Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

Um It's lovely that you're able to come back. So maybe, maybe with that, I will introduce Sy to everyone. So, Cii um Shaine did her radiology training in 2010 in George's. Um and then she did some subspecialty, pediatric body imaging fellowships um both in Great Ormond Street uh in London and the hospital for sick Children in Toronto. She's done a phd um funded by the MRC and the Royal College of Radiologists entitled to novel applications and refinements of ultrasound techniques and perinatal and infant death. And she's currently an academic pediatric radiology consultant. She's based in Great Ormond Street. She's an honorary consultant at Saint George's. She's got an N Hr Advanced fellowship, a very prestigious award, researching ways in which A I can be used to help radiologists better diagnose diseases. And if you've got an opportunity, do look at Suzie's website, it's the most amazing website I've ever seen. It's phenomenal so well done. I think we can all learn a huge amount from that Suzie. And then we're looking forward to hearing from what you're going to tell us today, which I think is entitled, Winning Hearts and Minds Change Management, relating to A I implementation. Thank you so much, Professor Gallagher for such a kind introduction and good evening everyone. So um this talk is something a little bit different. It's on A I, but I'm not gonna talk to you about rock curves or confusion matrices. I'm not gonna talk to you about hyper parameter or anything like that. Um I think now we're seeing that A I is coming more and more into clinical practice. It used to be something quite theoretical. And I think many of you who have probably been around the block and gone to a few radiology conferences are now seeing more and more vendors talking about implementing, deploying the A I. You may have heard of the A I deployment fund recently where the government was putting up 21 million lbs to try to get A I into clinical practice. And so now the conversation has moved very, very much from something that was very technical to something that's actually on the shop floor for all of us to start using. And I think sometimes when we think about that, we forget about the human factors involved. So this talk is a little bit different. I'm not going to quote numbers, facts, figures or any sort of stats to you, but it's a bit more of a qualitative exploratory, I guess social science kind of view about A I and change management and what we need to be thinking about when we're trying to deploy and implement new technology onto the shop floor. But with a sort of a I slant um to it. So let's begin um in a really good book that I read recently called Machine Platform Crowd. The authors thought of A I as the new electricity, that's what they kept calling it. And why did they say that? I think what they were trying to do was to draw the analogy of back in the day in the 18 100s when electricity came into existence, people didn't really know what to do with it. They knew it was powerful, they knew it was really good. They knew that it could change the way they did things, but they weren't really sure the best way of integrating it into their practice. If you like very similar to, I think how we see A I right now. So if you look on the, the screen on the left side, you'll see a steam powered factory where they had these big turbines that were being turned to move various machinery in the factories. But it was all steam powered and you had to have a huge extra outhouse built just for this big turbine and the steam engine to, you know, run these um big machines that you had on the shop floor. And when electricity came into power, people didn't really think of electricity as something new. They thought it's a replacement for these steam um turbines. So let's just bolt on another department, but put electricity in there. But they weren't thinking, you know, now we could have plug sockets, we can bring electricity to where it's needed. We can use it in different ways. We can use it for the light, we can use it to power heating, we can use it for the machinery. It's not just one thing for one purpose. That's a bolton. It's fully integrated in all aspects of the factory nowadays. But people didn't see it as that. They saw it as a bolt on. So really, the authors were trying to encourage us to say, let's not think of A I as a bolton or as an add on or as a replacement for something we already have. But think of new and interesting ways it can integrate into everything that we do. And I love this quote that Wayne Dwyer says, you know, when you change the way you look at things, the things you look at change. So do we really have to be thinking bigger about A I and not thinking of it as just, oh, it's going to diagnose something on a chest X ray. You know, what else can it use to change everything that we do and power? Most of the things that are the processes we don't like using it doing, you know, and um so when you think about all the ways A I can be used, um you can think about the whole patient journey and look at how in every single part of that journey, you can use A I to enhance or supercharge what we're doing already. So this is from a paper I've previously published where we looked at, you know, we can use A I as clinical decision support tools, generating differential diagnoses, generating best use of imaging, using it to power the IR for a criteria perhaps then booking a patient in for a time when they can also get a clinic appointment at the same time. Rather than having to have a very clunky system where you've got to go to one admin person, to another admin person all the way through to image acquisition, better enhanced images, few emotion artifacts, good post processing techniques and then you know down to reporting communicating results and maybe even something about prediction at the bottom. So there are many, many ways in which A I can change the whole of our health care system. And I think you need to start with that in mind because once you integrate one A I, it's not going to be long before you're going to start trying to integrate it into many many facets of your health care pathway. And so trying to start with thinking big thinking how we're going to scale this up rather than just how do I implement this one tool I think will really help um make your department even more A I enhanced for the future So that's the first lesson. And I think all of this in terms of change management comes to how do you implement A I as a strategy? And um one of the er things that the health Economic World Forum suggested was that you should have a strategy that defines the vision. So start with the vision for how you want your organization to be and then break that down into manage more steps. I know that sounds really obvious. But a lot of people start with the vision of just implementing one tool. And I think if you do that, you're never going to grow or expand or become the big hard hitting, you know, A I center you wanna be. So um that's just something I wanna instill from the start before talking a bit more about the nuts and bolts. And I think a lot of people maybe not so much now, but