Introduction

There are approximately 800,000 suicides every year, and for each successful attempt there are more than 20 suicide attempts (WHO 2014). Those who are at highest risk of suicide are men and younger age groups (Stene-Larsen and Reneflot 2019), and suicide was the second leading cause of death amongst those aged between 15 and 29 in 2015 (WHO 2018).

Suicidal behaviour is associated with feelings of hopelessness, depression, anxiety, an inability to imagine a positive future, and increased sensitivity towards being socially accepted (Bhar et al. 2008; Delam and Bazrafshan 2019; Rasmussen et al. 2018). Most people who die by suicide suffer from mental health issues (Hawton and van Heeringen 2009), most commonly mood disorders, and drug and alcohol abuse (Pompili et al. 2010). The lifetime risk of suicide for individuals suffering from a substance abuse disorder is between five and ten times the risk for the general population (Wilcox et al. 2004).

Whilst depression is most associated with suicide, this is not always the case (Baumeister 1990), and one study found only just over half of those who died by suicide were depressed (Bilsen 2018). Furthermore, many more people suffer from depression without ever committing/choosing suicide (van Heeringen 2000). Research has suggested that there are many different predictors or predisposing factors that contribute to suicidal behaviour, such as genetics (Wasserman et al. 2009), psychiatric disorders (Hawton et al. 2005), neurophysiological factors (e.g., serotonergic abnormalities; Bach and Arango 2012), low self-esteem (Yoo et al. 2015), family relationships (Wagner et al. 2003), a history of physical and sexual abuse, (Azzopardi et al. 2020), stressful life events (Sofronoff et al. 2004), and personality traits (Stefa-Missagli et al. 2020).

Moreover, the intensity of suicidal ideation in one study was found to be associated with neuroticism and extraversion (Stefa-Missagli et al. 2020). Neuroticism is a personality trait involving negative affective symptoms such as vulnerability, self-consciousness, anger, and anxiety, whereas traits of extraversion include thrill-seeking, outgoing, assertiveness, and expressiveness (Uliaszek et al. 2010). Nonetheless, no one factor, or theory, can truly capture the complex nature of suicidal behaviour (Mann 2013).

The Internet and Suicide

The Internet often plays a significant role in promoting and romanticising suicide, even portraying it as a heroic escape (Murray 2011). For example, the Internet has facilitated the dissemination of new methods for suicide, such as burning charcoal indoors (Cleary et al. 2018) and places to go for the suicide (such as the Aokighahara Forest in Japan [Takahashi 1988]).

In recent years, the Internet has gone one step further (Tam et al. 2007). The term cybersuicide is used to describe an instance in which an individual’s decision to die by suicide and actual death has been facilitated through the Internet (Fratini and Hemer 2020), whether through cyberbullying, encouragement from strangers, information learned online, or using equipment bought online (Chang et al. 2019; Daine et al. 2013).

A study, in which 22 participants who had attempted suicide were interviewed, discovered that 13 individuals had used the Internet to discover information on suicide, and the suicide attempts of eight participants were shaped based on Internet information (Biddle et al. 2020). As reported by Clarke and Lester (2013), the most common explanation for an individual’s choice of suicide method is convenience and accessibility. Tam et al. (2007) described how the Internet provides information on the locality and accessibility for suicide, making it easier than ever to die by suicide. For example, on Christmas Eve 2019, a 20-year-old was actively reading pro-suicide forums in the weeks beforehand and died by suicide using a legal chemical that she had bought online (Tweed 2021).

Furthermore, the Werther effect describes the phenomena of copycat suicides resulting from mass media coverage (Fahey et al. 2018). The media has been criticised for its reporting of suicides, sensationalising their coverage, and providing details of the method used and locality chosen (Ha and Yang 2021). The Werther effect is commonly demonstrated after celebrity suicides, and as a result, the WHO (2008) released guidelines for reporting suicides with sensitivity, including instructions not to mention the method or show the location where the suicide occurred (Posselt et al. 2021). The Internet, however, does not follow these restrictions (Fratini and Hemer 2020).

In today’s age, unsupervised access to the Internet at a young age is not uncommon (Prievara et al. 2019), and so it is a growing concern that young people are being desensitised to suicide and learning methods to use for suicide. The prevalence of suicide-related Internet use amongst suicidal individuals has shown to be significantly more common in suicidal youth than suicidal adults (Padmanathan et al. 2018).

