Estimating the scale of hospital admissions for people experiencing homelessness in England: a population-based multiple systems estimation study using national Hospital Episodes Statistics
Estimating the scale of hospital admissions for people experiencing homelessness in England: a population-based multiple systems estimation study using national Hospital Episodes Statistics
Background: people experiencing homelessness have substantial health needs and poor access to primary healthcare, resulting in high rates of hospital care. Housing status is not routinely recorded in English electronic health records, undermining service planning. We developed methods to estimate the scale of hospital admissions for people experiencing homelessness in England.
Methods: we analysed admissions for people experiencing homelessness using Hospital Episode Statistics for 2013/2014, 2015/2016 and 2017/2018. We applied multiple systems estimation Poisson regression methods to estimate total admissions and an inflation factor to correct for under-reporting. We calculated unadjusted admission rates per 1000 population per year and admission rate ratios compared with the housed population.
Results: we observed 34 790 admissions in 2017/2018, with total homeless admissions estimated at 176 342 (95% CI 164 031 to 188 654) (inflation factor=5.07 (95% CI 4.71 to 5.42)). The unadjusted admission rate for the 2017/2018 homeless population was 879.0 admissions per 1000 population per year (95% CI 817.7 to 940.4), 2.5 (95% CI 2.3 to 2.7) times higher than the housed population. Restricted to rough sleepers and hostel residents, the unadjusted rate was 3516.7 per 1000 (95% CI 3271.2 to 3762.2), with a rate ratio of 10.0 (95% CI 9.3 to 10.7) compared with the housed population.
Conclusions: we estimated five times as many hospital admissions for people experiencing homelessness than we observed directly. We advise caution when applying these inflation factors to other datasets because of methodological limitations in this study and sensitivities to local coding practices. In the absence of routine housing status recording, multiple systems estimation could facilitate improved service planning.
Luchenski, Serena April
3d27cd61-56ac-4562-ac90-ab366ccdd276
Böhning, Dankmar
1df635d4-e3dc-44d0-b61d-5fd11f6434e1
Aldridge, Robert
b6feac76-eb23-4ff6-ade6-563532adb617
Stevenson, Fiona
b031213f-8fd7-4871-8bf6-d2ab964f52ef
Tariq, Shema
dc761cdf-e36c-4e3b-b3a9-d8192c03b40a
Hayward, Andrew C.
e69c7f88-9d85-4e2c-bd76-23fd44fbbee5
Luchenski, Serena April
3d27cd61-56ac-4562-ac90-ab366ccdd276
Böhning, Dankmar
1df635d4-e3dc-44d0-b61d-5fd11f6434e1
Aldridge, Robert
b6feac76-eb23-4ff6-ade6-563532adb617
Stevenson, Fiona
b031213f-8fd7-4871-8bf6-d2ab964f52ef
Tariq, Shema
dc761cdf-e36c-4e3b-b3a9-d8192c03b40a
Hayward, Andrew C.
e69c7f88-9d85-4e2c-bd76-23fd44fbbee5
Luchenski, Serena April, Böhning, Dankmar, Aldridge, Robert, Stevenson, Fiona, Tariq, Shema and Hayward, Andrew C.
(2025)
Estimating the scale of hospital admissions for people experiencing homelessness in England: a population-based multiple systems estimation study using national Hospital Episodes Statistics.
BMJ Public Health, 3 (2), [e002978].
(doi:10.1136/bmjph-2025-002978).
Abstract
Background: people experiencing homelessness have substantial health needs and poor access to primary healthcare, resulting in high rates of hospital care. Housing status is not routinely recorded in English electronic health records, undermining service planning. We developed methods to estimate the scale of hospital admissions for people experiencing homelessness in England.
Methods: we analysed admissions for people experiencing homelessness using Hospital Episode Statistics for 2013/2014, 2015/2016 and 2017/2018. We applied multiple systems estimation Poisson regression methods to estimate total admissions and an inflation factor to correct for under-reporting. We calculated unadjusted admission rates per 1000 population per year and admission rate ratios compared with the housed population.
Results: we observed 34 790 admissions in 2017/2018, with total homeless admissions estimated at 176 342 (95% CI 164 031 to 188 654) (inflation factor=5.07 (95% CI 4.71 to 5.42)). The unadjusted admission rate for the 2017/2018 homeless population was 879.0 admissions per 1000 population per year (95% CI 817.7 to 940.4), 2.5 (95% CI 2.3 to 2.7) times higher than the housed population. Restricted to rough sleepers and hostel residents, the unadjusted rate was 3516.7 per 1000 (95% CI 3271.2 to 3762.2), with a rate ratio of 10.0 (95% CI 9.3 to 10.7) compared with the housed population.
Conclusions: we estimated five times as many hospital admissions for people experiencing homelessness than we observed directly. We advise caution when applying these inflation factors to other datasets because of methodological limitations in this study and sensitivities to local coding practices. In the absence of routine housing status recording, multiple systems estimation could facilitate improved service planning.
Text
Manuscript Luchenski et al. REVISED TRACKED CHANGES 8.9.25
- Accepted Manuscript
Text
e002978.full
- Version of Record
More information
Accepted/In Press date: 24 September 2025
e-pub ahead of print date: 28 October 2025
Identifiers
Local EPrints ID: 506338
URI: http://eprints.soton.ac.uk/id/eprint/506338
PURE UUID: 7a724127-aad1-480d-8ba0-d61949d9c829
Catalogue record
Date deposited: 04 Nov 2025 18:05
Last modified: 05 Nov 2025 02:44
Export record
Altmetrics
Contributors
Author:
Serena April Luchenski
Author:
Robert Aldridge
Author:
Fiona Stevenson
Author:
Shema Tariq
Author:
Andrew C. Hayward
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics