World Homeless Day: Preventing homeless deaths through harm reduction

Across the UK, people experiencing homelessness continue to die decades younger than the general population, often from preventable causes linked to poverty, trauma, substance use, poor health, and exclusion from support. The scale of these deaths should concern us all.

For the past year, my Churchill Fellowship has focused on a question that sits at the heart of this challenge: how can harm reduction approaches help prevent homeless deaths? Through field research in Canada, where progressive peer-led harm reduction approaches have been in place for over a decade, I have explored responses to the toxic drug crisis for people who experience long-term homelessness and drug dependency disorder.

The experience has reinforced something I’ve learned through more than twenty years working in homelessness and safeguarding: when people die while homeless, they have often suffered years of preventable harm beforehand. Hundreds of deaths occur every year in the UK where disadvantage, inequality, and exclusion overlap, and where professional systems struggle to respond in a coordinated, sensitive, and proactive way. Housing, health, adult social care, substance use services, safeguarding partnerships, and community organisations all play a role, yet too often they operate separately, and even at odds with each other.

One of the most striking lessons from my time in Canada earlier this year was the commitment to meeting people where they are. Harm reduction services recognise that people do not need to be abstinent, engaged with treatment, or free from complexity before they deserve support. Whether through street-based healthcare, overdose prevention initiatives or integrated outreach, the focus is on keeping people alive, reducing harm, and building trusting relationships. These approaches are grounded in evidence, but also in something very simple: the belief that every life has value.

This has important implications for safeguarding – the professional phrase for intervention in abuse and neglect impacting people at risk. Too often, adult safeguarding in England is triggered after a crisis has happened, with thresholds for help so high that people can be victimised and at risk for long periods before professionals respond. For people experiencing homelessness, a safeguarding approach that focuses solely on responding to incidents can miss opportunities to prevent harm much earlier.

We know from research, Safeguarding Adult Reviews and lived experience campaigning that stigma, service exclusion, and fragmented systems all exacerbate risk and make harm less visible. If we are serious about preventing deaths, adult safeguarding must become a shared, preventative responsibility that meets people where they are, including if they wish to continue to use drugs.

My work as a Safeguarding Adult Review author and with the Dying Homeless Project continually highlights the human cost of failing to act. In 2025, the Dying Homeless Project recorded 1,720 people who died while homeless in the UK, a figure that has increased every year since the project began. Every year, around 40% of these deaths are drug-related – far higher than in the general population. This points to missed opportunities to ensure that people experiencing homelessness have easy access to support that reduces harm, increases safety, and champions dignity. Importantly, these missed opportunities span the ‘whole system’, from the support provided by street-based outreach teams to local authority governance practices around homelessness and drug and alcohol support.

What would a whole-system harm reduction approach look like? It would mean frontline homelessness practitioners being trained and supported to offer harm reduction. It would mean housing and health services working together, with individuals, to explore and respond to escalating risks in sensitive and person-led ways.

In a whole-system response, social work teams and Safeguarding Adult Boards would take a more preventative approach to financial abuse, exploitation, and self-neglect, working with people from an earlier stage recognising the cumulative impacts of homelessness on wellbeing, self-worth, and access to safety. Importantly, it would involve embedding people with lived experience in decision-making about how drug and alcohol funding is spent, how local services operate, and how public health messages around drug-related harms are developed.

Ultimately, preventing homeless deaths requires collective leadership and accountability, within which challenging stigma should be a clear focus and priority, including in national law and policy. A whole-system approach to preventing drug-related homeless deaths would be bold, accessible, and unafraid to champion people who are so often blamed for their own suffering.

As my Fellowship continues, I hope to build on the learning from Canada and contribute to the growing conversation about how we can better safeguard people experiencing homelessness and drug dependency. It has been a pleasure to begin discussions with central government colleagues and pioneering grassroots organisations about how to transform my learning into useful tools for local and national action. The solutions already exist, and they provide evidence of the crucial role of overdose prevention support, peer support, and low-barrier rapid access to a wider range of help. The challenge in the UK is creating the conditions for this to flourish, through bold policy that draws pragmatically on evidence, not on archaic attitudes that push people further away from safety.

Following World Homeless Day, I want to honour the 1,720 people who died while homeless in the UK last year and make a call to action to prevent future deaths through a commitment to harm reduction. Doing this requires building whole-system approaches, with people with lived experience of homelessness and drug dependency at their heart.

As someone said to me during my trip to Canada, “It’s not radical to keep people alive”. I continue this work through my Fellowship with this in mind.