Episode 49 AI with Dr Keith Grimes

Debbie: Hello, and welcome to Inflammatory with Debbie.

Katy: And Katy. And today we’re really happy to be joined by a former GP and now a digital health and innovation consultant, Dr. Keith Grimes. Welcome to the podcast.

Dr Keith Grimes: Yeah, thank you very much for having me. It’s lovely to be speaking to you both

Debbie: Thank you. So, can you just give us a slight background to yourself? Obviously, Katy said you were a GP and then you’re doing something slightly different now.

Dr Keith Grimes: Yeah, sure. Well, actually, the slightly different thing that I’m doing now is the thing that I always wanted to do. But yes, I am GP trained. I was a frontline GP for about 24 years. So, I graduated in 1996, did the usual doctor training jobs, became a general practitioner and worked in that and actually stopped practicing in 2024. But the big thing is that I love tech. That’s my first memories and I’ve always tried to blend the two when I was a student and when I was a doctor, did a lot of it within the NHS. In 2018, I started working with a company which was doing telehealth and a lot of AI work. And when I got there, I realized that combination of medicine and technology and safety and general geekery was all in one place and I was extremely happy. And when I left, I set up myself as a consultant, then I set up my company afterwards. So, I’m a very geeky doctor that cares about getting technology to help people.

Katy: Brilliant. Recently we did our own episode of what we think AI will bring to the world of inflammatory arthritis. But it’s fantastic to have a guest on who’s working in it day to day and knows a lot more than we do about it. So could we just start with what is AI and what does it actually mean in healthcare?

Dr Keith Grimes: Yeah, well, I mean, it was great that you did your podcast as well and covered that side of things. And, you know, I might have some expertise in medicine and AI, but I have nowhere near the expertise that you have in the conditions that you live and deal with. So what does AI mean? AI is very broad. mean, when you talk to your podcast, there was a lot of different areas that you spoke about. And within that, people think of AI as maybe the thing in front of them, like your chat GPT or something like that. Artificial intelligence is a term that goes back to 1950. So, it’s been around for a long time and there’s a lot of different definitions. But the definition that I like is essentially it’s when a machine does something that a human brain usually does. So, thinking, knowledge, task work, solving problems and so on. And since the 1950s, we’ve had lots of different technologies anywhere from expert systems where you have like flow charts and rule-based systems through things like machine learning, which is a bit like statistics on steroids. Where it learns to make predictions from this. Then you’ve got neural networks and deep learning, more recent additions where they kind of simulate how your brain works with like the nerves passing messages on. And once we’ve got to that and all the recent power of computing, but also the large amounts of data, we start getting some really powerful tools, things like image recognition, speech recognition, and most recently generative AI and large language models.

So, AI is a very broad term. It’s good to think of it as a sort of tool bag, you know, and even older technologies are still used every single day. You know, you don’t just give up using a screwdriver because you’ve got a power screwdriver somewhere else. And it’s all about which tool does something useful for the important problems you want to solve.

Debbie: No, that’s a really useful way of thinking of it, And I think what would be really good is and I think what Katy mentioned previously on the podcast was about the admin side, because you’ve been in a GP as well, you must know, there is so much admin that goes into seeing one patient. do you think that will be able to speed, say, referrals up or things like that, that actually impact us day to day?

Dr Keith Grimes: Yeah, absolutely. I mean, the wonderful thing about being a general practitioner, you know, I became a doctor because I everyone has a sort of vision of what the doctor would do. And I liked the idea that I was a person that could help with kind of any problem. And then medicine that broadly fits into one of two areas, emergency medicine or primary care. Actually, oddly enough, I worked in emergency care in primary care. But, you know, you get to see all sorts of things. So, you got a broad range of clinical tasks, your diagnosis, management, treatment, all that kind of stuff. You get caring for patients as well and learning about them too, which is the, you know, some people sort of think of the softer side, but it’s the core of medicine is that, you know, understanding people’s story and helping them, make sense of it and what to do. But yes, you’re right. There is an awful lot of administration and chewing around looking for information too. And when I was working as a GP, large chunks of my day would be seeing patients, of course, but I suppose bigger chunks of my day would be ordering and chasing up investigations, writing letters, making sure that the right letters went to the right services, all those kind of things.

