FeaturesArtificial intelligence in healthcare: friend or foe?

Artificial intelligence in healthcare: friend or foe?

Could AI enhance your practice and give you more time to spend with clients? Or are there downsides? Journalist JO WATERS reports on the role of AI in the NHS right now, its wider potential – and what it all means for you.

Artificial Intelligence (AI) – the use of digital technologies to create systems capable of performing tasks commonly thought to require human intelligence – has arrived. It has been accompanied both by doomsday prediction stories and amazing case studies showing how it can help. While it may feel alien or scary to many, you may already be using some form of AI – such as personal assistant features on your smartphone (think Siri or Google assistant) or a smart thermostat to control your central heating.

When it comes to delivering healthcare, traditionally of course, that’s always been about face-to-face interactions between health professionals and clients, building relationships and trust. So, is there a place for AI in healthcare, and in community practice in particular? Could AI really affect job security and, overall, is it something to fear or embrace?

The government certainly has big plans for using AI in the NHS, announcing a £21 million fund to improve diagnostics in June, on top of £123 million already invested in 86 new AI technologies for supporting stroke diagnoses, screening, cardiovascular monitoring and managing more conditions at home (Health Education England (HHE), 2022). In addition, some £250 million has also been invested in the NHS AI Lab to accelerate AI adoption.


Professor Aldo Faisal, who leads Imperial College London’s UK Research and Innovation Centre for Doctoral Training in AI for Healthcare, emphasises that AI is not about replacing people with machines. ‘The purpose of AI for healthcare is to help reduce the unmet need for healthcare,’ he says. ‘It’s about helping healthcare workers deal with the onslaught of work and the rising demand caused by the pressures of the ageing population, the pandemic and elements of climate change.’

Professor Faisal says that without assistance from AI the NHS will be unable to keep up with rising demands of an ageing population and economic pressures. ‘AI in healthcare will focus on finding solutions, taking care of routine tasks, alerting people earlier and helping them to focus on the right things. Rather than AI replace people, what we are seeing is an increased need for the human element in interfacing between a person and AI.

‘AI is not about robots walking down the street and coming to your house for a home visit, rather it will be available on your smartphone or your wearables, such as an Apple Watch, that are able to support you and guide you 24/7. We have research projects where we can predict if someone is going to have a second stroke within 24 hours via wearables, and a current database of health and social care information which can be accessed on a dashboard by GPs and other community staff.’

Rather than AI replace people, what we are seeing as an increased need for the human element in interfacing between a person and AI

Professor Faisal is working on a research project with school nurses (SNs) in north west London to support children with eating disorders through an AI powered app, while they await psychiatric assessments. He said he would encourage community practitioners to get involved in similar projects as their contribution and insights are valued.


An HEE survey (HEE 2022) found that 77% of 56 AI technologies were being deployed in hospitals, with 23% in primary care settings and 7% in community care.

1: ‘Emotionally intelligent’
AI helps with mental health
Ross O’Brien is managing director of Wysa,
a conversational ’emotionally intelligent’ AI
system for mental healthcare. He says Wysa
is already being used in a pilot scheme by SNs
in the Midlands and is now available through
those Scottish health boards that commission it.
It has also been selected as an NHS accelerator
programme project.
‘As I’ve previously worked in the NHS as a
commissioner and provider in mental health
services, I was well aware of how long it can
take for young people to be seen by Child and
Adolescent Mental Health Services and how their
condition could deteriorate during that wait.
I know that SNs worry about this too and the
support they need.
‘Wysa is a digital AI tool to help fill this gap which
is free to children in areas where the health
authority has commissioned the service. It can
give highly personalised support, based on a
huge database of clinical information and using
natural language understanding and processing
to hold the conversations with people so it
doesn’t sound clunky or robotic.’
Ross says that some young people, particularly
those who are neurodiverse actually prefer
talking to AI, rather than a human as it is non-
stigmatising, and the app is one of the highest
rated apps in the ORCHA Health App Library, for
its efficacy and risk management.

