Drug breakthrough offers hope to arthritis sufferers: qualitative analysis of medical research in UK newspapers

Abstract Background Newspaper stories can impact behaviours, particularly in relation to research participation. It is therefore important to understand the narratives presented and ways in which these are received. Some work to date assumes journalism transmits existing medical knowledge to a passive audience. This study aimed to explore how newspaper articles present stories about medical research and how people interpret and use them. Design Qualitative research methods were employed to analyse two data sets: newspaper articles relating to ‘rheumatoid arthritis’ and ‘research’ from UK local and national news sources; and existing transcripts of interviews with patients with rheumatoid arthritis and their carers. Results Newspapers present a positive account of medical research, through a simple narrative with three essential components: an ‘innovation’ offers ‘hope’ in the context of ‘burden’. Patients frequently feature as passive subjects without attributed opinions. Few articles include patients’ experiences of research involvement. Patients with rheumatoid arthritis and their carers read articles about medical research critically, often with cynicism and drawing on other sources for interpretation. Conclusions An understanding of the simple, positive narrative of medical research found in newspaper articles may enable researchers to gain mass media exposure for their work and challenge this typical style of reporting. The critical and cynical ways patients and carers read stories about medical research suggest that concerns about newspaper articles misinforming the public may be overstated, but any effect on research engagement is unknown. Newspaper articles rarely present patients’ views or their experiences of research, and this can be conceptualized as ‘depersonalization bias’.

articles rarely present patients' views or their experiences of research, and this can be conceptualized as 'depersonalization bias'.

Background
In the UK, 87% of adults read newspapers, either in traditional paper or electronic format. 1 People use newspapers as a source of information about health issues including new medical research. 2 While media research has long illustrated the impact of newspapers on public knowledge, 3 there are recent concerns about the media's role in shaping the public understanding of health and medicine. 4 Newspaper stories have the potential to dramatically impact the behaviours of the reader, particularly in relation to stimulating medical research interest and participation enquiries, as demonstrated in the following examples. In March 2006, six men had severe and life-threatening reactions during a clinical trial of the compound TGN 1412, a treatment being developed to target leukaemia or rheumatoid arthritis (RA). They were in a private research unit at Northwick Park Hospital in North West London. Following these events, which generated huge media attention, there was an unexpected increase in general enquiries about participation to several charities and commercial research organizations. 5 A doctor who gave an interview to a leading Dutch newspaper about his trial of a new surgical implant, reported being inundated with 1000 enquiries from people wanting to take part in his trial. 6 Researchers have a duty to publish and publicize their work. To do this responsibly and effectively, they should be aware of how newspapers tend to present stories about medical research and the potential impact of these stories on patients and carers.
To assess the quality and accuracy of newspaper health stories, Robinson et al. 7 reviewed eight UK newspapers, finding a high degree of variability and raising concerns about medical journalism. Articles commonly failed to include information about risks and costs associated with treatments. Similar results have been reported in Europe, 8 the USA 9,10 and Canada. 11 Aggarwal et al. 12 explored the views of Euro-pean journalists reporting cancer research and described a 'schism', with limited communication and understanding between the cancer research and media communities. These studies take a linear reflection perspective, based on the assumption that medical knowledge is created by experts and transmitted to the uninformed public by journalists. 13 An alternative perspective is that journalists cocreate narratives through which information is understood and circulated. 13 Others involved in this cocreation include laypeople and professionals in, for example, media companies, health services, government, the pharmaceutical industry, charities and patient groups. Exploring the narrative structure of these health stories may help us to understand how they are received by readers and how they may influence decisions and behaviours around engagement with research. Previous work analysing narratives has focussed on specific health topics such as cancer [14][15][16] or particular events such as the Northwick Park trial. 5 The aim of this study was to explore how newspaper articles present stories about medical research and the ways in which people read them. We have used the example of RA research to explore this issue in depth.

