Introduction and state of the art

Suicide is, according to the World Health Organization (WHO), a major public health problem. Over 800,000 people commit suicide every year worldwide and it represents the leading external cause of death in Spain, according to the National Institute of Statistics (INE, in Spanish), with a ratio of 8.3 deaths for every 100,000 inhabitants. Suicidal thinking is observed at an increasingly young age, a situation that experts attribute, amongst other reasons, to the pandemics, harassment, and cyberbullying (Garcés-Prettel et al., 2023; Quinatoa-Santo & Larzabal-Fernández, 2021; Wake et al., 2021). Suicide is the fourth cause of death among 15-19-year-olds, according to the WHO (2000). A report issued by the WHO and the UN (2017) points out bullying as an integral part behind 200,000 yearly youth suicides.

All these figures verify that suicide is a highly relevant social problem, whose impact justifies the media spotlight. The presence of press news is ever increasing (Blanco-Castilla & Cano-Galindo, 2019), partly due to changes in recommendations from social and healthcare organizations. The traditional position had been to consciously silence suicide in the media, since experts pointed to a possible call effect (known in the technical literature as the “Werther effect”, the name of the suicidal protagonist in Goethe’s novel).

In the year 2000, the WHO published a document entitled “Preventing suicide, a resource for media professionals” where it is pointed out that the media may play a key role in preventing suicide (World Health Organization, 2000). The position of experts reaffirms this idea: psychology and psychiatry professionals request to give suicide more social visibility and require the cooperation of the media to contribute to its prevention with their information. This is what the technical literature calls “Papagen effect”, in reference to the character in Mozart’s “The Magic Flute” who is convinced by three childlike spirits to drop his plan given the alternatives to his death. However, diffusion must be properly focused (Philips et al., 1992; Hawton & Williams, 2002). The key lies, therefore, in how to inform.

Following the WHO’s change, varied codes with recommendations for its treatment in the media have appeared, but all of them focus on traditional support media and none includes the diffusion on social networks, despite their growth and great viewing capacity by audiences who are both consumers and producers of information (Ritzer et al., 2012). During the first half of 2022, the number of internet users reached 37 million in Spain, according to the General Media Study (GMS), of which 70% are social network users. As per the Observatory on social media of 2021, 85% of the Spanish internet surfers use social media, which are young people’s usual channels, amongst whom the suicide cases and attempts have alarmingly increased. These figures justify the need for specific studies which establish an appropriate approach to the media’s social networks.

The scientific literature on the treatment of suicide is scarce and practically inexistent with respect to the mass media publications on social networks. Some references are found on the media’s treatment of suicide and the WHO guidelines (Olivar-de-Julián et al., 2021; Durán & Fernández-Beltrán, 2020), on their inclusion in codes of ethics and manuals (Lois-Barcia et al., 2018), and on the coverage of deaths of celebrities by suicide (Niederkrotenthaler et al., 2020; Wake et al., 2021).

In all cases, the change of position by the WHO is assumed to have had a very limited impact on journalistic practice. Despite the publication of the “Guidelines for the treatment of suicide in the media, Support Manual for their professionals” by the Ministry of Health (2020), the “Recommendations to audiovisual media on the reporting of death by suicide” by the Autonomous Government of Catalonia (2016) and the Audiovisual Council of Catalonia, and the “Ethical code on the media treatment of suicide” (APIB, 2018), promoted by the Balearic Association of Journalists, the Balearic College of Psychology and the Helpline of the Balearic Islands, in the vast majority of cases, the mass media do not comply with the WHO guidelines (Rodríguez-Caro et al., 2021; Olivar-de-Julián et al., 2021). Durán and Fernández-Beltrán (2020) specify the principal errors in the lack of a global approach to the problem, the explicitness of the method and place, the use of non-expert sources, the simplification of causes, and the absence of support references (organisations and telephones).

