Social Science & Medicine 131 (2015) 289e296

Contents lists available at ScienceDirect

Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed

Patient involvement in drug licensing: A case study Nicky Britten a, *, Sarah Denford a, Faith Harris-Golesworthy b, Steph Jibson b, Nigel Pyart b, Ken Stein a a b

Institute of Health Research, University of Exeter Medical School, Veysey Building, Salmon Pool Lane, Exeter EX2 4SG, UK Peninsula Patient Involvement Group, University of Exeter Medical School, Veysey Building, Salmon Pool Lane, Exeter EX2 4SG, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 15 October 2014

Embodied health movements work on the boundary between lay and expert knowledge. Consumer groups, depending on their goals, may increase or decrease pharmaceuticalization. This paper reports a small case study about the retrospective evaluation of a specific second line treatment for type 2 diabetes by an existing patient involvement group. The group is part of a research collaboration between academia and the health service in England, and shares some characteristics of embodied health movements. We used the case study to explore whether an institutionally funded non activist patient group can make a more balanced contribution to drug licensing decisions than that made by either access-oriented or injury-oriented consumer groups, without being co-opted by an institutional agenda. The questions we wished to address were how this group evaluated existing mechanisms for licensing drugs; how they balanced scientific and lay knowledge; how they made their decisions; and how they viewed their experiences as panel members. The five panel members were interviewed before and after the panel discussion in July 2013. They were critical of current licensing processes, and used their own embodied experiences of medicines to evaluate expert knowledge. Their decisions on the panel were informed either by a balancing of benefits and harms, or by trust in experts. The case study suggests that such a group may have the potential both to balance the pro-pharmaceuticalization impact of accessoriented groups and to influence forms of pharmaceutical governance. © 2014 Elsevier Ltd. All rights reserved.

Keywords: UK Drug licensing Public involvement Pharmaceuticals Patient groups Health social movements

1. Introduction Health professionals, by the nature of their training and work, necessarily belong to social groups. The same cannot be said for patients who may have little contact with others in the same situation, although the internet is changing this. However there is an increasing range of social groups whose members share aspects of patienthood, such as self-help groups and disease-based campaigning groups. Perhaps the most well-known are those patient groups lobbying for access to new drugs, whose activities are often reported in the media. Such groups can accelerate regulatory approval (Carpenter, 2004), thus contributing to processes of pharmaceuticalization (Williams et al., 2011). In his work on the Food and Drug Administration (FDA) Drug Review process, Carpenter (2004) identified more than 3000 disease and patient

* Corresponding author. E-mail addresses: [email protected] (N. Britten), [email protected] (S. Denford), [email protected] (F. Harris-Golesworthy), [email protected] (S. Jibson), [email protected] (N. Pyart), [email protected] (K. Stein). http://dx.doi.org/10.1016/j.socscimed.2014.10.024 0277-9536/© 2014 Elsevier Ltd. All rights reserved.

advocacy groups in the US representing various medical conditions for which new drug applications had been submitted. Abraham has defined pharmaceuticalization as ‘the process by which social, behavioural or bodily conditions are treated or deemed to be in need of treatment, with medical drugs by doctors or patients’ (Abraham, 2010, p. 604). He considered the role of consumerism as a driver of pharmaceuticalization and noted that there are two types of active consumerism: one based on injury-oriented adversity to the pharmaceutical-industrial complex, and one based on access-oriented collaboration with it. The ability of such groups to achieve their ends depends on whether they support or oppose the interests of the pharmaceutical industry. Abraham concluded that although both types of group do influence processes of pharmaceuticalization, the pro-pharmaceuticalization consequences of access-oriented consumerism tend to outweigh the depharmaceuticalization effects of injury-oriented adversarial consumerism. HIV/AIDS activists were possibly the earliest of the accessoriented groups to achieve success and public notoriety in influencing drug regulatory and licensing decisions as well as the actual conduct of scientific research (Epstein, 1996). Epstein (1995)