Katy: Yeah. And that’s what the patient doesn’t see. So the patient can’t always appreciate what’s going on in the background.

Dr Keith Grimes: Yeah, but equally the doctors and nurses don’t see all the administrative work that the patients have to do. And yeah, and yeah, and.

Katy: Yes, very true. Yeah.

Debbie: That’s very true. Yes, I love that flip side of it. There’s a lot.

Dr Keith Grimes: Yeah, so everyone’s busy, everyone’s busy working away and flat out and then you come together at a certain time and and then yeah, we all know that sometimes that doesn’t work quite so well because you’ve got loads of great information way more than the doctor has very specific to you and the doctor has their own take on it and the nurse has their own take on it and yeah, so the consultation is often and particularly for people who have ongoing conditions is catching up on what’s happened, what’s happened elsewhere and so on, you know, because you get care in lots of different places.

Katy: Yeah. Yeah. And how, do you think it can actually help sort of how can technologies help with that bridging that gap between what the patient’s living day to day, what they actually talk about in an appointment and then matching that back to what their needs might be in the future.

Dr Keith Grimes: Yeah, so if you kind of like, it’s not all about kind of data and numbers and everything like that. But if we talk about that for a moment, I suppose that at simplest, you have data, you know, you’ve got all your notes and maybe people track their symptoms and their life. And, you know, as we all know, health is not just about the medical stuff, it’s about every aspect that contributes towards it. So so you may have all that information, some of it held in your head, some of it held on pieces of paper, some of it held in other apps and wearables, and so on. So you’ve got you as a patient have got all that information.

Katy: Yeah

Dr Keith Grimes: And then the healthcare system, doctor and the nurse have their own version of that too. And that will be in depth in many ways. I think the main problem is that you don’t also have access to all of that, if I’m honest. I think, you know, I’ve long been a proponent for patients having absolute access to the entirety of it. There are complexities in there, of course, but I think where we are now, we could be much further forward. If we made a decision that yes, the patients can have access to this because this is their story. This is their data. They need to understand this too. So yeah, you’ve got what you have and then you’ve got the doctor side, and we don’t have access to all of this. So, when they come together, there’s a little bit of sharing and knowing what the healthcare system wants to know is different to what you think they might want to know and vice versa. So that’s when that kind of choppiness happens. And if…If you don’t have many conditions, doesn’t matter quite so much because there’s less to share. If you’ve got something that’s rolling on, then that becomes increasingly complex.

Debbie: I suppose what also is very frustrating, I think, for some patients as well is when they say they have two different conditions, but they’re linked. Surely one hospital can see one thing and they should be able to see the other thing, but they don’t. it’s having then having to repeat your story the whole time. it’d be great if, just before we went into an appointment, there’d be one button where it says, when we last saw them, they had this and they have another condition. Instead of us having to repeat it all the time. That can take up half your time.

Katy: Quite often I’ll go into a consultant room and I’ll be like, surely you know about me by now.

Dr Keith Grimes: Yeah, I know. And actually, this is one of the great things about when general practice works well. And it really is under challenge that cradle to grave type relational medicine, where you have a named doctor and they’ve known you for a period of time. Now, sure, they still have to deal with all the paperwork and what’s coming in and so on. But everything else, like they know you, you know them. You don’t have to worry that those important moments in your life that maybe weren’t written down in the medical record are there, the highs and lows of life. And I’m sure when a person’s recounting their story, it’s like a mini trauma, isn’t it? I know I’ve got these things and now I have to run through it again and repeat how horrible things were, how great things were, whatever it was. And I’m trying to cram it all in and putting it in place. Now, this is, course, where artificial intelligence can come in. And I think people think of the NHS or any healthcare system as being like one big database. And so when I come in, it’s all available. I’m sorry, folks, it’s not, it’s all tucked away in different places.

Katy: Yep. And I imagine lots of different systems have been implemented at different times, just like in the business world, and those systems don’t know how to talk to each other.