There exists countless research on the influence of the internet and the social networks on present communication, but those focused on their responsibility towards suicide are rare. The Internet represents a new scenario as well as an additional concern for the prevention of suicidal behaviour, triggered by the influence of pro-suicide content (López-Martínez, 2020).

Social isolation has proven to be a risk factor that increases the chances of developing suicidal behaviour (Castillo-Sánchez et al., 2020), and the Internet has become an apparent socialisation space for many people. Adolescence is characterised by young people’s vulnerability and by the enormous influence that their cultural and social environments exert, (Monteagudo-Ródenas, 2002) and today, when we speak of our young people’s social and cultural contexts, we must inevitably talk about social networks.

Recent studies have shown that social networks may act as a highly significant means of suicide contagion (Yıldız et al., 2018; Carlyle et al., 2018; Abrutyn et al., 2020). Somehow, social networks are reproducing the Werther effect so feared by the mass media in the last century, thus provoking its self-censorship.

Particularly significant are the publications by «influencers» or famous people. The rise of suicide rates after media news concerning the self-deaths of public figures is a well-documented phenomenon, which has even led to the need for specific treatment in the WHO guide. Similarly, there is a strong correlation among the content of the messages from certain personalities in social networks which increase the risk of suicide by the general public (Fahey et al., 2018; Niederkrotenthaler et al., 2020).

The risks of the internet are obvious, but it also provides possible preventive support, above all, to detect suicidal tendencies in their phases of ideation (Castillo-Zúñiga et al., 2022; Lopez-Castroman et al., 2020; Rajesh-Kumar et al., 2020). Castillo-Sánchez et al. (2020) have conducted an analysis of the literature published between 2010 and 2020 on the subject and have developed, as principal techniques, the use of machine learning algorithms such as data mining, linguistic research, and text word count, among many others, which are used in a combined way at times. However, there is a paucity of academic studies on how social network platforms and their users must behave in relation to this issue, so that their online publications do not act as a driving element of suicidal behaviour. Cohen-Almagor (2022) warns about the dangers of hate speech on social networks, although suicide advice is scant. The salient Australian project #chatsafe (Robinson et al., 2018), provides some guidelines designed to support safe peer-peer online communication, especially among young people. It is a transfer to the web 2.0 of the guidelines for mainstream media (La-Sala et al., 2021), even though its model has not been expanded to other countries.

There is hardly any self-regulation regarding the online media treatment of suicide. By default, it can be understood that the limited norms included, both in journalistic codes and in guidelines issued by other organisations, would be extendable to digital information, but the specifics of the Internet, and especially of social networks, lead to the need of an ad hoc regulation and to the growth of specific studies pinpointing the suitable media approach to online social networks.

Methodology

The primary objective of this study is to pinpoint Twitter audience behaviour towards the selected media news on suicide, to assess how users respond to different treatments as well as what type of expressions or phrases provoke a reaction among the followers of the accounts studied. We seek to ensure that interaction is associated with the WHO guidelines and the Spanish organisations. Lastly, some action proposals are suggested in order to address the peculiarity of social networks given the absence of conduct guidelines.

Our first objective is to get to know the current status of the issue through the analysis of all the expert organisations’ codes and guidelines, as well as the existing research at a scientific or academic level, both nationwide and worldwide. Subsequently, an assessment of the conducted process has been made through a qualitative analysis and quantitative conclusions as well.

Both the qualitative and quantitative studies analyse the corporative profiles on the microblogging (Grahl, 2013) Twitter of the mainstream national media with more followers on this social network, according to the “top 20” media. We have selected two conventional general media or digital migrants (“El País” and “El Mundo”), two TV channels (RTVE and Antena 3), one free of charge (“20 Minutos”), two radio stations (Cadena SER and COPE), as well as two digitally native (“El Confidencial” and “El Español”). This study intends to examine the connection and the interaction between the selected media and their audience on the social networks (Correa et al., 2010), their participation (Otieno & Matoke, 2014; Gruzd et al., 2012), the type of content created by the media and the exchange produced with their audiences (Kaplan & Haenlein, 2010). To that end, firstly, from January 1 to June 30,2022, an advanced search was made on Twitter exclusively bounded to the above-mentioned media, with the word “suicide” and the synonymous expression “take your own life”, in order to verify to what extent they implement the WHO guidelines (2000) as well as all the subsequent codes in the informative treatment of suicide. Overall, 114 tweets were downloaded with the support of the platform «exportcomments.com», which also enables the interaction of each message (comments, likes and shared). However, eleven messages were excluded on using the concept suicide metaphorically, applied to sports and politics.

