P23 Working with routinely collected data in social care: What are the possibilities, complexities and considerations for researchers and commissioners/service providers?

Background

An understanding of the patterns of social care provision and outcomes for older populations will be increasingly important to the development of integrated care models. However, routinely collected social care datasets are difficult to access, and there is little guidance for researchers who are often more familiar with healthcare data. We collaborated with a local authority in the North-East of England to identify opportunities for service improvement by analysing their routinely collected social care data. Using this analysis as an example, we describe some of the complexities unique to social care datasets, key considerations for future researchers and potential benefits of working in this field.

Methods

Data were extracted from a local authority social care case management system, which contains details of the social care packages provided. We aimed to describe the care packages for older people (aged 65+), including care provided in care homes (with and without nursing), private households and assisted living facilities. Data quality checks were carried out to describe the completeness and validity of the data extract, and discussions with the data owners clarified the influences on the data and potential implications.

Results

We analysed 171,386 records from 38,191 unique individuals, across the last 40 years. Data completeness was close to 100% for the majority of variables, including costs and service type. We found variation in the average duration of individual care packages (median 41 days, IQR 260 days (14 - 274 days)). Patterns of care duration also varied by setting, from an average of 0.4 years for residential care to an average of 2.12 years for extra care. Additional analyses were conducted to understand the impact of variation in the nature of the care, including reasons for care packages ending, transitions between care settings, and the source of funding.

Discussion

There are many nuances to working with routinely collected social care data. Interpretation of findings requires an understanding of the context in which data were collected, in addition to different models of care provision and funding. We suggest some of the many possible impacts of such work and argue that developing work in this field is critical for future health and social care integration.