Dr Keith Grimes: Yeah, and medicine is a little bit different to other areas because there’s been great progress in areas like finance and so on. And so people get a lot done with online banking. But I’m sorry, financial people, that’s just numbers. I know they’re important numbers, but a number is a number. And when you write down if a person has a condition of inflammatory arthritis, there’s lots of different coding systems. There can be nuance in there. And of course, just because one person has psoriatic arthritis or psoriatic arthropathy, it’s not the same as the next person with a psoriatic arthropathy. So even just down to knowing what you mean gets complicated. And then you’ve got it stored in lots of different places from different perspectives. They don’t talk to each other. The problems that patients have, or people have in getting access to their information and having to go from place to place and asking for it and subject access requests if they won’t give it and everything. It’s not quite like that for the healthcare side, but it’s not far off. It’s not automatically shared around. So incomplete knowledge on that side too.

Debbie: That’s fascinating because when we were talking to Hafi, the guest we had on when she was diagnosed when she was pregnant, if she had access to her blood results before she was pregnant, she would have had her diagnosis of rheumatoid arthritis. But because she didn’t have access to that data being a pharmacist, she’s been medically trained, but she didn’t have access to that. Obviously the NHS app is there, but we do need to have that access just so for us, you know, this is our result. So why can’t we not see them? I think this is

Katy: I think people, when you’ve got access to the NHS app, it’s very different depending on your location as to what you can actually see in your app because different, yeah, it really varies. Cause I think, cause I can’t see like my blood results, but I think, can you Debbie? I can’t remember. feel like we’ve had this chat, but yeah.

Dr Keith Grimes: Yeah, varies, doesn’t it?

Debbie: Yes, no, I can and I can see when my appointments are as well through my hospital. I know you can’t do that.

Katy: Yeah. So I can’t see that. I can literally book a GP appointment and that’s it. And do a repeat prescription.

Dr Keith Grimes: Yeah. And I know in the more tools that you have, the more that you can do. So kind of the same with AI and agents. So, we might come onto that, but you know, you get more tools, they can do more too. but people are very smart and understand these things. And so, I went through the same thing, my NHS app access. I can’t get access to my full notes and get part of them. I couldn’t order prescriptions. Now I can. And of course, on the other side, the doctors and nurses are working their bums off trying to juggle everything as well. I think everyone’s trying. it’s difficult to resolve, isn’t it? And also, that horrible thing where a person, there two things, you know, getting the data quickly, like your blood results, that can be good, but equally, there’s lots of examples of patients getting blood results back before the doctor does. And so they get the information, without any kind of context and that can be scary.

Katy: Context is so important, isn’t it? Because if you start Googling or using a chat GPT or something yourself to try and make to understand them, that agent can only, it’s very different to your GP or your consultant running you through what this actually

Dr Keith Grimes: Yeah, yeah, a chat agent or something, yeah. And I’ve done some great stuff. Last year I started, give lots of talks to people in the healthcare sector about AI. And then I met Jules from the patient care collaborative. They teach medical students and gotten really well with her and brought her along to a hackathon where we talk about AI and they get to build things. And I’m really, really interested in patients or people having access to AI tools to build their own solutions. Now there could be risks in that, but people can solve the problems that matter to them. And I remember speaking to her and a whole bunch of her colleagues And when I was speaking to them it struck me that the thing that a patient with a chronic condition or an ongoing condition might have when they speak to a doctor it’s a bit like when you’re speaking to AI in the sense that you have all the information or a lot of information, they’ve got some and you’re trying to make sure that they know your story well enough to have all the context they need to really help you. Now, oddly enough, that’s exactly what happens with AI. Like if you go into chat GPT and you just, or any of those things and say, you know, can you tell me this? It’s a little bit like going into a doctor who hasn’t got any notes and saying, what’s wrong with me? You know, and they’ll make assumptions and bring things in, and it might not be right. So, what you want to try and do is give more information. And the more information, the more correct information, the kind of the better the AI works. It was kind of the same with doctors and nurses as well. So, it struck me that parallel there and this sort of eternal challenge that we’re all trying to do whichever part of the healthcare encounter we’re on is we’re trying to come to a common understanding.

