BackgroundMarginalised groups are not only excluded from health services but also from routine health statistics. A key challenge for public health is measurement of the health of marginalised groups to make their needs visible and focus prevention and treatment activities. Internationally, there has been much research into the health inequalities slope, whereby almost all important health problems gradually rise in frequency with increasing social deprivation. Geographical measures of deprivation based on people's postcodes are often used to assess the effect of inequalities on health (eg, dividing the population into deprivation quintiles on the basis of indices of multiple deprivation [IMD-5]). However, this notion of inequality does not adequately capture the experience of many marginalised groups who have more extreme deprivation. Furthermore, marginalised groups are often deemed too difficult to engage to allow collection of meaningful health data. In this study, we aim to compare the prevalence of chronic disease across population quintiles of social deprivation (housed population) and in homeless people; and develop a generalisable methodology for health surveys in marginalised groups by use of peer interviewers to achieve engagement and high uptake. MethodsWe undertook a cross-sectional health survey of homeless people attending 27 low-threshold services in London (eg, hostels, day centres, soup runs). Questions asked were directly comparable with key questions within the national Health Survey for England (HSE) and health-related quality of life (EQ-5D), allowing direct comparison across quintiles of deprivation in the housed population and in homeless people. Peers with personal experience of homelessness were used to engage participants and undertake the surveys to achieve high uptake and provide as representative a sample as possible. Data were merged with HSE 2010 data on a representative housed population. We calculated age-adjusted and sex-adjusted odds ratios (ORs) of chronic disease, multiple morbidities, and quality of life scores as measured by the EQ-5D to compare homeless participants with housed people living in areas in the poorest quintile of the IMD-5. Findings455 (77%) of 592 people responded. Reasons for non-response were not obtained, but this is a high response rate for a population survey. Most respondents were men (365 of 455 [80%]) and most were aged 16–44 years (261 of 455 [57%]) and were born in the UK (277 of 455 [61%]). Nearly three-quarters were registered with a local general practice (329 of 455 [72%]). Missing data was less than 1% across all variables. There was a gradual increase in risk of chronic disease across the quintiles of multiple deprivation in HSE. However, compared with housed people living in the most socially deprived areas, homeless people had higher risks of asthma, heart disease, stroke, and epilepsy. The adjusted OR for asthma was 2·46 (95% CI 1·78–3·39; p<0·0001), for heart disease was 5·87 (3·63–9·49; p<0·0001); for stroke was 4·93 (2·15–11·28; p<0·0001), and for epilepsy was 12·40 (5·10–30·16; p<0·0001). Conversely, the likelihood of diagnosed diabetes was lower in the homeless than in those living in deprived areas (adjusted OR 0·59, 95% CI 0·35–0·99; p=0·046). The adjusted OR for having two or more of these disorders was 5·12 (3·06–8·57; p<0·0001). Quality of life measures showed similar disparities. InterpretationOur data show that in comparison with the slope in health inequalities, the health experience of the homeless is more akin to a cliff, with homeless people experiencing a significantly disproportionate burden of morbidity. This method could act as a template for surveys in other marginalised groups. FundingThis study was possible as a result of funding from Knowledge Into Action (Registered Charity No. 1123566), with data management and analytical support provided through a National Institute for Health Research Programme Grant For Applied Research (PGfAR RP-PG-0407-10340). The study funders had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Read full abstract