AI kettles and fridges reduce hospital readmissions in NHS pilot

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

A pilot scheme using artificial intelligence in kettles and fridges to monitor discharged hospital patients in England has reduced unplanned readmissions by 77 per cent, its creators have said.

The Onward Care scheme, developed in partnership with Buckinghamshire Healthcare NHS Trust in south-east England, is the first of its kind to be used in the health service.

“AI is here to work with us,” said Jenny Ricketts, deputy chief operating officer for Buckinghamshire trust. “Not to do the job for us. People, especially the elderly, like human contact. The AI just makes it easier for us to do that.”

The system operates by linking AI electronic sensors on kettles and fridges to detect changes in drinking and eating patterns. Variances are then flagged with a member of the Onward Care programme who can arrange clinical help if needed.

Health leaders are increasingly seeking to use artificial intelligence to reduce the burden on the health system, with the demand for care set to rise as the UK population ages.

The pilot, co-led by facilities management and food services company Sodexo, has been running for 12 months in Buckinghamshire. Sodexo is in discussions about extending the scheme to other trusts around the country.

Adrian McCourt, managing director of the Onward Care service, said the pilot had supported about 140 people at home for 12 weeks after they were discharged from hospital.

He estimated that on average 40 per cent of “frail” patients who may not have fully recovered are readmitted to hospital within six months of being discharged. Under the pilot, he said, this figure has been reduced by 77 per cent.

A small number of hospitals are already using AI tools to help radiographers analyse X-rays, speed up bookings and referrals and deploy speech-recognition technology to take clinical notes.

It uses a box called Monica, similar to Amazon’s Alexa voice assistant, which is put in people’s homes and gives prompts to move if their levels of activity fall outside the norm.

Data on the number of steps taken or how long a person spends in a particular room is fed into a central dashboard.

“If a patient’s behaviour is changing we get a notification which prompts us to investigate the situation,” McCourt said. “We also have sensors on fridges and kettles, which we use as a proxy to understand whether hydration and nutrition is changing over time.”