Matthew Fenech is an Affiliate Consultant for AI in Health at a think tank called Future Advocacy. He lives with his wife, son and pet dog in Cambridge, UK. A doctor by profession who specialised in diabetes, Matthew has come a long way since his medical training. In our interview, we discover his thoughts on the application of AI in healthcare and its potential if implemented in Malta.
You originally started out as a doctor and you are now an Affiliate Consultant for AI in Health at Future Advocacy in the UK. What inspired you to move into this area of healthcare?
My interest in working in policy is driven by my tendency to think about the big picture. This was really highlighted by my experience of looking after patients with diabetes and obesity. Although as doctors, nurses, or other healthcare professionals we can look after individual patients—and we can do a lot of good in that way—I came to realise that, if we want to bring about real change, then the only way to do that is by working on systems.
How can work on policy help bring about change?
As much as we advise people who have diabetes about what medication to take, or try to help obese people lose weight, the best thing to do would be to prevent these conditions in the first place. When I refer to education policy, I’m talking about a policy that includes useful teaching about nutrition. A planning policy that ensures there are enough safe, open space for our children to exercise in, and fiscal policy that ensures healthy food is a cheaper option for families than unhealthy food.
How can AI aid the healthcare system?
By thinking deeply about the big picture in this way, I came to realise that technology such as AI could completely change how we deliver healthcare. If we take diabetes as an example, this is a chronic condition that people live with 24/7—they can’t have a doctor or nurse with them all day and all night. With this type of tech, patients with diabetes might feel more confident that they can get advice about their condition any time they want. And when you remember that all of healthcare is changing such that we have an ageing population with more chronic diseases, you realise that this model might be applied to many other conditions too.
As part of my work with Future Advocacy we identified five main ways that AI can be used in health and care settings:
- Optimising ‘back end’ processes such as procurement and logistics
- Supporting research, for example by helping identify molecules that might be used as drugs in the pharmaceutical industry
- Augmenting and assisting healthcare professionals in their work, for example by triaging X-rays or CT scans according to the urgency
- Delivering decision support and insights directly to patients, for example via apps on their smartphones
- Allowing the analysis of population-level data, for example enabling the prediction of infectious disease outbreaks.
What about implementing these ways in Malta?
There’s no reason to think that AI couldn’t be used for all these reasons in Malta too. A major opportunity would be to use AI-powered automation to take on routine and repetitive tasks that doctors and nurses do. Such tasks include form-filling, thus freeing them up to spend more time with their patients—what famed American cardiologist Eric Topol called “the gift of time” in his recent review for NHS England.
Are there any risks that come with bringing about these AI systems?
I am very excited about the great opportunities this technology presents, but I am aware that AI presents risks too. Unsafe technology does more harm than good. Doctors wouldn’t prescribe a new drug to you without being sure that it’s safe, and that the benefits for you outweigh its risks. Technological tools like apps should be no different. In our work at Future Advocacy, we identified 10 areas of possible risk for these tools, including changing relationships in healthcare, and the difficulty with understanding how some of these very complex tools come up with their recommendations.
What would Malta need to do to implement AI locally in healthcare?
Firstly, these tools need vast quantities of data in order to be developed. In the case of healthcare AI, this may well be personal medical information, and the only way we can source this data ethically is to ensure that the patients who provide it fully understand the risks and benefits of allowing their data to be used to develop these tools. Secondly, as with anything that has the potential to cause huge change, we need to make sure that everyone is fully on board and has ‘bought in’, otherwise we run the risk of these changes being seen as negative disruption, rather than positive improvements. In the case of patients, we need to do a much better job of patient and public involvement in research and development in Malta. In the case of doctors and other healthcare professionals, and I can say this as an ex-practising clinician myself: we tend to be a conservative bunch! As doctors, we’re trained from the very beginning of medical school to be sceptical, so if we’re told to use a new intervention, we immediately ask: ‘how is this better than what I am already doing?’ It’s only by involving healthcare professionals in the development of these tools that we can win hearts and minds and ensure their widespread acceptance.
What is the gadget you can’t live without and why?
For a technology policy specialist, I’m quite boring with my gadgets—I don’t have a thousand apps on my phone or a voice assistant like Alexa or Google Home at home. But one thing that I absolutely can’t live without is the internet. I’ve been known to become very grumpy when the WiFi goes down!
EDIT: Since the original interview, Matthew has taken up a new role as Medical Safety Lead at Ada Health – A healthtech company that develops a symptom checker app.