AI in Health: Board Governance That Complements the Health Policy Perspective

Byron Scott

Hello. Good morning, everyone. It’s a pleasure to be here today. And you know I appreciate the invitation by the alliance of you know, for health policy. I also want to thank Claire for the invitation and also the board chair of the Alliance Tanisha Carino. They thought it would, you know, be valuable for me to kind of. I think bring a different perspective. I’m not a health policy guy. I have a lot of respect for everyone that is involved in it. So I’m just going to kind of give you my perspective today around health policy, but also from the lens of primarily from as a corporate board director, I mean that’s going to kind of my lens, but also someone who’s practiced medicine for 27 years and worked in different, you know, types of organizations. So just kind of give you that kind of feel of things so. With that. Here’s our agenda, kind of kind of give you a little overview of me, of I think, Claire tried to highlight some of that and then we’ll talk a little bit about what AI is very high level. I’ll kind of do my best to kind of compare and contrast health policy and corporate governance, and then we’ll talk about, I think, some of the benefits of AI and the risk, and I know. The panels are going to really dive a lot deeper in this today, so I’ll just kind of really I think set the stage, you know for conversations. Later. So as Claire mentioned, you know a lot of what I do today is is serving on different type of healthcare boards both for profit and and non profit. You know spent years working at a health analytics company that got acquired into IBM Watson Health. So had experience kind of dealing with that in its early infancy in, in AI and then also served as a consultant at IBM. You know, during my career too. But my lens as a today as a corporate director, as we approach AI really comes from I teach at two universities I teach healthcare quality. So for those that are very familiar with the steep framework which is around the six aims of delivery of healthcare quality, I kind of always have that lens when I’m thinking about things around technology and AI. And for those who don’t know, the steep framework it’s, you know, delivery of healthcare that it’s safe, timely, efficient  effective, equitable and patient centered. You know those are things that I always try to instill. If you know when we’re trying to make decisions, are we doing what’s right. So couple of definitions. You know, artificial intelligence. It’s a technology that enables computers and machines to simulate human intelligence and problem solving capabilities. But I I do know for a lot of people in the healthcare community and even myself kind of think of it more as augmented intelligence. Meaning, you know, you. Kind of leverage AI’s, you know, technology to really augment you know what what humans need to do, you know, and at the end, you know, it’s got to be patient centered. So really AI from my perspective and I think a lot of others is, it’s a tool. To help clinicians and others involved in healthcare kind of deliver healthcare in a in a better way. So just want to kind of you know mention that because AI kind of means different things to different people. So just want to make sure I kind of clarify that. So as everyone knows AI it’s it’s not something that’s new it’s. Been going on for seven decades, you know, you know, essentially 7 years, but I think we’ve seen a, you know, just a huge evolution in healthcare over the past, you know, 10 years, mainly because of. You know, increased computing power also the ability for machine learning to have a, you know a lot more data sets to kind of train on. So you’ve really just seen this huge progression over the past couple of years. But again, Healthcare is not the only industry you know every business industry is is leveraging or using AI in some way. and I just kind of highlighted just a couple terms, real high level because people hear machine learning, you know deep learning, natural language processing and again you know it’s a kind of a a type of AI. So just very high level. We’re not going deep into that. We really don’t have time to do that you know today but just wanted to kind of set that framework. One of things I want to mention as we go through here today, you’ll see various references that I use and I think Claire is going to provide these my slides at the end. But on the very end, I actually have links to all my references and what I did. They’re all from peer review journals, except for I think one or two. So you know for further reading, for, for all of you, if you want to kind of dive deeper in some of the. The topics. So let me talk a little bit about corporate governance and kind of comparing contrasting it to health policy and it’s interesting. It is, you know, Claire and I were going back and forth by e-mail about this when I was kind of putting this talk together. And there’s really nothing published out there that’s comparing kind of. Healthcare, corporate governance and health policy. There was one article. It was actually a LinkedIn article that talked. About, you know, corporate versus political governance, but really nothing that kind of looked at health policy as a whole. So I kind of like took a corporate governance kind of you know journal, my perspective and my involvement in in corporate governance and then kind of created this slide and kind of some of my you know views and perspective so. Even the health policy may disagree with it, but this is Byron’s. You know, thoughts on this so. So one thing is obviously understand the accountability piece of it from a corporate governance standpoint, they’re shareholders. So as a corporate board, you’re your responsibility to those shareholders. Now, one caveat for nonprofit boards, they don’t have shareholders, they have stakeholders. And those stakeholders typically are, you know, the community is in the patient, so. You know, just kind of make sure you understand there is that subtle difference between a you know for profit and a, you know, nonprofit board. But but if you’re involved in kind of health policy, your stakeholders are really, you know, society as a whole. And you have just a huge, you know, wider, you know, span of influence. And when you kind of dive into the risk output acts of things you know in terms of managing risk corporations, really, they’re managing risk for their shareholders, their shareholders value. But then also for the corporation as a whole. And so they will sometimes for profit industries take risk to generate you know value. For the shareholders, I mean, that’s just the nature of it. It’s you