In Part 1, I looked at individual cases and personal stories. I wanted to give space to the human beings behind discussions about suicide, mental health and work, because statistics can never tell us everything about a person’s life, their relationships, what was happening around them, or what they were trying to communicate.
In this second part, I want to look at the research.
I’ve selected six pieces of research that approach suicide prevention from different but connected directions: work-related suicide and psychosocial working conditions; the Australian suicide-prevention system; lived experience and participation; disclosure of mental health conditions in workplaces; and what can be learned from coronial investigations into suicide.
I don’t think any one study gives us all the answers. Nor should research into suicide ever be reduced to a simple explanation of why one individual person died.
What these papers do, however, is broaden the questions we’re prepared to ask.
They ask us to look beyond the individual.
They ask us to look at work.
They ask us to look at psychosocial hazards.
They ask us to look at organisational responses.
They ask us to listen to lived experience.
And they ask whether systems intended to prevent harm are learning enough from the harm that’s already occurred.
——
1. Work-related suicide: looking at working conditions
Anthony LaMontagne and Tania King’s 2023 Work-related Suicide: A Discussion Paper, prepared for Suicide Prevention Australia, examines suicide from an occupational health and safety perspective.
The authors define work-related suicide as a death by suicide that’s wholly or partly related to, and caused by, work or working conditions.
That’s important because it shifts attention from asking only what was happening within an individual to also asking what was happening around that person at work.
The paper considers both individual case-based evidence and population-level epidemiological evidence. It identifies a number of potential work-related contributors, including access to means through employment, exposure to trauma, adverse chemical, physical or psychosocial working conditions, workplace environments where stigma discourages help-seeking, cultures that promote an extreme orientation towards work, and adverse experiences arising from work-related injury or illness.
The research concerning psychosocial working conditions is particularly significant.
LaMontagne and King examine evidence concerning job strain, job insecurity, low job control, poor workplace social support, workplace sexual harassment and workplace bullying. They conclude that evidence linking psychosocial job stressors with suicide has strengthened considerably.
They also raise the possibility that work-relatedness has historically been under-identified in individual investigations. Among the problems they identify are inconsistent collection of information about work circumstances and the possibility that medical explanations may receive greater attention while psychosocial factors are understated.
This is exactly the message I’ve tried to communicate to regulators, politicians and society, for years. This is why I’ve chosen to tell my story. What seems logical and common sense to me, is frustratingly hard to make those in the system open their eyes and see it too. What receives “greater attention” may also be seriously inaccurate, more harmful, discriminatory and can destroy the little agency a person has left. When systemic structures exclude the individual voice, it’s not only a human rights issue, it can cause the very risks it’s meant to prevent.
Case-based estimates considered in the paper generally place work-related suicide at approximately 10–15 per cent of suicides in the working population. The authors argue, however, that a fuller epidemiological accounting of multiple chronic workplace stressors could identify a larger preventable burden.
Perhaps the most important message for me is what they say about prevention.
The authors argue that workplace suicide-prevention practice has tended to concentrate on interventions directed towards individuals while paying insufficient attention to changing the working conditions that may increase suicide risk.
In other words, helping a distressed worker is important.
But so is identifying and controlling the hazard.
That is occupational health and safety.
——
2. Work-related suicide: the evidence continues to develop
LaMontagne and colleagues returned to this subject in 2024 in Work-related suicide: evolving understandings of etiology & intervention.
This later paper develops the occupational health and safety framework further and identifies six broad categories of potential work-related causes: workplace chemical, physical and psychosocial exposures; exposure to trauma through work; access to means through work; high-stigma working environments; environments that encourage an extreme orientation towards work; and adverse experiences resulting from work-related injury or illness.
They acknowledge that causal conclusions drawn largely from observational research will always contain uncertainty. But uncertainty doesn’t necessarily justify inaction where the potential outcome is grave and preventable.
That principle matters in occupational health and safety generally. We shouldn’t have to wait until every possible causal question has been resolved beyond doubt before considering whether an identifiable workplace hazard should be reduced.
The authors also acknowledge something equally important: work itself can be protective.
Good work can provide income, purpose, belonging, structure and meaning.
That matters enormously.
The argument isn’t that work is inherently harmful.
It’s that working conditions matter.
Safe, supportive and meaningful work can contribute positively to wellbeing. Harmful psychosocial working conditions can do the opposite.
Again, the authors conclude that prevailing workplace suicide prevention disproportionately emphasises interventions aimed at individuals and illness, while paying too little attention to working conditions that may increase suicide risk.
This isn’t simply a call for greater awareness.
It’s a call for prevention to move upstream.
——
3. What does Australia’s suicide-prevention system need?
Bridget Bassilios and colleagues approached the issue at a much broader policy level in their 2024 study, Key informant perspectives of suicide prevention in Australia.
