a huge proponent of clinicians like yourself who are willing to understand that the purpose of AI isn't to necessarily override their instincts or their best practices, but to guide, which I know we've mentioned earlier. You're listening to data nerds in the OR, a surgeon's journey toward value-based care. The podcast focused on data and data science as the keys to a better healthcare system. Ideas brought to life by the vision and experience of host Dr. Bruce Remshaw. Each week, Dr. Remshaw sits down with different players in the healthcare system to discuss how data is used today and what it could do if it were used effectively. How it could improve surgical quality, education, outcomes, and drive better patient care. Let's dive into the latest episode. Here's your host, Dr. Bruce Remshaw. Today we have a special guest because we talk about data all the time, but we actually have a real data scientist today, Lee McCormick. I want to thank you for joining us. Lee has very kindly over many years as I've tried to learn about how to apply data science principles to healthcare. She's been a resource for me and my partners over the years and spent a good bit of time and I've learned a lot from Lee and the other thing that I'm really passionate about and she's passionate about is improving healthcare for all people globally. And so she, not only is data scientists, I'll let her introduce herself more, but she's also an entrepreneur and passionate about improving our global healthcare system. So, Lee, thank you very much for being here. And I'd like you to just share with everybody about that yourself, your background and some of the passions you have about improving healthcare using data. Sure. So first off, thanks Bruce for having me. It's been a pleasure to work with you over the years. So it's exciting to continue that relationship focus on this podcast and beyond. So my background is in data science. I've spent my entire career wholly focused on what I would call population health data science AI, whatever kind of umbrella you want to put over it. That really has allowed me to see the evolution of data science throughout a very complex, exciting, sometimes maybe a dangerous side of AI. I consider myself lucky to be able to have traversed all that through my career. I think it really helps me understand the potential, the limitations, the pitfalls, and of course the hope that these technologies can bring, as Bruce mentioned, really in serving us as a larger population instead of these niche use cases that we have seen early in existence. I spent most of my career at Blue Cross Blue Shield of Tennessee focused on serving the enterprise everywhere from fraud, waste, and abuse to population health, to pharmacy management, to suburgations, really a jack of all trades and serving the enterprise with data science tools and some AI use cases early on. As Bruce mentioned, I am an entrepreneur. My passion eventually collided with my technical skills when I started to really focus on equity in AI. I went off and got my doctorate with a focus in health policy and helped them formatics really zeroing in on my research that centered on how do we create more equity in our AI solutions and health care by bringing in more diverse data. Once those passions really started to formulate an intersect, I really felt compelled to leave my corporate venture, if you will, and start on my entrepreneurial journey. And I know we'll discuss those efforts along the way, but I really think there's an opportunity, especially from innovators and entrepreneurs at this health equity space technology and health equity. There's a huge opportunity there, and I look forward to addressing that today. Great. Thank you, Lee. Yeah, I've got a few prompts that I want to go through that as I was learning this over the last 15 years of my career. In many cases, I'd learn something and say, wait a second, that's not what we do in health care. Why is this? If this is known, and one of the, one of the articles that I read that was eye-opening was a statement that was published by the American Statistical Association, where they actually felt compelled to write about this and say basically that we're not using statistics appropriately in health care. We try to make a very simplistic and use things like p-values to prove statistical significance as if that's a concrete fact. When in reality, we're supposed to be using statistics to gain insights and ideas, not so much a concrete fact. And in fact, they said specifically, p-values should not be used to measure or validate a hypothesis or model. It's not an appropriate use of p-values. So I was wondering if you could kind of elaborate on that and how important it is for us to understand this better in health care and not be using data tools inappropriately like we currently are. Yeah, so there's a part of me. I come from a little bit of academia, certainly in my research career and early on in my career, and then of course more applied AI and data science as I've matured in my professional career. I would have to say that p-values certainly have a place. We're not going to discount how effective they can be and really rigid research use cases, but I think agreeing with the comment made that we're not using them correctly. And that's really because when you think about data science and you think about AI, I really think it's the collision of creativity and insight generation with these statistical underpinnings. And to say that, really it means to go out on the limb and say we are testing things when we're testing data science and AI solutions, right? We don't have all the answers that are wrapped up in a single p-value, but rather we're building tools and solutions that have to balance both the statistical components and the math behind the scenes with its applicability in day-to-day life, right? So when you're thinking about academia and research, not all of those insights get applied. And I think that's really the delineation when we move into how data science gets applied in healthcare is we're consistently learning, right? And so is the data systems in the AI that are helping us along the way. And I really feel that something that we are getting better at in healthcare is how do we apply in a safe and secure way the things that we're learning, right? And so that we're not having to sit through rigorous clinical trials for things that we can learn on the fly, be respectful of the journey as it takes