Predictive risk project
NEWS: A new and improved version of PARR (PARR++) is now available. Click here to find out more:
Predictive risk project
The King's Fund, along with partners New York University and Health Dialog, has developed two ground-breaking techniques (PARRand the Combined Predictive Model) to help PCTs predict the risk of emergency admission and re-admission to hospital by identifying patients who are at risk of, but who have not yet entered, a spiral of emergency admissions.
Through using PARR and the Combined Predictive Model, PCTs have in a short time been able to reduce emergency re-admission rates and save costs in hospitals across the country.
If you would like to be informed when new versions of the PARR Tool become available, please enter your name and e-mail address here and click 'send'.
Risk prediction
PARR and the Combined Predictive Model are both free to download:
-
The Patients at Risk of Re-hospitalisation (PARR) tool, a software tool with an easy-to-use interface, that uses routine inpatient data to identify patients most at risk of future emergency re-admission to hospital. A new improved version of the tool, PARR++, was launched in November 2007. The tool is available for download or to order on a CDRom.
-
TheCombined Predictive Model, which links inpatient, outpatient, accident & emergency and GP data to identify the likelihood of patients across a whole PCT population being admitted to hospital in the next 12 months. Unlike the PARR model, the Combined Predictive Model does no have a user interface and is a 'string' of code that can be built by a programmer.
Why were these products developed?
Both PARR and the Combined Predictive Model were commissioned by the Department of Health and Strategic Health Authorities and developed by the King's Fund in partnership with New York University and Health Dialog to enable better management of patients with long-term conditions.
Such conditions result in significant morbidity and cost, and previous work by the King's Fund had shown that many emergency admissions are of patients with conditions such as asthma, diabetes, chronic obstructive pulmonary disease, epilepsy, cellulitis and sickle cell disease. More effective primary care, such as case management, disease management and health promotion, can reduce the risk of such patients being admitted.
PCTs face the ongoing challenge of allocating patients to appropriate 'at risk' groups - as set out by Department of Health guidelines - and specifically to identify the patients who would benefit most from intensive case management. PARR accurately identifes high-risk individuals, whereas the Combined Predictive Model stratifies the whole PCT population according to their risk of admission. These two products help PCTs to intervene and reduce future hospital admissions.
Literature reviewDuring the early stages of the predictive risk project, we conducted a literature review which used international studies to summarise and assess the principal approaches to predicting risk within the health arena.
- Predictive Risk Project Literature Review summary (Word, 259KB)
- Predictive Risk Project Literature Review Full Version (Word, 523KB)
Contacts
-
Natasha Curry, King's Fund:parrupdates@kingsfund.org.uk
-
Tracy Morton, Department of Health, Tracy.Morton@dh.gsi.gov.uk
-
Matt Siegel, Health Dialog:msiegel@healthdialog.com