Media content analysis of the introduction of a “soft opt‐out” system of organ donation in Wales 2015‐17

Abstract In an attempt to improve organ donation rates, some countries are considering moving from “opt‐in” systems where citizens must express their willingness to be an organ donor, to “opt‐out” systems where consent is presumed unless individuals have expressed their wishes otherwise, by, for example, joining an “opt‐out” register. In Wales—a part of the United Kingdom—the devolved government recently legislated for an “opt‐out” system. For the change to be effective, a public awareness campaign was critical to the policy's success. Using quantitative and qualitative content analysis, we explored media coverage of the change to better understand the relationship between the state, policy actors, media and the public when such policy changes take place. Our findings illustrate how a state communication campaign can effectively set the media agenda within which we saw a degree of interdependency created with the state using the media to promote policy, and the media relying on the state for credible information. Yet we also found that the media is not uncritical and observed how it uses its autonomy to influence policy setting. Over the period of study, we found that a change in tone and view towards deemed consent organ donation has taken place in the media. However, while this may influence or reflect public attitudes, it is yet to be seen whether the media campaign translates into behavioural change that will result in increases in organ donations.


| INTRODUC TI ON
In 2015/16, there were more organ transplants in the United Kingdom (UK) than ever, but despite increases the UK has one of the lowest rates of consent to deceased organ donation in Europe. 1 Targeted activity focusing on changing attitudes and behaviour has been taking place across the UK, but in 2015 the Government in Wales (which has devolved powers from the UK Government including responsibility for health) became the first UK country to introduce an "opt-out" system for consent to deceased organ and tissue donation through the Human Transplantation (Wales) Act 2013. Prior to this, an opt-in system of consent was in place. The new Welsh legislation provides an "opt-out" default where unless a person has explicitly registered or expressed their decision to "optout" of being an organ donor during their lifetime, all residents of Wales who are over 18 years with mental capacity are presumed to have given their consent to deceased organ donation. This is also called "deemed consent." It is however a "soft opt-out" system where family members are involved and asked to support the deceased persons decision made in life, whether it was registered on the organ donor register, expressed verbally or deemed (as opposed to a "hard opt-out" where families are not consulted). The legislation came into force on 1 December 2015 and aimed to increase the number of organ donations in Wales and save more lives through organ transplants and is part of a broader strategy to increase UK consent rates to 80% by 2020. 2 In order to be successful, the Act and the associated media campaign needed to inform the public of the change, create a situation whereby organ donation becomes the norm, and create a way for people to behave as the Act intended. 3 As one of only a small number of countries world-wide to have introduced a presumed consent system, the legislation constitutes a major change to the partnership and social contract between the Welsh Government and the people of Wales. 3 The Act is inevitably controversial and not everyone consulted agreed with the "soft opt-out" system and its principles. Some (eg, Ref. 4 ) warned that an "opt-out" system would result in a backlash from members of the public who currently support organ donation, which could have a negative impact on donation rates.
Others (eg, Ref. 5 ) argue that only a change to an "opt-out" system could improve UK organ donation rates and accordingly save lives, a conclusion supported by research that has modelled donor regimes. 6 Such arguments for and against were played out in political and public debates and across the media and continue to do so.
In this paper, we focus on this changing discourse around organ donation within the confines of the introduction of a "soft opt-out" system and in doing so examine the relationship between state policies, the media and public opinion. We review previous literature from studies of the media and its influence on organ donation then summarize the debates around public opinion, the media and the role of the state in public policy. We present the methods, then report and discuss our findings. In our analysis, we found close intertextuality and interdependence between the state and the media, observing that the health-related messages successfully framed by state organizations to promote their policies are used by the media to create human interest stories that appeal to readerships. We conclude that over the period of study, there has been a change of tone in media coverage towards support for the new organ donation policy, a change that is independent of irrefutable evidence that the policy is effective in Wales.