Debbie: AI isn’t going to go away. And some people can feel quite scared of it because it’s their data. we’ve seen what happens with, cyber-attacks on companies and, whether it would ever happen on the NHS, I think it’s happened in a few hospitals. So, people are worried about it. Healthcare professionals worried about that side as well when you talk to them.

Dr Keith Grimes (Curistica): Yeah, I think I’m very positive and optimistic about this, but I’m also not naive. I think there are things that people can rightly be anxious about too. So if we just talk about the healthcare professionals. Interestingly, when a doctor or nurse does their undergraduate study, even right now, there’s no compulsory element on AI or data science or any of this, you know,

Katy: Wow. Okay.

Dr Keith Grimes: So, if you go to medical students now can go through medical school for the next five years.

Katy: That’s mad, isn’t it? Really?

Dr Keith Grimes: and come out with no formal training in any of this. That’s the first thing to say. It’s crazy. That is not to say that they can’t get this training, but it’s often optional. And I’m campaigning to change that because, like doctor five years from now that could have cruised through without this. Imagine what it’s going to be like then. So, it kind of starts a little bit there. But yes, you’re right. Data protection is really important. Data is the raw material of artificial intelligence here and care in a lot of ways. So, we need to take care of it and data privacy and data security covered under the standards like GDPR and the data protection act. There’s a bit of mandatory training, but it still feels quite distant to people and maybe very distant for the general population too. So, we need to make sure the data is safe. And your general practitioner when you see them is the data controller. They’re the ones that bear the responsibility. And to give you a moment to think about this, if your GP is a partner in a practice, they’re part of an unlimited liability organization, essentially. And they’re responsible for the care of that data, which is why they’re so twitchy. Because if something goes wrong, they’re liable. And it’s an unlimited liability. Their houses can be at stake on this. So, if you wonder why doctors are tweaky about this, it’s because of that, right? Not many people appreciate that. Yeah. So, they’re more likely to be careful, you know, and that’s their anxiety. Is AI going to steal the data?

Debbie: Yeah, and I suppose in hospitals, it’s different because it’d be the hospital trust would be, the business entity that is then has the liability, not the GP, because obviously how GPs work, they have contracts with the NHS, but they don’t work directly for the NHS. I hadn’t thought of that before. But that’s, insightful. So, can you see the difference between those that work in primary care and those that work in secondary or tertiary care?

Dr Keith Grimes: Yeah, they can, and there are some practices which are configured differently as well. But, you know, that risk and maybe, you know, they’ve got so many things to do, they might not understand the full detail of this, you are going to default to being a bit more careful. Remember, also in bigger organizations and secondary care, they’ll often have a whole function within that organization that works hard on that. And so I think in secondary care, it won’t be quite dealt with, there’s still things that people have to do. And what they might say is like, you know, they’ve locked down this or I can’t do that. There’s a little bit more freedom within primary care, but equally kind of more anxiety. So yeah, when you’re dealing with AI, a good example would be, you know, like could a GP use chat GPT to put their patient notes into it, right? Now I won’t talk about whether it will do a good job of what’s being asked, but even there, when the data is put in and it’s got private information, special category data. If you put that just into chat GPT, one of two things can happen. Well, the first thing is that that data immediately gets moved off to the US because that’s where the models are. And data protection is very careful about when you move outside territories. But then the second thing about that is that they may use it to train. if that happened, like when you use chat GPT, you can turn that off and so on. But the worry is that if your personal information is then used to train a model, kind of baked into that model, and technically it can surface again from that model. So, you want to be careful. Now this applies to whether you’re a doctor or a patient. All right. So, you’ve to be careful yourself. And I would say I’m not going to stop you wanting to use it. think it’s amazing that people are using this, but it’s about using it safely. And so be careful about what you put into it. Maybe check the settings to make sure that the data that you’re putting in isn’t going to be used for training purposes.

Katy: Yes. Yeah. So we’ve got to be careful ourselves. If, yeah.