Those who are suicidal or suffering with mental illness are amongst the most reluctant help-seekers (Seward and Harris 2016). Those individuals commonly use the Internet as a source of information and help, mainly because of underlying shame and guilt, associated with suicidal ideation and behaviour (Lester 1998). Online exposure to graphic suicide-related content can move suicidal ideation and intention to actual suicidal behaviour (Daine et al. 2013). Furthermore, the stronger the suicidal feelings, depression, and anxiety that an individual experiences, the more they use the Internet (Sueki 2013). This creates a negative cycle in which the individual may find difficult to escape.

Whilst there are several websites online dedicated to crisis intervention and suicide prevention, in one study, participants, who were suicidal, described these sites as being impersonal, cold, lacking practical solutions and focused on information-giving rather than helping (Biddle et al. 2020). One individual requested a help from a crisis website because they were worried about committing contemplating suicide, and two days passed before they received a response. The participants argued that the assistance that they needed was live communication with another individual, which is possibly one reason why suicidal individuals gravitate towards using pro-suicide forums with individuals talking about the issue (Biddle et al. 2020).

Online Forums

There are many online support groups and pro-suicide forums online that allow people to reach out for help and talk to others. However, there have been reported cases in which these forums have encouraged suicidal behaviour (Niezen 2013). Some forums even ban users who attempt to discourage people from suicide—only support and assistance in terms of the process can be offered (Becker 2004). Many of these forums and chat room users view professional intervention in a very negative light, and Becker (2004) described them as anti-psychiatric.

One key individual in this area is that of William Francis Melchert-Dinkel who posed as a suicidal female nurse on support forums and manipulated at least five individuals into a suicide pact with him for his entertainment (Windgrove 2009). The victims hanged themselves whilst on video call to Melchert-Dinkel. Other cases of cybersuicide include a 17-year-old girl who had received advice from forum users about which suicide method she should choose (Becker et al. 2004), whilst a 15-year-old credited her inspiration for suicide on information learned online (Becker 2004).

However, these forums and chat rooms also have the power to help prevent suicide. Baker and Fortune (2008) interviewed users of self-harm and pro-suicide forums, and these users described the websites as empathic communities that gave them a new social identity in which they felt important. Thus, these groups can either mitigate or intensify a suicidal person’s ideations (Mokkenstorm et al. 2020).

Many suicides also occur online in forums, message boards, social media, and chat rooms that are not directly related to suicide, such as pro-drug sites and sites dedicated to freedom of speech (McSwain et al. 2012; Westerlund et al. 2015), as well as platforms such as YouTube and Facebook (Fratini and Hemer 2020).

Live-Stream Suicides

A recent public health concern is that of live-stream suicides, in which individuals allow social media users to watch their death as it happens (Fu et al. 2013). Cases of live-stream suicides have demonstrated methods of drug overdose, hanging, and shooting (Klein 2012), and they occur on many different types of social media, including Facebook, Periscope, Instagram, and YouTube (Fratini and Hemer 2020). In these cases, the suicide may be seen as a performative or dramatic act (Lester 2014, 2015). Every individual that dies by suicide is making a choice about how they want the audience to see them and how they want to be remembered, from decisions about what clothing to wear, the location of the suicide, and what final communications to have with or leave for others (Lester 2015). The audience also plays a role in suicides (Fratini and Hemer 2020).

The Influence of the Audience

The influence of the audience is paramount in online suicides (Phillips et al. 2019), whether the audience is one person or a large group. Fratini and Hemer (2020) conducted an analysis of live-stream video broadcasts and identified four main types of audience roles: the suicide “expert”, the silent one, the life saver, and the troll. The “expert” describes the role of those who critique a person’s suicide method, often offering suggestions to make it better or “right”. The silent one describes the individuals who do not comment but continue to watch or follow along. Life savers are those who attempt to intervene and discourage the suicide, and often these individuals will alert emergency services.

The role of the troll is predominantly suicide baiting, where the audience, whilst remote and anonymous, encourages the individual go through with the attempt (Fratini and Hemer 2020; Phillips et al. 2019). Research has suggested that individuals who have declared their suicidal intent whilst livestreaming experience an increase in their perceived obligation to go through with the suicide so as not to be humiliated (French 2020). Fratini and Hemer (2020) describe the life savers and the trolls as having the most influence on the suicidal individual.