AI in healthcare is most mature when it comes to medical imaging, says Pritesh Mistry, digital technologies fellow at the King’s Fund (an independent health think tank). ‘We’re seeing the use of AI in medical imaging to support staff to identify and measure features in medical images,’ says Pritesh.

‘AI can do this analysis and provide the results to a qualified member of staff to check and validate. In doing this, AI can help track and quantify complex changing aspects of a person’s health – in changes in moles and wounds, for example.’

‘Other forms of AI are being piloted and tested in the NHS and social care. For example, sound and motion sensors with AI to alert staff if a patient is moving in the night and at risk of a fall (NHS South West London, 2023), or analysing elective surgery waiting lists to help with prioritisation, or identifying patients who may benefit from support to improve their health prior to surgery.’

Various trials are happening in countries around the globe, including the UK, that focus on how AI can support screening for cancers, such as breast and prostate, with promising early results (Community Practitioner reported on initiatives in Wales and Sweden in the September/October 2023 issue, for example).

Pritesh says the advantages of AI in the scenarios he describes include reducing workload and some workflows or task time, providing better information, improving quality of care and changing how resources are allocated.

He sounds a note of caution, however, adding: ‘AI also has limitations. It is another tool: it’s a powerful tool but it can’t do everything. The important thing is how to bring people and AI together to change how care services are delivered while maintaining the important principles, including human contact.’


Could an AI powered robot or chatbot really assist you in consultations, prompt you to ask questions and then write up your case notes? Would you (and your clients) trust AI and do you have the digital skills and equipment to use it?

Thurrock Council, bordering east London
and Essex, worked with Solutions4Health, a
developer using AI, to produce an interactive
app with local healthcare professionals to
support families of 0 to five-year-olds on
breast and bottle feeding, immunisations,
sleep, play, mental health, diet and oral health.
The Ask Teddi app, which was launched when
Covid-19 restrictions were in place and parents
felt worried and isolated, was the result.

An evaluation, carried out by Swansea and
Essex universities, found the app was viewed
as a valuable, accessible and a useful support
tool, providing local information for new
parents by the professionals supporting them
(Local Government Association, 2022).

Keshav Sankla, director of AI and innovation
at Solutions4Health says: ‘Initially there’s
probably, [and] quite naturally, been some
fear about AI from stakeholders when they’ve
always done things in a particular way. But
involving HVs right from the start in building
the app meant they felt more in control of
how the app we developed [actually] worked,
the language it used and the quality of the
information it gave.

‘The moment they saw how it worked, they
changed their minds about the use of AI
and saw it could take some of their workload
burden away from them. Instead of having to
answer questions from parents at all times of
the day, parents could ask Teddi instead.’

Keshav says the AI app is now being used
by HVs in Slough in Berkshire and Barnet in
north London, while other apps have been
developed in a similar way with SNs for parents
of older children and teenagers.

‘AI has huge potential to reduce workload
for HVs and SNs, as well as improve access to
health information for parents, and if they get
involved right at the start they can end up with
a really useful tool,’ Keshav adds.

‘These are all difficult questions – AI has great potential to support CPs in their work, but there is a lot of fear and unknowns surrounding it at the moment,’ says Annie Hair, a senior health visitor in Glasgow who chairs the CPHVA occupational and professional group.

‘But the public absolutely still wants the human touch in healthcare – this was very clear from the evidence gathered during the Covid-19 pandemic.’

Pritesh says that while the impact of AI on CPs depends on the tools that are available, it does have the potential to improve how they communicate with clients.

‘It could also help them to reduce workload by providing summarised information that’s timely and relevant for the interaction, while also capturing information for record keeping – which would be logged after confirmation from the member of staff,’ says Pritesh.

‘AI could draw upon the various siloed data held within public services to identify support needs, high-risk individuals or potential safeguarding needs. It could help with assessment and diagnosis, enabling staff to do more or have more confidence in their decision making.’

Adds Annie: ‘It’s our job to build up trust and relationships with families over time and I don’t see how AI can do that part. In Scotland, our Healthy Child Programme is based around 11 contacts with families, and we must make in-depth assessments based on our knowledge, qualifications and experience. It’s only by doing those in-depth consultations that you can identify need and safeguard children in a child protection context.