Debbie: That is a really good point because I must admit, I don’t ever put any of my personal information in there. But I suppose sometimes people may do or scan their appointment letter saying, can you put this into the plain language, please, because I don’t understand. But then obviously, there is your NHS number, your date of birth, your address. There is so much there. So please be so careful when you use these tools. I suppose, you were saying that doctors then don’t have like medical students don’t have this training. But the general public haven’t had the training either. We don’t know what is safe what can we put in because we don’t know where this data goes.

Dr Keith Grimes: So how do we solve this? It’s like, well, it’d be great if we taught it at school and everything, but you know, that’s going to be tricky. I what can I do about this? So other people like me, and it’s maybe speaking here and being able to put it across. And the last thing I want people to do is I’m never going to use this again. I’m really scared of it happening. There are sort of small things, which is being aware that the data, could lose control of it, so to speak. When you use whatever tool you use, just check and see what’s gonna happen with this data. And if you’re at all concerned, just don’t put that data in.

Debbie: No, that is brilliant advice. Because yeah, I think there does need to be a huge educational piece just to make people aware. And I think when I used to work years ago as a civil servant and don’t put in an email the things that you would never want to be made public. And I think that probably works quite well with this. Don’t put anything in AI, whether it’s chat, GPT, copilot, anything like that, that you wouldn’t want to be public.

Dr Keith Grimes: Yeah. Though the equivalent, I mean, that’s good. The equivalent I heard for social media is don’t post on social media anything you wouldn’t be happy to just shout on the street.

Katy: Essentially that’s what you’re doing with social media, isn’t it?

Dr Keith Grimes: Yeah. Yeah. But because you’re not on the street, you don’t think about it that way. But with data, I think for anyone just a wee bit of care. But then, of course, we just said earlier on, but the more data these models have, just not for trained models, but when you’re discussing things and so on, it’s more helpful. So, there are ways and means. in fact, funnily enough, earlier on this year in the US, it’s not in the UK yet, both OpenAI’s ChatGPT and Anthropic brought out ChatGPT Health and Anthropic4Health. And their offering there was that for the users there’d be a sort of special section and anything that was uploaded into that would be handled in a more sort of walled off way or a secure way. And you can connect your wearables and other sources. Now it’s not available over here. People are still like, well, can you really trust them and so on? But they’re recognizing that. And for me, I was involved in writing a report to government about generative AI and healthcare for economic growth. there’s a lot of really strong things about the UK and the NHS that we could use here. It would be great if we could have within the NHS app or something similar, AI model of some sort, where you could trust that if you were using it in here that the data was held that way. And that’s nice because there’s nearly 40 million users of the NHS app. That could be a good way in. It’s not as easy as just doing it, but that could be quite a good way for people to start using it that way.

Katy: Yeah. That could be really powerful and to be confident that your data is protected because it’s run by the NHS. Now that’s really interesting.

Dr Keith Grimes: You know, there’s, so many things that we could do. Some are complex, but some are simple. I mean, if I’m honest, AI or otherwise giving people access to their data is probably more, cause it helps with the AI, but it helps with everything else. but on that side, that might be one way in and, there’s the healthcare side where AI can help them do their jobs. But again, I’m really more interested in like, but what can people do to help themselves. And part of it is helping the healthcare system help them, but a lot of it is just helping themselves.

Katy: Yeah. And also things like, so lots of people have wearables, use different apps to track, whether it’s tracking their calories, their exercise. How can that then feed in to maybe help, with research people’s future health to try and, if the NHS knows more about us in terms of what our daily life looks like then surely that can help in pulling together health plans for people that’s not just about the medications, but also around building better lifestyles.

Dr Keith Grimes: Absolutely. I suppose you can talk about the data and then AI. And so just having that data. So, if we didn’t have AI where it is right now, just having that. if you had, let’s say it’s your step count, let’s keep it really simple. It’s just your step count. Yeah. And that’s a proxy for your activity levels. And of course, in people with ongoing conditions, inflammatory arthritis, there’ll be times when it is better and times when it is worse. And what you may see is two things. Number one, things are worse. Your step count goes down or your step count has been going up and then you suddenly have more symptoms. know, that’s its crudest.