Another phenomenon observed in live-stream suicides is that, as the audience increases in size, the less likely the members are to assist when help is needed (Latané and Nida 1981). Zimbardo (1969) believed that anonymity and being part of a larger group encourages individuals to act differently from the way in which they would act otherwise, an example of deindividuation. When in a group, individuation theory states that individuals lose their sense of self-identity and, as a result, are more easily encouraged to behave in a socially unacceptable manner, such as suicide baiting. The influence of being anonymous and being a part of a larger group synchronises to enhance the effects of deindividuation, as the anonymity provides protection from social disapproval (Mann 1986; Mann et al. 1982).

However, despite there being numerous, high-profile cases of online-assisted suicide, there is limited research on this behaviour at present. Sharing one’s death with an audience over the Internet is a phenomenon that is not well understood, and so research into understanding the motivations and processes of online suicide is needed. Additionally, the role of the audience, or bystanders, is of importance as they have the potential power to save a life versus watching someone die. More qualitative research is necessary to explore the choices, reasons, and processes behind cybersuicide attempts (Lester 2015).

The Present Study

The present study is aimed at thematically analysing the final communications of cybersuicide victims, to identify common themes. The study focuses on the perspective of both the suicide and the audience or witness to the events, to better understand how the role of another online could influence a person to die by suicide. The role that the Internet plays in suicide will also be assessed.

Method and Data

To discover cases of individuals that have had their suicide facilitated by the Internet, a range of terms was searched online such as “suicide online”, “Internet suicide cases”, “live-stream suicide”, and “suicide and Internet forums”. There exists a very limited amount of qualitative public data to be analysed, but three very different cases were identified. To minimise any ethical issues regarding anonymity, and out of respect for the families of the victims, the names of individuals discussed are pseudonyms.

Connor and Molly

One case that was discovered through this research is that of Molly and Connor, aged 17 and 18 at the time, who had a two-year online relationship before Connor ultimately died by suicide. Police officers later discovered that Molly had sent him many messages encouraging him to choose suicide, reassuring him it was the right thing to do. Connor and Molly decided together that the best way for him to end his life was through pumping carbon monoxide into his car whilst he was inside. Molly later sent a text message to a friend admitting that, whilst Connor was in the toxic environment of the vehicle, he got scared and got out of the van and that Molly told him to get back in, knowing it would end his life (Chappell 2017). Molly was later found guilty of involuntary manslaughter (Orecchio-Egresitz 2020).

All SMS text messages between Connor and Molly over the course of 42 days, including the day of the suicide, and all messages Molly sent post-suicide are available publicly online. As a result of the extensive number of messages between the pair, over this long period of time, the main focus of this analysis will be on the text messages that were sent the day of Connor’s suicide.

Bradley

Bradley was a 21-year-old male who died of a multiple prescription drug overdose whilst livestreaming and messaging a pro-drug forum. Bradley spent most of his time online and was described as a computer expert. The identity of the suicide, as well as the chat log of the events that occurred during in the final hour of Bradley’s, was made public by his parents in an attempt to spread awareness of the behaviours that contributed to his death, as well as to prevent false information. Bradley detailed his experiences as he consumed a deadly mixture, several times over the lethal limit, of Methadone (an opioid) and Clonazepam, and Temazepam (both benzodiazepines), amongst other prescribed drugs, copious amounts of alcohol, and illicit drugs such as psilocybin and LSD. Those that were witness to the series of events initially praised him and encouraged him to take more, but they later realised he had taken it too far and feared for his life.

Matthew

The third case is Matthew, a 21-year-old who hanged himself and sent pictures every two seconds of the events to an open discussion forum dedicated to free speech. Matthew declared his intentions on the forum and received mostly encouraging comments from others, encouraging him to go through with it. Seven individuals sent encouragement and method tips, and only three were against the idea. Unfortunately, by the time others in the forum had alerted emergency services, Matthew was dead. The forum posts are still available publicly online, demonstrating Matthews chosen final communications before death.

Thematic Analysis

The method used in this analysis is thematic analysis, specifically the approach by Braun and Clarke (2006). This process involves six phases that the coder followed. First, the data must be familiarised, and this was accomplished by repeatedly reading the datasets several times, allowing time to process thoughts about the text. This was followed by the generation of initial codes in which the coder actively noted along each line of data the different codes that capture what is being said, felt, or implied at each point. These codes are the building blocks for themes (Clarke and Braun 2017) which describe commonalities within and across data sets. The third step is to actively search for and review these themes. Not all themes will be relevant to the research question, and so it is the role of the researcher to determine what themes are important and which are not (Braun and Clarke 2006). The next step is defining the themes, ensuring that the names for each theme capture what the data are reflecting, and then the write up can commence.

