Do housing interventions improve health outcomes?

Hopefully you are already screaming at your screen about the manifest ignorance of this blog’s title. You’ll perhaps be thinking of the appalling health outcomes experienced by those who are homeless. Or maybe taking a theoretical stance and remembering how the need for shelter forms the very base of Maslow’s hierarchy of human need.

For some aspects of housing policy we need no evidence. It is simply about, to paraphrase Mahatma Ghandi, how we as a civilised society treat our most vulnerable members. The fact that we still have too many people who are homeless or living in unhealthy housing shows we still have some way to go on this journey.

However if we are to persuade those from outside the housing sector to invest in housing interventions we will require evidence. The NHS has made some welcome initial forays into housing interventions, but the potential to do more is huge.

Of course we know some things already. I was invited to write this blog having tweeted a link to research (opens new window) reporting an impressive 39% reduction in hospital admissions from those living in homes which had been upgraded to meet national housing quality standards.

Like all research, this study had some methodological drawbacks. But taken with previous studies and systematic reviews it becomes clear that housing interventions such as improving thermal properties and increasing safety will improve health outcomes.

Whilst I am sure that there is sufficient evidence to justify action, I am also struck by the relative lack of evidence compared to many areas of clinical practice. This needs to change.

Firstly every housing intervention designed to improve health outcomes needs monitoring, and novel interventions need to be robustly evaluated. I’ve worked previously with funders who expected 10% of a project budget to be spent on monitoring and evaluation. This might sound a lot, but it can save you wasting 90% of your funds on things that aren’t working.

Secondly funders of health research need to direct resources towards housing interventions commensurate with the potential to improve population health. If a 39% reduction in hospital admissions is anywhere near being true then the level of investment in building the evidence base should be significant. Less than 5% of the UK’s research budget is spent on preventative interventions, and the proportion on social interventions such as housing will be tiny.

Finally more could be done to translate research into practice for health commissioners. This is a major expectation for clinical research (making clinicians change their practice as informed by current research) and should be no less so for housing interventions.

The renewed national Memorandum of Understanding on Improving Health and Care through the Home provides a great platform for collaboration and itself has stimulated new research and the dissemination of evidence.

So do housing interventions improve health outcomes? Once the basic housing requirements of a civilised society have been met, there is evidence that additional interventions can help maintain or improve our health. But this should not be a matter of faith. The recent DEMOS report on The Social Value of Sheltered Housing (opens new window) made clear the potential scale of the impact of housing interventions. They observed that residents in sheltered housing by definition have greater morbidity and are more likely to be admitted to hospital, but have a much shorter length of stay than the general population aged 75+ due to the social and housing support they receive. This probably saves health and care services £300m a year. Taken with other interventions such as falls prevention they estimated the saving to be nearer £486m annually.

Stories are important, but change at scale will require figures such as these to persuade health commissioners. To quote W. Edwards Deming  “In God we trust; all others bring data.”

(these views are my own and not necessarily those of my employer)


Add your comment

Leave this field empty