| Media and organ donation
A body of literature exists examining the influence of different sources of information on attitudes towards organ donation.
Matesanz and Miranda 7 identify mass media as having potentially positive and negative effects on public attitudes. They highlight historic moments when mass media was observed to have had a significant, if short-term impact on organ donation rates. For example, in 1980 after a prime-time UK current affairs programme questioned the validity of brain death criteria (Panorama, BBC) it took 15 months for donor referral rates in the UK to recover.
Conversely, the effect of increased organ donation seen in Italy following a highly publicized shooting of 7-year-old child and subsequent donation of his organs was seen to have a highly positive effect on public opinion. 7 A number of studies 8,9 have shown that the media is a primary source of information about organ donation-although based on research from the pre-mass-Internet age, the common findings that television has the greatest impact will inevitably be outdated. Nonetheless, these studies are helpful in consistently identifying specific types of communication that tend to have favourable or non-favourable effects on public opinion towards organ donation. Negative opinions have been shown to stem from mass media while positive attitudes are more likely to come from information given by health professionals and from discussions with family and friends. 8,10 The effect on actual transplantation rates was shown to be positively impacted by mass media in a content analysis of US television news over the period 1990-2005. 11 However, other research suggests that coverage in television dramas and documentaries along with sensationalist press coverage can have more negative impact, including the development of myths about black markets in organs, corruption within the medical community and preferential treatment of certain groups or individuals for transplant. 10,12 While studies have examined the influence of media on public attitudes towards organ donation and on organ donation itself, the relationship between public policy or legislation, the media, and public opinion has not been widely examined. This study aims to fill that gap and respond to previous studies examining presumed (also known as deemed) consent systems (eg, Ref. 13 ) where contextual evidence, such as media content, is highlighted as missing from debates.

| Media, policy and public opinion
Within the policy studies literature, media institutions are acknowledged as purveyors of information, as ciphers for competing ideas, and in some cases, policy agenda setting. 14 Actors within the policy process assume that media is a pervasive influence-whether it is or not-and success is often measured by the extent of media coverage of any particular agenda. 15,16 The media campaign around the Study of mass media content is often used to decipher the relationship between media and public opinion. It has long been seen as a means of monitoring the "cultural temperature" of society either from a humanist perspective in looking back to identify what it says about society and the culture producing it, or a behaviourist approach, looking forward to the effects that content produces. 16 This reflects debates about whether mass media creates public opinion, attitudes or perceptions, or reflects existing attitudes perceptions and culture 17 or as most scholars agree, does both.
When considering public policy however, the relationship between the state and mass media organizations needs to be examined to establish the antecedents of these constructions. In media studies, this relationship has often been assessed according to a propaganda model 18 contending that the state exerts a powerful influence over media through regulation, through censorship, and as a major source of "reliable information." The state relies on the media for the communication and legitimizing of its policies, and maintaining the status quo. 19 This dynamic influences what is emphasized and what is absent in media reporting through what Gramsci 20 refers to as the "negotiation of hegemony." The opposing view is that the influence of the state over the media is not absolute. 19 Hall 21 argues that mass media organizations have a degree of autonomy through which they can differentiate their views from the state and exert influence on politics and society independently as both drivers and reflectors of public opinion.
The media can also be seen as operating in at least two interlinked ways regarding public policy. 15 First, media influences the policy process by playing a gatekeeping role in whom is given status to comment on public problems and prescribe solutions. Journalists decide which policy actors (individuals or group that are directly or indirectly, formally or informally, affiliated with or affected by the policy process at any stage) are taken seriously as important players ascribing them with a particular "standing." Secondly, journalists and editors are more than gatekeepers in allowing policy actors a mouthpiece, but are themselves players in the policy process, shaping and framing the discussion. Media "frames" the ways in which issues are organized and understood in the public realm by organizing words, images and themes that are used to introduce and present a public policy issue. Frames are "a necessary property of a text-where text is broadly conceived to include discourses, patterned behaviour, and systems of meaning, policy logics, constitutional principles, and deep cultural narratives". 22, p. 37 In this study, we are interested in using text frames to examine who says what; through which channels; to whom; and most importantly, to what effect over the period of time since "opt-out" consent to organ donation legislation in Wales was first conceived, across the implementation phase (2015-2017). In presenting evidence, we therefore examine the relationship between policy actors, the state, the media and the public, to gain understanding of the dynamics of interrelation, assessing the extent to which mass media sets or reflects public policy agendas. 23 We draw upon Shoemaker and Reese's theoretical approach, 16 which focuses less on the process and effects of how media messages are given and received by audiences, but encourages us to investigate the factors inside and outside media organizations that affect content, incorporating the following objectives: • Review media coverage of organ donation prior, and post-implementation of the Human Transplantation (Wales) Act 2013.
• Examine relevant media content with foci on their newsworthiness, the framing of messages and the internal and external drivers of content.
• Investigate the factors which influence media content, and therein the relationship between the media, public policy and public attitudes.

