The NHS could use AI to help clinicians predict when patients are at risk of becoming long hospital stayers, the government has revealed.
The news comes after the NHS AI Lab worked with the Accelerated Capability Environment (ACE) to deliver a proof of concept that could risk stratify patients and allow health service professionals to avoid the negative outcomes associated with longer hospital stays.
The two organisations trialled their AI tool at Gloucestershire Hospitals NHS Foundation Trust after training it on 460,000 anonymised records.
The idea being that reception and clinical staff would have access to an immediate long stayer risk score, which would enable them to alter treatment paths as necessary and avoid known risk factors.
The key question for the team was if the initial patient data collected was enough to foresee risk.
After being delivered by ACE in just 12 weeks, the tool detected 66% of long stayers within the highest risk categories.
The government highlights the potential that can be gained from reducing long stays – in addition to the obvious health benefits, reducing the average stay by just a day could yield £1.7m in savings for the NHS in Gloucester alone.
Following its initial proof of concept testing, the tool was moved into an Alpha phase and integrated into the trust’s electronic health record system. The AI was then benchmarked against data from the pandemic to see if it could have benefited the health service.
The tool remained highly accurate, according to the government.
ACE is a Home Office function centred around solving the public safety and security risks of the rapidly evolving digital landscape.
Director of AI, imaging and deployment at the NHS AI Lab, Dom Cushnan will feature in our upcoming online magazine that is set to be published next week.
Image credit: iStock