Thematic analysis is aimed not just at summarising data but also at identifying, interpreting, and exploring key elements of the data. Thematic analysis was deemed most appropriate for this analysis because of its flexibility and ease of use, whilst still allowing for an in-depth, rich examination of the perspectives of multiple individuals (Nowell et al. 2017).

The study, because of the nature of publicly available secondary data and the anonymisation of the participants, does not go against any ethical guidelines and was given ethical approval before commencement.

Results

Through the process of thematic analysis, there were several themes that occurred in all three cases and some that were found in one or two. Two main themes were identified: the individualistic experience/behaviours of the suicide and the outside impact of those online who were witness to the events. The themes related to the suicides are pride, mental health and social issues, attention and self-esteem, and hesitation. The discovered themes related to the role of the online audience/witnesses are encouragement, doubt, manipulation, and excitement. The theme of hesitation was seen in both the audience and suicide. All spelling and grammatical errors within quotes are those of the utilised sample.

Pride

A common theme that was discovered between Bradley and Connor is that of a sense of pride. Bradley had tried over time to establish a reputation as being what he calls “hardcore” amongst the pro-drug community. Bradley took pride in his perceived tolerance to drugs and his extensive knowledge and collection of drugs:

don’t talk to me about doing enough drugs, that’s my personal stash.

On the night of his suicide, he claimed that drug use was his usual weekend behaviour, but the audience had previously accused him of lying. The last coherent words that Bradley typed before he died were:

I told u I was hardcore, fuck u, pusys, u are so fucking stupid.

Bradley felt that he needed to prove his audience wrong and make them believe that he had God-like drug tolerance. The sheer number of drugs in his system, particularly the opiate and benzodiazepines, was way above the lethal limit and, even though he claimed that he does this much every weekend, he would have overdosed much earlier if he had done so. Bradley, on this evening, decided that he had to prove himself and, despite being warned by viewers that he was going to die, he mustered up the strength to continue to take more drugs.

Connor was suicidal and had attempted suicide in the past several times. Molly forced pride issues onto him by saying “people who commit suicide don’t think this much and they just do it”. As discovered by Horne and Wiggins (2009), those that use the Internet as a form of suicidal self-expression do so to construct an identity as authentically suicidal, and they want to obtain validation of this identity from others online. Before his suicide, Connor would post YouTube videos online expressing his suicidal ideations and likely associated suicide as part of his identity. Thus, by saying what she wrote, Molly minimised his suicidal intent, and Connor may have felt that this was an attack on his identity and, therefore, assaulted his pride.

Mental Health and Social Issues

Both Matthew and Connor openly discussed their suffering in terms of mental health. Matthew described himself as overly vulnerable, with high functioning autism, but nonetheless described his life in a very positive light.

I have a good life, study at my own pace, own apartment, good income … poor social ability which makes me a somewhat lonely person. I think I have had a good upbringing despite a lot of hassle at school. Makes no sense that I want to take my own life? Well I know.

A forum user tracked down Matthew’s Facebook account whilst trying to get him help and shared his suicide note that he posted online. In which he wrote, in part:

[I] do not think my life is bad at all, it is actually really good. But sometimes it’s just too hard.

Despite describing all the ways in which his life is good and, as he also wrote in his suicide note, he loved his family more than anything, he could not find a good enough reason to live. Research has suggested that suicide occurs more frequently in individuals who have high functioning autism (Richa et al. 2014). The only issues that Matthew describes about his life are social-related, which is understandable as autism spectrum disorders affect social understanding, communication, and behaviour (Gittins et al. 2018).

Connor, in his messages to Molly, often detailed his struggles with social anxiety, feelings of hopelessness, and his constant need to compare himself to others. Even when he describes time living with his friends he says:

The thing is I was living with them. So I would have to stay in the social situation with limited break. And even on the first night there I couldn’t sleep and I felt like I needed to get away so I don’t get really depressed again.

It appears that Connor suffered during social interactions, telling Molly that he believed no one liked him and that the intense anxiety of being around others made him unable to communicate in person.

It is likely that Bradley was also suffering from mental health issues, as the drugs that were prescribed to him included Temazepam and Clonazepam, which are both benzodiazepines used to treat anxiety, panic attacks, and insomnia (Offidani et al. 2013). He describes using Methadone frequently, which is an Opioid usually prescribed to relieve serious pain or heroin addiction (Strang and Sheridan 2006). This suggests that, in all three cases, the suicides were suffering from mental health and social issues, which may be the reason that these individuals chose to die by suicide.