290

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

identified four tactics used by HIV/AIDS activists to achieve their goals: acquiring cultural competence in the language of biomedicine; establishing themselves as representatives of potential research subjects; yoking together methodological and moral arguments; and taking sides in pre existing debates, notably about pragmatic trials. These tactics were leveraged by highly dramatic publicity stunts which captured media attention. At first activists focused attention on the regulatory processes, putting pressure on the FDA to speed up the approval of new drugs; as they learnt more about what counted as ‘good science’, they later contributed to debates about the conduct of randomised clinical trials (RCTs). Activists advocated the use of surrogate outcome measures in clinical trials; they contested norms about what constituted suitable trial populations and placebo controlled trials; and took part in debates about pragmatic and fastidious trials (Epstein, 1996). They transformed scientific practices by creating new pathways for the dissemination of scientific information; sitting on funding bodies to determine which studies were funded; influencing the redefinition of AIDS to include HIV-related conditions affecting women; contributing to new regulatory and interpretive mechanisms by the FDA and NIH; and influencing arguments about how clinical trials should be conducted (Epstein, 1995). HIV/AIDS activists were mostly self-taught, although a few were already members of the scientific community. As time went on, the treatment activists became detached from the communities they claimed to represent, leading to a division within the HIV/AIDS movement between highly knowledgeable autodidact ‘lay expert’ activists and the ‘lay lay’ activists. Hess (2004) used the term ‘medical modernisation’ to characterise the ways in which modern scientific medicine responds to challenges to its epistemic authority from social movements and from complementary and alternative medicine. He argued that under conditions of medical modernisation, the distinction between lay (or alternative) knowledge and scientific knowledge, upon which the epistemic authority of medicine rested, is submerged in a more complex field of competing scientific networks and research programmes. This includes an emerging system of the ‘public shaping of science’ in which social movements have greater agency, and there is greater recognition of the legitimacy of that agency. Indirect public shaping of science takes the form of engagement with the policy process or funding decisions, while direct public shaping occurs when activists help to develop new research programmes to address their own goals rather than those of the dominant research programmes. Hess noted that further work was needed to understand the co-optation process of both patient advocacy group and CAM practitioners. In the context of increasing scientization of decision making in which the public is excluded from many important policy debates (Brown and Zavestoski, 2004), the work of Brown et al. (2004) on health social movements explored their potential for challenging medical policy and practice. Brown et al. distinguished between health access movements, constituency based health movements and embodied health movements, while acknowledging that the boundaries between these types are fluid. They defined embodied health movements (EHMs) as ‘organised efforts to challenge knowledge and practice concerning the aetiology, treatment, and prevention of disease’ (Brown et al., 2004, p. 54). Members of EHMs have a politicised collective identity born of a collective illness identity which turns a personal trouble into a social problem; a sense of grievance is central to the formation of EHMs. Since they depend on challenging medical and scientific knowledge and practice, EHMs are constantly engaged in boundary work and may be characterised as boundary movements in four ways: they attempt to redraw the lines demarcating science from non science; they blur boundaries between experts and lay people; they

transcend the usual limits of social movement activity; and they use boundary objects, such as scientific devices which are also used for political purposes. Members of these movements (for example, breast cancer activists) challenge science on the basis of their own intimate knowledge of their own bodies, in other words, their experiential knowledge. Members of embodied health movements may collaborate with scientists as well as challenging them, and the Silent Spring Institute in the US is an example of an institutionalised citizen-science alliance. In such alliances, scientists come to value the contributions of lay people and in the process, scientific norms are changed. In the UK, patient and public involvement in research has had some impact on the way in which health research is conducted (Staley, 2009). Brown et al. referred to the moral credibility bestowed by members' lived experience of the disease, which is unavailable to scientists. As members of embodied health movements learn about the science relevant to their conditions, the boundary between experts and lay people become blurred. Lay people may become more knowledgeable through the internet and reading scientific papers, increasingly available via open access publishing, while scientists may become advocates for the movements they collaborate with. Some groups teach science courses for activists, for example the Project LEAD (leadership, education and advocacy development) offered by the National Breast Cancer Coalition (Dickersin et al., 2001) and French AIDS associations (Barbot, 2006). A central question for those studying health social movements is that of conflict of interest, particularly for access-oriented groups (Hess, 2004). Unsurprisingly, many of them receive financial support from the pharmaceutical industry, leading to conflicts of interest (Hemminki et al., 2010). In a study of pharmaceutical sponsorship of health consumer groups in the UK, Jones (2008) concluded that although industry did not seem to have captured the policy agenda of these groups, their lack of disclosure about funding sources might undermine the willingness of policy makers to see them as legitimate spokespeople for patients and carers. Most of the social movements discussed above have been ‘bottom up’ movements driven by collective and politicised identities, with clear goals of their own in relation to treatment access, redress for injury, or better health care. Given the epistemic achievements of many social movements, the question arises whether broader public participation in pharmaceutical policy might have a wider contribution to make, beyond issues of access and redress. In particular, public participation is surely relevant in situations of long term treatments for chronic illnesses, in which patients are expected to take drugs for the rest of their lives. If those seeking access tend to focus on efficacy, and those seeking redress from injury focus on safety, what about patients who are not activists? Licensing decisions for such drugs do not currently take the views of end users into account (Eichler et al., 2012; Britten, 2008). In this paper, we wish to use a small case study to explore the question of whether a separately funded non activist patient group can make a more balanced contribution than that made by either accessoriented or injury-oriented groups, without being co-opted by an institutional agenda. By this we mean a balance between the proand de-pharmaceuticalization impacts of these two groups, and a balanced consideration of both potential harms and benefits. In this case study we explore the possibility of creating a different kind of patient group which might open up new arenas of public discourse relevant to pharmaceuticalization. We focus on the issue of drug licensing, as a context in which the potential benefits and harms of new drugs are formally evaluated; licensing decisions are directly relevant to pharmaceuticalization, as they may increase the repertoire of available drugs. The specific questions we wished to address were how this group of patients evaluated existing mechanisms for licensing drugs; how they balanced scientific and