I know back in the day when I first gave talks um about A I, people thought this was a technical thing. This was for people who are interested in informatics, people who are very, you know, computery and who were coding people and that A I wasn't for them. But when you think about implementation and change, you really don't need coding techie people involved, they can be but you don't need them. Just you need good people, people, right? It's um A I implementation or usage in an organization is not just a technical issue. It's a leadership and a management issue. And from that point of view, we need people who can speak to people, bring people on board, get them to change the way of thinking, get them to change the way of doing things. And we don't just need techie people for that. So I feel that everybody here has a role to play. So please don't feel alienated if you're starting out and you're on this call and you're thinking, oh, I don't think a I's for me or it's for someone else because everybody has a role to play in this. So as I said, just implementing technology isn't the only change when you implement technology, clinical care pathways need to adapt. People need to change the way you do. Things need to change your uh standard operating procedures, need to change. What happens when a mistake happens, needs to change as well. So really do think of this as not just one thing, like not just the tech, but a whole care pathway that needs to adapt. And also when you think about a whole care pathway adapting, you've also got to think of many unintended consequences as well. And you've also got to think about what change are you trying to implement and why do you want that to happen? So for example, I'm going to give you an example and um tell you how this could possibly make things better or worse for your department. Just imagine you decide to implement an A I in your department that looks at chest X rays because you want to save money for your department. So let's say you wanna save money outsourcing. It could be CT S. But in this example, let's say it's chest X rays. For simplicity, you've got a huge backlog of chest x rays that haven't been reported for months, maybe years. OK? And um as a symbolic gesture, the green X rays here are meant to be the normal ones. The yellow ones are the ones that are uh borderline, probably need some sort of follow up and the ones that have sort sort of black on them are bad. OK? And gray is another bad one, let's say. So there's a jumble in there, some normal, some needing follow up that and some that are definitely definitely abnormal that need urgent care. Um You decide that you're going to implement an A I that's going to get rid of all the normals and then you're only going to send all the um abnormals to some sort of teleradiology company or expert reporting place. So you screen out all of those normals with the A I and then people in house can report them in their spare time. There's no urgency, they're probably fine, but they need a report on them and then you will send the complex ones out. Um But then you know how long is it before the teleradiology, people realize that they're only getting the complex ones, they're only getting the difficult ones. It takes them more time. They're all about efficiency and time as well. And they start increasing the price and saying actually now we're going to give you a different price, one big price for the abnormals and a cheap price for the uh normals or, you know, many of these teleradiology companies actually use A I themselves. And so they're going to screen and go, hey, this is weird. We base our business model on a certain number of normals being in this um subset that we get from hospitals and we're not getting any normals. Now, from your hospital, what's going on there? This isn't part of the plan. Let's go back and revise that. And soon you think your cost cutting saving measure soon becomes actually um something even more expensive because the prices change or the market changes. Um Similarly, I've heard of another situation where someone's implemented an A I without much thought about the downstream consequences where let's say they wanted um it was for intracranial bleeding. So they had an A I that they implemented, that would prioritize anything with a bleed to the very top of their reporting pile. So all the normals would get shifted down to the bottom of their CT head on cold trauma, whatever sort of reporting list. So they implemented that and told everyone we've got this great new tool. It's going to pick up hemorrhage, you still have to report it, but it'll prioritize it to the top. They didn't tell their reporters that it wasn't going to look at any pediatric brains and that it wasn't trained on pediatric brains and that it wouldn't analyze anybody under the age of 18. So then they implement this tool on call. It's all gray, all of the ones with suspected bleeds get shifted to the top. But the people who were reporting thought that it was looking at the pediatric brains and that none of the pediatric brains had bleeds on them. So they didn't report any of them cos they thought they were all normal. And in actual fact, some of the pediatric brains did have bleeds on them, but they weren't flagged up by the A I because it wasn't trained to do that. And that wasn't communicated. And neither was there a warning to say this is not analyzed by air. So you've got to kind of think about what change are you implementing? What are the downstream effects? What impact does that have on the whole care pathway? Not just on your reporters or on the patients and what kind of uh mitigations are you going to have in order to make sure that harm doesn't come to people in your department and your patients and also you know, their families as well. So there are many reasons why A I initiatives will fail and this is outside the remit of this talk. But I just thought I'd mention it just so that you're aware, but you know, capability gaps, poor quality data, technology foundations, not in place, poor governance, lack of understanding about financing, lack of understanding about the management. Um I'm going to assume that when you implement your A I in your clinical practice, you've done your due diligence and all of this has been taken into account and you've looked at the evidence and you know what the use case is and I'm just going to talk a bit more now about the human factors and how to get people round to your way of thinking rather than, and any of these nitty gritty bits. But just to say they are all extremely important, which is why I bring them up and I mention them and I give you food for thought just um just for now, but do do bear these in mind. But for the purposes of this talk, we're going to be talking a little bit about um change management and how to get people around to your way of thinking. And I think the reason, well, the big reason I think why people don't want to change and don't want to come round to people's way of thinking is because they've been let down so many times before. There are so many times where people have said we're going to implement something. It's going to be amazing. It's going to be fantastic. You're