Main findings and interpretation

Although specific methods, and composite methods, have been associated with a higher risk of death and subsequent suicidal behaviour, epidemiological analyses of repeat hospital admissions for self-harm have concluded that ‘method of self-harm is fluctuating and unpredictable’.^{Reference Owens, Kelley, Munyombwe, Bergen, Hawton and Cooper9} The results of this exploratory study of first-person accounts suggest that, although indeed changing and complex, method choices are reasoned, personally and culturally meaningful, and a crucial element of the processes and planning involved in attempting (or desisting from) suicide. With very few exceptions, individuals who had contemplated suicide had considered a specific mean (or, more often, means) of taking their own life, even when they had never then engaged in suicidal behaviour. Almost 90% described a particular reason, or set of reasons, for considering or discarding such method/s.

To date, method choices have primarily been investigated in relation to completed suicide and suicide attempts.^{Reference Kõlves, McDonough, Crompton and de Leo5,Reference Biddle, Donovan, Owen-Smith, Potokar, Longson and Hawton21,Reference Rezaie, Hosseini, Rassafiani, Najafi, Shakeri and Khankeh22} This is the first study to explore patterns and motivations for method choices in relation also to suicidal ideation, which is a surprising gap in knowledge given that 10–20% of individuals experiences lifetime suicidal ideation.^{Reference McManus, Bebbington, Jenkins and Brugha2,Reference Nock, Borges, Bromet, Alonso, Angermeyer and Beautrais23} Suicidal thoughts are a known risk factor and immediate precursor of suicidal behaviour, and key to understanding the transition from ‘suicidal ideation to action’.^{Reference O'Connor and Kirtley14,Reference Klonsky, Saffer and Bryan15} In relation specifically to methods choices, suicidal thoughts provide some important insights into the acceptability and cognitive availability of different means of suicide. In other words, they can help us understand which methods are cognitively available to individuals in crisis (in a given sociocultural context), and what makes them more or less ‘attractive’ to them.

In this study, self-poisoning was the most commonly contemplated method of suicide. However, most respondents had considered multiple methods of suicide, particularly those who had also engaged in suicidal behaviour. These included a fairly wide range of potential methods and, especially among individuals who had never attempted suicide, relatively uncommon means such as jumping or lying in front of a moving object, drowning, handgun discharge and crashing of a motor vehicle. In England and Wales, self-harm involving a moving object, drowning or fall and fracture accounts, all together, for 12% of all suicides,²⁴ and only 1% of hospital presentations for self-harm.^{Reference Geulayov, Casey, Bale, Brand, Clements and Farooq10} Firearm use, although a common suicide method in countries with less stringent gun control such as the USA,^{Reference Ajdacic-Gross, Weiss, Ring, Hepp, Bopp and Gutzwiller25} is even rarer in the UK, accounting for less than 2% of all suicides²⁶ and 0.03% of self-harm hospital presentations.^{Reference Geulayov, Casey, Bale, Brand, Clements and Farooq10} Although these are only a small proportion of completed suicides and hospital presentations for self-harm, such methods tend to be disproportionately reported in the media,^{Reference Marzano, Fraser, Scally, Farley and Hawton11} which may reflect and account for their over-representation and over-availability in people's suicidal thoughts (in the UK, most suicides are hanging¹⁶ whereas most hospital presentations for self-harm involve self-poisoning^{Reference Geulayov, Casey, Bale, Brand, Clements and Farooq10}). Almost 10% of the participants explicitly identified exposure to a given method as a primary motivation for considering it as a means of suicide. Even more common reasons were perceived method ease, accessibility, speed and effectiveness, alongside the desire to minimise pain and violence.

However, the range of methods described in relation to actual suicidal behaviour was narrower, as were respondents’ motivations for using such methods. Over three quarters of respondents attempted suicide by self-poisoning, with many switching to this method having also considered more violent or unusual means. For almost half of the participants, choice of attempt method was primarily dictated by accessibility, and ease of method and the minimisation of pain were more common concerns than perceived speed or likelihood of death. An exception was jumping/lying in front of a vehicle, for which method effectiveness was the most frequently cited reason (see also Marzano et al^{Reference Marzano, Mackenzie, Kruger, Borrill and Fields18}).

In relation to all methods, lethality appeared to be a greater concern for male than female respondents, and impulsivity less common, as also reported elsewhere.^{Reference Cibis, Mergl, Bramesfeld, Althaus, Niklewski and Schmidtke27,Reference Värnik, Kõlves, Van Der Feltz-Cornelis, Marusic, Oskarsson and Palmer28} Together with the increased use of methods other than self-poisoning,^{Reference Bergen, Hawton, Waters, Ness, Cooper and Steeg29} this might contribute to the disproportionate risk of suicide in men.