Attention and Self-Esteem

A theme that appeared in both suicide chat rooms is the theme of wanting attention, potentially indicating low self-esteem.

Have always felt that I want to publish my suicide haha.

Matthew desired that people watch his suicide and keep a record of the pictures.

The FTP will of course be up until someone shuts down my computer, but do not think they are the first things they think of when they come in here. Someone can take a backup or something.

This may be an attempt to immortalise himself and gain notoriety on the Internet or just to obtain respect and attention from others. Bradley also seemed to be putting on a performance for his audience when he asked:

how many people are watching this nonsense?

In Bradley’s mind, it may seem like the risks of his behaviour are worth it as long as people are watching and remembering. Attention-seeking behaviour often stems directly from having a low self-esteem, suggesting that these individuals may base their self-worth on the attention that the get from others.

Connor openly talked about his battles with his self-esteem issues and would send lengthy messages to Molly.

I have no confidence and low self-esteem, and always feel like others are smarter.

Hesitation

Another theme identified in these cases is the admission of hesitation. Connor expressed several times to Molly that he was not quite ready to die by suicide and that he feared for his family.

Like why am I so hesitant lately. Like 2 weeks ago I was willing to try everything.

Connor’s hesitation, had it been met with emotional support and discouragement of his suicidal ideation, might have saved his life.

Matthew expresses that he has been hesitant to commit suicide through fear:

Well I’m not feeling so bad today, but the thing is, I’ve been suicidal for the past six months, its always been such a scary thing to take my own life…

Even after having decided that he was going to die by suicide, there was still a moment in which Matthew experienced hesitation. Moments before he hanged himself, he wrote:

Starting to feel like I’m changing and if I want to kill myself, I’ll have to hurry a little.

In Bradley’s case, some forum users potentially had the chance to save his life.

I am on line with 911. Is this the right choice?

NO, NO, NO, NO, NO, NO …

The individuals feared investigation into themselves, first for being active users of the pro-drug community and sharing their own drug experiences and second for encouraging Bradley’s behaviour. They also feared for Bradley, that if he did not die or if he had lied about the quantity of drugs he had consumed, that he would be penalised for his large possession and consumption of drugs. The drugs that killed Bradley were prescribed to him, but he also admitted to consuming psilocybin, marijuana, and LSD. None of these substances can cause death (Nichols 2016) and would not have contributed to this but nonetheless were still illegal at the time (Hall et al. 2020).

Encouragement

A theme found in all three cases is encouragement from the witness to the suicide. Connor was heavily encouraged by Molly. In their messages to each other, Molly pushes Connor relentlessly into thinking about suicide, telling him to do it as soon as possible and that it is the right thing to do.

Molly: I’m not gonna sleep until you’re in the car with the generator on …Do you promise [that you will kill yourself]?

Connor: I promise babe. I have to now.

Molly: and you can’t break a promise.

This is different to the encouragement that was being received by Bradley and Matthew. Bradley was encouraged by others, but mostly because they did not believe he was doing what he said he was.

I told him he was a pussy a minute ago as a joke and he ate more.

His behaviour was then reinforced with positive comments calling him “hardcore” and a “gangster”.

Matthew was also encouraged by the majority of those in the forum, many of whom also did not believe him. Some gave him tips on how to make his suicide more practical, and one even tried to hurry along the events, saying that he needed to go to work soon. Others said:

Good luck then.

I guess no one can stop you from doing that, but I hope you have a wonderful time in your next life.

Research has found that it is much harder for suicidal individuals to retract on their suicidal intention if there is an audience encouraging them (French 2020), as demonstrated in all three cases.

Doubt

One key theme throughout is the theme of doubt. All those who were witnesses, by video or message, to these suicide cases doubted that the actions of the individual were genuine or as serious as they were. Comments expressing their doubt in Bradley further antagonised him to take more. One witness wrote:

[Bradley] is either hardcore, ungodly, crazy or one hell of a poser.

Molly doubted that Connor would kill himself and expressed these doubts to Connor. Molly on the day that Connor died by suicide.

If you don’t do it now you’re never gonna do it. And you can say you’re going to do it tomorrow but you probably won’t.

In the case of Matthew, the other thread readers believed he was an Internet “roll”, and they did not believe that he would actually go through with the suicide. It also appears that Matthew doubted himself.

