The Department of Health conservatively estimates that the costs to hospitals of homeless patients is £85 million, 90% of which is accounted for by in-patient care. Proportionally, this is eight times higher than the cost of providing in-patient care for the comparable population of working age people. The average length of a hospital stay for homeless people is three times that of non-homeless people of the same age.
'Bay6' is a service run by Community Housing Aid, an Exeter-based charity organisation for homeless people. Bay6 employs four specialist housing workers to find accommodation for patients who are homeless or become homeless while a patient in the Royal Devon and Exeter Hospital (RD&E), Torbay Hospital or the North Devon District Hospital. Bay6 has been operating since October 2013 and was due to be funded until March 2015 by the NHS England Regional Innovation Fund. This has included funding an independent evaluation of Bay6 which is being carried out in collaboration between PenCLAHRC and the Public Health Directorate, Devon County Council.
The aim of the preliminary evaluation of Bay6 is to answer the following questions:
- How is the service provided and used by Bay6 and hospital staff? Is it provided and used differently in the three hospitals where it currently operates?
- What are the perceptions of hospital and Bay6 staff of the impact of the service on health and other outcomes for homeless patients?
- What are the perceptions of hospital and Bay6 staff of the impact of the service on the use of NHS resources?
- Are there any potential savings to Clinical Commissioning Groups (CCGs) and NHS Trusts arising from the implementation of Bay6, in terms of reduced readmissions, reduced delayed discharge and reduced emergency admissions? If so, what are they?
Questions 1 to 3 were answered through qualitative analysis of semi-structured interviews with Bay6 staff and NHS staff in the three hospitals where the service operates.
Question 4 was answered using analysis of Hospital Episode Statistics and other hospital data, conducted on behalf of the team by Gemma Hobson (data analyst in the Public Health Directorate of Devon County Council).
1. Service provision
"They’re my saviours. I know that sounds a bit cheesy but they are my saviours because before this, it was a struggle. It was a struggle at ward level." (Matron, City Hospital)
It was found that Bay6 workers have better housing knowledge and contacts than hospital staff which makes them more effective and efficient in finding accommodation. The reported impacts of this are to:
- Increase the chances of accommodation being found
- Avoid the likelihood of patients being discharged onto the streets
- Avoid delayed discharges
- Prevent hospital staff being diverted from providing care to non-homeless patients on the wards.
Hospital staff – and clinicians in particular - also emphasised how Bay6 has eliminated the time they would have spent finding accommodation for homeless patients and that the time saved has been redirected to providing routine patient care.
Hospital staff reported that the provision of suitable accommodation as an alternative to discharging onto the streets:
- Helps to prevents the patient’s post-discharge health deteriorating, and thereby avoids readmission
- Enables post-discharge medical care to be provided that will prevent readmission. (Post-discharge medical care includes district nurses changing wound dressings and the provision of support by community mental health services, services which we were told require the patient to have a stable home address)
- Improves the likelihood that patients comply with post-discharge medical care
- Enables patients to tackle their drug and alcohol misuse.
It was reported that pressure on beds and insufficient staff time to find accommodation increases the likelihood that patients are discharged onto the streets. Finding suitable accommodation can be frustrated or delayed by its limited availability and the complexities of a housing provider’s application process. Bay6 workers will continue to seek accommodation for patients even if they have been discharged onto the streets or into unsuitable accommodation.
2. Potential NHS cost savings
We analysed hospital data in relation to Bay6 service users, comparing their use of hospital services in the six months before and after using the service. We found that Bay6 clients (homeless people) are high consumers of hospital care – costing on average about £20,000 per year to the NHS.
For the 104 homeless client records we analysed, hospital use in the six months after using Bay 6 involved fewer A&E attendances (273 vs 367) and fewer A&E attendances by ambulance (187 vs 233). There were then also fewer inpatient admissions (157 vs 244) and fewer bed days (923 vs 2190) but more outpatient appointments (224 vs 191) after their use of Bay6 than in the six months before.
In terms of the cost of hospital care, total costs were about £90,000 lower in the 6 months after using Bay6 than in the six months before, a possible saving (if this is a causal association) of 20% of total costs (£309,638 vs £397,690). The pre- post comparison including the cost of the index admission episode in the pre-intervention six months shows that costs in the six months after use of Bay6 were less than half of the costs in the ‘pre-intervention’ six months.
The reported beneficial impact of Bay6 on the use of clinician’s time and patient’s post-discharge medical care and health outcomes has the potential to:
- Reduce hospital costs by preventing re-admissions and avoiding delayed discharge
- Avoids hospital staff time being spent finding accommodation for patients and enables it to be spent providing routine patient care
- Reduce NHS and other service provider costs by reducing drug and alcohol misuse.
However, paradoxically, there is also the potential for additional hospital costs to be incurred where some patient’s discharge is delayed until accommodation can be found.
Impact of Bay6 evaluation
We have been able to document a wide range of perceived benefits of the service. Some seem to be directly measureable, such as the reduction in hospital use in the six months following the appointment. In our sample of 104 homeless service users this was associated with a £90,000 (20%) reduced cost of hospital care in the six months after using the service. However, another perceived benefit – avoiding the wasted time of nurses and other clinical staff trying to arrange accommodation – may also prove to have a substantial economic saving as well as having implications for improved patient safety.
Our interim findings were shared at the Acute Medical Services Committee of NEW Devon Clinical Commissioning Group (CCG) in January 2015 and informed their decision to extend the funding of this service through the Royal Devon & Exeter Hospital until end of March 2016. By demonstrating the potential short-term cost savings of the service and identifying several less quantifiable benefits this CLAHRC-supported service evaluation will inform the ongoing commissioning of this service and similar services for homeless patients in other parts of the country.
We have produced a final report on our work (see link, above right) and will prepare a journal article to disseminate the findings. This will include directly sharing our findings with those running similar services in other parts of the UK.
Looking forward, we would like to use more rigorous evaluation methods to evaluate similar services in the longer term but this depends on such services attracting investment and existing in the long-term. We want our positive experience of collaborating with Community Housing Aid to be a basis for further collaboration between researchers and third sector organisations involved in commissioning and delivering health and care services.