This site is intended for health professionals only


High-risk patients to be identified via AI to reduce A&E attendance and GP referrals

High-risk patients to be identified via AI to reduce A&E attendance and GP referrals

NHS England has enlisted the services of an AI company to help identify patients at high risk of A&E attendance and hospital referrals by GPs.

During a three-year programme, ‘advanced AI screening’ will be used identify patients in North East London with long-term conditions who are likely to need unplanned emergency care and offer them ‘clinical coaching’ to help prevent this. 

Health Navigator, the company overseeing the programme in North East London, told Pulse they do not expect it to ‘create additional workload in primary care overall’, although it could do so in the older patients.

Previous pilots and studies showed that using the AI tool for preventative care in this way has resulted in patients aged 80 and over using primary care more but the company said this was balanced with a reduction in GP usage among patients below 80.

UCLPartners, which has collaborated with Health Navigator and North East London ICB on this programme, said that a pilot in Staffordshire showed a 26% reduction in GP referrals to secondary care.

The same pilot, monitored via a randomised controlled trial, also resulted in a 34% reduction in emergency attendances and 25% reduction in hospital bed days. 

In North East London, use of this new technology is expected to result in 13,000 fewer A&E attendances annually, based on forecasting models.

NHS England, which has sponsored and funded the programme, said teams in NEL are ‘harnessing the power of AI technology’ to support patients with conditions such as asthma or diabetes and help prevent frequent A&E attendances.

Chief executive Amanda Pritchard said initiatives using AI like this ‘help provide more personalised care and must be central’ to the Government’s upcoming 10-year health plan.

The company Health Navigator will use the tool to identify patients in North East London, rather than NHS healthcare professionals, and the tool will be ‘trained on local healthcare data’. 

Clinical coaches, who will help those patients with ‘self-management’ techniques via phone conversations, will also be employed directly by the company. 

GPs will then be sent ‘summary clinical reports’ of the interactions so they can be ‘written into the patient record’ to ensure transparency across the healthcare system. 

‘Clinical coaching does not replace GPs or existing care plans but works alongside NHS services to support patients in managing their health,’ UCLPartners told Pulse. 

On the potential impact on GP services, Health Navigator said: ‘The previous studies have shown that patients aged ≥80 years had increased primary care usage, with those aged <80 years experiencing a reduction.

‘The majority of patients with complex medical needs are aged >80 years, which may indicate that this group experienced prior unmet clinical needs and challenges accessing their GP, or a lack of knowledge of available services.

‘The intervention potentially facilitates and resolves these issues, enabling these patients to have increased appropriate contact with their GP. This may be considered an “ideal” outcome as care of chronic disease is optimised in primary care.’

However, there was a ‘corresponding reduction in primary care contact’ for patients under 80, since the ‘collective burden of chronic disease is lower’.

Health Navigator told Pulse this suggests the use of AI and clinical coaching can support these patients ‘without the need for their GP’s involvement or secondary care referral’. 

NEL ICB chief medical officer Dr Paul Gilluley said that the 15 million people in England living with one or more long-term conditions account for ‘50% of all GP appointments’. 

‘This new approach represents a landmark step in harnessing technology for preventative care to better support these patients before they reach crisis points,’ he said. 

UCLPartners chief executive Dr Chris Laing said the new project in NEL ‘provides a template for how the NHS can use modern technology to deliver predictive, proactive and preventive care’.

He added: ‘Our collaboration with Health Navigator and NHS North East London will not only enhance the lives of at-risk patients but will also relieve critical pressure on our healthcare system too, aligning with the government’s prevention priority.’

A recent report by Google claimed that greater use of AI could provide an extra 3.7 million GP appointments each week within 10 years.

And a study published in September found that one fifth of GPs are already using AI in clinical practice, with ChatGPT the most popular tool.

Pulse October survey

Take our April 2025 survey to potentially win £200 worth of tokens

Pulse October survey

          

READERS' COMMENTS [6]

Please note, only GPs are permitted to add comments to articles

Simmering Frog 13 December, 2024 4:22 pm

That’s a sharing agreement I’d not sign.

Dr Who 13 December, 2024 4:38 pm

No. Without funding and resources

SUBHASH BHATT 13 December, 2024 6:07 pm

Allow primary care to mange 24/7 365 service. Till this is not funded and done , all these AI things will never work to reduce an and e attendance..

So the bird flew away 13 December, 2024 6:41 pm

Don’t fall for the same old trick of snake oil salesmen this time pushing a crap AI product for their own profit-seeking in a further creeping privatisation. Everything claimed in this article re AI improving care and reducing attendances is not true, very poor science and so GPs shouldn’t accept the exposure to liability by engaging with it. Just fund real GPs.

David Mummery 15 December, 2024 10:58 am

Agree with STBFA

Simon Gilbert 17 December, 2024 7:58 am

The was an interesting Nuffield Trust conference about 15 years ago on this topic (admission avoidance). The researcher showed that every single intervention on highest risk patients led to a fall in A/e and hospital usage (and spend) over the next 12 month. They then revealed the ‘no intervention’ results which also demonstrated the same trend – reversion to the mean.

The description above is that the North East London scheme does include randomisation, so hopefully has a more rigorous outcome than an NHS ‘Pilot’, but it would be interesting to see the reduction in hospital admissions and overall (including primary care) resource usage above the control group in the headline figures also.

Pulse October survey

Take our April 2025 survey to potentially win £200 worth of tokens

Pulse October survey