Predictive modelling for social care: Next steps workshop

Predictive risk models for managing health systems were first developed in the United States in the 1990s. In recent years, many NHS organisations have begun using such tools including the PARR model that identifies which individuals in their population are at risk of a future unplanned hospital admission. These tools use historic patterns in the population’s data to make predictions at the individual level.

More recently, Nuffield Trust researchers have demonstrated that analogous models can be developed to identify which individuals are most at risk of starting intensive social care in the next year. These new models, which are thought to be the first of their type in the world, could help councils and NHS organisations to take earlier action to help people remain independent and stay in their own homes.

These new models for social care are thought to be the first of their type in the world

Emergency hospital admissions and admission to a care home are analogous in that both events are typically unwelcome to the person concerned, costly to society, recorded in routine electronic data and sometimes preventable. Following the publication of its feasibility study of models that predict future use of social care, the Nuffield Trust was contacted by a number of local authorities that expressed an interest in piloting the new models in practice.

This event report summarises the discussions that took place at a workshop in May 2011 for these local authorities as well as representatives of the NHS, two universities, and national bodies including the NHS Information Centre for Health and Social Care, the National Information Governance Board and the Audit Commission.

The event report will be of interest to policy-makers, commissioners across health and social care, and academics with an interest in this area.