open menu

The Administrative Data Research Network is an ESRC-funded project that ran from October 2013 - July 2018. It is currently at the end of its funding cycle and is no longer taking applications. Administrative data research will be taken forward in a new project, to be launched later in 2018.

Morbidity and mortality among people experiencing severe and multiple disadvantage: a cohort study using cross-sectoral data linkage

Research overview

People who experience homelessness, imprisonment, drug addiction, or severe mental illness tend to have very poor health, and are at high risk of dying early. There is increasing evidence to suggest that these experiences overlap to a great extent in the population: this overlap is often referred to as ‘severe and multiple disadvantage’. However, previous research has typically studied these experiences in isolation, so we know very little about how this overlap affects health.

 This study aims to link together a number of existing databases from health and social services in a secure and confidential way. This will allow us to investigate how homelessness, imprisonment, drug addiction, and severe mental illness interact in important ways to affect health, and to identify the most common causes of ill-health and death among people with more than one of these experiences. Our findings will help inform the design and delivery of services for this vulnerable population and provide the basis for future studies to develop tailored interventions to better meet their needs.


Severe and multiple disadvantage (SMD; also known as multiple and complex needs) has been recognised as a priority area for the national Homelessness Prevention and Strategy Group in Scotland, as well as local Health and Social Care Partnerships (HSCPs) and third sector organisations such as the Scottish Drugs Forum and Glasgow Homelessness Network. Recent UK trends, such as rising rates of homelessness, rough sleeping, and imprisonment, suggest that the number of individuals affected by SMD may increase in future.

There are long-standing concerns from service providers and policymakers alike that health and social services are failing to meet the needs of people experiencing SMD. At present, service responses to homelessness, offending, substance use disorder, and mental illness are frequently characterised by specialisation and fragmentation, creating barriers to access and hindering retention in care.

By helping to address key knowledge gaps regarding the health burden, service utilisation, and resource usage associated with SMD, this study will help identify priority areas for service design and planning, inform resource allocation decisions, and provide a foundation for the development of specific interventions and models of care for this group. For instance, our results will help quantify the potential benefits of preventative spend – particularly when developing and evaluating new models of care and tailored interventions.

This study is particularly timely in the context of health and social care integration in Scotland: with this process still at an early stage, our findings have the chance to shape how services for this population are planned and delivered. Finally, our pioneering methods will provide proof of principle of the value of linked administrative data for investigating SMD. In doing so, this study will lay the foundations for future work, including the development of individual and service level interventions and the evaluation of policy changes, such as welfare reform and Housing Options.

Data souces

This project will bring together data from the following sources to identify the population of interest:
  • HL1 - people applying for statutory homelessness support, collected by Scottish local authorities and collated by the Scottish Government
  • PR2 - people received into Scottish prisons, collected by the Scottish Prison Service and collated by the Scottish Government
  • Scottish Drugs Misuse Database - people being assessed for specialist drug treatment 
  • a subset of Prescribing Information System data, relating to people prescribed opioid replacement therapies like methadone and buprenorphine
  • a subset of SMR01 and SMR04 data, relating to people admitted to hospital for drug-related causes
  • a subset of SMR01 and SMR04 data, relating to people admitted to hospital for severe mental illness (such as schizophrenia)
Records from these sources will then be linked to the following datasets to investigate health outcomes over time among people with the experiences of interest:
  • Unscheduled Care Datamart - for ambulance call-outs and A&E attendances
  • SMR01 and SMR04 - for admissions to hospitals
  • National Records for Scotland death registrations - for mortality
  • Patient Level Information and Costing System (PLICS) - for secondary health care costs

Date approved

June 2017

Research team

Dr Emily Tweed (lead researcher)             MRC/CSO Social and Public Health Sciences Unit, University of Glasgow

Dr Vittal Katikireddi Srinivasa                   MRC/CSO Social and Public Health Sciences Unit, University of Glasgow

Professor Alastair Leyland                        MRC/CSO Social and Public Health Sciences Unit, University of Glasgow

Dr David Morrison                                   Department of Public Health, University of Glasgow

Dr Oarabile Molaodi                                 MRC/CSO Social and Public Health Sciences Unit, University of Glasgow

Professor Glen Bramley                            I-SPHERE, Heriot-Watt University

Professor Suzanne Fitzpatrick                   I-SPHERE, Heriot-Watt University

Page last updated: 08/01/2018