For the qualitative analysis, we start from an adaptation of the texts by Herrera-Damas and Requejo-Alemán (2015), who propose a classification of strategies for use on Twitter (information dissemination, promotion, engagement, participation, contests, and active learning), to which a new category is added, awareness. We also draw on the proposal by García-Avilés and Arias-Robles (2016) on visual formats and journalistic genres on Twitter. To complete this classification, we have added the tone to analyse the proximity and the creation of links between emitter and receiver.

Our methodology concludes with a Delphi, a hugely effective method to analyse complex topics and gather consensus views (Landeta-Rodríguez, 2002; Reguant-Álvarez & Torrado-Fonseca, 2016; Baladrón-Pazos & Correyero-Ruiz, 2019). During the month of September 2022, a questionnaire was sent to the seven major national organisations which work on suicide prevention, and on family and survivor care. The sample of this panel of experts is integrated by the entities collected in the guide published by the Spanish Government to which we have added two more: The “Teléfono de la Esperanza” (the Phone of Hope), which has been operative in this field for over 50 years, and the ANAR Foundation, focused on childhood and adolescence in light of the rise of cases in that segment of the population. A second online closed response questionnaire was forwarded to complete the Delphi (November 2022), where four questions were included, separated into two different pathways, with a view to unify positions.

The questions focused on whether the media should report a suicide or not, which data should be included on an item of news on the media’s social networks, and justifying the reasons for the answer. Among the data being reviewed to decide whether they must be included or not, are the allusion or not to the method, the reference to the exact place, the victim’s identification whether anonymous or famous, age, gender, municipality, farewell notes, images or photos of the victim, concepts like epidemics, data to contextualise the problem, expert sources, statements from relatives, neighbours or similar, the trigger when it is a social issue (harassment, etc.), testimonies of survivors, help resources, victim’s screenshots on social networks, associating mental health with risk of suicide, or linking with myths.

The organisations that nswerred our questionnaire, in both phases, were the confederation “Salud Mental España" (Mental Health Spain), the “Teléfono de la Esperanza ASITES” (ASITES Phone of Hope), Papageno, AIPIS, Després del Suïcidi Associació de Supervivents DSAS (After Suicide Association of Survivors), the “Fundación Española para la Prevención del Suicidio” (Spanish Foundation for Suicide Prevention), and the “Asociación de Investigación, Prevención e Intervención del Suicidio” (Association of Suicide Research, Prevention and Intervention). In addition to the two rounds of questions, a further open one and another with closed responses, some telephone conversations were carried out with several of the mentioned entities in order to shape some responses and expand information.

Analyses and results

Results in the social profiles of the media

Of all the tweets examined, 28.43% were shared by Cadena SER radio, 17.65% by the newspaper “El Mundo”, 16.67% by “20 Minutos”, 11.76% by “El País”, 10.17% by “El Confidencial”, 9.8% by RTVE, 7.84% by Antena 3, and only 1.96% by COPE. The single media that reached most retweets on its publications on suicide in that period was “El Mundo”, with 62.28% of the total amount of the messages analysed; followed by “El País”, with 15.93%, and, distantly, by Cadena SER, with 9.31% out of the total. Something similar occurs regarding the number of comments and tweets marked as favourites, and in percentages much alike.

