How will AI technologies be adopted in health and care?

Dr Rishi Das-Gupta

Chief Executive, Health Innovation Network in south London, part of the AHSN Network, and Digital Transformation and Innovation Lead, AHSN Network

While adoption of AI in healthcare is undeniably complex, AI has the potential to tackle some of the NHS’s most pressing issues.

There has been an explosion of interest in artificial intelligence (AI), in part due to large language models, such as ChatGPT. As the innovation arm of the NHS, the AHSN Network is engaging in the early adoption of AI in healthcare.

Although the need for transparency about the algorithms used, the datasets they’re modelled on and how the technology performs in real life — with good reason — add to the complexity of adopting innovation into the NHS, there are several examples of where AI is already having an enormous impact and delivering benefits for clinicians and patients. 

Increasing clinical capacity with AVT  

An innovation identified by Network colleagues in south London uses ambient voice technology (AVT) to capture patient-clinician interactions in real time, automatically creating a summary of key points that can be added to patient records. If utilised, with consent, across the 1.75 million daily patient appointments, this would save thousands of clinical hours and dramatically change the capacity of the NHS.  

Identifying eligible patients with imaging decision support 

Increasing the number of patients receiving mechanical thrombectomy (MT), where a blood clot is removed from the artery through a catheter, is a key NHS priority. Network colleagues in Oxford are evaluating AI brain imaging technology e-Stroke, for speedier detection of stroke patients suitable for thrombectomy and have, so far, supported the spread of the technology across 73 hospitals. Here, detection rates have risen from 1.5% to 8%, compared to the overall national figure of 2.9%, leading to more rapid lifesaving treatment of stroke patients and resulting in the prevention of stroke-related disability and death.  

If utilised, with consent, across the 1.75 million
daily patient appointments, this would
save thousands of clinical hours.

Prioritising those most in need with risk assessment tools 

A trial of a risk-stratification system, created by C2-Ai, one of the AI specialist companies supported by our Network colleagues in the northwest of England, found it accurately predicted the risk of mortality and complications for patients listed for planned surgery and reduced emergency admissions by 8%. It also helps hospitals identify up to 90% more harm than current processes.  

Bringing AI into the NHS at scale  

Whether it’s tackling long waits for treatment, improving diagnostic accuracy or saving clinical time, we are excited by the potential that AI provides to transform the experience of clinicians and patients across health and care systems — and we will continue to support and drive the identification, evaluation and adoption of these innovations at scale.