‘I can see that AI might be useful in some contexts such as mass population screening programmes, but I don’t think it can replicate those sorts of in-depth consultations we do so well.’

Perhaps some of the current AI projects in community practice will provide insights into how it might be used to help in the future. These projects include a conversational ’emotionally intelligent’ AI system for mental healthcare being used by SNs in a pilot scheme (see box 1) and an interactive app to support families of 0 to five-year-olds in a number of areas including breast and bottle feeding, immunisations, mental health, and
diet (see box 2).


The Topol review (Health Education England, 2019) identified a need to equip NHS staff with the skills to use cutting edge AI technologies.

Pritesh says: ‘Broadly speaking, the skills needed will depend upon the technology being used and the role of the individual. There are parallels to existing complex technologies like MRI scanners where staff don’t need to know the details of the machinery in minutiae – but they need to be able to use the tool and have confidence in what it’s able to do and not do.’

The AI and digital healthcare technologies capability framework (Health Education England, 2023) offers competency statements on the AI skills that staff are likely to need.

Some young people, particularly those who are neurodiverse actually prefer talking to AI, rather than a human
as it is non-stigmatising

However, Pritesh stresses that improving the general baseline digital skills of staff – even before considering advanced skills that AI use might require – has been identified as something that requires urgent action. A recent House of Commons report, Digital transformation in the NHS, (House of Commons Health and Social Care Committee, 2023) stated that: ‘The government must also address the long-recognised digital skills gaps in the NHS and social care workforces. We recommend the Department of Health and Social Care work with NHS England and Skills for Care to design training in which digital skills are embedded throughout.’

A YouGov survey of 4000
adults and 1000 NHS staff,
commissioned by The Health
Foundation (2021) found that
both groups harbour similar
reservations over whether
automation and AI in healthcare
is a good thing. While 40% of
the public said they felt more
positive than negative about AI
in healthcare, 37% were more
negative than positive. When it
came to NHS staff, 40% were
more positive than negative
about AI and 38% said they were
more negative than positive.


Professor Sir John Bell, Regius Chair of Medicine at the University of Oxford and an adviser on Sir Patrick Vallance’s review into the regulation of AI and other digital technologies, recently told a podcast that AI is going to be a ‘major disruptor in health’ (The Health Foundation, 2023).

Sir John said: ‘We’re closing in on a situation where [AI’s] going to be probably used pretty substantially across a whole variety of healthcare domains. We’re not that far away from getting the answer to… simple questions that would normally come from a healthcare professional – provided to people through large language models like ChatGPT… – that would provide us with a pretty accurate view of what symptoms might be related to and what [we] may need to do after that.’

With the significant leap in AI in 2022 due to the launch of ChatGPT, Pritesh says it’s difficult to predict whether the AI field could become stagnant again – or, indeed, leap forward.

‘Looking at the AI capabilities that are already demonstrated to be feasible – such as image analysis – we can anticipate these will improve, becoming more accurate and faster, which will have more applications and impact. With any transformative technology it will impact staff, so it is likely there will be changes to staff roles and responsibilities.

Improving the general baseline digital skills of staff – even before considering advanced skills that AI use might
require – has been identified as something
that requires urgent action

‘This might affect staff in medical imaging services in the short-term but expand to other areas as AI capabilities and evidence continue to develop,’ Pritesh adds. Experts such as Professor Faisal are clear that AI in healthcare is not about replacing people. Yet financial institutions have been less upbeat. In March, the investment bank Goldman Sachs estimated that 300 million jobs worldwide will be lost or diminished by AI, with healthcare being one of the many industries affected. (Goldman Sachs, 2023). And professional services network PwC (formerly Pricewaterhouse Coopers) has predicted that although AI has the potential to contribute to global GDP by US$15 trillion by 2030, this ‘will come at great human cost’. However, PwC analysts also predicted that healthcare may emerge relatively unscathed as it has a ‘greater reliance on social skills and the human touch’ (PwC, 2023).