to get there. But I really think that those are two different paths, one of which is again most more rigid than the other, but there's a place for both. Can we define p-value for those that are not that majors? Because I think that may be an opportunity. Sure. Well, I won't give the textbook definition of p-value, but really a p-value is a way to measure the significance of a finding when you're building a research project. You start off with your research project having a hypothesis in mind, is this treatment going to work? Is it going to outperform the gold standard or the status quo? And so your p-value really, if you set up your study correctly, is the determining factor on, is that hypothesis true? Is it false? Or is there is it inconclusive? And so that's how we use p-values. Really, it's just a statistical measure of the likelihood that you're actually dealing with something that's rare, or sorry, circumstantial, versus something that really was brought upon by your research and the study that you put in place. Yeah, and the common, almost every journal in healthcare will want to see a p equal to 0.05 or less, which means that you have a 5% or less chance of it being not what you think based on the data. And that's just an arbitrary thing, right? It's not like some golden rule that somebody said, oh, this is going to make or break you. It's just realizing that all these statistics and analytics, they're not perfect. And it's just arbitrary. So just understanding that is, I think, really important. And like you said, Lee, I think we traditionally put a lot of effort into traditional clinical trials and that kind of science. And we've undervalued real world data and learning and improving in the real world. And that's where the data science, I think, and system science is really a new way of thinking. And I could say that it took me a long time to learn this over literally over a decade. My brain's struggling with this because it's so different than what we do in healthcare. But I think it's so important. And it really isn't about proving statistical significance in a concrete fact. It's, I think of it as learning how to improve through data and insights. And if this is what our outcomes are today, they could always be better if we can look at our data and measure things. And it took us a long time. But we learned if we're going to have a sustainable healthcare system, we need to measure what matters and what matters the most in outcomes is value based outcomes. So we can put financial data with patient clinical data. So we can, at the same time, lower costs and improve patient outcomes, eventually, you know, working our way back to prevention. When we look at a clinical process, and we're looking at the data, we ultimately want to see how can we make people healthy? And whether that's with drugs or devices or surgery or non-surgical nutrition, whatever it is, we want to understand that so we can make things better. And just to really pile on there, there's a famous saying, I'm sure someone far smarter than I stated it and I'm missing who it was. But all models are around some models are useful. And I really think that's where we start to see this separation between if we just discount something because of a p-value, then we may be missing out on just how useful it can be. Now, again, and we're going to get there in this conversation. It's how do we keep improving that usefulness? That we as a healthcare system should strive for. But I think that's really sums up how we should be looking at these really rigid models and how to create use out of something, even if it's not the most accurate, according to a p-value. Yeah, and I think that brings up in the article that I'd like to talk about when I'm giving talks about this. There's an article that was published and it was essentially just one data set. It was a European soccer football data set and it was about use of red cards and the skin tone of the players. And so the data set was collected to see if there was a bias in darker skin players getting more red cards. And so it was one data set. They recruited 29 teams of expert data scientists. 29 different teams were given the same data set and they asked them to tell us in your analysis, was there bias against darker skin players? And 2019 said 29 different analytical tools and they generated 29 different answers. And about 70% did say that there was bias and about 30% said there was no bias. So when you see expert data scientists doing the same data set and having 29 different answers, you realize that these tools are not perfect and they should be used to gain insights, not conclusions and use for improvement. So if you want to comment on that, that to me that really helped me understand. Okay, there's no one right tool and there's no one right answer, but we should be using these tools to gain insights. That's exactly right. And I think a lot of people when they hear bias, whether that's in your data or in the AI itself and the outcomes, it kind of makes them feel, oh, this must be something I shouldn't use or I'm going to set aside for fear that it's going to cause more harm than good. Bias is going to exist and that's just because we are human and we generate the vast amount of data that feeds these algorithms. I think our job as data science practitioners specifically in health care certainly in other verticals, but everyone kind of interacts with the health care space and we all know its value in really creating a quality of life for all of us is that we have to find ways to mitigate that bias, to understand it, to address it, and then to ensure that we're applying insights with that in mind. We're never going to be able to completely get rid of bias. We are human, but it's how do we gain the enough trust to walk alongside that bias, remediate it, refine ways to really improve the insights and leverage them accordingly. Yeah, and the way I've learned to do that is to do that as a small diverse team. So the people doing the work, in this case, taking care of patients, it's providing those insights to that team and letting that diverse perspectives in that team interpret those insights and then apply the insights that they gain to continuously try to improve what they're measuring, which hopefully more and more will be value based outcomes. I want to talk about a current state in AI and algorithm development because again, I think there's a lot of misunderstanding. I've read research