… you are just a joker, go and hang yourself.

You will never dare, you are too cowardly.

To which Matthew replied:

Haha it remains to be seen. Both for you and me

French (2020) noted that livestream suicides put witnesses in a situation that they find too difficult to get out of. They often doubt the suicidal intent (e.g., “You will never dare”). In each of these cases, the suicides were not believed by the audience, thereby encouraging them to step up and go through with the act.

Manipulation

In the case of Connor, Molly used emotional manipulation in order to convince Connor to kill himself. Whilst this theme is not present in all three cases, it was of great importance for Connor’s suicide.

Molly: I’m serious, you can’t even wait til tonight. You have to do it when u get back from your walk.

Connor: Thank you [for] still being here

Molly: I would never leave you, you’re the love of my life, my boyfriend. You’re my heart, I’d never leave you.

Throughout the series of text messages, Molly only expresses her love towards him when he is agreeing to kill himself. When he is demonstrating hesitation, she instead treats him with hostility.

Connor: I am gonna eventually. I really don’t know what I’m waiting for … but I have everything lined up.

Molly: No you’re not ... Last night was it. You keep pushing it off and you say you’ll do it but u never do. Its always gonna be that way if u don’t take action.

Molly was probably Connor’s only confidant during his difficult times, and restricting her affection to only when Connor has agreed to die by suicide is emotional manipulation. Furthermore, Molly threatened to Connor that she would get him help if he did not kill himself:

you just need to do it … or I’m gonna get you help

Connor had in the past insisted that he does not want to see a professional about his mental health, made Molly promise to keep it secret, and made it clear that failing to keep it secret would terminate their relationship. However, Molly uses this against him, pushing him towards suicide.

Excitement

A theme on all three cases is that of excitement in the witnesses. Whenever Connor expresses that he is going to go through with the suicide, Molly starts to get excited, “I can’t believe this is really happening. You’re actually gonna do this … I always knew you were special”.

After Matthew had die by suicide, there were posts from individuals watching the series of events expressing:

this is uncomfortable as hell but at the same time very exciting!

This level of excitement across all three themes may suggest that the lack of face-to-face communication makes the events seem more like entertainment than real life.

Finally, the themes are presented in Figs. 1 and 2 below.

Fig. 1
figure 1

Venn diagram of the themes associated with the victims

Full size image
Fig. 2
figure 2

Venn diagram of the themes associated with the audience

Full size image

Discussion

This study is one of the very few qualitative analyses on online-assisted suicide from the perspective of both the audience and suicide, with the goal of highlighting key themes to obtain an understanding of why and how people commit cybersuicide. The three suicide cases used for this analysis are all drastically different: a voluntary hanging, a persuaded suicide, and an “accidental” overdose. Nonetheless, there were many common themes across all three cases. We identified hesitation, mental health and social issues, and attention seeking behaviour and low self-esteem in the suicides. We identified encouragement, doubt, and excitement in the witnesses or audience. Hesitation was also observed in the audience in Bradley’s case. Two less common themes were also identified: pride in the cases of Bradley and Connor and manipulation by another in Connor’s case.

In terms of the role of the audience, a common theme in all three cases is that of encouragement. As reported in previous research, people online are often too suspicious of someone’s suicidal intentions, and this can lead to suicide baiting and trolling (Fratini and Hemer 2020; Phillips et al. 2019). This highlights the lack of knowledge of suicide and suicide prevention in the general population and the stigma associated with suicide.

As well as the role of the troll, the role of the suicide “expert”, as seen in research by Fratini and Hemer (2020), was also present in the cases of Connor and Matthew, as the audience recommended better ways to die by suicide or complained that the method chosen by the potential suicide was not effective. Both trolls and suicide “experts” played a role in the suicides, whereas they could have done more to prevent them. The data suggests that the majority of those who were baiting the suicides were doing so because they doubted that the intentions and acts of the suicides were genuine. Research has found that it is much harder for suicidal individuals to retract on their suicidal intention if there is an audience encouraging them (French 2020), and this can be seen in each of the cases used in this analysis.

The group mentality of those within the pro-drug community is that the use of drugs is good. By hiding their true identities, being part of a large group, and being remote, it could be assumed that those witnessing Bradley overdose were encouraging him as a result of deindividuation and the pressure of social identity in the context of the pro-drug platform (Coppolino Perfumi et al. 2019). Furthermore, behind the disguise of anonymous social media profiles, the responsibility to offer support diminishes, particularly in larger groups (Markey 2000). This finding was supported by the present study.

