A person-level risk stratification model, driven by system-wide data insights, provides a shared and embedded view of admissions risk across multiple user groups.
The primary focus of High Impact Change Model: reducing preventable admissions is to reduce admissions to hospitals by remaining in or returning to the community.
A1: A person level risk stratification model, aligned to admissions risk
- Have an agreed system approach to Population Health Management, with clear leadership responsibilities.
- Understand risk at an individual level.
- Frame the right questions up front to be specific to hospital admission avoidance – see Putting this into Practice section for advice on clarity around admissions impact.
A2: A data-led approach joining together insight from across the system
- Join data from across health, social care and wider sources such as public health, housing and wider local government.
- Develop predictive models around a person’s individual risk, combining service activity patterns, demographic factors, and other local variation.
- Be mindful of underrepresented groups in the data and other sources of potential bias.
A3: A shared view of risk, embedded within system ways of working for multiple user groups
- Use outputs from the risk stratification to agree system interpretation and agree priorities.
- Couple digital tools with skills, infrastructure and support to make the outputs interpretable and usable at a local level by individual practitioners, and to support service redesign.
- The outputs should be easily refreshable and able to demonstrate any further changes in people’s risk, and the positive impact of interventions.