes work. an enduring truism about the limits of law: legal authority does Competing interests None. not necessarily command social legitimacy. A parliamentary Provenance and peer review Not commissioned; externally peer reviewed. majority may allow legislation to be passed, but that does not Open Access This is an Open Access article distributed in accordance with the equate to a societal seal of approval or to securing the trust and terms of the Creative Commons Attribution (CC BY 4.0) license, which permits confidence of patients, citizens, healthcare professionals and others to distribute, remix, adapt and build upon this work, for commercial use, researchers. Securing a social licence may require something provided the original work is properly cited. See: http://creativecommons.org/ other than a legal mandate. licenses/by/4.0/ Although England has remained firmly on the legislative path, different solutions to the same challenges relating to the use of REFERENCES personal medical information have been found north of the 1 Department of Health. Best research for best health: a new national health research border. Rather than legislative change, actors operating within strategy. The NHS contribution to health research in England. London: Department of Health, 2006. the existing legal frameworks have worked together under the 2 Liddell A, Ayling M, Reid G. Innovation, health and wealth accelerating adoption Scottish Health Informatics Programme (SHIP) to develop and diffusion in the NHS. Department of Health, 2011. mechanisms of principled proportionate governance—building 3 HM Treasury and Department for Business Innovation and Skills. The plan for on public engagement exercises—to deliver responsive, risk- growth. Crown Copyright. 2011. based approaches to data linkage for health research. Consent, 4 The Academy of Medical Sciences. Personal data for public good: using health information in medical research. London: The Academy of Medical Sciences, 2006. anonymisation and authorisation are all available as governance 5 Taylor M. Information governance as a force for good? Lessons to be learnt from tools to be deployed, as appropriate, in any given data linkage care.data. Script Ed 2014;11(1):1–8. proposal after an assessment of the risks and tolerances 6 Pollock AM, Roderick P. Trust in the time of markets: protecting patient information. involved. Lancet 2014;383(9928):1523–4. 7 Dixon-Woods M, Ashcroft RE. Regulation and the social licence for medical As the Scottish example shows, this does not mean that research. Med Health Care Philos 2008;11:381–91. unanimous social consensus is required for all developments. 8 Clark S, Weale A. Information governance in health. Nuffield Trust, 2011. Rather, a social licence for research will require, as a minimum, 9 Caldicott F. Information: to share or not to share? The Information Governance that certain conditions of social engagement have been Review. Crown Copyright. London: Williams Lea for the Department of Health, respected. Genuine dialogic engagement that might result in a 2013. 10 Taylor MJ. Health research, data protection, and the public interest in notification. broad licence must be distinguished from more narrowly Med Law Rev 2011;19(2):267–303. focused public relations exercises that seek to ‘capture’ the 11 Department of Health. Research governance framework for health and social care. public, that is, to persuade the public of the legitimacy of deci- 2nd edn. https://www.gov.uk/government/uploads/system/uploads/attachment_data/ 33 35 sions already taken by experts, and from simple ‘awareness- file/139565/dh_4122427.pdf (accessed 24 Oct 2014). 12 Doyal L, Tobias JS. Informed consent in medical research. London: BMJ Books, raising’ information exercises. While legitimate disagreement is inevitable, if a social licence is to be maintained, both the final 13 Walley T. Using personal health information in medical research. BMJ 2006;332 result and process used to achieve that result must be one which (7534):130–1. reasonable citizens can at least recognise as defensible on the 14 McHale J. Reforming the regulation of health research in England and Wales: new grounds that it reflects common social values and goals. challenges: new pitfalls. J Med Law Ethics 2013;1(1):23–42. 15 Harris J. Scientific research is a moral duty. J Med Ethics 2005;31(4):242–8. Trust and confidence in research governance or the ‘social 16 The Academy of Medical Sciences. A new pathway for the regulation and licence for research’ depends upon ideas about the public good governance of health research. The Academy of Medical Sciences, 2011. http:// that are not straightforwardly synonymous with the aim of www.acmedsci.ac.uk/p99puid209.html increasing the UK’s gross domestic product. What patients care 17 Hughes EC. Men and their work. Glencoe, Illinois: Free Press, 1958. 18 Gunningham N, Robert A, Thornton D. Social license and environmental protection: about as patients cannot be equated with what patients care why businesses go beyond compliance. Law Soc Inq 2004;29(2):307–41. about as citizens who are part of a much wider social endeav- 19 Allen G. Public affairs and mining in Australia–a rescued license to operate. J Public our. If care.data is to succeed, patients need to have the confi- Aff 2011;11(4):382–6. dence that their medical records will be held securely, 20 Rooney D, Leach J, Ashworth P. Doing the social in social license. Soc Epistemol anonymised appropriately, and that secondary use of this per- 2014;28(3–4):209–18. 