going to love it. You're not going to notice it's there. I mean, how many times has like a new ri system or a new pack system been promised to you and then it's not always turned out the way you think. And it's always been a bit painful and maybe it's been better in the end, but it's always been a bit painful. And I think as a species we just wanna avoid pain at all costs, even if we know it's going to turn out better in the end. There is some pain in the beginning. And so it's all about mitigating that and helping people through that and not lying and being realistic about what the benefits and drawbacks are going to be. So, um in business change management models are actually quite commonly talked about, but I don't, well, I've never come across them at all in medicine until I was asked to deliver a talk on the topic. And then I had to read about it quite a lot, but essentially a change management model for those of you that don't know is a system that strategically assists institutions, companies bodies in navigating through periods of transformation. It's not, I'm not saying that this is a step by step way of how you must do things. It's more of a framework, more of a guideline, more of things to think about and to help you understand where along a pathway you may be and why you're probably getting into trouble because there may have been things you haven't thought about. We all know health care is complex dynamic and you can't just have one size fits all. Um But as I say, it's really something that is to guide you. Um it's not something you need to do, as I say, step by step. And in a different hospital, you may have a different model that works in a different guideline because that works with that culture. Whereas in others, it may work differently. But there was a systematic review that was actually written, er not that long ago in 2021 about different change management models and how they have been applied to health care. And so I was going to talk to you about the two very, very most common ones that kept being quoted time and time again. And also one that I quite liked that I read about. So the three that I'll talk to you about are Mao's resistance model, Lewin's model and Kotter model. And I'll talk to you a little bit about what they say and how you can use them and why they're important. Um II don't want to make this really dry and boring. So I'm going to try as much as I can to give you some funny anecdotes from my own experience of implementing change and A I and how these have really applied and how you can use them too. So let's start with Mao's resistance model because I think that is the most um unusual of the other two. The number two and number three are quite similar, just phrased in a different way. But number one is, I think, slightly different. So um Mao's resistance model says something different. They don't say you have to manage change. They say you've got to be managing the resistance if you like. So it's not the change, that's the problem, it's the resistance to it. That's the problem. And that is what you've got to be addressing. In essence is what MAOA says. So you start with different levels and um level one is, I don't get it. Level two is I don't like it. And level three is I don't like you. And as someone who's quite keen on pleasing people, it really pains me that actually in this model getting up to the fact where people don't like you is actually your objective, believe it or not. But that is that is the objective to get people to not like you so that you are no longer needed in this particular framework. And in this particular change management process and with it, you're supposed to go through different levels of support, which I'll talk through. So let's start with level one, level one, you say, hey, I've got this cool A II want to implement, I think everyone should be using it, let's do it and level one is people are not necessarily against it, they just don't really get it, they don't really understand it. So at this level, it's all about delivering facts, figures, ideas, clarifying confusion over what people don't get and imparting to them the critical information they need. And I think the biggest, biggest mistake people make at this level is having one message and not adapting it to the different stakeholders that you will need to speak to. And I put many, many of the stakeholders down at the bottom. In this little chart, you may have information teams, legal teams, local champions who may be the radiographers who may be trainees, who may be the consultants or maybe the subspecialty leads um operations lead it lead, maybe there's a research component to it, maybe there's an integration component and of course, you're going to have to speak to many members of the executive team and finance when you implement this. But having one message is good, but having one message that you're very, very fixed on that you don't change your language or change your vocabulary or adapt to helping people understand what they don't understand is is a big problem. Many people have very different ideas about what A I is and what it can do. And so really coming to people that you're going to speak to on those different levels and finding out what they know versus what you want them to know is so important. I feel that a lot of people and I'm generalizing now, this probably doesn't hold to everyone. But I find that a lot of people senior in management really think A I will replace doctors. They, they genuinely believe that. Um And they think that A I, anything that A I can do to replace some radiologists would be amazing. Um On the shop floor level, people are very scared because they don't really understand what it's going to do, whether it's going to cause them more problems, whether it's going to cause more harm, patients aren't sure if this means they're not going to see a doctor anymore or if they're going to just have a robot that churns them out their results. Um I think radiologists do realize it's not going to replace their jobs, but they do wonder if it's going to add to their jobs, whether it's going to be harmful for their jobs, whether they're suddenly going to have multiple pop up screens come up and they're not really sure which ones they're going to have to acknowledge or not acknowledge or whether the user experience is going to be difficult. So at that point, it's about what, what is this thing that we're going to implement and why, why do we need it and how is it going to help? All right. Level two is now they get it, they understand it. It's, it's ok, but they just don't like it, right. They just don't like it. It's either like they don't like the product. They don't know why you've chosen this, this tool over another tool. They don't know why they weren't involved in that decision making. Um, they, they just don't like the pathway, they don't like the new clinical pathway. They don't like how they're going to report findings that come up. Um, at this point it's not survival, that's at stake. It is survival that's at stake in some places. But it's more just, um they, it can be very emotional at this point. And I think this is the hardest part in getting people to speak and