papers about the analysis of algorithms being used there generated in what I call some processes. In particular, there's a large EMR company that generated an algorithm for sepsis and the analysis was that it was terribly inaccurate and led to many more false positives than it should have creating a lot more work for clinicians and alerts. And in my understanding, data science, if you generate an algorithm for a subprocess that includes many different contexts, it's likely to be very poor quality because for sepsis, you're talking about some process, it's not the whole patient process. And in a sepsis algorithm, you have an 80 year old woman with euro sepsis with a 20 year old motorcycle accident victim who has a compound fracture and wound sepsis and you're putting that data together totally out of context. So that was one of the things I saw that we shouldn't be doing algorithms around subprocesses out of context. The other common flaw I see in healthcare is generating algorithms. Let me lightly comment on that first because I think the other example will lead into our discussion that'll be more in depth about your company in addressing bias. What do you think about the concept of generating algorithms or not in context and they're in subprocess? Yeah, I think it's a great point. Now I will stand on my soapbox all day long around the value of having data scientists that really understand healthcare context, clinical context, the workflows in the administrative processes of healthcare, depending on where you're focused. I am not a believer in the statement data is data or data, whichever way you want to put it. I believe that even context from a data science perspective is really key. And all of a lot of times you're not going to find a data scientist that has enough knowledge to understand the difference in multiple types of sepsis. And I think that's really where we start to say we have to be building alongside these practitioners who are very entrenched in the use case that we're building for. They're also going to be the users of that solution because they start to provide the context not only from a subject matter expertise, but your model in those analytical solutions are only as good as how they're getting leveraged. And so if you don't bring those people in to provide that context, you really start to lose sight of what you're building. You can have the most accurate model from a statistical perspective, which kind of goes back to what we were to say. But if you don't have the right context of the problem or how it's going to get applied, I mean, you're ultimately going to miss the vote. And I think that we as data scientists, we are very technically savvy and we pride ourselves on that. But that's only one piece of the equation. And I really think that kind of letting that guard down and starting to really absorb the knowledge from our clinical partners is key to really making AI as fruitful as it can be in healthcare. Yeah, that's a great point. And it's really, again, it took over a decade, but I don't think there's any way I could have learned this without working closely with a team of data scientists. And without those data scientists, I never could have learned what we learned together. So that's very important. The other fly, which is probably more common than anything is aggregating data from lots of sources. There are a couple of issues with that. One is the quality of the data isn't really checked. It's normalized, but you know, the quality checked is a problem because when you aggregate data from lots of different local environments, you don't know what the quality issues are in each of those local environments. These aggregated data sets are generating population health algorithms, other guides for hospitals and clinicians. And unfortunately, there are, again, dozens, if not hundreds of articles that I've seen that show that these algorithms are generating averages, which means they're unintentionally harming minorities in marginalized subpopulations. And so that's a, to me, a pretty huge issue in healthcare. And I think you've seen that too, Glee. Certainly. I'm a firm believer read that for AI to deliver on all of its promises in healthcare, which holds a lot of potential. We as data scientists must be training those AI solutions on as diverse data and as varied data as the patients that this healthcare system serves. And sadly, a lot of that means tapping into data that may not live inside of our own systems, right? So we can think about geographical differences and different pockets of maybe disease progression that some health systems or some pairs have more data and or less data and and how valuable it could be to really start creating more representative. Corpuses of data or for buying data, I suppose it's called and to train these AI systems on. Now, that in and of itself is really a culture shift that healthcare is going to have to face. Right now, we're still sadly dealing with silos of data in our own systems. But eventually, we're going to have to start breaking down those barriers so that we're not, like, as you said, Bruce, causing harm to marginalized populations because we don't have those marginalized populations in our data systems, right? We need to start working with other data systems to bring that diverse data in. It's going to be, again, large culture shift to a huge security, a privacy effort in order to share those data, keeping that patient privacy in mind. But I think it's something that we as a healthcare system would benefit greatly from in using these AI tools. Yeah, and I think this is a good point in our discussion to maybe dig a little deeper into your company. Platypus, because I think, as we've talked recently, especially, you've got this passion and you're building this global network potential. It sounds like a marketplace for sharing this so that we can reflect all different types of people and different variations and different subpopulations. Maybe go into some of your effort there. Sure. So just at a high level, and then, of course, we can couple click into anything that we'd like. But platypus is a technology that's really built to identify, remediate, and monitor bias in healthcare AI. And we do that primarily by creating tools and technology to help organizations figure out what is the best path forward in remediating that bias, right? So again, we can't remove it all together, but moving closer and closer to AI that's more generalizable, that's