From this data arises the question about what type of messages, with what tones and intention and, above all, what phrases they use to achieve such levels of interaction. As far as the audience is concerned, in this case, it is interesting to assess which messages encouraged the followers’ reaction. To that end, we adapted the classification by Herrera-Damas and Requejo-Alemán (2015), and a new category was added, awareness. Most messages intend to spread information of interest and current affairs (32%) or raise awareness (24%). It is noteworthy that 24% of the texts examined could be classified under the category of promotion, since they are, fundamentally, books or radio and TV phone calls, related to the topic.

Concerning the attitude of the audience from the media studied on Twitter, it is significant to highlight that most users opt to interact more in the awareness intention. Over 60% of the retweets, comments, and favourites took place in this area, despite the fact that most of the messages focus more on diffusion, which is the second most commented topic.

On the other hand, according to the methodology of García-Avilés and Arias-Robles (2016), the journalistic genres mostly used by the media on Twitter are the news (73%) and the interpretative (12%). A total of 8% are references to news and information on current affairs, and just 7% are based on the genre of dialogue.

As for the most used format, we highlight photo news, headlines, some live programs, and information with key data and statistics. The media that share most photo news are “20 Minutos”, “El Mundo” and “El País”; whilst those publishing most messages while going live are “Cadena SER” and “El Español”. In these cases, photo news is only used for sharing images together with the advertisement of the telephone number 024 to prevent suicide or with photos of celebrities who confessed a suicide attempt.

Our attention is now drawn to the most journalistic part of these messages shared on Twitter, where just a fourth part mentions expert sources, although it must be stressed that just 2% of the shared news use politicians, neighbours, and relatives as sources to address suicide.

The majority of messages in the sample analysed are focused on prevention (52.43%), people’s testimonies who tried to take their lives, fundamentally celebrities (18.45%), and description of facts (16.5%). The approach used the least is the victim’s identity profile, an issue not additionally recommended, although some media used it at particular times. In few cases, it was decided to share information on support, for instance, psychological (6.8%), or counselling (4.85%).

“La Cadena SER” (22.8%), “El Español” (11.4%) and “20 Minutos” (6.2%) are the media whose tweets focused more on an approach to suicide prevention; whilst other media, like the newspaper “El Mundo”, opted for the description of facts in the orientation of the messages published.

The interaction between the mass media’s audience and their followers regarding the diverse approaches of the media on Twitter about suicide is crystal clear, and most people opt for sharing, commenting, or marking as favourites those messages related to prevention. Thus, for instance, the tweets with preventive information and content which was retweeted represent 78.72% of the sample, whereas those getting some comments was 62.48%. On 80.64% of the cases they were marked as favourites.

On the contrary, when it comes to analysing the interaction of the audience, the type of messages highlighted are those referring to messages with non-recommendable phrases, such as “frustrated suicide attempt” (11.65%), or “mentions the person’s name”, when the person is not a recognised face (11.65%). Instead, expressions such as “commit suicide associated with a crime”, “suicide-driven person” or “has committed suicide” were hardly used.

Discussion and conclusions

The experts’ change of guidelines has fostered a rise in journalistic publications on suicide, although still insufficient in light of the large number of cases. This situation is also reproduced by the mass media’s publications on their social networks: compared to the roughly 2,000 people who lose their lives to suicide every six months in Spain, we have only detected 141 tweets among the profiles of 9 mass media over the first six months of 2022. In their communication on social networks, the media reproduce the modus operandi of their respective news channels. Thus, the mostly used journalistic genre is the news (73% of tweets) and its own prevailing formats being: photo news, headlines, live, key data, and breaking news. It should be emphasised in positive terms that the predominant approaches are prevention (52.43% of the shared messages), and the testimonies of people who tried to commit suicide (18.45%). The predominance of the communicators’ responsibility is also observed in the behaviour of audiences, who focus their interaction on the publications concerning prevention (78% of the retweets and 62% of the comments), as well as the testimonies (12% of the retweets and 22% of the comments).

