NHS AI Lab has developed a proof of concept to identify patients at risk of unnecessary long hospital stays.
It has worked with the Accelerated Capability Environment (ACE) – the Home Office body focused on digital and data in public safety and security – and AI specialist Polygeist to deploy the tool at Gloucestershire Hospitals NHS Foundation Trust.
The long stay stratification tool uses an AI model that was trained on 460,000 anonymised records from the trust to identify people at risk from becoming long stayers from the initial data collected on them.
This produced an immediate long stayer risk score that ACE said could be available to all reception and clinical staff to help avoid known risk factors, so they can take action that can reduce the chances of a patient becoming a long stayer.
The proof of concept was delivered in 12 weeks and the tool detected 66% of long stayers in the highest risk categories, ACE said. This could provide benefits for the patients and significant savings for the hospitals.
Now in alpha
It has now been moved into a limited closed alpha service phase and integrated with Gloucestershire Hospitals’ electric health record system via APIs. This has allowed for further testing against Covid era datasets, anonymously checking patients who had already been discharged to see if the tool could have helped. ACE said it was highly accurate in this context.
Extended stays in hospitals often lead to negative outcomes, such as an 11% mortality rate compared to 5% for all admissions and 23% chance of becoming unwell again after discharge compared with 1%.
ACE said many long stays are avoidable and interventions such an increase in walking and daily physiotherapy are known to reduce overstays.
