4 Discussion

4.1 High-risk groups
4.2 Interventions
4.3 Interventions in Indigenous populations
4.4 Data availability and study limitations

4.1 High-risk groups

The review identified a raft of evidence identifying those groups that are at elevated risk of suicide. These correlate to some degree with those identified by the LIFE Framework as being supported by Level A ('strong') evidence. However, as the definition of Level A evidence in the LIFE documentation extends to robust evidence other than systematic reviews and meta-analyses, it is unsurprising that not all were captured in the current review. In addition, the LIFE analysis concentrated on risk factors, rather than at-risk groups, and utilised statistical data in addition to peer-reviewed literature, further accounting for differences.

These LIFE Framework risk factors include family history of suicide and psychopathology, negative life events and low coping potential, marital status (with people who are divorced, widowed, separated and single identified at higher risk), low socioeconomic status and unemployment, and neurobiological, psychological, social and environmental factors.

Interestingly, the current review suggests good evidence for groups not identified in the LIFE Framework being at increased risk of suicide: these include culturally and linguistically diverse populations and LGBTI communities (identified in the LIFE Framework, but with a cited lack of good evidence), youth and current or recent prisoners.

These inconsistencies highlight the need for more robust data collection processes and further research in this area.

4.2 Interventions

4.2.1 Universal approaches

As noted above, universal approaches to suicide prevention target whole populations with the aim of reducing risk factors or enhancing protective factors. The types of universal interventions identified by the current search include restricting access to means of suicide, guidelines for media reporting of suicide, and school-based suicide education and awareness programs.

Reducing access to means

Restricting access to the means of suicide is a feature of most national suicide prevention strategies, and was highlighted by each of the four papers that reviewed broad-spectrum approaches to suicide prevention as one of the more effective interventions to date.^57-60 Three additional retrieved studies reviewed the literature specific to means restriction as a suicide prevention method.^61-63 These studies add support to the above conclusions that restricting access to means reduces the number of suicide deaths.

The studies by Mann et al⁵⁷ and Beautrais et al⁵⁸ report that restricting access to certain suicide methods has been shown to reduce the number of suicide deaths by that method, including in the case of firearms, pesticides, medication such as barbiturates and analgesics, compulsory use of catalytic converters in motor vehicles, construction of barriers at jumping sites, and the use of new, lower toxicity anti-depressants. Although the authors acknowledge that substitution of method can occur, restriction of a common means of suicide in a given country can lead to a reduction in the overall suicide rate. For example, this has been the case following the restriction of firearms in parts of North America and Canada, and barbiturate restriction here in Australia. However, restricting common means of suicide may be easier in some cases than others: for example, it may be easier to restrict access to pesticides than to hanging (the most common method of suicide in Australia).[5] Furthermore, in order for means restriction to be an effective strategy access to timely and accurate national mortality data is required so that interventions can reflect current trends. Top of page

Implementation of media guidelines

The reporting of suicide in the media has the potential to precipitate suicidal behaviour in vulnerable individuals. For example, a recent systematic review by Sisask et al⁶⁶ concluded that there was an association between media reporting of suicide and actual suicide. The type of media exposure rather than suicide reporting per se, likely explains the link: sensationalising suicide, explicit descriptions of means and the portrayal of suicide as a legitimate solution to problems by the media may contribute to increasing suicide risk in some vulnerable individuals.⁷⁹ Pirkis et al⁸⁰ report that media items on suicide are more likely to be associated with suicide if they appear on television rather than radio, and if they report on actual suicide, rather than suicidal ideation or attempted suicide.

Some countries have implemented reporting guidelines in response to these findings, and the reviews by Mann et al⁵⁷ and Beautrais et al⁵⁸ highlight that reporting guidelines may be used to achieve accurate and non-sensationalist reporting of suicide. Beautrais et al⁵⁸ outline that there is potential to work "...collaboratively with media to disseminate factual, accurate evidence and information about suicide and mental health in a non-stigmatising manner and to promote knowledge and information about suicide prevention".

Bohanna et al⁸¹ recently undertook a systematic review assessing evidence for the effectiveness of media guidelines on the reporting of suicide. The authors concluded that guidelines are able to prevent imitative suicide, however the awareness, use and opinion of such guidelines by journalists is generally low. Given this, more training and collaboration between media and mental health agencies is likely to improve the implementation, and thus effectiveness of such guidelines.

