The number of patients who fail to turn up for a doctor's appointment has been cut by nearly a third in a trial using artificial intelligence (AI) that could save the cash-strapped NHS a fortune.
No-shows account for more than a million missed GP appointments every month - equal to one in 20 scheduled consultations.
This adds to pressure on a healthcare system already creaking from the demands of an ageing population.
Now a Leeds-based software firm has developed an AI tool that predicts which patients are most likely not to turn up to their appointments.
The company, TPP, uses machine learning models to analyse patterns of patient behaviour based on a range of factors.
Its algorithms analyse demographic details including age, gender and appointment booking history.
Younger male adults from a lower socioeconomic background in city areas are most likely to miss GP appointments, Dr Chris Bates, TPP's research and analytics director told the Mail on Sunday.
He said patients living closest to GP surgery were more like to skip a consultation than those living further away, though it is not clear why this is the case.
Armed with the information on the type of patient most likely to skip an appointment, GP surgeries can take action to nudge them into turning up, such as sending them additional text alerts just before the appointment is due and by giving them a telephone call.
The idea is not to try to deprive patients of an opportunity to see their doctor based on an AI profile but instead would encourage them to come by giving them additional reminders.
'There's been too much hype about AI in healthcare and not enough delivery – we're changing this,' he said.
'This is a brilliant exemplar of what machine learning will bring to healthcare, helping to tackle real problems for frontline teams and patients,' Dr. Bates added.
Missed appointments cost the NHS £216million a year – enough to pay 2,325 full-time doctors - according to its own figures.
The scale of the crisis prompted former Prime Minister Rishi Sunak to consider slapping a £10 fine on patients for each consultation they missed but the idea was quietly dropped.
A successful trial in Norfolk of the technology is now being rolled out across the country.
'Since we started using the report, we've seen our DNA (Did Not Attend) rate decrease by 30percent, equating to hundreds of appointments since the start of the pilot,' said Karen Bell, surgery manager at Trinity and Bowthorpe practice, which has over 11,000 patients.
Patients flagged as red or amber risk of not attending are sent text the night before their appointments, she added.
TPP is also using AI to help oncologists detect ovarian and other cancers in patients much earlier than was previously possible.
'The DNA algorithm is just the start – we have a suite of AI solutions about to launch, focussed on the NHS's key clinical and operations priorities,' Dr. Bates added.
