GPs are using artificial intelligence to spot early warning signs of patients most at risk.
The system looks for patterns to assess chances of emergency admission to hospital.
GettyGPs are using artificial intelligence to spot early warning signs of patients most at risk[/caption]
It was used to analyse the records of 400,000 residents of care homes in Somerset.
The tech cut the number of visits to A&E by as much as 60 per cent.
The pilot also reduced falls by 35 per cent and ambulance call-outs by 8.7 per cent.
In one case, with simple changes to the way a 91-year-old sat and slept, care workers were able to prevent potential future problems.
Now the software, called Brave AI, will be rolled out to 30 areas from Gloucestershire to Cornwall over the next few months.
Nurses, pharmacists, therapists and doctors can use the information to reach those in need.
NHS Somerset digital boss Dr Matthew Dolman said it would help stop staff being overwhelmed by workloads.
He added: “How do you identify those people most at risk out of populations of, say, 10,000 in one community?
“This tool will allow that.”