Katy: Yeah, because it’s then working out what works for you and what doesn’t.

Dr Keith Grimes: Yeah. So, you have a lot of data and what do with it? Well, the first you sit with it in front of you and you try and work it out and that’s very hard, right? Or you can use all the versions of AI, but let’s say it’s Apple and you have Apple health and you can see it on a graph and you can kind of work it out. But what you really want is like, why can’t I just get that into my medical record? And then when I see the specialists, they know that and they consider it all. Then you think on the specialist side and you’re thinking, right. They already struggling to keep on top of all the official stuff. And then there’s this other stuff that we’re not entirely sure how significant it is. They’re like, I can’t be coping with this. And that of course is where AI comes in because AI has the ability to look across much larger pieces of data and make associations and predictions. So that would be the ideal world is that all the data is in place. have a very competent AI model that can say, you know, or Katy’s come in and we’ve got all our medical history and our medication and everything, but she’s also been tracking her diet and she’s also been tracking a step count and actually looking here, switched, her diet has changed in some way that we know from research might affect her inflammatory state and so we’ll discuss that. It’s that kind of situation. The bit before that, other than getting the data up there, is the research needed to turn it into something that we’re confident on.

Katy: Yes, yeah.

Dr Keith Grimes: Yeah, and so, there’s the UK gene banks, there’s our future health and so on contributing towards that can be really helpful participating in trials. And believe me, academics and researchers are desperate to try and recruit people to start looking at these areas and then finding like one narrow area. because you’re talking about rheumatoid arthritis, they’ve used AI for predicting biologics response, identifying new subtypes, x-rays and so on, you know.

Katy: Brilliant. Yeah, because that’s really important being able, like if you can get a patient and straight away you get them on the right medications, but based on their history and their biomarkers, then that’s an absolute huge win, isn’t it, for the future so that that person doesn’t have to go through like lots of different medications that don’t work. And then

Dr Keith Grimes: For my words, yeah. Trial and error.

Katy: I guess in the long run for the NHS, they’re not having to deal with the complications of joint damage later down the line because hopefully that person’s well looked after from the start

Dr Keith Grimes: Yeah. yeah, and actually that’s another. For Arthritis, inflammatory arthritis. I finished my training in 1996. I mean, we are in a different place now. I remember and still saw patients who had the consequences of long term, not only the inflammatory arthritis, but also the consequences of taking medications that were very broad. but now we have incredibly powerful drugs that if they work, by golly, they work. But if they don’t work by golly, they don’t work, you know, and you’re right. You want to get to the right place quickly. And I think a lot of effort is being done there. And there’s a lot of hope that AI can help with that. Not only to find new drugs and, know, match people very quickly and say, well, this person, because of the specific information and their, genetic subtypes and their pharmacogenomic and everything like this one’s much more likely to work than this, this, this helps. But then there’s also people tracking and then so we started them on something and these more subtle things that suggest that it’s not working not ones that take a while like my symptoms are improving or side effects are emerging but other measures can be picked up saying we’ve got you on this we wouldn’t be expecting changes just yet but there are some things happening that suggest that this might not work for you or not suit you it’s that so you’re not aware of it AI could be really good at that

Katy: And how far down the line do you think this sort of thing is likely to happen? Don’t want to put you on the spot, but I am.

Dr Keith Grimes: Well, there’s kind of two answers to that. No, I’m very happy to say that. Okay, so we’re in a really interesting time because the time it takes to go from like the drug idea to the real world is measured in 10, 15 type years. And there’s very good reasons for that about trials happening and so on. But if there’s something cool now and I had a condition, I’d want it I want it to be safe and everything. I want it right now so it’s traditionally take a long time. AI is being used a lot in the drug design discovery and development phase as well. identifying new compounds is happening much faster because of AI searching through them. Trials are being optimized so they can happen faster. get the information back faster. Some aspects of trials can be done in a synthetic way where you don’t actually have to have people, but you can actually run them in a simulated fashion. So, all of that can kind of press down. The time from the idea through to the drug being safe to be going out there. There’ll still be some other bits, but those bits that are done physically can be done in a much more efficient way. So that all comes down. Again, you’ve got new drugs that are kind of being tweaked as opposed to being discovered. You know, that accelerates everything. All the regulatory work, AI can help with that. So, the kind of paperwork can be done. So new drugs will take less time to come to you. Yeah, they still might take some time, but they’re going to take less time. So that’s part one.