The researchers interviewed 24 key informants drawn from government departments and agencies, peak bodies and suicide-prevention leadership, researchers and people with lived experience.
They asked what Australia was doing well, what remained problematic, and where the opportunities for improvement lay.
There were acknowledged successes: leadership and funding, growing community and political awareness, greater recognition of collective lived-experience voices, and movement towards whole-of-government and whole-of-system approaches.
But significant challenges remained.
Participants identified limitations within the service system, workforce issues, problems in defining and building the suicide-prevention sector, and the continuing need to strengthen the evidence base.
The opportunities they identified are just as important.
They included moving towards wellness rather than crisis-driven models of care, incorporating lived experience and co-design throughout policy planning, service development and evaluation, improving collaboration, and investing properly in data, research and evaluation.
I think the distinction between prevention and crisis response is critical.
If our systems only respond when somebody has reached crisis, we have to ask what opportunities existed before that point.
- What could have been changed?
- What harm could have been prevented?
- What social, economic, workplace or institutional circumstances were contributing?
- Who already knew?
- Who had been told?
- What could reasonably have been done differently?
A whole-of-government approach to suicide prevention should allow those questions to be asked.
——
4. Lived experience must have a genuine voice
Dreier and colleagues’ 2021 paper, Development of an online suicide prevention program involving people with lived experience: ideas and challenges, provides a very different but deeply important perspective.
The researchers describe the development of 8 Lives, an online suicide-prevention program developed collaboratively by researchers, clinicians and people with lived experience of suicidality, suicide attempts and suicide bereavement.
The people with lived experience weren’t simply consulted after the program had already been designed.
They helped shape it.
They participated in decisions about its concept, structure, content and design. They reviewed written material. Some shared their experiences through video or written accounts. Their feedback was incorporated into the program itself.
That’s genuine participation.
And there’s something in this research that I find especially important.
The authors distinguish between a duty of care based upon actual risk and inappropriate overprotection.
They identify transparency, autonomy and clear structures as important when working with people with lived experience of suicide.
Autonomy matters.
A person doesn’t cease to have a voice because they are distressed.
They don’t cease to understand their own experience.
They don’t cease to have dignity.
The lived-experience participants themselves influenced how disclosure was approached. The program didn’t push people to disclose. Decisions about what to reveal remained with the individual, and participants contributing videos retained significant control over their own material.
There’s a lesson in that which reaches well beyond the development of an online program.
If we say that lived experience matters, then we have to actually listen to it.
Not reinterpret it until it becomes convenient.
Not speak over it.
Not strip the person of autonomy.
Listen.
——
5. What happens when an employee actually discloses?
The 2025 Australian study by Ivan Volkov, Dianne Sheppard, Andrea Kirk-Brown and Pieter Van Dijk, Workplace Outcomes After Mental Health Disclosure: The Critical Influence of Supervisor Reactions, brings this discussion directly into the workplace.
It examines what happens when workers do something they are so frequently encouraged to do: they tell somebody that they’re experiencing a mental health condition.
The researchers analysed the experiences of Australian employees who had disclosed a mental health condition to their immediate supervisor.
Supervisor responses were grouped as positive or supportive, negative or unsupportive, or ambivalent.
Positive responses included empathy, attentive listening, concern, validation, information and tangible assistance such as workplace adjustments.
Negative responses included judgemental or condescending behaviour, minimising or denying symptoms, avoidance, refusing assistance and subsequent negative treatment.
And the response mattered.
Workers who experienced a positive supervisor reaction reported an average work-ability score of 7.53 out of 10, compared with 5.52 among workers experiencing a negative reaction.
Among workers receiving a positive response, 85.7 per cent intended to remain with their organisation over the following 12 months. Among those receiving a negative response, only 36.7 per cent did.
Workers experiencing positive reactions were also considerably more likely to feel comfortable talking openly about their mental health at work.
The researchers found that some negative disclosure experiences involved stigma, discrimination and denial of assistance and, in the most severe cases, termination of employment.
The authors are appropriately cautious about causation because the study is cross-sectional.
But their conclusion is still powerful: it’s not simply the decision to disclose that matters. How the disclosure is experienced matters.
That changes the responsibility contained in our mental-health messaging.
We cannot continually encourage workers to speak without also preparing workplaces to hear them.
Disclosure cannot be treated as the end point.
It’s the beginning of a response.
——
6. What do we learn after somebody has died?
The final paper brings these issues into the coronial jurisdiction.
Jennifer Schulz (formerly Moore), Kate Diesfeld and Christine Forster examine suicide within the Australian and Aotearoa New Zealand legal professions and ask what lessons can be distilled from coronial investigations.
Their paper examines five coronial cases involving members of the legal profession and judiciary.
