Talk like a health and care economist

Data science — the collection and analysis of data to extract meaningful insights

Discrete choice experiment — a quantitative research method for understanding people’s preferences and decision-making processes, in which participants are asked to choose between hypothetical choices

Integrated management dashboard — an interactive platform that allows a user to visualise a range of datasets and extract meaningful insights

Out-of-hospital care model — any type of care not undertaken in a hospital setting

Stakeholder — anyone who uses (e.g., patients, carers), provides (e.g., healthcare professionals, social workers) or commissions (e.g., local authorities, government departments) a service

Step-down services — services for patients discharged from hospitals, typically integrating health and social care needs

“People experiencing homelessness commonly face a range of health and social challenges,” says Dr Michela Tinelli, a health and care economist at the London School of Economics and Political Science. “For instance, exposure to the elements during rough sleeping results in increased vulnerability to illness.” The stress of living on the streets, coupled with social isolation and discrimination, means that people experiencing homelessness are also vulnerable to mental health problems. Homelessness makes it much harder to access healthcare, education and employment opportunities, which compounds these problems. This highlights the need for services that address these issues.

“People experiencing homelessness need access to healthcare and mental health services, community clinics focused on the needs of the homeless population, and services to connect individuals with housing and employment opportunities,” says Michela. Out-of-hospital care models play a crucial role in providing these services, and Michela is exploring how to optimise these models to provide the best possible care.

Out-of-hospital care models

Out-of-hospital care models provide health and social care services in the community, rather than in hospitals. For example, step-down services address the unique needs of people without stable housing by providing transitional care facilities for people leaving hospital. Often, if a person experiencing homelessness is discharged from hospital with no further follow-up, they will remain homeless. Step-down services aim to address this. “The primary goal of step-down services is to provide individuals with a supportive environment that promotes recovery and stability,” explains Michela. “This approach recognises the interconnectedness of healthcare and social factors for improving well-being and quality of life.”

The importance of data science

Michela and her team used a mixture of qualitative and quantitative methods to evaluate out-of-hospital care models used in England. “We conducted qualitative interviews with people experiencing homelessness, healthcare professionals and other out-of-hospital care model stakeholders,” she says. “These discussions allowed for in-depth exploration of personal experiences of and insights into the services.” Quantitative methods included surveys of participants’ experiences and analyses of outcomes (such as rates of readmission to hospital or return to homelessness) and service delivery costs.

Michela also used discrete choice experiments to understand what people experiencing homelessness want from out-of-hospital care models. These surveys asked participants to choose between two hypothetical situations, to understand their preferences and what trade-offs are acceptable for them. “For example, we asked how long people would be willing to wait to receive care in their own home, compared to receiving care immediately in temporary accommodation,” explains Michela. Data science methods such as these will help to inform the rollout of services that cater directly to the needs of the community. For instance, the discrete choice experiments uncovered some interesting results. “Unsurprisingly, participants strongly expressed their aversion to returning to rough sleeping, compared to accessing step-down services,” says Michela. “By synthesising their preferences, we deduced that the ideal care service model encompasses a housing support worker visiting the participant’s own accommodation three to four times a week, for a duration of ten to twelve weeks.” Crucially, this suggests that a stable, private living arrangement – having their own flat, rather than living temporarily in a hostel – makes people experiencing homelessness more likely to access health and social care services.

Michela’s results suggest the impacts of out-of-hospital care models are strongly positive. “We collected data from 1,254 homeless patients, many of whom experienced improvements in quality of life and felt they were treated with dignity and respect through these services,” says Michela. “Only 7% returned to rough sleeping after accessing step-down services, compared to around 77% before these services existed.”

Developing dashboards

Using the insights gathered from her results, Michela developed integrated management dashboards to serve as tools for out-of-hospital care model stakeholders. “The objectives of these dashboards include monitoring the performance of out-of-hospital care models, assessing the quality of care received by patients and optimising the allocation of resources,” says Michela. The interactive dashboards allow users to visualise data through clear and customisable presentations. “They empower the monitoring and evaluation of current events, allowing for immediate action based on present insights (what is happening now) instead of relying on published reports with outdated data (what happened in the past),” she explains. “The dashboards provide actionable insights for stakeholders. They facilitate data-driven decision-making to enhance the effectiveness of out-of-hospital care for people experiencing homelessness.”

There is strong support for the integrated management dashboards as operational and management tools. Stakeholders report they are useful for enhancing evidence-based decision-making. Michela is now in conversation with local health and care providers around England to provide the dashboards to out-of-hospital care model stakeholders throughout the country.

From research to policy

Thanks to the data science research of health and care economists such as Michela, services and national policies will improve the health outcomes for people experiencing homelessness. “Our analysis offers practical guidance for policy development,” says Michela. “It informs policymakers where to focus their efforts and resources to align with the preferences of those accessing and implementing these services.” Furthermore, these results will help to optimise out-of-hospital care model services, improving quality of care and tailoring support to those in need. “Our outputs will support decision-making for the health and well-being of the population at the national level.”