Katy: Yeah.

Dr Keith Grimes: But part two is that AI can do a lot of the science itself. Now you’ve got examples of that. And AI can help with almost everything else as well. And so, I know coming into this, some people have waited a long time for a diagnosis. But getting to that definitive diagnosis sooner doesn’t require new drugs, just requires us understanding the information that we have or identifying those tests that could tell us sooner. AI could do that very quickly. a person may be labelled in one way that actually might have something slightly different. So, a solution might come in a bit sooner as well.

Katy: Yeah.

Dr Keith Grimes: And then at the same time, there are syndromic conditions, know, like those where people struggle to identify, issues like chronic fatigue and long COVID and everything, AI is being pointed directly at trying to solve the foundations of that too. And so we see great progress and across a whole bunch of disease classes where we’re now understanding a thing. And we thought it was this, but it’s actually this. then, and we can look back at drugs that we thought weren’t going to be useful all of a sudden, like, hang on a second, this ancient drug that we thought did one thing actually works really well over here. So, then we have lots of new science and everyone’s overwhelmed by all the new science. AI helps us sift through the new science. And in terms of my role in all of this, other than being giddily excited about it all, the work of my company is about, well, this is all great, but at some point someone has to test and make sure it’s safe. If we can make the proving that it’s safe, rigorous, but faster and easier. Well, that’s another part of the journey that gets kind of shrunk down and that we can watch to make sure that things are safe. And I’m a clinical safety officer. The role of a clinical safety officer is very kind of specific to technology. for a patient, if your question is, know, who’s the one who sort of can assure that this is all safe as a clinical safety officer?

Debbie: Well, that’s been absolutely fascinating. Thank you so much, Keith. think we’ve learned a lot, Katy, AI It’s not going to go away and how we embrace it safely is obviously very important, but it’s been absolutely, fascinating and so insightful.

Dr Keith Grimes: Thank you.

Debbie: Thank you. So please don’t forget to rate and follow the podcast wherever you get the podcast from. Please sign up to our regular newsletter. We are at inflammatoryarthritis.org where you’ll get all the information about the latest podcasts, any IA news, research and events. Please follow us on social media to continue this conversation. it’d be fascinating to hear what you think about AI and from our perspectives on our previous episode as well. We are on Blue Sky, Instagram, Facebook and LinkedIn. Please do follow, like and share. But until next week, Katy, it’s goodbye.

Katy: Goodbye.

In this insightful interview, Dr. Keith Grimes, a former GP and digital health innovator, explores the transformative potential of AI in healthcare. From data sharing and patient access to AI-driven diagnostics and drug discovery, discover how AI can revolutionize patient care while emphasizing safety and privacy.

 Key Topics Discussed

 What AI means in healthcare

The administrative burden in healthcare

Fragmented health records and data sharing

Patient data ownership and transparency

Personal health data and wearables

AI in diagnostics and drug discovery

Safety, privacy and data protection

Education and digital literacy in healthcare

Key Comments & Insights from Dr. Keith Grimes

• “AI is best thought of as a toolbox—different tools for different problems.”
• “Patients should ideally have access to their entire medical record, it’s their story.”
• “Healthcare data isn’t one big database; it’s stored in lots of separate systems.”
• “The more accurate information AI has, the better it can help but privacy must be protected.”
• “AI could help identify which medications will work best for a patient much earlier.”
• “We’re entering a time where science and medicine could move much faster.”

How to Connect with the Podcast

Website:
Inflammatory Arthritis – https://inflammatoryarthritis.org

Social Media:

• BlueSky, Instagram, Facebook, LinkedIn

Follow, like, and share to continue the conversation around inflammatory arthritis, healthcare innovation, and patient experiences.

Newsletter:
Sign up via the website to receive updates on:

Events and community news

New podcast episodes

Research developments