What’s particularly significant is the attention the authors give to structural influences.
The cases documented work-related experiences including overwhelming workloads, exposure to distressing material, vicarious trauma and other occupational pressures. The authors argue for greater attention to the broader workplace context rather than conceptualising wellbeing primarily through individual ill-health.
And then there’s the question of prevention.
Coroners don’t investigate deaths merely to reconstruct what happened. Coronial systems also have an important public-health and preventive function.
Yet none of the coroners in the five cases examined made statutory preventive recommendations.
In several cases, the coroners considered changes already undertaken by the workplaces sufficient. In another, the coroner considered it difficult to identify particular actions that would have prevented the death.
Schulz and colleagues argue that there are strong reasons for coroners to consider evidence-based recommendations, particularly given the public-health purpose of the jurisdiction.
They also make a profoundly human point.
Research concerning bereaved families has found that recommendations can matter to families because they can give meaning to what happened through the possibility that lessons from one death may help prevent another.
That’s prevention after tragedy.
It cannot undo a death.
But it can refuse to learn nothing from it.
——
What these six pieces of research tell us together
These six papers don’t say exactly the same thing. They shouldn’t be made to.
They examine different populations, different questions and different levels of prevention.
But when I place them beside one another, I see a consistent movement away from an exclusively individualised understanding of suicide prevention.
Look at the environment.
Look at working conditions.
Look at psychosocial hazards.
Look at what happens when someone discloses distress.
Look at whether the person is actually heard.
Look at the systems surrounding them.
Listen to lived experience.
And when somebody dies, learn.
This doesn’t mean individual support is unimportant. Of course it’s important.
Nor does it mean that any individual suicide can simply be attributed to one event, one workplace or one cause. Suicide is complex, and these papers themselves reflect that complexity.
But complexity cannot become an excuse for looking away from preventable contributing conditions.
That, for me, is one of the strongest messages in this research.
There’s a difference between responding to distress and preventing the conditions that contribute to it.
There’s a difference between encouraging somebody to speak and creating an environment in which speaking is safe.
There’s a difference between saying that lived experience matters and actually allowing people with lived experience to influence decisions.
And there’s a difference between investigating what happened and using what was learned to prevent it happening again.
When we talk about suicide prevention, we need the individual supports.
But we also need to look upstream.
We need safe workplaces.
We need psychosocial hazards to be taken seriously.
We need managers and organisations capable of responding appropriately when someone tells them they are struggling.
We need policy informed by evidence and genuine lived experience.
We need regulators and other institutions prepared to examine systems and structures.
And we need to be willing to learn from people who have died, from the people who loved them, and from people who survived circumstances that could have ended very differently.
Prevention cannot only mean asking a person in distress to do something.
Prevention also means asking what we - workplaces, institutions, governments, regulators, professionals and communities - need to do differently around them.
That’s what the research tells us.
——
References
Bassilios, B., Dunt, D., Krysinska, K. et al. (2024). ‘Key informant perspectives of suicide prevention in Australia’. BMC Public Health 24, 3449. https://doi.org/10.1186/s12889-024-20943-6
Dreier, M., Baumgardt, J., Bock, T. et al. (2021). ‘Development of an online suicide prevention program involving people with lived experience: ideas and challenges’. Research Involvement and Engagement. 7, 60 https://doi.org/10.1186/s40900-021-00307-9
LaMontagne, A.D., & King, T.L. (2023) Work-related suicide: a discussion paper. A report prepared for Suicide Prevention Australia. https://www.suicidepreventionaust.org/research-grants/research-outcomes/
LaMontagne, A.D, Åberg, M., Blomqvist, S. et al. (2024). ‘Work-related suicide: evolving understandings of etiology & intervention’. American Journal Industrial Medicine. 67(8) 679-695. https://doi.org/10.1002/ajim.23624
Schulz (formerly Moore), J., Diesfeld, K., & Forster, C. (2026). ‘Suicide in the Australian and Aotearoa New Zealand legal professions: distilling lessons from the coronial jurisdiction’. Psychiatry, Psychology and Law, 33(4), 1122–1138. https://doi.org/10.1080/13218719.2025.2497786
Volkov, I., Sheppard, D.M., Kirk-Brown, A. et al. (2025). ‘Workplace Outcomes After Mental Health Disclosure: The Critical Influence of Supervisor Reactions’. Journal of Occupational Rehabilitation. https://doi.org/10.1007/s10926-025-10352-w
Further research
Dr Karl Andriessen is a suicide-prevention researcher whose work includes suicide prevention, bereavement following suicide and other causes of death, and help-seeking. His UNSW research profile provides an extensive list of his publications and continuing research in these areas.
Dr Karl Andriessen — UNSW research profile and publications
https://www.unsw.edu.au/staff/karl-andriessen