Responsible media reporting is a feature of the majority of suicide prevention strategies across the world, including in Australia. However, there remains much to be done to promote responsible and informed media coverage of suicide by maintaining, implementing and promoting the use of existing media resources, and by working collaboratively with media professionals and outlets, around the dissemination of factual information about suicide and mental health (and its prevention) in a non-stigmatising or non-sensationalist manner. There is additional evidence generated by Australian-based researchers who report that the implementation of media guidelines here in Australia has led to an increase in both the number and quality of media reports about suicide,⁸² adding more weight to the value of such guidelines. Top of page

Awareness programs

Awareness programs were highlighted by each of the four reviews that looked at the broad-spectrum interventions for suicide prevention.^57-60 However, the effect of these programs on rates of suicide is unknown. These programs usually target the general public with the aim of increasing education and awareness of suicidal behaviour and mental illness, improving understanding by recognising established risk factors and reducing stigma by challenging unhelpful perceptions or myths. The overarching aim is to increase treatment seeking, support and recognition for those with suicide risk and mental illness and, by this mechanism, influence suicide rates.

General population education and awareness programs appear somewhat effective in changing attitudes but fall short of translating attitude change into behaviour change. For example studies in the UK, Germany, New Zealand and Australia indicate these programs have some effect on attitudes regarding causes and treatments for depression however they do not appear to reduce suicidal behaviour or increase help seeking.⁵⁷

There is some evidence to suggest that more success may be had with programs that are targeted to specific groups (e.g. adolescents) and with those particularly aimed at suicide prevention as opposed to depression or mental illness more generally.⁵⁸ Some have suggested that school-based suicide prevention programs are effective in terms of increasing students' knowledge in relation to suicide.⁶⁵ However, within their review, Ploeg and colleagues also note the potential risk for such programs in increasing hopelessness and maladaptive coping behaviours in males specifically.⁶⁵ Cusimano and Sameem,⁶⁴ in a more recent systematic review of school-based suicide prevention programs concluded that knowledge, attitudes and help-seeking behaviours were increased as a result of such interventions; however, insufficient evidence exists to determine the impact they have on actual suicide rates.

There is limited evidence to support the use of large-scale suicide awareness programs in schools, although the review conducted by Cusimano and Sameem cited above⁶⁴ did report some potential benefits of such programs. However, it is worth noting that concerns continue to be expressed regarding the possibility of iatrogenic effects of such programs, particularly among already vulnerable youth.⁸³ To our knowledge, to date no studies have examined the potential for negative effects, which has led some to recommend that broad-based awareness programs should focus on mental health promotion and not suicide itself, until evidence exists to demonstrate their safety.⁷³ Top of page

4.2.2 Selective approaches

As noted previously, selective interventions target subgroups whose members are not yet manifesting suicidal behaviours, but exhibit proximal or distal risk factors that predispose them to do so in the future. The types of selective approaches identified in this review are GP education, gatekeeper training and screening programs.

General practitioner education

The relationship between depression and suicide risk is well established.^84,85 In addition, depression is often under-recognised and under-treated by GPs, yet many people who die by suicide have had recent contact with a GP.⁸⁶ Thus GP education programs are based on the premise that improving the ability of GPs to identify and treat depression will lead to a reduction in suicide rates. These types of program have been tested in a number of countries around the world and have been shown to lead not only to increases in rates of anti-depressant prescribing, but also (often substantial) decreases in rates of suicide, leading Mann et al⁵⁷ to conclude that such education programs 'represent the most striking known example of a therapeutic intervention lowering suicide rates' (p2067).

These programs were also cited by Beautrais et al,⁵⁸ Feltz-Cornelis et al⁵⁹ and Nordentoft⁶⁰ as one of the most effective suicide prevention strategies to date. Top of page

Gatekeeper training programs

'Gatekeepers' can include clergy, first responders, pharmacists, geriatric caregivers, and those employed in schools, prisons, sports clubs and the military.

Gatekeeper training was identified as an effective preventative approach by three of the four reviews of broad-spectrum interventions retrieved by the current search^57,58,59 Two further relevant studies were also retrieved: one reviewed gatekeeper training across all populations,⁸⁷ while the second reviewed all school-based interventions (including gatekeeper training programs).⁷³

These programs generally focused on raising awareness of risk factors for suicide, increasing confidence and perceived skill when working with at-risk populations, policy changes to encourage help-seeking, improving the availability of resources, and stigma reduction. However, some of these programs (i.e. those conducted in institutional settings such as the Norwegian Army and the US Air Force where the roles of gatekeepers were formalised, and pathways to treatment were readily available) also promoted organisation-wide awareness of mental health and suicide and facilitated access to mental health services. It was these types of programs that reported the greatest level of success in lowering suicide rates.⁵⁷ That said, others reported increases in knowledge, confidence and perceived skill when working with at-risk people, however their impact on rates of suicide-related behaviour and overall suicide rates remains unknown.^73,87

The current search also retrieved two gatekeeper interventions in Indigenous populations that have been formally evaluated (see 'Interventions in Indigenous populations').