getting people to talk to you because you're going to ask people, why don't you like it? What's wrong? And they're not gonna say what they really feel because no one wants to be the one that's against a new technology. No one wants to be the old dinosaur who doesn't get it. No one wants to come across as being difficult and everyone wants to be polite and wants to get with the program. So you're not really encouraged to speak out in a lot of health care organizations about why you don't like something. If you can't verbalize it with facts or figures or evidence, um You're not really told that you can say emotionally why you don't like something. It may just be, I don't like the fact that, you know, I'm not talking to as many of my colleagues anymore. Or maybe my clinical colleagues are bypassing me or seeking advice from other people. Or maybe I feel left out because I'm not being asked about A I opinions or people are doing the research on the A I and are not taking part in that. So I think this point is quite difficult to um address but really, really important. Um So um II love this quote. Um Don't ignore domain expertise or create greater silos when you implement A I, every time I fire a linguist, the performance of my speech recognizer goes up said one NLP researcher, I think you can say, you know, this is someone who doesn't really recognize the importance of an expert. You do need experts and you do need radiologists to tell you why they don't like something, even if they can't really verbalize it in sensible terms. And even if it's an emotional reaction, even if it's something there that's not quite working, it's really important that people with the expertise who are on the shop floor are encouraged to speak out. Um You don't want to create silos where you've got, you know, the worker and the people who are on the shop floor doing the reporting who are telling you it's not working out for some reason, it's just not happening. The care pathway is not working, it's there are delays um and they're giving you feedback, but you ignore it and then you have different people who are could be radiologists as well, but who are labeled the A I team. You know, the thought leader, the disruptor, they get to supervise, ensure A I delivery, speak to A I companies get invited out to A I dinners test the model, get invited to give lectures and then they become the cool popular crowd. And the rest of the people who are actually using the air on the shop floor who are finding problems with it are like, oh the annoying people, the people who are complaining, the people who are just not getting with the program, the people who are like, you know, being sidelined. So you, you really wanna make sure that doesn't happen right from the start because that can happen very easily. And I've had a situation where, you know, we've been implementing an A I tool and hopefully publishing about it and analyzing the results. And someone who is using the A I tool has said from now on, I'm no longer um integrating or using the A I unless my name gets put on the paper that comes out because I feel I'm contributing. But um I'm not being acknowledged. And so you have people like that, they may stop using the A I or they may just ignore the A I or they may revolt because they don't feel they're being acknowledged in some shape or form or their opinions aren't being taken on board. So I know that sounds laughable and that might seem really ridiculous, but that does happen and people do protest when they're not really happy. So you don't want that to ruin the use of the A I in people to then not even engage. And that, that can be a way that it happens. So, you know, it's very important as well to use um empowering language when you're implementing and using change. So for example, frame things for learning and see things as a learning experience for the department rather than you know what, but I want a promotion. I wanna be the cool person that gets invited to lectures. Therefore, this needs to be executed because I think the language you use when you have that mindset versus sharing and caring and learning is very different. For example, you know, in talking about the purpose of your project, you know, rather than saying we're going to implement this because we don't want people to get ahead of us, maybe say, you know, we're helping patients get their results faster or we're learning about how to innovate or we're driving change and you know, planning a new pathway together. So let's learn how to do that together. Um And you know, instead of an a leader saying listen to me, I can get this done. I've done this before. This is what you need to do, you know, try to frame it as I need your help, I need your advice on how we can better implement this tool. So all of us win. And as I say, you know, don't be a whole stay in your lane. But more, what can we do as a group to learn from issues that arise and welcome these issues that arise, you know, and say, you know, if there are problems come to come to the forum, come to the discussion points, have some ideas about how you think it could be improved. And then maybe as a team learning project or a team learning goal, we can, you know, talk about our experiences, write about it, you know, maybe make a podcast about it, talk to other teams about it and you know, perhaps being an example for other departments that want to implement the same tools that we've implemented. So that's um more about level two. It, I think it's the hardest one to get over. But if you can, then you can get on to level three, which is um now I don't like you. So you may be the A I lead, you may be the person driving the change or one of many people driving the change. And then up to a point, people get the A I solution, they start liking it, they can get how to use it, but they just don't like you. And that's more of a trust and confidence in management and leadership. Now it's no longer a resistance to the idea. They are resisting you as the leader or the people you repres represent, maybe you represent senior management and they don't like that. Maybe they do like you as a person. So it might not be super personal. But um there's a bit of mi trust or lacking confidence. They're like, hey, you know, what are you getting out of this? Are you suddenly a shareholder for this A I company? Are you suddenly like in their pocket? Are you suddenly getting paid on aurarium to speak on their behalf? Maybe we don't like the fact that you're now leading this change in our department because of all the conflicts you've got could be one situation in which case, it probably is time to move on or to figure out what that mistrust is. It may be that you need to go back to all the stakeholders in level one, reengage rehab the conversations, find out what it is they don't like or mistrust about the organization, the leadership and maybe that's when you need a shake up in that particular area. But as I say, by that time, you're probably doing really well and you've gotten to level three, which is resisting most of the levels of change. So basically, the levels of support are first of all intellectual, giving them the understanding because they don't understand