Health and homelessness in Scotland: research

Chapter 1: Introduction

Building upon analysis undertaken in the United States of America ^[1] and by Fife Council and NHS Fife in 2015 ^[2] , this reports links health and homelessness data for the first time at a national level in Scotland. This research combines Scottish Government homelessness data ( HL1) with six health datasets from NHS National Service Scotland covering Accident and Emergency attendances (A&E2), Inpatient admissions ( SMR01), Outpatient appointments ( SMR00), Prescriptions ( PIS), the Scottish Drugs Misuse Database ( SDMD) and Mental Health admissions ( SMR04), together with information about deaths from National Records of Scotland.

1.1 Background

Health inequalities across Scotland are well known. Many sources provide evidence that individuals in more deprived areas have worse health outcomes compared with those from less deprived areas:

• The Long-term Monitoring of Health Inequalities ^[3] show that people born in the 10% most deprived areas in Scotland have considerably shorter healthy life expectancies than those born in the least deprived areas (26.0 years lower for males, 22.2 lower for females). This difference has been stable for several years. Furthermore, premature mortality rates were 3.7 times higher for people in the most deprived areas.
• Individuals living in the most deprived areas account for twice as many attendances to Emergency Departments (A&E) as those in the least deprived areas. The likelihood of being admitted following an Emergency Department attendance also increases as deprivation increases. The difference in attendances could be for a number of reasons including poorer health, more complex social needs and service provision in areas of higher deprivation ^[4] .
• Another study suggests that patients in the most deprived areas had an overall prescription rate for antibiotics that is 36.5% higher than those in the least deprived areas ^[5] .
• Patients from more deprived areas in Scotland were more likely to experience a general acute stay related to drug misuse. The highest rates were observed among those in the most deprived areas ^[6] .

Homelessness in Scotland, is also well documented. Under the Housing (Scotland) Act 1987, people can apply to their Local Authority ( LA) for assistance if they are homeless:

• In 2016/17, LAs received just over 34,000 applications for assistance. Of these, LAs made 28,000 assessments where the household was assessed as either homelessness or likely to become homeless within two months (threatened with homelessness) ^[7] .
• The number of assessments where the decision was either homeless or threatened with homeless in Scotland has fallen from a peak of 44,000 assessments in 2009/10 to 28,000 in 2015/16. This reduction is mainly due to the impact of the introduction of Housing Options services in Scottish local authorities which have a focused on homelessness prevention.
• At the end of March 2017, approximately 11,000 homeless households were in temporary accommodation within Scotland.

People assessed as homeless are likely to be among the most deprived in Scotland. As people from more deprived areas are known to have poorer health outcomes, and there are many homelessness applications per year, it is important to understand the impact homelessness and health have on each other.

Health outcomes and homelessness are known to be related. Many studies have been conducted looking at the specific health issues of people experiencing homelessness:

• Homeless people are among the most vulnerable and socially excluded in our society and often find it difficult to access the help they need ^[8] .
• Many homeless people present to health services with multiple morbidity including drug or alcohol dependence, mental health and physical problems such as tuberculosis and breathing difficulties (Department of Health 2010) ^[9] .
• Homeless people have higher rates of premature mortality than the rest of the population, especially from suicide and unintentional injuries, and an increased prevalence of a range of infectious diseases, mental disorders, and substance misuse. Although engagement with health services and adherence to treatments is often compromised, homeless people typically attend the emergency department more often than non-homeless people ^[10] .
• Other studies has shown that homeless populations, individuals with substance use disorders, sex workers, and imprisoned individuals experience extreme health inequities across a wide range of health conditions, with the relative effect of exclusion being greater in female individuals than in male individuals. ^[11]

As deprivation is known to have an adverse impact on health outcomes, poor health outcomes related to homelessness may be the result of the many factors associated with deprivation.

A 2008 study by Dr. David Morrison ^[12] showed that homelessness itself is an independent risk factor for certain health outcomes. The findings include:

• The health of homeless people in Glasgow, measured through hospital care and deaths, was consistently poorer than that of the most deprived non-homeless local populations.
• This could be partly explained by poorer health at the point of becoming homeless but an estimate was also made of the additional hazard of homelessness itself.
• Homelessness is an independent risk factor for deaths from specific causes.

This implies that homelessness influences health outcomes in addition to deprivation.

1.2 Study research questions

This study aims to expand on these findings and examine the relationship between health and homelessness for the first time at a national level in Scotland. The study will adopt a similar methodology to that used by Dr. Morrison, and match people from Local Authority homelessness applications ( HL1) to the non-homeless Scottish population across on age and sex. Furthermore, the study will create two controls groups, one from people living in the 20% least deprived areas, and one from people living in the 20% most deprived. In this way, the study can better understand the relationship between health outcomes and homelessness, and their relationships with deprivation.