In all three cases, there were moments of hesitation that, if handled appropriately, could have saved a life. Matthew and Connor both displayed hesitation when the time was getting closer to their intended suicide, and had someone reached out to them or contacted emergency services, they could have been saved. Bradley provided his address should anything go wrong. If those in the chat not been so scared of the consequences of calling for help, Bradley may have lived. With the knowledge that those who die by suicide online have moments of hesitation prior to going through with the act, there is room to educate others into helping prevent the suicide. Social support is a protective factor for suicidal individuals, and if audiences are active enough to provide social support and attempt to call emergency services, lives can be saved (Ma et al. 2016).

Each of the individuals demonstrated some level of mental health issues primarily anxiety and interpersonal disorders. Their attention-seeking behaviour aligns with past findings that those with traits of extraversion and neuroticism are more likely to suffer from suicidal ideation (Stefa-Missagli et al. 2020). Each of the cases appeared to have social anxiety and depression but also was partaking in risky behaviours, putting themselves out there in public display and demanding attention, thereby demonstrating traits of neuroticism and extraversion.

The theme of excitement in the witnesses in all three cases suggests that online interactions are experienced differently remotely, as witnessing a suicide in person is a traumatic experience (Robb 2018). This aligns with the interpretation of suicide being a staged performance (Lester 2015), and the reactions received from the audiences appear to be almost typical of those in an entertainment situation, such as films or TV, rather than in real life. In Bradley’s and Matthew’s cases, in which there were several people witnessing the events, the suicides felt like a performance to the witnesses, and this suggests that attention-seeking is an important part of cybersuicide for the suicides. Suicide and mental illness are often stigmatised in society and viewing suicidal ideation and suicide attempts as attention-seeking further stigmatises these behaviours (Li et al. 2015; Sudak et al. 2008). It is crucial that any awareness projects for the public are sensitive to this, and attempts are needed to destigmatise reaching out for help.

One feature in Connor’s case is that he was emotionally manipulated into dying by suicide. This theme was not present in the other two cases, and his situation involved a private communication rather than a public display. Connor and Molly had a long-term online relationship, sand she had much more control over his emotions than the audiences to Matthew’s and Bradley’s suicides.

Bradley’s case may not be considered a genuine case of suicide, but rather an accidental overdose. Nonetheless, researchers have discovered that often accidental overdoses are a hidden expression of suicidal ideation (Heale et al. 2003). Those who partake in illicit drug use with a lower self-esteem and a stronger feeling of hopelessness are significantly more likely to “accidentally” overdose. Moreover, the incidence of suicidal ideation and attempts amongst those who use opiates is 14 times that of the general population (Wilcox et al. 2004). The extreme consumption of drugs suggests that Bradley knew he was risking his life, yet nevertheless proceeded to take more drugs even after becoming incomprehensible and blacking out. He was, therefore, deemed an appropriate case of cybersuicide for this analysis.

Whilst it may seem an obvious decision to not allow unregulated sites that encourage free speech pro-suicide beliefs (and in this case as well pro-drug forums), they can have an important strength. Murray (2011) detailed how one user of pro-suicide forums said that they function as a diary for those fortunate enough to find them and provides a place to open up in a non-judgemental environment, without shame, and to voice their inner turmoil. Censoring this content removes the possibility of talking through one’s problems with other like-minded people, and there have been many cases where communications through media such as this has reduced suicidal ideation and saved lives (Davis and Lewis 2019; Murray 2011).

Moreover, online suicides do not happen only in pro-drug and pro-suicide forums, but across regulated social media websites such as Facebook (Fratini and Hemer 2020). The problem is not the platform, but the skills that individuals carry with them in terms of being able to communicate and assist when a person displays suicidal intent.

The present study comes at a crucial time as the world struggles with a global pandemic when normal daily life has shifted online for the most part. A large-scale study, nationally representative of the UK, demonstrated a significant negative impact on anxiety, depression, and loneliness from this (Li and Wang 2020). A recent survey found that more than one-third of firms that changed to remote working because of the pandemic believe the practise will remain in place at their company long after Covid-19 is no longer a threat (Bartik et al. 2020). Furthermore, online teaching at schools and universities will remain in place in some areas of the world as society adapts to involve the Internet into all parts of daily life (Singer 2021). Whilst the efficiency of the Internet is clear, this increased use and dependency on the Internet will have an impact on cybersuicide and other risky and dangerous behaviours, and so research is essential to prevent losing more lives online.

