Kent hospital first in NHS to use AI to catch infections

TL;DR:

  • East Kent Hospitals NHS Trust is the first in the country to use MEMORI, an AI tool that flags patients at risk of infection before it develops.
  • The software analyses routine data — blood tests, blood pressure, temperature, medications — to generate an individual risk score.
  • Staff say it frees clinicians from trawling notes, and 85% of clinicians using it reach for it every day.

At Kent and Canterbury Hospital, clinicians are now working with a fully licensed AI medical device that turns information already sitting in patient records into an early warning. Built by UK health-tech firm Sanome in collaboration with the trust, MEMORI produces a score indicating whether a patient is likely to develop an infection “before we actually see an infection develop”, according to ward manager Julie Jones.

Built with clinicians, not to replace them

The trust is at pains to frame the tool as support rather than substitution. Nurses, healthcare assistants, consultants and therapists helped develop it to fit real clinical practice. “It absolutely is not” replacing staff, Jones said, describing the aim as giving clinicians time to “actually be with our patients” rather than interrogating IT systems. Sanome founder Benedikt van Thüngen — who started the company after his father developed sepsis — said catching deterioration early lets staff “do what they’re really good at, which is providing care”.

Looking forward

The deployment is a concrete example of the shift the NHS 10-year plan envisages from “analogue to digital”, and it lands amid a broader public-sector push: the NHS is making AI-assisted app triage a priority in a £10bn tech drive, while Lancashire says AI is giving social workers time back. The test now is whether a single-trust pilot becomes a repeatable model. Chief clinical information officer Dr Mike Bedford called it “just the beginning” — the real prize is a licensed, trusted tool that other trusts can adopt without rebuilding it from scratch.