During the analysis, the following research questions arose to describe the relationship between health and homelessness:

• How does health prior to the first homelessness assessment influence homelessness?
• Does the point at which someone becomes homeless have an impact on one's health? Is a crisis with a health component involved?
• How does homelessness influence health?
• Is there a relationship between health, homelessness and deprivation?

In order to answer these questions, the study tests the following null hypotheses:

• Prior to becoming homeless, all three cohorts (the homeless cohort, the 20% most deprived cohort and the 20% least deprived cohort) use health services the same.
• At the point of homelessness, all three cohorts use health services the same.
• Following being assessed as homeless, all three cohorts use health services the same

These null hypotheses – which we are seeking evidence to reject - assume that use of health services does not change with either homelessness or area-based deprivation.

1.3 Overview of the report

Chapter 2 gives an overview of the study design, an overview of the datasets used in this study and their limitations. The methods used in the analysis are explained.

Taking each health dataset at a time, chapters 3 to 8 quantify usage of the above health services by these three groups. The differentials in usage are highlighted. Chapter 9 analyses the deaths dataset.

Chapter 10 focuses on individuals rather than the particular datasets. It quantifies the proportion of people in the cohort who have multiple conditions - issues with drug or alcohol dependence, mental health conditions, or a combination of these.

Chapter 11 concludes the report and brings together the findings from the previous chapters.

1.4 Study history

In total, this report is the culmination of over three years of work involving Communities Analysis Division of the Scottish Government, all 32 Local Authorities in Scotland, National Records of Scotland ( NRS) and NHS National Services Scotland ( NSS).

The study has made use of the Scottish Informatics and Linkage Collaboration ^[13] – the national infrastructure for data linkage in Scotland. This has involved use of eData Research and Innovation Service ( eDRIS) research co-ordinators, the indexing team at NRS and the national Safe Haven (the environment in which this analysis was conducted).

In order for this work to commence, a number of key Information Governance documents needed to be completed. These are available on the Scottish Government website ^[14] . This included an application to the Public Benefit and Privacy Panel ( PBPP) for Health and Social Care , covering the privacy, security and ethical aspects of the project. The timescales for the project were as follows:

• November 2014: Fife Council presents their analysis at Scottish Government and CoSLA Homelessness Prevention and Strategy Group. Question asked whether this can be taken forward for Scotland.
• April 2015: Initial application submitted to SG Analytical Leadership Group including Privacy Impact Assessment
• December 2015: Public Benefit and Privacy Panel Application submitted. Approval with conditions granted Feb 2016. Enabled access to health datasets and access to de-identified data in the National Safe Haven.
• November 2016: Data Processing Agreements with all 32 LAs in Scotland and NRS signed, enabled indexing of personal identifiable data on homelessness.
• November 2016: Data Sharing Agreement signed between SG and NSS ISD for Homelessness Data
• December 2016: NRS: Indexing Completed
• February 2017: Final Conditions met for PBPP
• April 2017: Health and Homelessness datasets prepared by NSS ISD
• May 2017: Analysis commenced in NSS National Safe Haven
• June 2018: Results published

1.5 Acknowledgements

The authors would like to acknowledge the support of Fiona Campbell and Dionysis Vragkos at eDRIS, the staff of the 32 LAs for providing data for this project and the team at Information Services Division for the provision of the health data. Thanks also goes to the Scottish Government Analytical Leadership Group for providing funding for this work and the Scottish Government Communities Analysis Division for the supply of the homelessness data. Thanks go to Dave Clark and Ken Humphreys at the NRS Indexing Team for their role in de-identifying the personal data and for the construction of the cohorts.

The authors would also like to thank Esta Clark (Head of Design for Scotland's Census 2021) and Amy Wilson (Head of Scotland's Census 2021 and Director of Statistical and Registration Services at NRS), for generously allowing us time to work on the project.

In addition the authors would like to thank Professor Glen Bramley and Dr Emily Tweed for their input from an academic perspective. Particular thanks also go to Adam Krawczyk and Joe Jobling from Scottish Government Communities Analysis Division and Marion Gibbs from the Scottish Government Homelessness Team for their invaluable input on the draft report.

1.6 Re-use of the data

Applications to re-use the data in this study should be made to Communities Analysis Division ^[15] and the Public Benefit and Privacy Panel for Health and Social Care ^[16] .