Limitations

In the case of Matthew, one key limitation that cannot be ignored is that the webpages had to be translated into English using Google translating. Google translator has shown to produce somewhat flawed translations, but still comprehensible (Groves and Mundt 2015).

With the ages of the victims in the present study being 18, 21, and 21, these findings may not be representative of the experiences of older adults. Young people are believed to be more vulnerable to suicide because of their immaturity which affects impulse control, peer pressure, rational decision-making, and the capacity for long-term planning (Casey et al. 2011).

Nonetheless, research using young adults is important considering that suicide is the second leading causes of death globally for those aged between 15 and 25 (WHO 2018). Young people are much more likely to use the Internet for self-destructive purposes (McCarthy 2010). Young adults today have been raised in a digital environment, and they use the Internet and social media more than older adults (Ofcom 2020). Thus, they are most at risk of the abuses of the Internet (Hunsaker and Hargittai 2018).

A further limitation of the current study is that the sample was limited to males, potentially a result of the gender differences in suicides rates. (Many more men die by suicide than do women [Hawton and van Heeringen 2009]) Research by Clarke and Amerom (2008) suggested that depressed males are more likely to discuss suicide via online blogs than are women. It may be the case that males are more likely to die by suicide using the Internet than are females and that women have different cybersuicide processes and experiences to those of men. This is something to be explored in the future.

With these issues considered and the use of only three cases for this analysis, there are some issues in terms of generalisability. Nevertheless, there is a lack of data on this topic online that can be analysed. The sensitive nature of suicide, including the desire to prevent copycat cases, means that the data are often not made public (Fahey et al. 2018). In the present cases, Connor and Molly’s messages were made public only due to high-profile criminal investigations. Bradley’s mother made her son’s messages public to raise social awareness, and Matthew’s forum remains available online for anyone to read. In normal circumstances, personal data, such as direct messages, would not be made publicly available to protect the families of the suicides and to respect their privacy. Despite the small sample, the qualitative method of thematic analysis allows for rich, in-depth exploration of individuals experiences (Braun and Clarke 2006), and it is essential to understand suicide in-depth at an individual level to be able to make progress in the field of suicidology (Hjelmeland and Knizek 2010).

Future Direction

At the present time, there are many more social media platforms available than those used in the present analysis (e.g., TikTok use has been associated with suicide [Manzar et al. 2021]), and the wide variety of different media and audiences online make it impossible to make blanket statements about the Internet without taking these other platforms into consideration. Future research, therefore, is needed to analyse online suicide cases across different types of social media to grasp a better understanding of why and how people die by suicide online and how different audiences can provoke or prevent suicide. It would also be of use to further study the role of the audience when it consists of only one person rather than many people, as in the case of Connor.

Whilst there have been many cases of Internet ‘trolls”, it is important to treat all indications of suicidal intent as if they were real. A study by Phillips and Mann (2019) examined cases where an individual threatened to die by suicide whilst live streaming, of which they identified 26 cases, and they found that the individual followed through with the threat in 92% of the cases. There needs to be an increase in public awareness into the rates of online suicide and how it can be prevented, and one of the most important things is to take every threat seriously, whether that involves getting involved in trying to help or simply avoiding making comments that might further antagonise a vulnerable individual into dying by suicide.

Future projects should attempt to destigmatize mental illness and seeking professional help, but not necessarily destigmatise suicide (Sudak et al. 2008). As seen in the past, the media’s representation of suicide can influence copycat suicides (Fahey et al. 2018). A stance must be taken that does not attempt to normalise or justify suicide (Sudak et al. 2008).

Conclusion

Suicide is one of the leading causes of death across all ages (WHO 2014). With technology and the Internet becoming more and more integrated into everyday life, it is important to find an effective online system for preventing Internet-assisted suicide. This study has shed light on different themes in cases of online suicide, such as the hesitation experienced by the suicide and the excitement experienced by the audience. Audiences encouraged those online to die by suicide, often because they did not believe them. This seems to result in pride issues for the suicidal individual, and the lack of in-person communication made the suicides appear more like entertainment than real life for the audience and witnesses. More research in this area is needed, including the study of suicidal females and older adults online and the use of other online platforms. Public awareness is required for suicide prevention so that society is better equipped to help those in need and reduce the stigma of receiving treatment.