���ϳԹ�

The Hospital Frailty Risk Score (HFRS) helps identify older patients (aged 75 and over) who may be frail and more likely to experience poor outcomes during or after a hospital stay.

, led by the in collaboration with , has found the HFRS is effective at predicting extended hospital stays in patients of all ages—not just older adults.

Data scientists in the University’s Centre for Healthcare Modelling and Informatics and the Artificial Intelligence and Data Science Centre analysed adult patient records from Queen Alexandra Hospital in ���ϳԹ�between 2010 and 2018. 

They discovered the HFRS is a more reliable predictor of long hospital stays than other common measures, especially for stays longer than 21 days. It was particularly effective for patients over 44 years old, but also useful for younger adults in most cases. 

The paper, , says using HFRS on people of all ages could help hospitals identify high-risk patients earlier, ensuring they receive appropriate support and care.

The data analysis was carried out by ���ϳԹ� PhD student Huda Kutrani. She said: “Hospital stays can be stressful for patients and costly for healthcare providers. Knowing in advance which patients are at risk of staying longer than expected allows hospitals to allocate resources better.

“By incorporating the HFRS into routine hospital assessments, healthcare providers could improve patient outcomes, reduce pressure on hospital beds, and lower costs.”

Professor Jim Briggs, Director of the ���ϳԹ�’s Centre for Healthcare Modelling and Informatics (CHMI), added: “This is important work because it extends a useful and easily-adopted tool to all adult patients, not just the very elderly ones it was designed for. This helps ensure that healthcare is provided equitably and appropriately.

"This study also shows the importance of using big data analysis to develop new insights into patient care. We were particularly pleased when one of the original designers of HFRS complimented us on our work."

The UK estimates that each year 350,000 patients stay in acute hospitals for more than three weeks. In-patients with frailty may have a longer length of stay and increased health resource use.

The researchers hope that hospitals will adopt the HFRS more widely to enhance patient care and hospital efficiency. Further studies could explore how to integrate this scoring system into daily hospital practices for maximum benefit.

Claire Spice, from , said: “Identifying people in hospital who may have frailty is vital so that we can meet their clinical needs, improve health outcomes and enable appropriate service development.

“Clinical tools are useful but not always applied consistently - automated information-based measures have the potential to highlight groups of people at the most risk quickly and consistently.  Most frailty identification tools have not been researched in younger people so this study really extends the potential utility of the HFRS in further research.”

Last year, AI experts from the University and healthcare professionals within worked together to help prevent a common and painful complication in advanced kidney failure treatment.

The developed an AI model to predict which patients are most at risk of their blood pressure dropping during dialysis; a condition known as intradialytic hypotension (IDH).

“The collaboration between the University and the Hospital has, over 20 years, led to many world-leading developments in the monitoring of deteriorating patients”, said Professor Briggs. 

“We are confident that hospitals are a safer place now than before our research began."