what's going on, emotional, supporting them through. You know, maybe their um own emotional changes that are happening and feeling left out and wanting to be involved. And then the last one is more personal. And so those are the three levels of CH resistance. And hopefully, if you can tackle those with your levels of support, you can get to implementing your change in a smooth seamless manner. Now, the second model that I'll come to is very similar to the third and this is very, very much a very loose framework. It's more of a visual framework that I've placed here. It's only one slide and this model of change is very much about you are in a certain situation where you're like a frozen ice cube. Things are this the way they are. Um And then you prepare to change by you unfreeze so you melt, you melt, you think about how you wanna change, you wanna change the status quo, adapt to what you want, find out what works. And once you've got what works, you then freeze into a different shape, maybe a pyramid and then it's all about sustaining that change. And this is just getting you to think actually, you know what you're frozen, you're going to unfreeze and it's going to be a period where you don't know what shape you're going to take. And that's OK because you're figuring out the shape that this is gonna take and you're gonna do it with your team. And then once you know what shape it works and takes, then you solidify that and you make sure that works and people get with the program and that is the new system. So this is essentially it, it's more of a visual concept, this Lewins three stage model, what I will go on to next, that's a little bit more um fine print and telling you step by step is this cou eight step model. But you can visualize it in three stages like the previous slide by breaking it up into creating climate for change, then actually en enabling and engaging the change and then implementing and sustaining it. So we think about implementing change as just making the change. But there's a whole bit of work where you have to prepare people before you even talk about change or making the change and getting them on board. And this is what I think this model does very nicely. So we'll go through it. Um It's a bit more prescriptive this model, but some people may prefer that and that's why I give you the different options so that you can pick what works for you. So um creating climate for change, you wanna establish a sense of urgency. Now, I think this is a little bit, this was a little bit weird when I was reading about it because it's all about more energizing people to do it. It may not be an urgency for the change. But if you say you wanna change things and then you take ages to do it, people then forget that you wanted to do it in the first place. So you've got to energize people and you've got to have some sort of shortish time frame in which you think this is gonna happen. So be a whole like let's motivate people, let's do this. Um One of the more powerful ways of doing this is creating personal patient stories. So it may be that you might have a situation in your hospital where an NG tube was misplaced and it caused a huge amount of harm and damage and sadness to the patient's family and to the people involved and to the staff who were involved in that case. Um And so, you know, creating an urgency is to say, hey, you know, we don't want this to happen, let's stop now, we need to do something. So this never happens again. We need to improve lives, you know, or you could talk about the money. I don't know that that drives many people in the NHS necessarily, but it could be that a cost saving that might be something that you drive urgency for finance teams. But it may be more about patient care, patient understanding, sometimes it may be competition. So I know that in some places, competition is a big for change. So it might be that the hospital down the road has implemented an A I or is struggling to implement A I and you wanna like get ahead of them. So that might be establishing a sense of urgency for change. So then once you've got that and people are like, yeah, let's do it, let's save money, let's save lives. Let's beat the hospital down the road, then um you can create a coalition. So now this is more about um thinking about the change, finding out the knowledge about what specific change we're going to do and finding out who in the department has the expertise to be your expert panel. You know, maybe they've got authority, maybe they've got it expertise, maybe they've got leadership qualities, maybe they've got um maybe you need people from different subspecialties like radiology or chest or cardiac or maybe you need the surgeons to be on board. So now you're going to create your like, I guess expert panel, right, the senior people who are going to help you drive that change and then you're going to develop a clear strategy and this has to be really clear, not waffle. It has to be something that is explainable in 60 seconds or less motivational, emotional. And, you know, it's got to really involve the outcomes and the behaviors you want. Maybe it's about faster turnaround times faster reporting or less backlog or maybe it's about getting diagnoses faster or maybe it's about getting a, a patient to a biopsy faster. So what are the behaviors you want? What are the outcomes you want? What's the driver that you want to change and make it very obvious why this has to happen because I think a lot of the time people feel like people are changing for changes sake, but not really for any benefit. So it's really important at that stage to have that in your um short snappy sort of motivational vision right now, um in the next sort of engaging and enabling change, um communicate the vision to people and now you're going to go to your stakeholders. So which staff are going to be most effective? What are their fears? Can you alleviate them? Do they understand um what you're trying to do? What's in it for them? What what are they going to gain from this? Um There may be nothing to gain and you kind of do need to be honest about that because if you're going to lie, people are going to not trust you if you wanna do this again in the future, um can you get them to support the change? Maybe it's learning about a I maybe it's the opportunity to just be part of a new tech revolution in the department. Um Maybe you could create a motivational story behind the vision like the poor patient who died or you know, came to harm and how you can improve that for future patients. Maybe you could get some patient advocates on board and they'll help you talk about their vision for it as well. So now you've got the communication of the vision and then I guess this is the part where they talk about empowering people to act. So what that means, I guess is, um, you're always going to get resistance, you're not going to get people who are going to be on board with this. Um, you're going to have people who are not happy with it, but it's more about rewarding people who are with the program who are on board with you. Maybe people who wanna help, you can have a position on your little team. Maybe it's about, you know, helping them, empowering