Patients across Wakefield District set to benefit from new data tool that predicts who will require hospital admission

A groundbreaking new data tool is set to transform personalised healthcare support in Wakefield District by predicting who could end up in hospital – helping teams support patients before they require emergency care.

The NHS Wakefield Business Intelligence (BI) team has designed and implemented a risk stratification model that enables local Neighbourhood Health teams to proactively identify patients at greatest risk of unplanned emergency admission.

The predictive model was built using the sophisticated Wakefield District population linked data model which pulls together information from across the NHS, primary care, social care and also non-NHS sources such as housing.

The tool creates a holistic picture of local population health and supports more effective, joined-up care. The risk stratification approach is a key enabler to helping clinicians, social care teams and voluntary sector partners work together more effectively, ensuring the right patients are seen at the right time.

The tool has been designed to segment patients with Chronic Obstructive Pulmonary Disease (COPD), Dementia and those receiving end-of-life care into high, medium and low risk of unplanned emergency admission.

Several different algorithms were tested and trained to pick out the patterns that led to emergency admission. Each algorithm was tested on unseen data and who was likely going to be admitted in an emergency in the next three months.

The models were found to have extremely high rates of precision and what in data science is referred to as high ‘F1 scores’ – ranging between 86 and 94 per cent in accuracy across the cohorts.

Dr Colin Speers, Director of Integrated Community Services and Digital Lead for Wakefield, said:

“This is a powerful step forward in making our decision-making truly clinically led and data-driven. The Wakefield BI team has created an innovative resource that puts actionable insight directly in the hands of frontline teams.

“This is about turning data into better patient care and improving population health.”

The development of the tool reflects Wakefield’s commitment to integrated neighbourhood health, which brings together multidisciplinary teams to provide preventative care for defined populations of about 30,000 to 50,000 people. It also aligns with the NHS ambitions to embed population health management and use data to support more proactive models of care.

Shaukat Ali Khan, NHS West Yorkshire Integrated Care Board Chief Digital and Information Officer, said:

“The BI team has created a resource that makes data truly actionable.

“It helps ensure our clinicians and partners have the information they need to make the best decisions for patients. It’s a great example of how BI and frontline teams can work together to improve patient outcomes.”

Early feedback from colleagues in primary care has highlighted how the tool is already supporting multi-disciplinary teams in prioritising caseloads as part of the integrated neighbourhood health programme. Wakefield colleagues are also sharing the best practice tool with wider teams across the whole of West Yorkshire.