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

lay knowledge; how they made their decisions; and how they viewed the experience of becoming part of a patient panel. 1.1. The case study We used a case study design based in the English National Institute of Health Research (NIHR) Collaboration for Leadership for Applied Health Care and Research for the South West Peninsula (known as PenCLAHRC, http://clahrc-peninsula.nihr.ac.uk). This is a ten year, nationally funded, collaboration between academia and the National Health Service in England aimed at generating research evidence which is relevant to clinicians and patients. The epistemological context of the case study is that of Evidence Based Medicine: one of PenCLAHRC's activities is running training courses in evidence based practice and clinical decision making. Although these courses were originally intended for health professionals, they now include patients and carers who learn alongside clinicians and academics. Public involvement (PI) is a mandatory requirement of many NIHR funded programmes, and the Department of Health funds a national advisory group called INVOLVE to support PI (www.invo. org.uk). PenCLAHRC funds a patient and public involvement (PPI) group known as PenPIG (Peninsula Patient Involvement Group) which contributes to many of its activities. One of the purposes of PenCLAHRC is to promote the take up of evidence based treatments and interventions in everyday clinical practice, and its leaders recognise that this is facilitated by the active engagement of both professionals and patients in prioritising, conducting and implementing research. Members of PenPIG are committed to working with clinical and academic professionals to help health care research become more relevant and accessible to people who use health services. At the time of writing PenPIG consists of 15 people who have lived experience of many health problems, often concurrently, as well as those caring for family members with health problems. Members are recruited from a range of local networks, outreach activities, community events and word of mouth. They are drawn from a wide geographical area spanning over 4 h' driving time including remote rural locations. PenPIG members are diverse in terms of age, gender and social circumstances but not in terms of race; the population in the surrounding geographical area is predominantly white. They contribute to many PenCLAHRC activities including developing the on-line research question generation tool, prioritising research questions, commenting on and helping to write grant applications, reviewing patient information and consent documents, joining individual projects including clinical trials as co applicants, recruiting new staff, teaching medical students, presenting papers at conferences, and sitting on the PenCLAHRC Management Board and Stakeholder Panel. These experiences have also been useful to PenPIG members in dealing with their own health problems and their own health professionals. PenCLAHRC is thus a research collaboration between health professionals, academics and patients in which the contribution of patients is highly valued. PenPIG is not part of a social movement, but a group created by PenCLAHRC. Its members share some characteristics of EHMs inasmuch as they all have experience of long term health problems, either as patients or carers. They do not all share the same collective illness identity, but have a shared identity as ‘service users’, that is people with considerable experience of using health services. Like EHMs, they operate at the boundary between scientific and lay knowledge. Eleven members of PenPIG accepted an invitation to take part in open discussions about the problem of ‘non-compliance’, at the start of which the first author summarised the academic literature and sought their views. The aim of these lively discussions was to bring fresh thinking, from patients' perspectives, to the medically

291

defined problem of ‘non-compliance’. As a result, one member emailed the author of a Danish paper about compliance and the concept of liminal space, and he said that the ensuing ‘exciting exchange’ added meaning to his membership of PenPIG by showing how expert and lay understanding could be brought together. Gradually the idea of creating a patient drug licensing panel emerged as a separate project, using the licensed drug liraglutide as the exemplar drug. Five members of PenPIG were able to take part in the panel, and will be referred to as panel members. In January 2010 liraglutide (a glucogen-like peptide-1, GLP-1) was approved by the FDA as a second tier treatment for type 2 diabetes. It is injectable, with a flexible treatment regimen due to its prolonged half-life and once daily dosing schedule independent of meal times. According to the Summary Review of the FDA's Center for Drug Evaluation and Research (CDER) which reviewed liraglutide, the claimed benefits of liraglutide were improved glycaemic control and weight loss as well as convenience for patients (CDER, 2010). Its potential harms included thyroid C-cell tumours, papillary thyroid cancer, and pancreatitis as well as a long list of gastro-intestinal side effects. During the review process, liraglutide was presented at a public advisory committee meeting in April 2009. This committee focused on thyroid neoplasms particularly Ccell neoplasms as well as cardio vascular safety assessment (CDER, 2010, p. 4). In April 2012 the Public Citizen's Health Research Group petitioned to have the drug withdrawn from the market because of the potential side effects of the drug (BMJ, 2012). We considered this to be a suitable topic for a case study, in which members of PenPIG could make their own assessments of the benefits and harms of liraglutide using publicly available documents from the FDA website. The intention was to elicit patients' views of a licensing decision that had already been taken. Members of PenPIG knew the outcome of the FDA licensing decision and how the original expert panel had voted. We were not attempting to replicate the original decision, but rather to explore it from patients' perspectives. Since then, several papers about the adverse consequences of liraglutide have been published (Gale, 2013; Cohen, 2013) including an editorial about how to help ‘patients make sense of the risks of taking GLP-1 agonists’ (Montori, 2013, p. 1). In this case study we focus attention on a very different context to the one faced by patients with life threatening illnesses lobbying for new treatments, or those seeking redress from harm attributed to pharmaceutical treatments. Diabetes, at least in the short term, is not life threatening, and there are several other available long term treatments. The case study thus consisted of the retrospective evaluation of a specific second line treatment for a long term condition, by an existing PPI group which is part of a research collaboration between academia and the health service, and independent of the pharmaceutical industry. This project was incidental to the main work of PenPIG, and was not part of the institutional aim of PenCLAHRC: neither PenPIG nor PenCLAHRC has a strong agenda around drug licensing, and there was no conflict of interest. The authors (researchers and panel members) organised a patient licensing panel meeting which took place in July 2013. It was audio recorded, and one of the researchers interviewed panel members before and afterwards to explore their views on an individual basis. At the end of the first interview, panel members were asked if they would approve liraglutide or not. In the second interview, panel members were asked if their vote had changed over the course of the panel meeting. Panel members had two roles, firstly as research collaborators designing the project and discussing the outcomes, and secondly as study participants. The patient panel consisted of the five members of PenPIG with voting rights. Panel members have various long term conditions; four of them have diabetes (either type 1 or type 2) and one had a diagnosis of