21 Department of Health, London. NHS Constitution—the NHS belongs to us all. sonal data is in the public interest: the conditions of the social licence need to be respected in ways that go beyond compliance 22 Armstrong N, Dixon-Woods M, Thomas A, et al. Do informed consent documents laid down in a legal framework. Necessary—and hopefully suffi- for cancer trials do what they should? A study of manifest and latent functions. cient—conditions for social licence include: (i) reciprocity, Sociol Health Illn 2012;34(8):1230–45. 23 Dixon-Woods M, Ashcroft RE, Jackson CJ, et al. Beyond “misunderstanding”: which must begin with sound two-way communication, (ii) non- written information and decisions about taking part in a genetic epidemiology exploitation, which must exclude the spectre of disempower- study. Soc Sci Med 2007;65(11):2212–22. ment, and (iii) service of the public good, which need not 24 Dixon-Woods M, Tarrant C. Why do people cooperate with medical research? exclude a wealth agenda so long as there is confidence that Findings from three studies. Soc Sci Med 2009;68(12):2215–22. research governance and information governance systems can 25 Dixon-Woods M, Wilson D, Jackson C, et al. Human tissue and ‘the public’: the case of childhood cancer tumour banking. BioSocieties 2008;3(1):57–80. hold researchers, and others with custodial responsibility for 26 Hill EM, Turner EL, Martin RM, et al. “Let’s get the best quality research we can”: medical information, to account. public awareness and acceptance of consent to use existing data in health research: a systematic review and qualitative study. BMC Med Res Methodol 2013;13(1):1–10. Twitter Follow Mary DixonWoods at @MaryDixonWoods 27 Kenyon S, Dixon-Woods M, Jackson CJ, et al. Participating in a trial in a critical situation: a qualitative study in pregnancy. Qual Saf Health Care 2006;15 Contributors PC conceived of the project and drafted the article. MD-W and GL (2):98–101. reviewed several drafts, suggested new literature and commented on and revised the 28 Dixon-Woods M, Tarrant C, Jackson CJ, et al. Providing the results of research to structure and intellectual content. participants: a mixed-method study of the benefits and challenges of a consultative Funding Graeme Laurie holds a Wellcome Trust Senior Investigator Award approach. Clin Trials 2011;8(3):330–41. (WT103360MA) and is Co-Investigator on the 10-funder Farr Institute of Health 29 Callaway E. UK push to open up patients’ data. Nature 2013;502:283. 408 Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 Current controversy 30 NHS England. Better information means better care, 2014. 41 Brown I, Brown L, Korff D. Using NHS patient data for research without consent. 31 Health and Social Care Information Centre. General Practice Extraction Service Law Innov Technol 2010;2(2):219–58. (GPES) customer requirement summary NIC-178106-MLSWX. Health and Social 42 Thomas R, Walport M. Data sharing review. Copyright Richard Thomas and Mark Care Information Centre, 2012. Walport, 2008. 32 Sciencewise. Big data: public views on the collection, sharing and use of personal 43 Ohm P. Broken promises of privacy: responding to the surprising failure of data by government and companies. Sciencewise, 2014. http://www. anonymization. UCLA Law Rev 2010;57:1701. sciencewise-erc.org.uk/cms/big-data 44 Francis R. Report of the mid Staffordshire NHS Foundation Trust public inquiry. 33 Raman S, Mohr A. A social licence for science: capturing the public or London: Stationery Office, 2013. http://cdn.midstaffspublicinquiry.com/sites/default/ co-constructing research? Soc Epistemol 2014;28(3–4):258–76. files/report/Executive%20summary.pdf 34 Garrety K, McLoughlin I, Wilson R, et al. National electronic health records and the 45 Macfarlane A, Pollock A. Opting out of care.data is not the answer. 2014. http:// digital disruption of moral orders. Soc Sci Med 2013;101(0):70–7. www.statslife.org.uk/opinion/1210-opting-out-of-care-data-is-not-the-answer 35 Davidson S, McLean C, Treanor S, et al. Public acceptability of data sharing 46 Haddow G, Laurie G, Cunningham-Burley S, et al. Tackling community concerns between the public, private and third sectors for research purposes. Scottish about commercialisation and genetic research: a modest interdisciplinary proposal. Government Social Research, 2013. http://www.scotland.gov.uk/socialresearch Soc Sci Med 2007;64:272–82. 36 Wellcome Trust. Towards consensus for best practice: use of patient records from 47 Ipsos MORI. Public Dialogue Workshops Report prepared for the Health Research general practice for research. London, 2009. Authority. Ipsos MORI, 2013. 37 Stone MA, Redsell SA, Ling JT, et al. Sharing patient data: competing demands of 48 Hunn A. Survey of the general public: attitudes towards health research. Health privacy, trust and research in primary care. Br J Gen Pract 2005;55(519):783–9. Research Authority, 2013. 38 Perera G, Holbrook A, Thabane L, et al. Views on health information sharing and 49 Chalmers I, Glasziou P, Godlee F. All trials must be registered and the results privacy from primary care practices using electronic medical records. Int J Med Inf published. BMJ 2013;346:f105. 2011;80(2):94–101. 50 McCartney M. Care.data: why are Scotland and Wales doing it differently? BMJ 39 medConfidential. What’s the story? 2013. http://medconfidential.org/ 2014;348:g1702. whats-the-story/ (accessed 7 Nov 2013). 51 Sethi N, Laurie GT. Delivering proportionate governance in the era of 40 Murtagh M, Thorisson G, Wallace S, et al. Navigating the perfect [data] storm. eHealth: making linkage and privacy work together. Med Law Int 2013;13 Norsk Epidemiol 2012;21(2):203–9. (2–3):168–204. Carter P, et al. J Med Ethics 2015;41:404–409. doi:10.1136/medethics-2014-102374 409

Journal

Journal of Medical EthicsBritish Medical Journal

Published: May 23, 2015

Keywords: Research Ethics Informed Consent Confidentiality/Privacy Information Technology Regulation

There are no references for this article.