If AI was used to cut jobs, it would be short-term thinking, says Loy Lobo, past president of Royal Society of Medicine’s Digital Health Council and a visiting lecturer at Imperial College Business School. Loy says demand in healthcare is going to more than double in the next 20 years, fuelled by rises in the ageing population and the prevalence of long-term conditions.

Health Education England
publishes roadmap into use of AI
in the NHS
The NHS AI Lab
The Health Foundation
podcast AI in health care:
hope or hype?
GOSH: What do young people
want from AI in healthcare?
AI network at Imperial
College London

Loy says the way forward is for CPs to work alongside AI assistants, something that already happens in areas such as customer services.

‘If you want a more consistent service of the highest quality, wouldn’t it be best to have AI assistants in your consultations?’ asks Loy. ‘The assistant can gather all the information required for your job automatically and prompt you to ask questions or take action in real time during the consultation. At the end of the interaction, it can write up your notes for you. This technology already exists in other industries.’

Unite (health) lead professional officer Ethel Rodrigues says: ‘AI is set to be a tool that can be valuable, but will never replace the wide range of experiences, creativity, intuitive decision-making processes and touch human beings bring. It is very important that this tool is used responsibly and ethically. Data protection must not be compromised. Organisations using AI should be prepared to invest in digital training and development of the workforce.’

Loy concludes that more healthcare staff will be needed, not fewer, while those who acquire the skills to work effectively with AI could earn higher wages.

Have you come across AI in your work, or are you surprised by some of the CP AI success stories here? Do you think it will be a force for good or have strong reservations? Whatever your view, and whatever your experience, we want to hear from you. Email editor Aviva Attias aviva@communitypractitioner.co.uk


Community Practitioner. (2023) Professional update. Community Practitioner: 10. 

Community Practitioner. (2023) Research Digest. Community Practitioner: 39. 

Health Education England. (2023) AI and Digital Healthcare Technologies Capability framework published. See: https://www.hee.nhs.uk/news-blogs-events/news/ai-digital-healthcare-technologies-capability-framework-published (accessed 18 October 2023). 

Health Education England. (2022) HEE publishes roadmap into use of AI in the NHS. See: hee.nhs.uk/news-blogs-events/news/health-education-england-publishes-roadmap-use-ai-nhs (accessed 18 October 2023). 

Health Education England. (2019) The Topol Review: Preparing the healthcare workforce to deliver the digital future. See: topol.hee.nhs.uk/ (accessed 18 October 2023). 

House of Commons Health and Social Care Committee. (2023) Digital Transformation in the NHS. See: committees.parliament.uk/publications/40637/documents/198145/default/ (accessed 18 October 2023). 

Goldman Sachs. (2023) The Potentially Large Effects of Artificial Intelligence on Economic Growth. See: key4biz.it/wp-content/uploads/2023/03/Global-Economics-Analyst_-The-Potentially-Large-Effects-of-Artificial-Intelligence-on-Economic-Growth-Briggs_Kodnani.pdf (accessed 18 October 2023). 

Local Government Association. (2022) Thurrock. Early Years artificial intelligence powered wellbeing support. See: local.gov.uk/case-studies/thurrock-early-years-artificial-intelligence-powered-wellbeing-support (accessed 18 October 2023). 

NHS South West London. (2023) AI in care homes to reduce falls and hospital stays. See: southwestlondon.icb.nhs.uk/news/ai-in-care-homes-to-reduce-falls-and-hospital-stays/ (accessed 18 October 2023). 

PwC. (2023). How will automation impact jobs? See: pwc.co.uk/services/economics/insights/the-impact-of-automation-on-jobs.html (accessed 18 October 2023). 

The Health Foundation. (2023) AI in health care: hope or hype? See: health.org.uk/news-and-comment/podcast/ai-in-health-care-hope-or-hype-with-professor-sir-john-bell-and-dr-axel-heitmueller (accessed 18 October 2023). 

The Health Foundation. (2021) Switched On: How do we get the best out of automation and AI in health care? See: health.org.uk/publications/reports/switched-on (accessed 18 October 2023).

Image | Shutterstock


Latest articles

More articles