292

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

borderline type 2 diabetes. None of them is a member of any other group that focuses on drug related issues, and none receives funding from pharmaceutical companies. Two are unemployed and three are retired. Like the PPI representatives studied by Thompson et al. (2012), panel members had a breadth of experience from their previous lives. One panel member had represented disabled people at work tribunals and was familiar with the impact of drugs and their side effects. Another had worked as a clinical research coordinator for a major pharmaceutical company and had been involved in setting up and monitoring RCTs. Four panel members own their own homes and one is renting. Two have university degrees. Two do not receive any benefits, two receive state pensions and one is in receipt of several state benefits. Each panel member had attended a PenCLAHRC training course in which they had learnt about RCTs and carried out critical appraisal of an RCT using the Critical Appraisal Skills Programme. The panel also included two nonvoting researchers acting as facilitator and notetaker. As agreed beforehand, panel members were provided with two documents containing information about the potential benefits and harms of liraglutide. This included the detailed 62 page Summary Review of the evidence surrounding the safety and effectiveness of liraglutide based on the manufacturer's submission to the FDA (CDER, 2010), and a four page summary of that report written by the researchers. PenPIG members were asked to read as much or as little information as suited their time and capability. They were not asked to read anything about subsequent developments, as the intention was that panel members would only have access to information available to CDER panel members at the time of the original decision. Despite this, one panel member was drawn to read exhaustively about the work of the FDA in general and about liraglutide in particular. He went well beyond the short list of benefits and harms written by the researchers, and learnt about the stance taken, role and background of the original FDA panellists and their resources. He had previously given a talk to fellow members of PenPIG entitled ‘A snort from the sty’ in which he had concluded that the level of involvement in PenPIG went beyond interest towards active practice, and that this transition was accommodated by different levels of participation. In other words, he felt that his wider information seeking was consistent with the ethos of PenPIG. The panel met approximately one month later. At the start of the meeting, a public health physician experienced in health technology assessment presented information about liraglutide, the evidence surrounding its effectiveness, the potential benefits and harms, and information on how FDA members and the expert panel voted. The patient panel was then given the opportunity to ask questions, after which the presenter left the room. The panel discussion, including voting, lasted approximately an hour and a half, and was facilitated by one researcher while the other took notes. The panel discussion was taped and transcribed verbatim. In the results section, PenPIG members are numbered P1eP5 and researchers are numbered R1eR3. Data from the interviews and panel meeting were analysed manually using a thematic approach aimed at identifying the issues raised by panel members (Green and Thorogood, 2004). Two researchers independently read transcripts and identified themes. Two main themes were identified: the decision and the panel process. Subthemes within the decision theme related to interpretations of benefits and harms relating to: the evidence; the experts; and their own experiences. Subthemes with the panel process were: disagreements and apologies; rules and remit of the panel; and panel members' freedom to speak their minds. The researchers discussed the themes, and developed an initial framework. This framework was checked against the data, and refined when necessary. Charts were developed for each theme in the framework, and relevant text from the transcript was copied or

summarised under each theme in the chart. Charts were then used to compare data within and between cases. Panel members subsequently discussed the analysis and interpretations with the researchers both face to face and by email. The researchers avoided giving their own opinions, both during the panel meeting and subsequent discussions with members of PenPIG, as the purpose of the exercise was to elicit panel members' perspectives. 2. Results All the panel members (participants) had read at least one of the two documents given to them, and reported that hearing the presentation by the public health physician (R3) consolidated their existing knowledge. After the presentation, the panel discussion was lively, and panel members had no reservations about disagreeing with one another. In this section we present extracts from the panel discussion. 2.1. Assessment of FDA licensing decision The panel members made critical comments about the FDA licensing process and the way in which the FDA made decisions about the licensing of drugs. This included the validity of decisions made on the basis of incomplete evidence, the extent to which advisors and decision makers could be considered to be experts, and the extent to which the questions that were asked of the advisory committee were appropriate. Panel members asked questions about the relevance of animal models for humans, and whether patients recruited to the phase three trials were representative of the wider population. They also asked more general questions about drug licensing processes, post marketing surveillance, and the different roles of various regulatory agencies in the UK, US and Europe. Panel members queried the cumulative effect of taking liraglutide over a long period of time and asked whether harms increased with prolonged use. In the absence of long term evidence, this led to questions about the extent to which it is possible to make a decision about the safety of drugs based on the data available at the time of licensing. After the presentation some panel members asked questions about the extent to which it was or was not possible to draw conclusions based on the available evidence: P3: I also wondered about the lifelong risk of taking liraglutide, and whether the effect was cumulative? R3: Ok, that's a very good question. And importantly, the answer to that is I do not know. And wouldn't care to speculate. P3: So there's nothing in these papers to tell us that? Another participant commented on the delay between the end of a trial and identification of the long term effects of a drug: P2: Nobody can really say what's going to happen until those trials have finished, until 30 years down the line. Another area of critical comment and disagreement between panel members was the value placed on expert opinion. Some participants did not consider the experts to be experts due to the limited evidence available to them: P2: Now, you know, sit back and wait and see the result before you call yourself a specialist in that subject One participant was of the view that the experts know best and decisions should be made on the basis of their opinion.