them to alleviate the fears of other people who are maybe not on board or maybe, you know, it might be, hey, you know, if you're interested in A, I, let's have a little A I study group, let's have an education group, let's have a journal club, get people on board, get them to be empowered with decisions or, you know, maybe get them speaking to the vendors, maybe making them a part of the decision process. So, and praise the people who want to be involved. So I guess that's what they mean by empowering people to act, give them some decision making uh capabilities and then creating short term wins. So setting short term goals being unambiguous visible, meaningful. And when we implemented the chest X ray tool in Epsom heal in, in southwest London, we um celebrated every 100 cases that the A I analyzed or every good pick up that the A I analyzed, there would be a little whatsapp group where we'd like, go, hey, you know, that was a really cool case. Let's talk about it. Let's discuss it. Or maybe, you know, having um some pitfalls and being, oh, let's learn from that. Let's have a little like discussion, a meeting where we can learn about the A I pitfalls. Um Also, you know, every time we got invited or submitted an abstract on our experience to one of the local conferences, that would be a, you know, meaningful goal that someone could go to. And we tried best as possible to get different people to present so that we shared the successes all around. And we've also audited our progress and you know, shared in the um success and also the failures and some how some of that has happened and then finally implementing and sustaining. So once you've actually implemented, you've got it going, people are on board with it, don't give up. It's, it's very quick to be like, hey, we've done it now, we can forget about it and not think about it anymore. So um you want to avoid complacency there, remind people of the long term goals and visions say, you know, it doesn't just stop with this. There will be other A I tools we want to implement or there'll be other adaptations to this we wanna change. Are there any issues with the A I tool you wanna resolve them? Quickly. So for example, it's really important that customer service and tech support is quick when you implement. Because if you implement something, people tell you it doesn't work and then they don't get any feedback on it, they're going to very, very quickly lose um interest in whatever you've implemented. So maintain that momentum of change and make sure that if there's an issue, you contact the team who implemented it very smoothly and they do something about it. Um Otherwise people will get bored and lose interest. And then finally, this bit I thought was quite interesting was institutionalizing the change. So once you've got a new care pathway, a new way of doing things, you know, the A I way, the digital way, whatever it is, um remind people that this is the way we do things around here. This is what we've always done. Now, this is our new status quo. Um and new people who come on board, like new members of staff who don't know about it need to be supported in taking that on because they may not have used this and where they've worked. And if they don't know how to use it, they may just ignore it or not get with it. So, um you know, empowering staff to support that, to have this ongoing education of how to use things, reinforce the new way of doing things and then um any new improvements or audit or learning that comes from the A I should include the whole department who are involved in that. So that's what that whole eight steps is all about. So, um I think that brings me nicely to my take home messages. Um The big things I wanna impart are think big with A I, you know, it's the new electricity, it can integrate into the many, many different areas. Um, not just in one bit, which is what I think most people are focusing on, which is the diagnostic bit. Um, it's not a bolt on, it can be, um, something that just integrates seamlessly. So, think big with it, think of how you're going to scale that up from the starter, manage resistance from the start as opposed to the change. What is it that people don't like about it? What are their fears? What are their worries challenge that not the whole, like, er, change, you know, because people might be ok with the change. I think a lot of them are quite familiar now with A I, but it may be some of the worries, they're not saying because it's emotional because they can't quite verbalize it because people aren't encouraging them to verbalize it. That might be hard to do. So, manage that and really do find out from people in the department, what they truly think it may be that you have to have one on one conversations rather than real team departmental ones because people won't say it in a group and I think it's so important that you have a compelling vision and reason for the change. Remember, there may be um ways in which you want to improve patient care and you think A I will help it. But A I is not the answer to everything A I is just one tool of many tools that may, may help. So is A I really the thing that's going to be the most cost effective beneficial change you want to implement. I think if the answer to that is yes, then you can create a compelling vision and reason. If it's not, if it's just literally more education or telling people how to report better or, you know, informing people of guidelines, they need to be following, you know, an A I is not gonna help you with that. And if you say it is, then people aren't going to believe you. So make it believable, make it compelling, really, really explain why A I can do this, that nothing else that you can try will do it. Um Because then I think people will come on board with that messaging. And as I say, I've given you a few change management models to think about. They are a very loose guide. It's not a one size fits all, but it may give you some ideas of things that you can think about when bringing people on a journey with you. I would say that key elements for all the strategies are exploring the need for change and also people's fears around it. Crafting the right messages for the right people. You may have one vision but changing the wording around that for different people you speak to is very, very important because you need to bring them on board and you need to create moments for celebration and goals um to keep that sustainability because it's a long slog and I think a lot of people have a lot going on. So having moments of joy and a break and moments where you connect with other people in this shared goal is really important. And um yeah, very, very good to have. So, thank you very much for your time and attention. Um I hope that was helpful and handy and I'm happy to take any questions, Suzie. That was wonderful. That was very clear, very compelling. Uh I hope that all of us will be able to think about how we change in our own practice. Um If that was as much a talk, I guess on psychology and sociology as it is on radiology. And so I've got a couple of questions and please put those in there. If you've got other questions in the audience, please put them in the chat and