Preventing premature deaths among people who are homeless

Published: 6 March 2019

In this blog, Dr Emily Tweed discusses how we can use evidence to help us prevent premature deaths among people who are homeless.

Published on 6th March 2019

Dr Emily Tweed is a Clinical Lecturer in Public Health and CSO Clinical Academic Fellow, based at the MRC/CSO Social and Public Health Sciences Unit. Her work, funded by the Scottish Government’s Chief Scientist Office, is examining the feasibility of using record linkage between administrative datasets to understand the health of people with overlapping experiences of homelessness, imprisonment, substance use, and serious mental illness.

Last year, in the days leading up to Christmas, there was a news story that might have given us pause as we hurried between last-minute shopping expeditions, Christmas parties, and school nativity plays – or might have gone unnoticed amid our hectic preparations. On 20th December, the Office for National Statistics (ONS) published its first official estimate of deaths among homeless people in England and Wales, and the figures made for grim reading: an estimated 597 homeless people died during 2017, a 24% increase since 2013.

These figures were just the latest in a series of attempts to quantify the health disadvantages of homelessness. Since April 2018, the Bureau of Investigative Journalism – who collaborated with ONS on the trial of official statistics – had been compiling a database of people who die homeless in the UK, using reports from the public, charities, and local journalists. Their dataset is marked by a unique regard for each person’s individual story, often including heart-wrenching details of their personalities, hobbies, and the circumstances which led to them losing their homes.

In Scotland, summer saw the publication of a long-awaited study of healthcare usage and mortality among homeless people, led by analysts from the Scottish Government and National Records of Scotland. This study took a different tack from ONS and the investigative journalists: instead of starting with data on deaths and working backwards in time, it used local authority data on people applying for homelessness support and then, using data linkage, followed people up over time using their healthcare records. It found that, compared to people living in the most socioeconomically deprived areas of Scotland, people who’d been homeless at some point between 2002 and 2016 were around twice as likely to die. This suggests that homelessness is associated with a substantially increased risk of premature death over and above the effects of poverty, a finding substantiated by other studies.

The challenge now is to tackle this tragic excess through effective action. This was the message that came out loud and clear from a recent workshop convened by the Faculty for Homelessness and Inclusion Health in Scotland to discuss preventing deaths among homeless people, including perspectives from the third sector; primary care; health & social care partnerships; and academia.  Achieving change will also be central to the agenda of a roundtable held tomorrow by the Bureau of Investigative Journalism to mark a year since the launch of their ‘Dying Homeless’ campaign.

So what can be done?

Looking at these data, I see another piece in the jigsaw of evidence demonstrating that our life circumstances profoundly shape our chances of good health: the high risks of poor health and premature death associated with homelessness are an extreme manifestation of how social inequalities give rise to health inequalities.

Looking at the specific conditions named as the cause of death in these recent studies, we know a lot about what works: there are numerous systematic reviews examining the treatment options for opioid dependence, liver cirrhosis, or heart disease. But we know that adversity and disadvantage mean that people are much less likely to be able to access and benefit from forms of prevention and treatment we know are effective.

Moreover, the medically-defined cause written on the death certificate only takes us some way towards understanding the complex constellation of circumstances that led up to each tragic event. Decades of evidence from ‘lifecourse’ research tells us that our health is shaped by multiple influences across our entire lives, from conception onwards. This is very much a theme of the forthcoming Hard Edges Scotland project, which shows that homelessness, offending, substance use, mental ill-health and domestic violence are commonly preceded by adverse experiences in early life, disrupted schooling, exclusion from the labour market, and other disadvantages (watch this space for the final report). If we’re to address this excess of deaths, we have to start much earlier than you might think.

Tackling the causes of the causes

We need to look to the “causes of the causes”: those wider social, economic, and political factors which shape our lifecourse. Most of the research to date about how to improve the health of people experiencing homelessness or other forms of social exclusion has focused on actions at the individual level: for instance, enhancing people’s access to healthcare services. Yet the biggest wins are ‘upstream’ – actions which would prevent homelessness from occurring in the first place, ensure it is short-lived, or minimise any adverse effects. As the Centre for Homelessness Impact’s Effectiveness Map shows, there is a real lack of this kind of evidence in relation to health. 

This kind of evidence is difficult to get: it’s much easier to evaluate the impact of a case management worker than national legislation on homelessness entitlements, for instance. However, there is hope on the horizon. Large-scale datasets like the HL1 local authority data used in the Scottish Government project; those we are collating in Glasgow City on homelessness, substance use, and offending; and potentially the new national rough sleeping data system commissioned by Scottish Government, may help support rigorous evaluations at lower cost in future. In Wales and England, new developments in data are also promising and the ONS have made a commitment to build on their recent work, including with the release of local level data last week.

We also need to think broadly about what constitutes data, and who we should collaborate with, taking an inter-disciplinary approach that draws on a range of methods and ways of knowing. Although the focus of my work is quantitative (looking at numbers and risks), if we are to tackle the problem of homelessness and its impact on health, understanding people’s stories, the processes of services, and the workings of policy are just as important. The better use of evidence, in all its forms, is essential if we want those overall numbers – and the lives they represent – to change.

Disclaimer: The views expressed in this blog are those of the author.


First published: 6 March 2019