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

P4: To my mind, their [the experts'] knowledge should be good enough knowledge to sort of say well, yes I agree with this drug or I don't agree with this drug. Others had concerns about the questions that the experts were being asked. The CDER Summary Review reported the outcome of voting at the Endocrinology and Metabolism Advisory Committee meeting, who were asked to vote on four questions including the following: 3. Assuming the remainder of the risk:benefit data are acceptable, do the available data on thyroid C-cell tumors permit marketing of liraglutide? 4. Assuming the remainder of the risk:benefit data are acceptable, do the available data on papillary thyroid cancer permit marketing of liraglutide? (CDER, 2010, pp. 33e34). In the CDER panel, the vote on question 3 was split down the middle while the vote on question 4 was a unanimous yes with one abstention. PenPIG panel members did not accept the hypothetical nature of these questions. P3: You look at the questions they were asking … And I thought how can you have a question like that where you're just saying ‘well actually, if we pretend that's alright, can we then go on and accept this?’ And the following questions are the same you know. And that concerned me, the types of questions they were asking. Although the panel members were critical of the FDA licensing process, they all agreed that patients should take part in licensing decisions.

2.2. Balancing lay and scientific knowledge Brown et al. (2004) claimed that science is an inextricable part of embodied health movements as their members seek to influence research agendas and the search for new treatments. The drug licensing process involves the assessment of data from clinical trials, animal experiments and other sources of scientific data. As well as questioning the relevance of short term trial data, members of PenPIG brought their own experiences to bear on the evidence about liraglutide. Panel members drew on their experiences of benefits and side effects of other pharmaceutical treatments for diabetes, as well as their hopes for improving their control of their diabetes with liraglutide. The interpretation of trial data was filtered through their own history; people focused on the data consistent with their previous experiences. These lived experiences were considered to be critical to the licensing decision, but would most likely be unavailable to the experts. On the whole, the panel focused more on the harms than the benefits; during the presentation R3 reminded panel members also to consider the benefits of liraglutide. When considering the potential benefits of treatment, one participant drew on the experience of his wife who had recently started taking liraglutide: P4: Within a week and a half, my wife started on 0.6 of liraglutide, now she's on 1.9. And this week she's noticing a difference. Another participant had high hopes for what liraglutide could do for him: P5: I'm talking now as a patient, the mechanism looked bloody great. You think cor that's fantastic, and given the impact of, for me, on managing my HbA1c, I would have approved it.

293

However, one participant had failed to see any beneficial effects on her own blood sugar levels whilst using a similar drug: P3: I've been on exenatide, which is a very similar drug, same kind of family of drugs, had horrendous nausea, didn't lose any weight whatsoever, and it had no effect whatsoever on my blood sugar. Most participants had previously experienced side-effects whilst on similar treatments. One participant had previously had cancer, and felt that any increased risk of cancer due to a drug treatment was unacceptable: P3: Because I'd had cancer, even if only one person out of that thousand got cancer, to me that's one person too many. Participants suggested that the ‘experts’ are not experts in the use of the treatment, and do not know what it is like to experience adverse drug reactions, what an impact this can have on a person's quality of life, or what it is like to live with ‘minor’ side effects such as nausea. Whilst these side effects were considered minor by the experts, for one participant, they were so unbearable that even death was preferable: P1: metformin made me feel so dire that, you know, I would have gladly died than continued taking it. Panel members were particularly concerned about the risk of side effects because they thought that doctors might not be fully informed about, or might not adequately inform the patient about, the harms associated with liraglutide. Panel members also spoke of the possibilities and uncertainties associated with liraglutide described in the data. The person who had had cancer and had not benefitted from taking exenatide was not convinced that the drug would be effective, but thought that the risk of cancer was considerable: P3: I was kind of alert really to the problems, that the cardiovascular risk, the cancer risk, and the fact that I didn't think it was going to work anyway. In contrast, one participant who was facing the possibility of being prescribed liraglutide in the near future focused on potential benefits: P5: A couple things R3 said, particularly a study I did look at which showed the importance of … a huge significance for a decrease in haemoglobin level of 1%, and the difference that that makes. And that was why I would have approved liraglutide. Participants varied with regard to how they used the advice of the experts. In some cases, the data from the experts appeared to be used selectively to support the panel members' decisions: P4: Because you look at the last slide that R3 showed, which is where I made most of my decision from, which was looking how most of the doctors voted, and I broke it all down, and you look at the endocrinologist and the diabetology people and they all voted, near enough, for it. P3: What the endocrinologist was saying particularly sort of interested me because they seemed to be saying no, you know, about whether they felt this was acceptable from the point of view of thyroid C cell tumours.