Methods
Our research comprised two separate elements, both using qualitative methods: 1. Analysis of newspaper articles relating to both RA and research. 2. Analysis of existing transcripts of interviews with patients with RA and their carers.
We used LexisNexis, 17 a database of UK newspapers to search online and print versions of the 10 national newspaper groups with the highest circulation 1 (see Table 1 We identified articles containing both 'rheumatoid arthritis' and 'research'. These search terms were chosen to capture articles about any type of research with relevance to RA. Where possible, we also searched the websites of the individual newspapers for the same time period using the search term 'rheumatoid arthritis'. As the focus of the study was on representations of research, we excluded several kinds of article that were identified by these search terms, such as items about charity fund-raising events or 'research' referring to information-seeking behaviour. In addition to the above, we also used an existing data set of transcripts from interviews with a total of 42 people, with 20 identified as containing talk about medical research in the news media. Each interviewee was either a person diagnosed with RA or a friend or family member with a physical and/or emotional role in the person's care. Participants were identified and approached at secondary and tertiary rheumatology clinics. Face to face, semi-structured interviews were conducted, audio-recorded and transcribed verbatim. These data were collected between 2012 and 2014 as part of a project exploring how people learn about RA, which included a focus on sources of information. All interviewees gave written informed consent prior to participation and the study was approved by NRES Committee North West -Cheshire (ref 12/NW/0148). We searched the transcripts for mention of 'news', 'paper', 'article', 'report' or 'headline' and began analysis work on the texts identified. Concurrently, we also reviewed all the transcripts to confirm that all relevant data had been captured in our initial search.
The analysis was conducted according to the standard procedures of rigorous qualitative analysis. 18 Intellectually, we drew on ideas from the analysis of media in discourse analysis 19 and ethnomethodology. 20 In the early stages, we employed procedures from first-generation grounded theory, open and focussed coding on subsets of data both independently and coming together as a group to discuss, challenge and modify codes. 21 Later, we developed tables, or 'frameworks' to express our emerging concept, which we used to interrogate wider sections of the data, checking for 'fit' and searching for cases to challenge or modify our concept. 22 We analysed data from the newspapers and interviews separately, but with each piece of work informing the other. Throughout the process, we undertook memoing 21 (recording and developing analytic notes about ideas emerging from data analysis), mapping 23 (creating visual representations of relationships among codes and categories) and deviant case analysis 24 (sampling for and in depth study of anomalous cases).

News sources
Our search yielded 1024 articles from all 10 of the national newspaper groups (see Table 1),  Table 1).
Article type or main focus Table 2 shows the type or main focus of the articles in the data set. Our analysis work focussed on the Trial of New Drugs or Medical Technologies category as most relevant to medical researchers publicizing their work. The set of categories reflects all the contexts in which medical research was mentioned. The categories were generated through an iterative process of grouping and regrouping the articles. The categories appeared uniformly across the 10 year time period, with the exception of the Northwick Park Trial category, which was concentrated in a 2 year period following the event in 2006. The most common category was Lifestyle and Alternative Medicine. There were some differences in the distribution of articles between the local and national data. The national press published all the Obituaries and far more articles about Lifestyle and Alternative Medicine and Business.
The local press published more articles relating to Infrastructure. There were no major differences between the Newcastle and Birmingham data sets.

Positive and negative narratives
We carried out in depth analysis on the 100 articles in the Trial of New Drugs or Medical Technologies, focus RA category and identified two distinct narratives. The dominant narrative was a positive one, evident in 91 of 100 articles, while the second, less common narrative was a negative story of harm caused by clinical research. The negative narrative was apparent in only nine of the 100 articles and usually related to just two stories. The first story was the heart attack and death of a phase one trial volunteer in Edinburgh. The second story was the suspension of a repurposing trial of the drug Vioxx and the subsequent withdrawal of the drug. This negative narrative echoed reporting of the Northwick Park trial, previously analysed by Stobbart et al. 5  The positive narrative of medical research, present in 91 of 100 articles, was characterized by three essential components: 1. Innovation described the idea at the beginning of a study or the discovery at the end of a study. Novelty was stressed with words such as 'breakthrough', 'revolutionary', 'pioneering', 'new approach' and 'world's first'. 2. Hope referred to an optimism that the innovation may eventually contribute to a reduction in the burden. 3. Burden was expressed as both personal suffering and societal cost. Emotive language was used to highlight the extreme nature of the personal burden, for example 'crippling', 'debilitating' and 'constant agony'. Dramatic figures were employed to show the cost on a societal level, for example '700 000 sufferers', '1000 a year have to give up work' and 'cost to the UK economy of the condition is at least £4.75 billion a year'.
In each article, these components were narratively related in the following format: An innovation offers hope in a context of burden. The following example from The Daily Telegraph (16 June 2008) was one of the shortest articles in the data set and as such contained the three essential components and little else: Arthritis Hope A new 'smart' drug (innovation) could save (hope) thousands of rheumatoid arthritis patients from years of pain and disability (burden), research has shown. Tocilizumab, which is expected to be launched in Britain within six months, is being hailed as a major breakthrough (innovation).

