Staff retention is a challenge shared by providers of adult social care. 400,000 people leave their jobs each year. High turnover disrupts the continuity of care for service users, is hugely costly, and leaves significant workforce gaps. Skills for Care is looking to translate its knowledge of factors influencing retention at a national and regional level into insight which supports decision-making at a bespoke organisational level.
Skills for Care manages the Adult social care workforce data set (ASC-WDS), on behalf of the Department of Health and Social Care. All adult social care providers are invited to contribute - the data goes back 10 years and covers 55% of the workforce. By analysing this data, Skills for Care has been able to highlight a range of factors influencing workforce retention. These include:
provision of training
proximity of place of work to home
tenure of the service’s registered manager (as a proxy for effective leadership).
The aim of this project is to identify those changes which an individual organisation could make at a local level that most influence improvements to their retention figures - and so improve the working conditions and morale of their staff.
The Skills for Care analytics team deployed a machine learning model on its longitudinal data to analyse the stayer/leaver status of individuals at the end of a 12 month period. Trained on 50,000 pseudonymised individual staff records - comprising data on fields such as demographics, experience and employment terms - the model can identify the key factors which affect whether an individual stays with or leaves an organisation. This data has then been aggregated at an organisational level, enabling the model to generate recommendations on improving retention.
The next step is carrying out user research to explore how to present the insight and recommendations so that organisations can understand the data and take action. The tool outputs could range from a simple set of recommendations and signposting of further resources to a more sophisticated interface allowing users to model the impact on organisational retention of, for example, offering increased pay or investing more in professional development.