294

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

2.3. Decision making At the start of the discussion panel members had already decided how they were going to vote, and did not change their minds as the discussion progressed. This was confirmed during the second interviews. All panel members voted for or against approval of liraglutide, with no abstentions. Three panel members voted against, and two voted for, approval. As the number of panel members was an odd number, the result was unambiguous. In the original CDER panel meeting, the number of panel members was an even number, and the chair had to use her casting vote. Voting in both panels was very close; PenPIG members did not all vote in the same way, suggesting that they had no shared agenda about the licensing of liraglutide. However the panel was too small to draw any strong conclusions. Of the two members who recommend approval of liraglutide, one was considering liraglutide for the treatment of his diabetes. Decisions were made on the basis of personal experience; trial data; and the views of the experts. The panel emphasised that the day to day side effects, such as nausea, were not given enough consideration by the experts. Three main patterns of decision making were apparent. The first route “trusting the professional” refers to one panel member who used the advice of the experts to make their decisions. Little or no attention was paid to the other factors. The panel member using this route to make a decision voted to approve liraglutide: P4: I'm not being funny; the diabetic specialist nurses wouldn't be giving this drug out if they felt it wasn't good for you.

honestly if you've got the sort of effects that I had I'm not surprised that people lose weight.

2.4. Views about the patient panel During the course of designing the case study and after it was over, panel members said that they enjoyed the experience. From the outset of the project, they said that patients should contribute to licensing decisions. During the panel meeting, there was disagreement about the amount of prior preparation panel members should have done, and the role of personal opinion. The first issue that emerged during the discussion was how much information panel members could (or should) have read. There was considerable variation in how much they had read. Some panel members were confused by the complexity of information in the longer document, and chose to focus on the shorter summary document to inform decisions. P1: I went really from the shortened version, um, because to be perfectly honest, going through the full page version, I ended up not really knowing which way was up … And I can't digest that amount of information; I don't have the amount of knowledge. Nor did I really want to. One panel member had read a lot more than the information provided, and some of the panel discussion focused on the impact that this might have had on his and others' decisions.

The second route “benefits outweigh harms” describes those who considered the harms, but ultimately thought that they were outweighed by the possible benefits. These panel members voted for approval.

P3: I did wonder whether there had been some sort of contamination really, by people not actually reading what they were asked, and going beyond it, and checking other things.

P5: My views as a patient, as a person with diabetes is that … the doubling of relative harm with liraglutide in respect of dizziness, fatigue, nausea, diarrhoea, vomiting, dyspepsia, and gastritis are all very salient to what I've experienced, and I take into full account. Having been diagnosed with cerebral vascular disease about 13 years ago, including global amnesia, minor stroke and TIAs, and now diabetes, I'm particularly concerned to distinguish between cerebral vascular symptoms, which you know, straight off to death and mild stroke, um, that might precede a stroke, and diabetes symptoms. So I'm fully aware that there are harms, but on the other hand, I am also aware of the importance to me of reducing or keeping low, my blood sugar level.

The panel had a heated argument about the legitimacy of reading beyond the two documents provided, the main issues being the impossibility of reaching a decision on such limited evidence, and different interpretations about the panel's working methods. A related area of disagreement was about the role of patients' own views on the panel. One participant described how she thought that the purpose of the patient panel was to provide her views and opinions, rather than be concerned with the views of the experts.

The final route “harms outweigh benefits” describes those who focused on the harms of liraglutide with little or no attention to the benefits.

P2: It's our opinion. Our opinion as PenPIGs. Nobody else's. They weren't invited to join this panel. It's our view. And as far as I'm concerned, your vote wouldn't count because you're looking at everybody else's view. That's my opinion. And I have to say that.

P1: I think if it was a new oral medication that gave um, a much better outcome and the HbA1c was, or whatever it is, you know what I mean. If that was um, significantly decreased, then yes I think it would be useful and I would consider it and approve the drug and hope for the best as far as the toxicity is concerned … Um, but I couldn't actually see that there were that many benefits for anybody. Indeed, these participants suggested that claimed benefits may not actually be benefits: P1: Um, I'm a bit sort of dubious about the, what is put down as being a benefit of um, liraglutide being weight-loss, well, quite

3. Conclusion Debates about the relative benefits and harms of new medicines, and the decisions to license them or not, are of immense public importance. Pharmaceuticals save lives and ameliorate disease; pharmaceutical companies are major contributors to national economies; but pharmaceuticals are also the fourth leading cause of death in the USA (Light, 2010), and a cause of environmental pollution whether ingested or not. This small case study suggests that institutionally sponsored patient groups may be able to make a more balanced

