Predictive Modelling for Unplanned Care in the North East and North Cumbria
|Researchers Involved:||Prof. Graham Towl, Dr Camila C. S. Caiado, Dr Rachel Oughton|
|Main disease areas impacted:||Planned/unplanned care, dementia|
|Key Partners:||Dr Ian Briggs, CDDFT, South Tees NHS Trust, South Tyne NHS Trust, Darlington Borough Council, NECS|
|Start and end dates:||Oct 2016 – Dec 2018|
A collaboration to produce statistical models that can be routinely used by appropriate health/local authority/other analytics teams to produce daily forecasts up to six months in advance with the pertinent associated uncertainties and variations in Urgent and Emergency Care.
This project will facilitate planning and management in hospitals, walk-in centres, and GP practices. As more health, local authority and other relevant data become available, the models will be enhanced to track ‘at-risk’ cohorts coming through unplanned care. For this initial project, the focus will be on the frail elderly with an emphasis on dementia, and on alcohol-related attendances.
Why is this research project important?
This project will investigate strategies for optimal planning and management for urgent care arrivals and staff reconfiguration, planning and workload predictions, and identification and planning for at-risk cohorts such as the frail elderly.
We will bring together expertise from academia, healthcare providers, and regulators to deliver practical modelling, planning and decision-support tools and expediting the integration and germane availability of information across health and social care services in the North East.
Our approach will lead to an efficient allocation of reducing resources within the system through prevention, early intervention, targeting at-risk cohorts, optimised care management, and improved patient outcomes. We will address issues related to the information governance, information technology, and legal frameworks necessary for the deployment and implementation of this strategy.
The unprecedented levels of change in policy and funding, the shifting and ageing demographics of the population served, and ever-increasing workforce pressures are of substantial concern to health and social care sectors. Consequently, the health and social care system is facing a range of immediate problems, and pressures which require a significant transformation of existing care delivery systems and networks at both local and national scales.
Academic collaborations bring innovation and research that can facilitate and optimise the current environment, and support planning for a more robust health and social care system. Facilitating data and information sharing across healthcare providers, local authorities, and academic partners will be essential to deliver efficient and reliable solutions whilst complying with data sharing regulations.
What data are being used in this project?
The project will use datasets provided by the NHS Trusts, Darlington Borough Council, and GP practices with all A&E and planned/unplanned appointments in the past few years. Anonymised and pseudo-anonymised data provided.
In addition, the project will access external datasets from the Metoffice and Wunderground weather data.
What methods are you using to conduct this work?
- Build prediction models for attendance using linear and Bayesian models
- Scenario planning for new models of care
Who will benefit from your research?
- Medical staff
What will be the intended outcome of your research project?
We expect to observe:
- Earlier diagnosis and specialist referral leading to better health outcomes;
- Improved patient outcomes in terms of service satisfaction resulting in improved patient experience;
- Improved staff satisfaction in terms of the professionalism of service provision, and a less stressful environment;
- More efficient deployment of appropriately trained staff where and when needed, and a more efficient use of funds for planned and unplanned care;
- Relationship building, system testing, and better integration of staff with academia, e.g. possibilities for postgraduate qualification, registration, research and development support
- Reduced costs and improved services that will assist in managing financial cuts. Ultimately, there may be less time lost by employees across the region in sickness days. There will also be scope for involvement with local businesses for support with deployment and management of the tools developed.
Are there any early findings?
We have received initial datasets and developed preliminary models which have been shared with the trusts to inform healthcare planning.
We are now working with the trusts to refine the models, processes and infrastructure.
Any comments/additional information you would like to report?
We will produce guidelines and recommendations for the implementation of the information governance, information technology, and legal framework necessary for data and information sharing for projects involving both data analysis and research.
From the academic perspective, we will develop and test a transferable methodology for predicting patient flows, and design decision support tools that are accessible to all relevant partners.