It is not clear from the literature exactly which gatekeeper or educational programs are the most effective, or indeed which components of training programs lead to the best results. Therefore specific recommendations cannot be made regarding which programs should be funded. However, together these findings have led both the Senate Inquiry into Suicide in Australia⁸⁸ and a more recent Parliamentary Inquiry into Youth Suicide in Australia⁸⁹ to recommend an increase in the delivery of gatekeeper training across the country, including in rural and remote areas. Top of page

Screening programs

Screening programs generally aim to identify at-risk individuals in order to direct them to appropriate treatment. These programs either directly screen for suicide risk behaviours (e.g. suicidal ideation) or for known associated risk factors (e.g. mental illness). Both Mann et al⁵⁷ and Beautrais et al⁵⁸ indicate that some success has been noted with both types of screening programs in terms of increasing the number of individuals identified as being at risk. However the effect of screening and increased identification on the prevention of suicide remains unclear. This is highlighted by the findings of a systematic evidence review prepared for the US Preventative Service Task Force,⁹⁰ which established that no published studies have yet investigated the effect of suicide screening on actual suicide rates.

Pena and Cane⁹¹ and Robinson et al⁷³ have both reviewed types of screening programs used to detect adolescents at risk of suicidal behaviour, the latter specifically focusing on those used in a school setting. It was concluded that overall, screening programs are able to successfully identity students who are at risk of suicidal behaviour who may not have otherwise come forward for help. However, more research is needed in order to determine the long term benefits of such programs, the referral pathways subsequently utilised by students, and the take up rate of mental health services as a result of the screening process.

Lapierre et al¹¹ report that screening for depression is one of the most popular suicide prevention approaches used within elderly populations, although not all programs measured suicide-related outcomes. Oyama and colleagues⁹² found that community-based depression screening programs in Japan amongst the elderly are associated with a reduced risk of completed suicide.

While the effectiveness of screening programs requires additional research in various populations and settings, at present there is no evidence to suggest that screening for suicide risk is harmful or increases the risk of suicide or suicidal behavior.⁹³ Top of page

4.2.3 Indicated approaches

As described previously, indicated interventions are designed for people who are identified as already beginning to exhibit suicide-related thoughts or behaviours. Types of indicated approaches identified in the current review are pharmacological interventions, psychological interventions and follow-up care post suicide. While there are approaches that show promise in this area, there is clearly a need for more intervention research that specifically targets people demonstrating suicide-related behaviour, in order to ascertain the effectiveness of these approaches in reducing relevant outcomes in those at risk.

Pharmacological treatment

There has been debate in the literature over the past decade as to whether pharmacological interventions reduce the risk of suicide-related behaviour. A limited number of reviews were identified in the current search that assessed the effect on suicidal behaviour of pharmacological treatments used to treat a variety of mental health disorders, including Major Depressive Disorder (MDD) and Attention Deficit-Hyperactivity Disorder (ADHD).

In the past, concerns were raised about whether the use of Selective Serotonin Reuptake Inhibitors (SSRIs) to treat depression may increase suicidal risk in some patients. The current search identified one study that found that the use of SSRIs by adults may reduce the risk of suicide in this population group,⁹⁴ however, this systematic review included observational studies only. Systematic reviews of trials using duloxetine and fluoxetine to treat MDD in adults have found no evidence that the approach increased the rate of suicidal behaviour or ideation in participants^95,96 and in some cases pharmacotherapy has been associated with greater improvement, and faster resolution of suicidal ideation than placebo.⁹⁶