| ME THODOLOGY
We used what can be described as a "summative" content analysis 24 where we start by identifying and quantifying particular words and content within the text with the purpose of understanding the contextual use of the words or content. We categorized the media message related to soft opt-out organ donation for valance 25 (positive, neutral, negative tone) and assessed its newsworthiness through the twin dimensions of deviance and social significance. 26 We then counted the frequency of news stories across time periods. Finally, we qualitatively explored the discourses within the narrative messages that the stories contained. This facilitated investigation of meaning by bridging the gap between texts and readers' interpretations of them. 27 We were therefore able to examine the preference given to some arguments by journalists and look for sources and the explanatory frameworks which underpin them. 28

| Sampling
A purposeful approach was taken in selecting all articles from key media. This consisted of news texts published either online, or available online that refer to organ donation policy in Wales during the period January 2015 to October 2017. Search terms deployed included "Wales" AND "Human Transplantation" OR "Transplant" OR "Organ Donation" OR "Organ Donor." A focus on textual material was made as time and resources forbade analysis audio or video sources. This is an acknowledged short-coming of the investigation, although many written reports mirrored or reported on audio and video sources, which mitigates the potential impact of this limitation.
The dates were chosen to include coverage of the Act from its implementation in 2015, thereby excluding speculative discussion that preceded it. Within the sample are news and opinion pieces, as well as public reaction to them in the form of user-generated comments.
The importance of the latter in shaping or reflecting public policy has grown in importance as the boundaries between "old" and "new" media become less clear and the public become both consumers and producers of content. 29 However, some research 30  on user-generated commentary as representative of public opinionwhich it is clearly not-and its influence. While it has been shown to be widely read and popular, user content is also found to be treated very sceptically by the public. 31,32 In this study, a systematic search of online news sources was undertaken and focused on obtaining relevant findings for the objectives of the study and within the parameters set out above. Sixty items were identified, and texts were imported into NVivo (v.11) software for analysis (Table 1).

| Data analysis
Our analysis began with searches for occurrences of identified words and messages within the sources (Figure 1). While their frequency is interesting, to understand the underlying contexts of messages within the sources further information was needed. As well as the subject-specific messages, other key content variables were recorded including media weighting to allow high circulation, high rating or highly influential media to be scored higher than small, less important media. This is particularly important in the case of a small country such as Wales where UK-wide and predominantly London-based media have been shown to have greater reach and impact than regional, Wales-based media. 33  Weighting calculated from the country website ranking (www.alexa.com) of each news source then ranked by its size relative to the other sources in the list. If more than one source has the same rank, the top rank of that set of values is presented. Higher numbers indicate greater weight of influence.
TA B L E 1 (Continued) lower than it). Also recorded were the number of "shares" and where appropriate, the number of user-defined "comments" although inconsistency in the availability of these variables meant that we did not account for these in our final analysis.
We examined each news item headline and assigned a valance

| Procedure
Search results (online news articles) were saved as pdf files. The files were skimmed for content not relevant for the research question, such as side-bar advertisements. Data were imported into the data analysis software and articles were assigned variables using a source classification sheet that included publication, date of publication and UK site ranking. Each news transcript served as the unit of analysis for investigation.
The coding scheme was developed with key concepts highlighted based on the objectives of the study, text-frequency analysis and an initial review of content (Table 2)

