Predictive risk project

As part of the long term conditions agenda, The King’s Fund, along with New York University and Health Dialog Data Service (an American company specialising in health data analysis), has been commissioned by 28 strategic health authorities and the Modernisation Agency to produce a risk prediction system for use by PCTs to identify patients who are at high risk of readmission to hospital. The project started in April 2005 and is due to end in December 2005.

The outputs from the work will be:

1. a comprehensive literature review summarising and analysing international evidence of risk prediction methods.  Completion date June 2005
2. a tool that PCTs can apply to HES data to predict high risk patients. Completion date July 2005
3. a tool that PCTs can apply to HES and local routine data to predict high risk patients. Completion date January 2006

Predicting risk

There is, at present, no robust way to stratify PCT populations by potential risk. In the context of this project, ‘risk’ relates to the likelihood of hospitalisation. Two main approaches have been used to date in the UK for predicting risk of hospitalisation: one using quantitative data (for example one or more hospitalisation in the last year, plus the use of multiple medicines) and the other is qualitative (clinician “hunch”). The recent Evercare pilots in the UK have used the former method to identify older people at high risk of admission but have found it, so far, not to yield a high predictive accuracy. Research studies in the UK and abroad have also indicated that clinician ‘hunch’ is not a highly accurate method. Managed care organisations in the US have undertaken a great deal of research into the various methods of predicting risk.

The project

In response to the need to identify high risk patients, the Department of Health and Modernisation Agency put out a tender for such work, with Essex SHA acting as lead commissioner for the other SHAs. The King’s Fund successfully bid for the work and is working with New York University and Health Dialog Data Service, a US-based companythat has significant experience executing predictive risk modelling and applying its findings.

This project aims to learn from US, and other, international experience in order to produce an algorithm for PCTs in the UK to use to identify those most at risk of hospitalisation. Several PCTs are involved in the project as pilot sites and are providing the necessary data to undertake the work. Once patients are identified, PCTs can then develop interventions that are appropriate to this high risk group with the aim of preventing a further admission to hospital. The developed algorithm will be made available to PCTs across the country.

Contacts

For more information about this project, please contact:
Kate Gill, Essex SHA: kate.gill@essexsha.nhs.uk
Natasha Curry, King's Fund: n.curry@kingsfund.org.uk
Chris Hall, Health Dialog: chris.hall@healthdialog.co.uk
John Billings, New York University: John.billings@nyu.edu


Using Data to Drive Health System Performance: NatPaCT learning commissioned from Ovations 
The development of skills and competencies to better analyse and manage information is increasingly important to PCTs to drive decision-making for planning and commissioning health services and improve service availability.  It is also increasingly important to manage, forecast, and predict impact to the local health system as modernisation is contemplated and undertaken.

A learning programme in improving analysis and management of information has been developed by NatPaCT in partnership with Ovations.  It will provide development opportunities to PCTs to extend their capability and capacity for information management for planning, commissioning and delivering appropriate services.  Increasingly this is also a key skill for driving service and system improvement for managing patients with long term conditions.

The module has been developed to cover five topical areas:

• Using data to drive strategies – an introductory session to develop an understanding of the key principles in the use of management information.
• Organising to effectively use information to manage and create results – This session will provide information that can be taken forward to develop staff and community buy-in for the use of information in management.
• Data collection – This session provides an overview of the process of data collection.
• Data analysis and methods – This session will focus on how key data are organised for analysis, types of analysis used to identify performance improvement opportunities, and how to drive further evaluation around selected objectives and/or tactics.
• Strategies and tactics using information for planning and actions –Without well analysed, well-planned approaches to development of strategies and related actions, goals and targets cannot be accomplished.