In children and adolescents, the use of pharmacological interventions to treat suicidal behaviour is even less clear. Recent research has suggested that in children and adolescents, certain classes of drugs (most notably SSRIs) may actually increase this risk.⁹⁷ The review by Barbui et al⁹⁴ found that SSRIs increase the risk of suicidal behaviour in adolescents, despite observing the opposite trend in adults. In addition, a meta-analysis assessing the effectiveness of paroxetine found a higher incidence of suicidal behaviour in young people aged between 18 and 24 years with MDD specifically, compared with adults over 24 years of age.⁹⁸ However, whilst acknowledging that there may be an increase in the risk for suicidal behaviour in the younger population when prescribed SSRIs, Bridge et al conclude that the benefits of medication may outweigh the risks in certain age and chronicity profiles.⁹⁹ Top of page

Psychological treatment – cognitive behavioural therapy (CBT)

One of the first systematic reviews to assess the effect of CBT on suicidal behaviour was conducted by Van der Sande et al,⁶⁹ who investigated randomised controlled trials involving suicide attempters. They found that CBT had a protective effect on future suicide attempts in this population, but noted the large variability in the quality of trials.

Tarrier and colleagues⁷⁰ conducted a systematic review and meta-analysis of CBT-based interventions aimed at reducing suicidal behaviour in adults and adolescents, demonstrating that these interventions effectively reduced suicidal behaviour and proximal suicide risk factors. For example, CBT was found to be more effective than minimal treatment and treatment as usual but did not demonstrate superiority to other active interventions. The authors also concluded that interventions aimed directly at reducing some aspect of suicidal behaviour, as opposed to targeting an associated risk factor (e.g. depressive symptoms), are likely to be more successful at reducing specific suicidal behaviour.

Robinson et al⁵⁶ conducted a systematic review of interventions aimed at young people with a history of suicide attempt, suicidal ideation or deliberate self-harm. One trial indicated positive findings for CBT when compared to treatment as usual, showing that CBT effectively reduced the incidence of deliberate self-harm and suicidal ideation. While the evidence base for CBT delivered to young people with suicidal behaviour is small, these results indicate that it is a promising intervention strategy.

While it is acknowledged that reductions in measures of depression and suicide-related behaviour do not translate directly to reduced suicide rates, CBT-based interventions appear promising in ameliorating a range of suicide-related behaviours (e.g. ideation, plan, attempt) in those at risk. These effects appear strongest over the immediate to short term following treatment, and to reduce over time.⁷⁰ Given that suicidal risk likely fluctuates over time, intervention programs need to consider strategies around follow up and maintenance of treatment gains over the medium to long term. It should also be noted that high-risk suicidal people are frequently excluded from research for a number of reasons,¹⁰⁰ making it difficult to draw conclusions as to the efficacy of a range of indicated interventions, including CBT, on this population.

Follow-up care after suicide attempt

The studies by Mann et al,⁵⁷ Beautrais et al⁵⁸ and Nordentoft⁶⁰ each identified follow-up care as a promising suicide intervention strategy following a suicide attempt, presenting evidence to suggest that it can reduce the number of subsequent suicide attempts and possibly the rate of completed suicide.

Follow-up strategies have the potential to keep high-risk individuals linked with services, facilitate future service access and assist with maintaining adherence to treatment regimes. Components of the follow-up strategies outlined within the reviews have not been fully identified or evaluated, however examples may include regular, brief telephone or letter contact, coordinating follow-up appointments, linking services and providing information about emergency access to mental health care. Top of page

4.2.4 Postvention

Postvention refers to a range of strategies that can be delivered following a suicide event, generally targeting individuals or groups who have been affected by, or are in close proximity to a suicide. The aim of these strategies is to reduce the distress experienced in response to a suicide and the risk of suicide contagion.

Szumilas et al⁷² reviewed the literature on suicide postvention programs, including school-based, family-focused and community-based programs. The authors examined data evaluating these programs and concluded that there was no protective effect on suicide rates or suicide attempts. However, some program components that had positive effects on knowledge, help-seeking and psychological distress were highlighted, namely gatekeeper training, counseling for survivors and the provision of outreach support at suicide sites.

Cox et al⁷¹ looked specifically at postvention strategies delivered in response to suicide clusters in young people. The results of this review indicate that few evaluations of postvention responses have been conducted, making it difficult to draw firm conclusions about the effectiveness of these strategies on the reduction of suicide risk or completed suicide. However, some promising strategy components were described, including the development of a community response plan, educational/psychological debriefings, provision of both individual and group counseling to affected peers, screening of high-risk individuals, responsible media reporting of suicide clusters and promotion of health recovery within the community to prevent further suicides. The Australian government has recently produced a set of community guidelines¹⁰¹ that provide assistance to communities who wish to develop and implement a community response plan such as those noted in this review: these represent a new initiative and are yet to be evaluated.