Along these lines, we can highlight the widespread practice of sharing more messages on Twitter with the purpose of raising awareness, and also issuing information and data which reinforce the necessity to tackle this matter on the part of the political and health authorities. In this regard, it is noteworthy that most of the tweets have been to remember the helpline 024 (22.33%), to make an urgent appeal for a National Plan of Prevention (4.85%), or even to specify the recommendations for the treatment of suicide in the media (3.88%).

As for the non-recommended expressions mostly used by the media on Twitter, the ongoing references to the idea of “frustrated suicide attempt” (11.65%), or mention the victim’s name (11.65%) play a prominent role. Even though the level of non-compliance of the codes by the media is low, it suffices for it to happen from time to time so that, around social networks, their impact becomes widely amplified by the multiplier effect of the audiences. Thus, the wrong expression like «frustrated suicide attempt» gets the most retweets (54.15%) and comments (46.68%). For that reason, it is especially critical that the media exercise extreme awareness of their responsibility when they publish or share content on social networks.

There is unanimous agreement amongst the associations in viewing the elements that must be specially avoided, such as any allusions to the method, to the exact place or the municipality, or to data that allow the victim to be identified: name or surnames, or screenshots from the victim’s social networks. The unanimity is also absolute when regarding it as inappropriate to mention farewell notes, to allude to concepts such as epidemics, or to use statements from relatives and neighbours.

Among the data that must be collected, which have been agreed on by the associations, are the reference to support resources and the helpline, as well as the inclusion of expert sources. This last aspect has much room for improvement.

In other aspects there are differences of opinion between associations, especially about whether the inclusion of specific content related to the victims’ age and gender is convenient, which can be relevant to observe their socio-demographic profiles. There are those who understand that their inclusion provides a warning to highly vulnerable target groups, while between the opposing organisations it is argued that both the person’s anonymity and intimacy must prevail.

As reflected in this study, that inclusion in the media is higher around social networks than the one they already have in their traditional formats and, therefore, they must accomplish a far more rigorous communicative practice when publishing and sharing suicide-related content. Therefore, we conclude with a series of guidelines for the managers of social profiles in the media, which may also serve as a reference to guide audiences in rating and sharing that content:

• Communication on social networks must act to show the true dimension of the problem of suicide. The current media coverage is undoubtedly insufficient. It is essential to arbitrate communicative formulas that reiterate the seriousness of the problem. An expression such as “the number of victims so far this year is already X”, which is generally used for gender-based violence, may be useful to create public awareness of the true scope of the problem.

• We advise social publications to include data related to the victim’s age and gender insofar as they may serve to visualize the greater impact between specific groups who require some special attention.

• We also deem it positive for social publications to include references to the trigger cause (whenever it is reliably proven), and its relation to mental health, as society can thus become more aware of the origins of the problem.

• Both the victim’s and the family’s preservation of privacy must be a fundamental principle when carrying out any publication on social networks about an act of suicide, although a balance must be sought with the diffusion of data that may be significant from a demographic standpoint, or from the viewpoint of the “news value” or the media interest in news items. We agree with the associations that it is not convenient to publish the place nor the municipality, although the Autonomous Community can be reported.

• The media must avoid publishing or sharing any information from victims’ social profiles, as well as personal pictures or those which may lead to their identification.

• Prevention must be the inspiring principle of the communicative practice on social networks and it must be repeatedly promoted. Therefore, information on the helpline 024 must be included in all social publications on suicide.

• It is convenient for the media to widen the number of expert sources consulted to report on suicides and resort to associations and professionals associated with prevention. It is recommendable that the main associations are labelled on social media publications to publicise them.

• There is a strong need that the media, in their work to create opinion and train citizens, take a stand in favour of the approval of a National Plan of Suicide Prevention and contribute to claiming health improvements that help reduce the number of victims.

• The media must routinely check their online publications to eliminate those that do not comply with the guidelines of responsible action and thus hinder their possible negative effect from perpetuating themselves.