we can take those as we come along. One is a sort of philosophical question. Actually, humans have evolved because of change. We're here because of change and yet we seem intrinsically not to like change. There seems a bit of a conflict there. Why do you think humans in general do not like to change what they do? Well, it's fear, isn't it? The fear of the unknown? The fear of something worse happening. I think that whatever you do, staying put, at least you know what to expect. And you've sort of in a way, found a way to manage that. You've got your routine, you've got your ways of coping with the stress. It may not be great. You may be burned and you may be on the brink of burn out, but at least, you know it and you know how to handle that issue, I suppose. Whereas, like, you always worry that if you change, it could be worse. Like, you know, if you change house, if you change jobs, if you change partners, if you change projects, there is always that small, small risk of things being so much worse. And I think we're very, very happy as a human species to deal with a lot of rubbish because we don't want to have to face the fact that things could be worse even if that is only 1% chance. But I feel like that happens a lot of the time. And I think the way to sort of bring people around is to make them feel safe through that change, supported through the change told that, you know, the worst thing that can happen is that we just don't implement it or we go back to how things were, you know, we don't have to do this. You are safe space it, we're not forcing this on you. That's the other thing I think people have had things forced on them so many times that it causes a lot of fear and resistance. You know, we're going to do this, we're going to implement this pacs, we're going to implement this soliton, we're going to implement this, ri you don't have a voice, you're just going to have to get with it. And I think that builds up a huge amount of resentment, a huge amount of resistance and a lot of fear. So I think maybe one of the things that we can do to help people is to say, you know, genuinely and, and genuinely mean it like, you know, there is something we can do to make things better, but it doesn't have to be this way. If you are saying as a group, you hate it, that's fine. We can go back to it and, and making people feel supported in that I think is important. OK, great. So can I ask you a practical question now? Because a lot of this was theory and you touched upon you introducing new elements of A I in your own practice. Could you give us an example of how you've taken something introduced it and the things that you've done along the way in order to make that, that work to convince the people went over those hearts and minds that you talked about. Yeah, absolutely. So, um, I'm part of the South West London A I group and in one of the trusts and Southwest London, Epsom and Saint Helier, they've implemented an A I tool that, um, looks at chest x rays and it can pick up many, many different pathologies. But, um, they had to from the very beginning decide which care pathway that was going to be implemented into because there are many care pathways with chest x rays. So as a team, um the chest radiologists were involved, the general radiologists were involved. Um governance integration were involved and um the way they decided to use the chest X ray tool from the start was based on an audit that had already been done in the department showing that there was a very, very big lag between when a chest X ray from a GP showed a lung cancer to when a CT chest was being done. So this was already a problem that had been flagged up by their own department by their radiographers in various audits um acknowledged by everyone in the hospital as a problem that needed solving. So that was the first start. Find a problem that everyone acknowledges not a problem you think exists, that no one else believes exists or a problem you think that has to happen. Find a problem that everyone is on board with solving, right? So that was the first thing. There was an audit, there was evidence it was a problem. They then brought in a A I solution. Er various vendors had chest X ray tools that could look at lung cancer. They looked at this as a team. We brought the stakeholders on board that I mentioned, invited the vendors to give various presentations of how this could be used to flag up lung cancer for earlier detection to streamline to same day. Chest ct and um through that discussion, various different people had different opinions about the different vendors. They together came up with a score sheet of what kind of points they thought were very important that each vendor needed to promote or have. They then asked the vendors to come back again. They shortlisted the top three, gave their presentation again, gave each one a score based on this checklist that they designed and then um together picked the vendor they wanted. So then now they picked a vendor, they acknowledged the problem, they knew the care pathway they wanted to implement it into. Um There was then a lot, a lot of discussion about how to integrate it. That was when all the integration, the data DPA forms, the information and governance had to happen. Um Finance at this point were also involved, but this was a free trial that was being instigated. So they were OK with that. Um once that was all involved, then it was actually putting it into the practice on the shop floor. And in this point, this was a lot about winning hearts and minds about how to use it practically day to day. And a lot of it was to do with radiographers because they were going to be the first point of call of using it. The A I would flag an abnormality on the chest X ray as soon as the radiographer took the X ray and then they had to talk to the reporter, a radiologist or a reporting radiographer to get them to report it and then implement the same day CT you know, pathway and that, that took a lot of work. So first of all, they were happy flagging up and speaking to the reporters to say whether or not they needed the same day, CT and if this was a real finding or not by the A I, what they weren't so happy doing was speaking to the patient and breaking the bad news in a way. So we then had to talk, think about a lot of work about how we supported them in that. There has been talks, people have come in to help train radiographers on breaking bad news, how to inform the patients. There was a big body of work that then had to be done with the GPS who were referring, the patients leaflets were given to the G to the patients and to the GPS about your X ray may be seen by an A I which may flag a finding which then mean you have act the same day, which is a good thing because you're getting streamlined. But you may find it a little bit jarring to be told by a radiographer. You might have to wait around in the waiting room for a few hours for this extra test. So, you know, a lot of things we didn't always think about ahead of time, but then we did and then we had to go back and speak to people and bring them on board because then they weren't happy. So there, there was a whole piece of work that was done about communication