McDaid et al⁶⁸ conducted a systematic review of interventions for people bereaved through suicide. They found some evidence for the benefit of a four-session cognitive behavioural family intervention for first-degree relatives and their spouses compared with no intervention. At 13-month follow up, participants reported fewer maladaptive grief reactions, and less perception of being blamed for their relative's suicide.

In addition, Robinson et al⁷³ reviewed the available literature specific to school-based suicide postvention programs, again noting a considerable lack of evidence pertaining to postvention strategies. Top of page

4.3 Interventions in Indigenous populations

Each of the articles retrieved by the separate search for interventions undertaken specifically in Indigenous populations are single studies,^75,76,78 of which two have key methodological limitations.^75,78 This means they are of limited utility in terms of informing suicide prevention initiatives among Indigenous Australians.

In addition, the Closing the Gap Clearinghouse¹⁰² recently released a resource sheet summarising policies and programs aimed at reducing suicide and suicidal behavior. This information captures key learnings from work undertaken specifically in Indigenous populations as well as more generalised programs that may be suitable for adaptation into Indigenous contexts. The Clearinghouse concludes that there is evidence for 'community programs that focus on the social, emotional, cultural and spiritual underpinnings of community wellbeing' in suicide prevention, and that 'a number of effective non-Indigenous-specific programs have been shown to be culturally appropriate and acceptable to Indigenous people'.¹⁰²

It is widely accepted that there are significant differences not only between Indigenous and non-Indigenous suicides, but also between different Indigenous communities,^103,104,105 meaning that interventions that are appropriate for one community may not be for another.¹⁰⁶ This raises questions about the generalisabilty of research findings from any single study. As a result, it has been recommended that a series of 'community-specific strategies' should be developed and evaluated¹⁰³ that can respect and accommodate the unique nature of different Aboriginal groups and communities, and that employ appropriate methodological approaches. Top of page

4.4 Data availability and study limitations

A number of limitations to the current review need to be taken into account when interpreting the findings. Firstly, the current search was restricted to systematic reviews and meta-analyses, as it was beyond the project's scope to retrieve all articles pertaining to high-risk groups for suicide and suicide-related interventions. However, the systematic reviews conducted in the area of suicide prevention that are highlighted used a thorough methodology and, as such, were likely to have picked up key articles of interest.

Due to time constraints, the only database searched was Medline. However, past experience suggests that this database would have most likely included all the articles that met the inclusion criteria for the current review.

A further limitation relates to the identification of studies reporting on programs designed to reduce the risk of suicide in Indigenous populations: no systematic review articles were retrieved by our search. However, because the level of risk is high among this population and the relevance of this issue to Australia's suicide prevention program, we conducted a separate search that retrieved three articles.

Other limitations exist that are not specific to this review. One such factor is the absence of standardised definitions of the key outcomes of interest and the use of variable terminology and outcome measures across those studies included in the reviews cited here, which can make it difficult to generalise findings across studies.

Further, as has been previously reported,^107,108 there is in general a lack of intervention studies assessing suicide-related outcomes, in particular suicide itself. Although suicide is widely regarded as a significant public health problem, it is, in statistical terms, relatively rare, meaning that in order to conduct research with suicide as an outcome, very large studies with long follow-up periods are required. As a result, studies are often not adequately powered to measure suicide as an outcome and therefore frequently report on proxy outcomes such as suicidal ideation, suicide attempt and deliberate self-harm, and in some cases help-seeking behavior, identification of at-risk individuals, entry into treatment, and antidepressant prescription rate. Whilst this is not unusual in suicide research there remains a need for large, adequately powered studies that can examine the effects of interventions on rates of suicide.

The issue of sub-optimal data collection and recording processes has been the subject of much discussion in Australia and internationally.^109,110 In 2010, the Senate Inquiry into Suicide in Australia (The Hidden Toll: Suicide in Australia)⁸⁸ also raised concerns regarding the accuracy of suicide reporting in Australia and outlined a number of factors that may impede the accurate identification and recording of possible suicides, noting the consequences of any under-reporting for the understanding of risk factors and provision of services to those at risk. The Senate Inquiry recommended a program of reform designed to improve the accuracy of suicide statistics across the country and the timely dissemination of these data, in order that suicide prevention programs can be responsive to current need. The Government responded positively to these recommendations in the Commonwealth Response to The Hidden Toll: Suicide in Australia.¹¹¹