Characters and voices
There were recognizable patterns to the presentation of groups, such as patients and scientists, in the Trial of New Drugs or Medical Technologies, focus RA category and a key difference was the presence or absence of their views. Characters were described without any attributed quote or opinion, while voices were people, other than the journalist, giving a view. Characters appeared in descriptions of research activity, representing either the research team or the research subjects. For example: The Newcastle team will test the jab in eight people with the condition. Characters were identified solely in terms of role (e.g. 'volunteers'), or with additional reference to location and/or institution (e.g. 'scientists at Utah University in the United States'). Research team characters were always presented as active (e.g. 'scientists found'; In contrast with the actions of the research team, even though the subject is presented as active, their action has no impact on others. Both articles also contain several examples of passive subjects. Voices appeared with either a direct quote or an attributed belief or thought. Voices were either the research team or a commentator and the functions of each are described below. When a voice was a named individual, their credentials as a worthy spokesperson were confirmed in terms of their role and/or their institution. The following examples demonstrate the briefest and the most elaborate credentials from the data set, respectively: Voices were sometimes groups rather than individuals and in such cases, there were no credentials (e.g. 'doctors say'; 'researchers hope'). There was a tendency for articles to mention an anonymous group first and later a named representative from that group.
Several functions were performed by the voices in articles: presenting the research; endorsing the research; and providing context. These functions overlapped, although patient voices never presented the research. The typical format was for the research team to present the research and a doctor or charity representative to endorse the research. Depending on the stage of the project, the team presented the research in terms of their aims or results, for example: . . .says Dr Jane White head of clinical research at the company. . . 'The first patients to have the implant, who were aged between 39 and 82, experienced a 60 per cent drop in pain levels'. (The Daily Mail, 24 October 2006) Commentators were not part of the research team and were usually a medical professional, academic and/or a representative of a charity (e.g. 'Prof Alan Silman, medical director of Arthritis Research UK charity'). Patients were sometimes commentators, but only very rarely and in circumstances discussed below. The primary function of any commentator was to endorse the research.
Commentators endorsed research with varying levels of enthusiasm, indicated by a number of factors (see Fig. 1): firstly, the degree to which the research was described as exciting; secondly, the extent to which the research was welcomed in a general or a specific way; lastly, the presence or absence of qualifiers, such as 'ifs' and 'buts'. The qualifications themselves clustered around just four categories: timing; uncertainty of outcome; limited application; and limited access to eventual treatment. The level of enthusiasm expressed did not appear to relate to the identity of the commentator or the stage of the research, with equally enthusiastic endorsements for the start of phase one and two trials as for the conclusion of phase three trials. The endorsements of Professor Alan Silman, the most frequent commentator in the data set, varied in enthusiasm. In essence, patients frequently featured in articles only as passive subjects with no attributed opinions. The views of the research team, other professionals and charity representatives were presented in strikingly uniform ways.

Patient voices
Patient voices were relatively rare in the Trial of New Drugs or Medical Technologies, focus RA category, although more common in the Newcastle data: 7/80 National; 5/15 Newcastle; 0/5 Birmingham. When present, the key role of these patient voices was to contextualize the research. A patient voice was very often presented as a discrete entity, at the beginning or end of the article and sometimes as a separate piece with its own headline. Patient voices illustrated the burden of arthritis (an essential component of the positive medical research narrative described above) with descriptions of everyday struggles and losses. This served as a frame of reference for the presentation of innovation and hope. Furthermore, patient voices tended to end their stories with statements of hope as these final sentences show: A lot of research is going on into the condition in the North East, which offers hope for sufferers. Hope was commonly associated with the research, but occasionally referred more generally to the impact of medications, as in the second example. In addition to contextualizing the research, some patient voices also endorsed the research. They did this with varying levels of enthusiasm, following the same pattern as for medical professionals, academics and charity representatives (see Fig. 1). Compared to other endorsers, patients tended to give endorsements with low specificity, for example: There was a standard format to the presentation of patient voices, which gave biographical information and, as already discussed, illustrated the burden of arthritis. We termed this the 'minimum patient-hood' and found only two exceptions to the pattern. In the first instance the article was primarily about the patient's experiences of RA, with research a secondary focus. In the second instance, the patient was fulfilling two functions, both as a charity representative endorsing the research and as a patient illustrating the burden. The minimum patient-hood comprised the following biographical information: first name; last name; age; location; occupation; and disease onset. This information suggested that the person was real and enabled the reader to imagine the person. and life before the disease (e.g. 'was sports mad as a child'). These details were often present in longer articles and served to give a clearer, more relatable picture of the individual. Information about family may have been a proxy for occupation when a person was not in paid employment. All 12 articles with a patient voice presented patients talking about their experience of the disease, but five of these patients made no mention of research involvement. In the remaining seven articles patients did mention research participation, usually highlighting positive outcomes, for example:

Research into new treatments for other conditions
The Trial of New Drugs or Medical Technologies, focus not RA category, included treatments for a range of conditions including cancers, heart disease, Alzheimer's disease, multiple sclerosis and Crohn's disease. These studies were captured in our search because they either drew on knowledge from RA management or the work had potential future application for the treatment of RA. To understand whether our findings were trans-ferable to other diseases, we tested our ideas against this data set. All our key findings were apparent in this category. There were far more positive than negative stories. All the positive stories included the narrative: An innovation offers hope in the context of burden. Few of the articles had a patient voice and even fewer included patients' experience of research involvement (see Appendix S1).

How newspaper articles are read
We have shown how newspaper representations of research and trials tend to fit a common formula. However, how are these newspaper articles understood and used by the public? To explore this we analysed existing interview transcripts from patients with RA and their carers. Our sample comprised 20 individuals, 10 of whom had a diagnosis of RA and 10 were friends or family members with a physical and/ or emotional role in the person's care. The relationship between the patient and carer was most commonly spouse, but sometimes child, partner, sibling, friend or child-in-law. Interviewees gave their age range, with these varying from 20s to 70s, but most commonly 60s. 12 of the 20 interviewees were female.
People with arthritis and their friends and relatives discovered print media resources either through actively looking for such material, just happening upon it, or being directed to it by someone else. These other people included carers, family and friends alongside broader social networks like friends of friends. I get phone calls as soon as there's something in the paper. Because the people who know I've got it, they'll think I've missed it, and that will help me. (RA patient, F, 70s) They did, at times, ascribe others motives for directing them to or physically giving them articles. These ranged from demonstrating acts of care to, in one case, something like guilt, when another patient with well-controlled RA contacted an unwell patient with news of an innovation, which may enable the unwell patient to also improve. The findings suggest that people are not uncritical in their reading and can be cynical about how innovations are described by the press. Massive headlines of the Daily Mail or the Daily Express, or something like that, for a miracle cure for rheumatoid arthritis . . . And you think 'oh yeah, oh yeah, take that with a pinch of salt'. (husband of RA patient, 60s) All interviewees undertook work to contextualize the stories presented, such as asking those with relevant knowledge, say within the pharmaceutical industry, to translate the potential impact and timeline of an innovation. Thus, people can read articles with knowledge of the potential for newspaper articles to be partial or exaggerated.
The data show how people engage with and use the content of the articles in a variety of ways. Some read stories of innovation out of general interest in keeping updated, to see 'if there's something recent' (wife of RA patient, 60s). Some want to learn more about the personal stories of other people with RA or to generate a sense of hope for their own future. Others used the information within articles to assess their own care and treatments, for example noting that they have already 'had that' specific drug. They also reported discussing the stories in clinic, with some patients taking a copy with them, or in the case of a report on an ongoing trial, asking about potential eligibility, or even contacting the trial team directly. Alongside this, articles were discussed with or given to carers, to inform them, or they were 'filed', either literally or metaphorically.