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

contribution to drug licensing decisions than access-oriented or injury-oriented social movements, in part because they consider both the potential harms and potential benefits of new drugs. Such groups may be able to challenge the terms of the licensing process as well as offer their own perspectives on scientific evidence; in our case study the group was not sponsored by the pharmaceutical industry or the regulatory body but by an academic-health service collaboration. This kind of group is unlikely to have a predictable influence on pharmaceuticalization, one way or the other, and may therefore balance the pro pharmaceuticalization influence of accessoriented groups. Although our panel members felt strongly that patients should have a significant role in drug licensing, they were aware of potential negative consequences. They noted that those who shout loudest may be more likely to gain access to their preferred medicines, but that this might deprive other patients of effective treatment. The limitation of this work is that it is based on a single small case study, in a particular geographical, epistemological and institutional context, about a single drug. Patients and patient groups in other contexts and about other drugs might respond differently to the opportunity of taking part in this kind of panel. Although an institutionally sponsored patient group may have no obvious conflict of interest, it could be argued that members might become ‘medicalized’ or turned into ‘professionalised patients’ (Ives et al., 2012). This could take the form of panel members adopting a biomedical worldview and devaluing their own experience. The panel members had all been trained in evidence based practice by PenCLAHRC, but they invoked their own knowledge and experiences during the panel discussion. Some felt strongly that their lived experience was more relevant than expert knowledge. Like members of EHMs described by Brown et al. (2004), panel members were engaged in boundary work and used boundary objects. For example, during the training courses panel members were taught about the PICO format (population, intervention, control and outcome) for framing research questions. One of them subsequently used the PICO in a consultation with her own GP, to persuade the GP to prescribe a different treatment than the one originally planned. Members of PenPIG have also been critical of PenCLAHRC processes, arguing that meetings should be organised in a more user-friendly fashion. Thus, unlike some of the PPI representatives studied by Thompson et al. (2012), members of PenPIG have been able to challenge and criticise. The case study suggests that patient groups can be supported to take their place at the decision making table without losing their critical focus or devaluing their own experiential knowledge. One of the goals of NIHR in general and PenCLAHRC in particular is to normalise PPI in health research, and this is an agenda which members of PenPIG are happy to sign up as it does not compromise their shared identity as users of health services. Because the work is focused on epistemic practices rather than the delivery of particular services, there is less room for conflict. This stance distinguishes PenPIG from health social movements, perhaps because unlike health social movements, PenPIG was not born of grievance. The panel members made a cogent critique of the licensing process. They questioned the validity of making decisions about drugs intended for long term use in the absence of long term safety data. This is a critical question for drug regulators (Abraham, 2007). Like the earlier AIDS activists (Epstein, 1996), panel members questioned the validity of trial populations, asking if they represented the patients for whom the drug would be prescribed. They objected to the use of hypothetical questions during the CDER review process. Panel members also disputed the expertise of scientific experts with no first hand experience of the disease in

295

question or of its treatments. In the FDA Summary Review, CDER panel members made assertions about patients' preferences, for example the convenience of less frequent dosing, seemingly on the basis of speculation. If such claims are going to be made, panel members felt that those with experiential knowledge should be there to make them or dispute them. Members of PenPIG are not part of embodied health movements but share some characteristics of EHMs. Their epistemic work is similar to that of EHMs, working at the boundary of health research and the lived experience of patients. They were knowledgeable, but varied in their capacity to read complex documents such as the CDER Summary Review. Our panel members were drawing on their embodied experiences of medicines, and were judging scientific data on the basis of their first hand experience of the effects of other similar drugs, or other drugs for the same disease. They did not reach the same decision about the licensing of liraglutide as the FDA panel; their pattern of voting suggested that they did not have a shared agenda about liraglutide. Some members of PenPIG based their decisions on their own experiences, and all tended to focus more on potential harms than benefits. Voting on the panel was based either on the panel member's own assessment of the balance between benefits and harms, or was influenced by the voting of the medical specialists. This diversity would likely be replicated if patients were represented on licensing panels. Our case study may be seen as an example of evidence-based activism focussing on pharmaceutical governance and aiming to change the rules of the game (Rabeharisoa et al., 2014). It suggests that an independent group may have the potential both to balance the pro pharmaceuticalization effect of access-oriented groups and to influence forms of pharmaceutical governance. It could also be seen as the public shaping of science in an indirect way but without co-optation (Hess, 2004). Liraglutide and other GLP-1 drugs are now attracting critical attention; one commentator concludes that the hiding of the potentially harmful effects of these drugs reveals problems in a system which subordinates public interest to commercial secrecy (Gale, 2013). This echoes Davis and Abraham's (2013) assessment of the approval of pioglitazone and rosiglitazone, also for diabetes; they concluded that the commercial interests of manufacturers were prioritised over public health. Although some regulatory bodies involve patients and the public in advisory roles, we suggest that involving patients in licensing decisions might go some way towards addressing this democratic deficit. At the end of their book about pharmaceutical regulation, Davis and Abraham (2013) consider the question of democratic accountability and citizen-participation in regulatory decision making. They argue for a more sophisticated, critical and differentiated understanding of patients' needs and patients' rights, partly on the grounds that it is not in the general interest of patients if well resourced groups can mobilise political support. They suggest that we need to ask which types of public are involved and how they relate to other organised interests; they also suggest that the impact of public participation on specific regulatory outcomes needs to be scrutinised. We hope to build on this case study with other, larger, panels and thus contribute to further exploration of these issues. Acknowledgements This article presents independent research funded by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) for the South West Peninsula. The views expressed in this publication are those of the authors and not necessarily those of the NHS, the