know, you’ve heard the term risk appetite. I mean that’s just something that boards have to do to really kind of leverage and and generate return on investment again. Doesn’t necessarily apply for nonprofit organizations, but if you look at it from a health policy perspective, you know you’re trying to design, you know, policy from a risk standpoint, really, to eliminate risk, you know, to a patients health. And also, you know, kind of eliminate anything around, you know, patient safety. So, you know, quite different. But I think you know, as you look at health policy as a whole, you you want to make sure that the policy doesn’t, I think stymie innovation, especially around AI. You know, there’s a lot of opportunities I think with it, do we need guardrails? Absolutely, I think so. But I think we have to be very careful and we’ll talk more about that, you know, in a few minutes. But from a compliance standpoint, you know, obviously you know both from a health policy they have internal you know compliance. You know, guidelines they have to follow in the corporations, they, you know, have to do themselves. But one subtle difference, I think is, you know, organizations have to follow laws and policy, whereas health policy is kind of developing them and setting them, you know, so it it’s kind of comparing those kind of two different. So just wanted to kind of lay that. You know, kind of groundwork, you know, I look at, you know, both corporate governors, health policy kind of working together. They’re kind of complemented, you know, have one kind of doing things in one higher level. And then you have, you know, corporate governance kind of doing yet another. So I just want to kind of lay that, you know, framework. So when we talk about benefits of AI and this is where I think which would be, you know, pretty interesting and and this is by no means an inclusive list, I could probably put 30 things up here and there’s no particular rank order. So just want to make sure that you know that’s clear. But I think you know, when you look at benefits of AI, I think administrative efficiency is one thing that I think everyone’s looking at and and why is that? Well in healthcare, what are some of the two biggest issues? You have physician and nurses burnout, the EHR was a big part of that. Honestly, I spent 2/3 of my career. Paper charting the last third you know using an EHR and I can, understandably, you know, agree with some of the challenges. So when you look at generative AI as a as a way of summarizing clinical notes, I mean huge value and I think you know, you’ll find that there’s a lot of healthcare organizations, provider organizations really diving into this. I know the hospital where I served on the board. We’re actually doing some pilots around this. You know now like other health systems. So that’s where I see really a huge benefit. You know, with really kind of diving into to AI. But it’s interesting there was a KPMG did a survey this year kind of a poll survey of of corporate board directors now not just healthcare but just you know all corporate board directors around administrative efficiency and boards really expect their organizations to really focus on leveraging generative AI. 70,69% of those corporate board members say they want to see, you know, generative AI really try to bring some operative, you know, efficiency. So there’s going to be a, I think a lot of folks on this. But again, I think it’s a win win for healthcare this in my mind is I I think you know a top reason to really look at the benefits. Obviously there’s a lot of research and you know things that are going on around, you know, imaging and AI. You know, I think eventually we’re going to get that in such a way to help, you know, really improve diagnosis and, you know, detection, you know, prevention. But what I’m most excited about is actually looking at. You know what we can do around quality? I’m a quality person, so if if AI can improve outcomes, improve safety, improve quality, then that’s where we need to be investing our time and energy as a whole. So you know, and then I’ll talk about, you know, fraud prevention. So as many of you know, a lot of organizations, healthcare and non healthcare are using AI tools to do audits, very common in healthcare industries using hospital audits are being done, you know by the auditors. Using AI tools. So you know when you’re looking at claims, there are some health insurers that are using it. So I I think that’s one benefit obviously to prevent fraud and abuse and waste. So absolutely kind of you know I I think embrace it, but in a couple of minutes when we talk about the risk, I’ll bring this topic back up because. I think there is, you know, another caveat to to this in a second. And and then you know, clinical pharmacology, obviously drug development, clinical trials, I look at AI as a way with clinical trials to really you know increase access, increase the diversity of of you know the patient populations for clinical trials. So you know huge benefits there and again. This is not a. You know what I would say an an inclusive list, it’s just a couple of things I kind of in this 20 minutes decided to kind of focus on and and highlight. So we talk about risk. So again, this is not an inclusive list, but I will say I put date and privacy and security is number one for a reason. You know I I will tell you. Every one of my board meetings and then two of my boards have committees. One’s a information Task Force Committee, the other ones a science and technology we talk about. Cybersecurity risk and AI all the time. I mean, this is front and center in our minds every day. No one wants to. I think end up being on the news. I mean, everyone’s aware what happened last Friday with the crowd strike and I think these are the concerns with just everything being digital with technology and there’s still a lot of unknowns. With AI I mean is it going to improve, you know, reduce risk. We pray and hope so. But the concern is just, you know what we don’t know is is whether or not it could actually make us more, you know, vulnerable. So. So this is, you know, what I say when you talk about risks and vulnerabilities. The number one thing to kind of focus on. You know, I was having a conversation a couple of months ago with our CIO at our health plan and so. You know, he has a very kind of I think conservative view about AI in his mind, he says. You know, we have to make sure we get the right people first, the right processes, get those right and the right foundational technology before we even dream about laying AI on top of it. So you know, and he said it will be very selective, but we’re going to get. These other things. First, before