Risk calculator to prevent delayed discharges in hospitals

New research could significantly reduce overcrowding in emergency departments – potentially saving the NHS millions of pounds

Image of an ambulance

The calculator is designed to reduce overcrowding in emerging departments (image: Makizox)

We have based our model on data routinely collected in all hospitals which means it has the potential to be adopted across the NHS. This problem is not going to vanish and in the wake of COVID-19 it is more important than ever to find solutions.

Md Asaduzzaman, Associate Professor in Operational Research

In a collaborative project between University Hospitals of North Midlands (UHNM) NHS Trust and Staffordshire University, experts have developed a prediction model to identify patients most at risk of taking up hospital beds longer than needed.

A new study, funded by the North Staffordshire Medical Institute (NSMI), led by the core team of Dr Andrew Davy (UHNM) with Associate Professor Md Assaduzzaman (Staffordshire University) and Mr Thomas Hill (UHNM) details an eight-variable predictive tool which can calculate the probability of a patient experiencing a delayed discharge at the time of admission.

Delayed discharge from hospital is one of the major challenges facing the NHS and has increased considerably over the last decade. According to 2019 data, delayed bed days cost the equivalent of £27,000 each hour and the additional pressures of COVID-19 have since intensified the problem.

Dr Andrew Davy, GP Lead for Research and Development in A&E at UHNM, explained: “A delayed transfer of care occurs when an adult inpatient is medically ready to go home but is unable to because other necessary care, support or accommodation is unavailable. These delays can have serious implications such as mortality, infections, depression and reductions in patients’ mobility and their ability to undertake daily activities.

“It also has a knock-on effect on patients in A&E departments who cannot move into ward beds until current patients are discharged. This bottleneck effect on flow causes significant overcrowding within emergency departments and other emergency portals, which results in increased mortality, poor patient outcomes and significantly higher consumption of hospital resources.”

For the development of the predictive tool, Md Asaduzzaman, Associate Professor in Operational Research at Staffordshire University, explained: “Administration and clinical data from the Royal Stoke University Hospital’s emergency department was analysed for the study, covering a three-year period from 2018 to 2020. The researchers used information routinely collected when patients are admitted to hospital from A&E to identify several demographic, socio-economic and clinical factors associated with patients experiencing a delayed transfer of care or not.

“Age, gender, ethnicity, national early warning score (NEWS), Glasgow admission prediction score, Index of Multiple Deprivation decile, arrival by ambulance and previous admission within the last year were all found to have a statistically significant association with delayed transfers of care.”

The prediction model and digital toolkit is currently being piloted at Royal Stoke University Hospital with Thomas Hill, Technical Business Intelligence Analyst/Developer at UHNM, developing how the scoring system is visually displayed on A&E’s live dashboards, to ensure patients at high risk of delayed transfers of care are flagged early to reviewing teams. Further variables, felt likely to be causative in delayed discharge, are currently being reviewed. The researchers believe that eventually, this predictive model could easily be rolled out across the country.

Sarahjane Jones, Associate Professor of Patient Safety, said: “Better discharge planning would reduce a huge cost burden to the NHS and we believe that this paper could have a major impact on managing patients.

“Understanding who is most statistically likely to experience a delayed discharge could help hospitals target patients for proactive discharge planning early on in their care journey. This could be achieved by alerting internal teams such as therapists earlier on as well as external partners, enabling more timely provision of community care plans and placements in residential and nursing homes.”

Building on this study, the research team now hope to improve the accuracy of the risk calculator and work with local authorities to better understand the logistics of patient aftercare. Dr Keira Watts and Dr Simon Lea from UHNM Research and Innovation, who have been involved in the development of the project, are keen to take this work forward.

Dr Lea said: “This work is incredibly important for the wellbeing of our patients and for the delivery of our services and we are pleased that this work, funded by a North Staffordshire Medical Institute grant, has the potential to have such significant impact. We look forward to working with Dr Davy and Prof Asaduzzaman on future grant applications and research projects.”

Associate Professor Asaduzzaman added: “We have based our model on data routinely collected in all hospitals which means it has the potential to be adopted across the NHS.

“This problem is not going to vanish and in the wake of COVID-19 it is more important than ever to find solutions. We must develop a well-designed patient care pathway model for vulnerable patients, incorporating all stakeholders including acute care hospitals and social care centres alongside local governments.”

Read the full article – A predictive model for identifying patients at risk of delayed transfer of care: a retrospective, cross-sectional study of routinely collected data

Latest news

Top 5 UK University

StudentCrowd University Awards 2022

87% of Research Impact is ‘Outstanding’ or ‘Very Considerable’

Research Excellence Framework 2021

Top 10 for Job Prospects

StudentCrowd University Awards 2022

Top 15 for Student Satisfaction

Complete University Guide 2022

68% of Research is “Internationally Excellent” or “World Leading”

Research Excellence Framework 2021

6th for Social Inclusion

The Times and The Sunday Times Good University Guide 2022