in that point. And then there was also a, a big body of work about responsibility. So this is another thing once the A I flagged a an an abnormality whose job was it to then arrange for the same day CT it was going to, it was, and it is the radiographer who took the X ray to go round to CT to say, hey, you know, we've reported this. It's flagged up, the patient needs to come round for act. But a lot of people were then saying, hey, you know, this is taking time out. I'm not getting through as many people that I'm scanning. Usually there's a big waiting room backlog. You know, the CT radiographer isn't happy because now there's extra patients on their CT list and that's causing issues with relationships between the CT radiographers and the X ray radiographers. So, you know, you, you do have to um worry about that as well. I mean, there were free slots that were always, always set aside in CT for the same day. CT slots. It wasn't like extra work that was just being put on them. I mean, it was extra, but there was slots available for that. So um it was just about the messaging and how to communicate that and how to be more polite through that. So um that's still a bit of a challenge because no one likes to suddenly have extra scans, you know, just dumped on them. But absolutely. So it sounds like communication is a very key thing there. So we got a message um on the chat here from Holly Sing. Have you any experience or tips in differing needs for different change management with shorter stroke, rotational employees such as trainees versus longer term employees like senior permanent staff. So short term versus long term needs to change, to win hearts and minds. I think that's really difficult. Um I think partly because the change management models I've been reading don't differentiate between people who are there short term versus long term. They assume people are there for the long haul and for the long term, which is why it's all about building trust, loyalty communication. I think for the short term. It's very much about just imparting knowledge and saying, hey, this is our culture. This is what you need to get on board with. You know, if you're here for a short term, it's more about, here's the information you need. This is the way we do it. If you have any questions, ask us for the long term, it's about slowly making this part of the culture. So I'm sorry, but you know, I guess if you're trying to implement change with short term employees, it's going to be a lot harder because they're not going to be around for much longer and it's going to be more about praising them or giving them very short term wins. So it might be short term bonuses or goals or, you know, maybe it might be giving them some educational sort of experience and saying, you know, now you've used this tool, you can go away and talk about it. You know, you've got this experience rather than saying, hey, you're invested in this and you're invested in making it work really well. Um I don't know, but if anyone has any ideas of how you can invest people or make them actually buy into a change when they're only there for one or two months on a rotation, that would be very interesting. I think for a lot of people, it's just getting through the work when they're there only for a few months. I mean, I wanted to talk about generic skills and taking those away. I mean, we hopefully all work from NHS. Most people will continue to work for some NHS organization even if they move between them. So taking those skills between organizations in the same structure could, could be one of the ways in which we could convince them to do that. Um And of course, a lot of short term things you've mentioned already, some of them, but learning new things, getting points in the CV, giving presentations can be ways in which they're able to get something out of it without necessarily that long term at long term buy in. Um Could I maybe just change the conversation a little bit now and move towards regulation? I mean, do we need change management there? Do we need to convince the regulatory authorities in order for change? Because much of this has been about um changing people who are, who are the users, the end users. But actually there's a big piece about regulation. All of this, isn't there? What are your thoughts on that? Um Yes, the regulation does need to change. Um I think there are some changes that are happening but it's very slow. I think um the problem with the regulation is it's a very reactive beast. It's not something that's more proactive. So a lot of the times things happen and then the regulations come into play. It's very hard for them to get you know ahead of the schedule if you like. So, um, and a lot of the discussion now is all about regulating how we use foundational generative A I models like chat GPT in clinical practice. Um, but II mean, I don't have a huge amount of experience in dealing with the regulatory bodies myself, although I hope too soon, but, um, I don't really have much experience in that or dealing or talking to them, But II do agree that some things need to change there and we probably need to engage with them as a specialty to make sure that they understand our needs as well. They very much turn to us as the experts to understand what the needs are there. Good. Ok. Well, we've just got a minute or two left. So if there are no further questions, thank you, Suzie so much. That was very informative, very inspiring talk actually about how we can change. And you were talking mostly about A I here, but it actually it applies to so many things across medicine, doesn't it to make that change? Um How do we convince people that this is really what's required? And I think in the long term, most people have the same interests, we want to make sure that patients have the best care that uh we can do the best for all of our patients that come through. But how do we instigate that is the key thing? So I think that's been very inspirational and hopefully everyone will have taken away some ideas to change their own practice. Like I guess I will end by saying we're not at level three of Mara's level yet. So we still like you have, we haven't got to the point that we don't like you and hopefully that will not happen, but that was great. Thank you so much on behalf of everyone. We really appreciate your time and effort in coming. Thank you so much. It's a pleasure to be here and I really appreciate the invite. I hope that was helpful for you all. And I agree you can apply this actually to any change. It might be, I don't know um a new CT um protocol, for example, could also be in this change management as well. Any technology really. But yeah, I hope it's useful. We've got lots of thank you in the chat there and speaking of change, it would be useful to hear feedback on metal. So this has been instigated in the last few forums, this this um platform and it'd be nice to hear some feedback as to whether it is useful or not. So please do send it to Ellis and copy me. And if you think there are, if it's going well and if it's not going well as well, what we can do to change it and maybe we can follow Suzie's guidance then, right? Lovely to see you all. Cheer. OK.