we start, you know, kind of laying AI on it. Because you know, if you don’t have the people process and the foundational technology first, all you’re going to do is lay another piece of expensive technology on it, and you’re going to end up in the same, you know, bad outputs. So, and that’s something that, you know, during my career in consulting with IBM. Really talk to clients about, you know, they always wanted the new technology, the shiny new toy and. I said, well, what are your processes? Do you have the right processes first? So that’s I I think the number one thing and I think as we approach AI you know to really reduce you know and mitigate privacy and security issues. I think that has to be kind of a foundational thing that we absolutely need to. You know I think focus on 1st. So everyone’s heard about, you know, issues around, you know, transparency. You know, obviously people always want to know, you know, what’s underneath, how this tool was developed. You know, the whole black box, you know, issue and and I think there’s absolutely has to continue to be you know most importantly I think healthcare providers involved in. Development of of of AI tools to make sure we get it right. You know and also to kind of mitigate issues around bias which we’ll you know kind of talk about in a minute you know too. But informed consent is interesting, you know, because you know there’s times that all of us we go to a doctor and you’re asked to sign a form to get treatment, not a one of us reads through those forms. You just sign them. I mean, I do the same thing when I go see my providers. So. And believe me, I’m pretty sure that there’s things in there about AI these days I honestly think it should probably be on the top of the page in bold print and what you’re consenting to do, especially as it it means to your data and information. So that’s one thing I absolutely think we can do an absolute better job. You know, it even goes to the simple things like wearables. I mean that data is going somewhere and if you agree to do it, great. But let’s just make sure it’s real clear and where you’re trans,parent about informed consent you know. So for me it’s important because as a physician, you know, I never did any type of procedure, even a minor, without really kind of either getting a verbal and or written consent explaining what I’m going to do and why. So I I think, you know, when it comes to AI, we really need to make sure in this any type of data that we’re really making sure. Patients understand, you know what’s going on and what’s being, you know done. And this week, some of you may have saw this, that the Mayo Clinic they’re doing a pilot for nurse charting using AI and and they made a very public statement about we’re getting patient consent first before they progress down this path. So it was kind of very nice. To kind of see that kind of approach with it. Now data monopolies. You know, obviously it’s a it’s a concern. You know that if there’s a, you know, a few companies organizations and I’m not just picking on them, but you know again who controls your data, who’s you know kind of, you know dominating the data market is a concern. Some of you may have seen a few months ago. The DOJ is actually announcing, you know, some investigations into some of the large companies around AI dominance. So it would be interesting to see kind of what transpires with that. And again, you know it, it’s the conversation of who owns your data. I think as a patient its your data, but you also have issues around the hospital claiming they own the data EHR companies claiming they own the data. You know there’s you know these questions and issues, but in the end you know I firmly believe patients. You know, own it, but it’ll be interesting to see kind of how that plays out. You know, around AI going forward. Bias so. You know, AI has to be trained on diverse data sets and you know, there are some insurers that are using, you know, AI to, you know, look at claims denials. And most recently, some Medicare Advantage members. You know, CMS, you know, the senators on the hill, some of you may know this. We’re asking really to look at it because. There’s a lot of elderly and very sick people having AI was denying claims and so. So the big question is, you know obviously is is that tool, was it trained on diverse data sets? I mean I don’t know the answer, but these are the questions that you want to make sure that there’s not, you know, inherent bias in these tools. Especially if you’re using them, you know for things like, you know, denials and claims. So these are the things I think you’ve got to kind of watch. You know get fixed and and the whole conversation around biases can go in so many directions, you know, but that’s just kind of one thing that where AI is being used, you know, currently by, you know, some health insurers. So Claire asked me to kind of, you know, do this and it and I kind of put some thought on it. And so, you know, she. Said hey, what does one ask of the health policy community? Would you consider, you know, saying and so obviously I would say, you know, foster it, you know, policy, you know, that is, you know, fosters innovation and development that can benefit but also protect patients. I mean, in the end, we have to make sure whatever we do. We protect patients, protect their data, protect you know, from a safety standpoint. And and then the one thing to remember and I will say this, you know is that there are a lot of you know corporate board members that are both on for profit and non profit. A lot of my peers I have a lot of them that you know serve on you know many, many boards in this country and and they take this seriously. I’m I’m telling you this is something that keeps us all up at night. Not a one of us ever wants to be involved in our company or organization. You know, having a cybersecurity or an issue and AI especially so, I will say from a policy standpoint, you know, just, you know, believe that you have another partner that’s really kind of looking out from healthcare from that standpoint. Both from the provider, I think the health plan you know side so just kind of keep that in mind as you guys are, you know developing policy that there are other partners out there that are really I think looking at it lens trying to from the bigger you know society view and and protecting things. So again want to say thank you for your time and. You know, here are the references links, so she’ll send them out and you know, like I said, they’re all from peer review journals. And if you know, if you need some extra reading or some of these topics, you can kind of dive in. So appreciate your time. Thank you.