Recipient perspective of organ donation
Recipient_Gift "generous," "receiving," "gift of life," "giving" and synonyms The role, and changed role of the family of the deceased Family_role "Family" AND "conversation," "discussion," "debate," "veto," "refusal," "override" and synonyms TA B L E 2 Coding scheme and 30% neutral (eg New opt-out system in Wales aims to revolutionize organ donation. The Guardian 25/11/15). When we applied influence weights calculated from the popularity ranking of each news website, a variance was noted (49% positive, 18% negative and 33% neutral), suggesting that highly ranked news providers were more circumspect in their coverage.
Adding a temporal dimension to the analysis (see Figure 2) highlights the spike in media interest in organ donation that took place following the implementation of the Act in 2015, but also enables clear shifts in media tone towards the policy in Wales to be observed.
Previous research 9,23 studying news stories and organ donation has suggested a link between negative media coverage of organ donation and sensationalism-as represented by measures of deviance. We found no relationship in our data with 57% of articles coded as either "not at all unusual" or "somewhat unusual" in their content, while only 12% contained "extremely unusual" information.
Social significance was greater with 60% of content coded as having moderate or major social significance. Combining the deviance and significance variables to create a measure of newsworthiness as suggested by Shoemaker and Cohen, 35 we observe over the period of study a steady increase in the complexity of stories ( Ministers have said more than two-thirds of people in Wales are now aware of the changes and they hoped a "leap in consent rates" would follow (BBC News 01/12/15) A number of news organizations were observed struggling to explain the change using unfamiliar terminology. The terms "soft opt-out," "presumed consent" and "deemed consent" were used frequently but not always contextualized or explained, and this lack of understanding led to some negative commentary while the implications of an organ donation register where people could both "opt-out" and "opt-in" were not fully understood. In particular, the notion of a "soft opt-out" that retained a central role for families was not always presented: Opt-out organ donation will treat bodies like "clapped A key message frame that emerged from the initial scoping of content was the notion of organ donation as a "gift" that is regarded as having wider societal benefit and driven by an individual's desire to help other people. 38 In a number of cases, the law change was pre-  A further topic considered in the literature and which appeared as an important message frame in the media was around the role of family within an "opt-out" system. Study of this area 40,41 suggests that this is often an area of conflict of interest and control between potential donors, families, medical professionals and potential transplant recipients. In adopting a "soft opt-  The UK Government subsequently followed Scotland by announcing a consultation in England on a "soft opt-out" in December 2017 and supported a private members bill proposing "deemed consent" legislation in March 2018. This might be perceived as representing a major change in culture and public opinion towards organ donation. Across time, we found fewer dissenting opinions in the media although it is unclear whether this is genuinely representative of societal pressures or represents a "negotiation of hegemony" 20 between the state and the media. Interestingly, in interviews with non-donor relatives other research 44 has found that some family members feel pressurized by media to "give the gift of life" resulting in feelings of guilt and selfishness when they choose not to.

| CON CLUS ION
Media content provides a window into changing public discourse and political agenda setting related to the issue of organ donation consent systems. Looking at content over time allows for a nuanced understanding of the relationship between policy makers, implementing institutions, and the public, while also analysing the ways in which each party shapes press coverage.
Our analysis charts an observable change in tone and view towards deemed consent organ donation with a public health communication campaign successfully setting the agenda and informing the debate. Over the three years of study, there is evidence that the principle of deemed consent and the "soft opt-out" system has become increasingly newsworthy, portrayed in the media as a public good, even with limited evidence of actual impact. Furthermore, discussion of the wider moral and ethical issues that dominated coverage prior to implementation has largely disappeared.
If, as the literature suggests, the media both reflects and influences public opinion, then we see in the evidence a change in attitude towards the Human Transplantation (Wales) Act 2013 and more importantly, towards organ donation itself. However, content analysis is only the very first, preliminary step in the investigation of changing public perceptions and attitudes. While the media has a central role in communicating to the public, people do not absorb messages uncritically and further research is needed to confirm these findings, and more critically to investigate whether changes in opinion translate into behavioural change that result in increases in organ donations.