296

N. Britten et al. / Social Science & Medicine 131 (2015) 289e296

NIHR or the Department of Health in England. We thank Jim Harris and Lynn Tatnell for their contributions to the patient panel. References Abraham, J., 2007. Drug trials and evidence bases in international regulatory context. BioSocieties 2, 41e56. Abraham, J., 2010. Pharmaceuticalization of society in context: theoretical, empirical and health dimensions. Sociology 44, 603e622. Barbot, J., 2006. How to build an “active” patient? The work of AIDS associations in France. Soc. Sci. Med. 62, 538e551. Britten, N., 2008. Medicines and Society: Patients, Professionals and the Dominance of Pharmaceuticals. Palgrave Macmillan, Basingstoke. British Medical Journal, 2012. News: diabetes drug should be removed from market, says consumer group. Br. Med. J. 344, e3259. Brown, P., Zavestoski, S., McCormick, S., Mayer, B., Morello-Frosch, R., Gasior Altman, R., 2004. Embodied health movements: new approaches to social movements in health. Sociol. Health Illn. 26, 50e80. Brown, P., Zavestoski, S., 2004. Social movements in health: an introduction. Sociol. Health Illn. 26, 679e694. Carpenter, D.P., 2004. The political economy of FDA drug review: processing, politics, and lessons for policy. Health Aff. 23, 52e63. Center for Drug Evaluation and Research, 2010. Summary Review, Application Number 22-341. Food and Drug Administration. See URL: http://www. accessdata.fda.gov/drugsatfda_docs/nda/2010/022341s000sumr.pdf. Cohen, D., 2013. Has pancreatic damage from glucagon suppressing diabetes drugs been underplayed? Br. Med. J. 346, f3680. Davis, C., Abraham, J., 2013. Unhealthy Pharmaceutical Regulation: Innovation, Politics and Promissory Science. Palgrave Macmillan, Basingstoke. Dickersin, K., Braun, L., Mead, M., Millikan, R., Wu, A.M., Pietenpol, J., et al., 2001. Development and implementation of a science training course for breast cancer

activists: project LEAD (leadership, education and advocacy development). Health Expect. 4, 213e220. Eichler, H.-G., Abadie, E., Baker, M., Rasi, G., 2012. Fifty years after thalidomide; what role for drug regulators? Br. J. Clin. Pharmacol. 74, 731e733. Epstein, S., 1995. The construction of lay expertise: AIDS activism and the forging of credibility in the reform of clinical trials. Sci. Technol. Hum. Values 20, 408e437. Epstein, S., 1996. Impure Science: AIDS, Activism, and the Politics of Knowledge. University of California Press, Berkeley. Gale, E., 2013. Incretin therapy: should adverse consequences have been anticipated? Br. Med. J. 346, f3617 (10 June 2013). Green, J., Thorogood, N., 2004. Qualitative Methods for Health Research. Sage Publications, London. Hemminki, E., Toiviainen, H.K., Vuorenkoski, L., 2010. Co-operation between patient organisations and the drug industry in Finland. Soc. Sci. Med. 70, 1171e1175. Hess, D.J., 2004. Medical modernisation, scientific research fields and the epistemic politics of health social movements. Sociol. Health Illn. 26, 695e709. Ives, J., Damery, S., Redwood, S., 2012. PPI, paradoxes and Plato: who's sailing the ship? J. Med. Ethics. http://dx.doi.org/10.1136/medethics-2011-100150. Jones, K., 2008. In whose interest? Relationships between health consumer groups and the pharmaceutical industry in the UK. Sociol. Health Illn. 30, 929e943. Light, D.W. (Ed.), 2010. The Risks of Prescription Drugs. Columbia University Press, New York. Montori, V.M., 2013. Helping patients make sense of the risks of taking GLP-1 agonists. Br. Med. J. 346, f3692. http://dx.doi.org/10.1136/bmj.f3692. Rabeharisoa, V., Moreira, T., Akrich, M., 2014. Evidence-based activism: patients', users' and activists' groups in knowledge society. BioSocieties 9, 111e128. Staley, K., 2009. Exploring Impact: Public Involvement in NHS, Public Health and Social Care Research. INVOLVE, Eastleigh. see. www.invo.org.uk. Thompson, J., Bissell, P., Cooper, C., Armitage, C.J., Barber, R., 2012. Credibility and the ‘professionalized’ lay expert: reflections on the dilemmas and opportunities of public involvement in health research. Health 16, 602e618. Williams, S.J., Martin, P., Gabe, J., 2011. The pharmaceuticalisation of society? A framework for analysis. Sociol. Health Illn. 33, 710e725.