Of note are also some of the differences between respondents who had contemplated but never engaged in suicidal behaviour, and those who had previously attempted suicide. The former were more likely to mention wanting to minimise the impact on others, especially family and friends, and, in smaller numbers, to be in an isolated location where one's attempt and body would not be discovered. This points to the presence and impact on bystanders and loved ones as important ‘dissuaders’ in relation to suicide. Indeed, these were also among the most frequent reasons for not attempting suicide by a specific method. Other common deterrents were fears of surviving with injuries and of ‘not getting it right’.

Interestingly, impulsivity was not a more common theme in those who had attempted suicide per se, but was cited more frequently in relation to suicidal behaviour than ideation. This finding lends support to the idea that individuals who attempt suicide may not have significantly elevated trait impulsiveness, compared with ‘ideators’; however, they may have higher impulsiveness when in a negative state.^{Reference Millner, Lee, Hoyt, Buckholtz, Auerbach and Nock30}

Strengths and limitations

We analysed rich descriptions of people's choice of methods in relation to suicidal thoughts as well as behaviours, with strong interrater reliability, and no prompting, structure or limit to the answers that could be provided. This exploratory, inductive approach is rare with samples as large as the current study's, and allows for appropriately powered statistical analyses, as well as more nuanced, in-depth analyses of particular groups or methods, including the different images, myths and cultural scripts that exists around – and against – specific means of suicide (as presented elsewhere in relation to railway suicide^{Reference Marzano, Mackenzie, Kruger, Borrill and Fields18}).

However, findings were based on a self-selected, predominantly female sample, and may not necessarily be representative of all individuals who consider, attempt or indeed die by suicide, within the UK and more widely. National and cultural variations in suicide method choices suggest the need for replication with broader and more diverse samples,^{Reference Ajdacic-Gross, Weiss, Ring, Hepp, Bopp and Gutzwiller25} in a wider range of community and clinical settings.

Given the exploratory nature of this study, we did not gather systematic information about potentially important factors such as prior psychiatric and family history, nor about the sequence or temporality of suicidal thoughts and behaviour. This limits the conclusions that may be drawn in relation to method switching, escalation and substitution, and prevented us from making inferences about age differences in method choice. Further, longitudinal studies could usefully investigate how the frequency, intensity and intent of suicidal thoughts and behaviour may affect method choices in different groups and communities. A more structured approach to gathering information about deterring factors could also offer important insights into how best to prevent, and ‘dissuade’ from, specific means of suicide, in different contexts and locations.

Implications

Previous research has concluded that ‘people's risks or needs [cannot be] based simply on the method of harm’,^{Reference Owens, Kelley, Munyombwe, Bergen, Hawton and Cooper9} or the potential lethality of that method. The results of this study also suggest that the methods people consider when contemplating suicide are often not the ones that are then enacted in a suicide attempt, nor are they necessarily chosen for the same reasons. Nonetheless, exploring method choices is far from a fruitless activity, and can usefully inform preventative initiatives.

Consistent with earlier literature, our findings suggest that cognitive and physical availability are key drivers in the choice and prevalence of suicide methods. The former appears to be particularly relevant in the context of people's suicidal thoughts, with methods perceived to be ‘easy’ and lethal featuring highly in individuals’ so called ‘ideation menu’.^{Reference Pharoah31} The latter (i.e. the accessibility of a given method), becomes especially important in the context of a suicide attempt, and impulsivity relatively more common. This clearly underscores the need for preventative measures that restrict access to means and delay impulsive behaviour, such as the erection of barriers at high-risk locations and reduced pack sizes of paracetamol.^{Reference Yip, Caine, Yousuf, Chang, Wu and Chen32} Avoiding depictions and descriptions of suicide methods in the media may help limit the cognitive availability of specific means, particularly for methods portrayed as lethal, easy, quick and painless. This may be especially important in the reporting of celebrity suicides. A recent meta-analysis found that when the suicide method used by a celebrity was reported, there was an associated 30% increase in deaths by the same method.^{Reference Niederkrotenthaler, Braun, Pirkis, Till, Stack and Sinyor33}

In clinical settings, exploring the decision-making around specific methods may help challenge unhelpful myths and misconceptions, identify areas of ambivalence and hope, and develop appropriate safety and treatment plans. At public-health level, factors known to attract individuals and subgroups to lethal means of suicide may be challenged via targeted communication and media strategies, and common ‘dissuaders’ reinforced. However, the risk and unintended consequences of different approaches to means and ‘myth-restriction’ need careful thought and evaluation, whether in the context of means-restriction counselling^{Reference Bryan, Stone and Rudd34} or as part of wider public health approaches. Further research is needed in this area, but the risks and ethics of how and where research findings are disseminated and ‘translated’ need careful attention to avoid reinforcing the acceptability and availability of (lethal) means of self-harm.