Discussion
We know that newspapers are read by a high percentage of the population 1 and they are a source of information about health and research. 2 Furthermore, it is likely that newspaper articles contribute to many people's perceptions of medical research, which in turn influence decisions and behaviours around engagement with research. 5,6 Our results show that newspaper articles of numerous different type and foci discuss research related to RA (see Table 2). 10% (100/1024) of articles relate specifically to research into new drugs and medical technologies to treat arthritis. This is a substantial amount given the breadth of articles reviewed. With the notable exception of the Northwick Park trial, the narrative is overwhelmingly positive and uniform with the same essential components and narrative presented in all positive stories: An innovation offers hope in a context of burden.
There are well documented concerns about the accuracy and completeness of many reports. [7][8][9][10][11][12] None of these articles assess the ways these stories are read and they tend to assume a linear perspective and a passive readership. Kitzinger 25 investigated audience understandings of AIDS reporting and found that even groups with no special knowledge of AIDS, presenting themselves as entirely dependent on the media for their information, were nevertheless drawing on other sources of knowledge such as friends, family and personal experience. These results coincide with contemporary theories of 'active audiences', reading with resistance as well as alignment to dominant ideas. 2 Health information seeking, including from printed sources such as newspapers, is a component of health literacythe capacity to obtain, process and understand basic health information and services to make appropriate health decisions. 26 Recent work builds on the concept of health literacy as an individual risk or asset, 27 describing 'distributed health literacy' 28 as a potential shared resource of individuals' skills within social networks. Our analysis of interview transcripts suggests that patients and carers read newspaper articles critically and draw on other sources, such as friends or family with relevant knowledge, for interpretation and verification. These findings support the theories of both 'active audiences' and 'distributed health literacy'. The potential for newspaper articles to misinform the public may be less of a concern than the cynicism with which some patients and carers viewed newspaper reports. This cynicism may be influenced by the striking uniformity of the narrative, as well Understanding the essential components and narrative of positive stories about medical research (An innovation offers hope in a context of burden) may promote mass media publication by helping researchers to 'package' their research results in an appealing way for journalists. Conversely it may also enable researchers to challenge the status quo and argue a case for presenting stories about research which do not fit this narrative. Examples of this would be stories comprehensively addressing risks, uncertainty and timescales or stories highlighting patients' experiences of participation. An awareness of the power of newspaper reports to impact upon behaviours, particularly to generate large volumes of participation enquiries, 5,6 should enable researchers to anticipate and make provision for this eventuality.
The rarity of patient stories and patient voices in newspaper articles about trials of new drugs or medical technologies is both surprising and notable. During the late 19 th and early 20 th centuries, there was a shift in newspaper content away from the simple reporting of macro events and towards the inclusion of personal stories to illustrate the impact of these events on the lives of ordinary people. This trend of personalization was described by Hughes 29 in 1940 and has been explored more recently as one of many forms of popularization by which journalism engages audience attention. 13 In contrast, our results suggest a 'depersonalization bias', conceptualized by Hallin and Briggs 13 to explain journalists accepting linear views of the flow of health information and thus 'privileging administrative perspectives at the expense of looking seriously at lay experience and understanding'. Kitzinger 30 identified a similar privileging of official sources in media reporting of risk. There is evidence that patients seek out others' stories related to health and illness to use as a resource, for example, when identifying options 31 and making decisions. 32 It is therefore likely that patients also look for stories about medical research involve-ment to explore options and make decisions about this. We found further evidence that patients look for personal stories in our interview transcript analysis.
We have explored the narratives present in the text of newspaper articles about medical research and the ways in which newspaper articles are read by patients with RA and their carers. Our analysis was developed using articles focussing on RA research and tested using articles about research in many other conditions. We identified the articles relating to other conditions from our original search for RA research. These articles all mentioned RA, which may have biased this sample towards immunological research. Our analysis of representations of medical research does not extend to the pictures in newspaper articles or to other sources of information such as television, radio or the internet. Thus, our conclusions cannot apply to these alternative sources. A focus on newspaper representations has been noted in many areas of media studies, with less known about the reception of newspaper articles. 2

Conclusion
Newspaper articles relating to research into new drugs and medical technologies are overwhelmingly positive and contain a simple narrative with three essential components: An innovation offers hope in a context of burden. An awareness and understanding of this narrative may enable researchers to gain mass media exposure for their work and challenge this conventional style of reporting.
Our analysis of interview data on the interpretation and using of newspaper representations of medical research is a valuable contribution to a neglected area of study. Patients and carers read stories about medical research critically and sometimes with cynicism. This cynicism may be a consequence of the uniformity of the narrative, as well as possible inaccuracies or incompleteness. The effects of this cynicism on research engagement should be explored.
Patient voices and patient stories about research involvement were extremely rare in arti- cles relating to research into new drugs and medical technologies. This finding conflicts with the historical trend in newspaper reporting towards more personal stories and the evidence suggesting that people would seek and use such stories. Further work is needed to explore the production of newspaper articles about medical research, which may shed light on the lack of patient stories and voices.