more impactful. We're also creating that connective tissue for organizations to safely and securely share that representative data. So that as we start to identify bias that may exist, we now have a path to overcome it by sharing and really designing insights that are more not only representative, certainly equitable, but also bring more value, right, to those that have invested in them, to those that are receiving care based off of those insights. I think that's our mission at platypus and we're excited to be really again at that intersection of health equity and bringing a scalable technology to the space. Yeah, I think that's great. One of the one of the most exciting things as I learned the science was how important it is for us to collaborate together and share. We can't do this and you mentioned this earlier in a silo. You have a limited ceiling when you do that, but when you collaborate and you share knowledge, you share algorithms and you network those algorithms, then the potential continues to go up and it gives us the highest predictive ability for the things we want to do like match the best value treatment to the right patient sub-populations. It really is exciting what you're doing. When I learned it under the term of ensemble model for learning, and really essentially it is instead of aggregating all the data and doing one algorithm, you use these algorithms in these different disparate local environments and you network them together. I learned some examples. I learned when they trained IBM's Watson to play jeopardy, they had to train Watson to understand the English language very well and very quickly respond. Otherwise, they couldn't win. Instead of doing the best practice, one best algorithm, they actually networked dozens of algorithms, NLP algorithms. You didn't get just one. You got many different ones. I've also seen the Netflix competition, Netflix prize or through the K-Go competition where the teams that win are the teams that collaborate and develop the algorithm but then network algorithm with other teams and it's combining those algorithms that gives us the highest potential for predictive ability. Yeah, and I think too that's when you see ensemble modeling, which again is just the ability to draw insight from a lot of different models versus depending on one. It's really a step towards how do we start to leverage knowledge from across different either entities that have a separate model built on their data or separate types of models, maybe built on the same data. And I think that's really a step in the right direction is starting to leverage these ensemble models. We're seeing a lot of federated learning coming to fruition and healthcare, which federated learning just means a lot of different organizations can build the same model, right? The same use case, if you will. They'll build an AI model to attack that particular use case. And then they bring all of their insights into a single, I'm going to call it like a chili pot and someone stirs up that chili pot. And so you're getting all the goodness from all of the different organizations. And I think that federated approach is a huge step in the right direction because we're starting to learn from one another versus, you know, just pointing the finger at one another in terms of who is doing it better. I don't see AI as a competitive advantage anymore. I see really impactful AI as the competitive advantage and your ability to contribute to that more robust AI as a competitive advantage moving forward. Yeah, it allows us to raise all boats together. And again, that's what really really excited me about the the sciences that we all need each other. And it's a global, it's a global opportunity. If we don't do that, we limit our ability to learn and improve. And healthcare should be the place where we do this. It's the only industry where we're supposed to learn how to care for each other better as human beings. The other kind of concept I wanted to talk through a little bit was the need to continuously update the data models. I think in healthcare again, we have hundreds of algorithms that have been approved by the FDA. And from what I've seen, they're all static algorithms that are not being updated with new data. And how important it is to continuously update the algorithms with new data so that it keeps up with what reality is is constant change. There's always things new and different in our bodies. Take the pandemic, for example, you know, emergence of virus is normal in our role. That's a natural thing. And if we don't continuously update the data, we will, you know, a static algorithm from what I understood, it's going to get worse over time because it's just going to not reflect reality. Correct. All AI models undergo drift. And so if you're not regularly opening up that model, understanding the data that fueled that model has drifted or changed, or if there's the opportunity to add new data or add new features in terms of predictors and the ability to improve the predictability of the outcome itself, then you're really doing yourself an injustice, right? That's when we really start to see AI going from potential to possible harm as it may evolve. I am a huge advocate for right now we do see the FDA sort of monitoring these models on inside medical devices and those types of things, but eventually moving that more into things that help clinical decision support and that admin processes because ultimately we need to hold data scientists accountable for what their models not only do the second that they put them into re-wife into them and put them into play, but also what happens to them 12 months down the road because it's not the same model. So important for us to have the clinicians, the people doing the work caring for patients to understand the data and the data science and give them the tools to be able to, like you said, the decision support. Where I was considered most of my career as a honey expert and I thought I knew what the right answer was, which mesh to use, which technique to use. As I learned the data science principles, I was like, wow, I really don't know what's the right best value treatment for this patient with their situation, what's the population that they fit into. And it really changed my career over time. I used to think it was important for me to prove that the way I did surgery, which was mainly laparoscopic surgery, that I wanted to prove that my way was the right way. And as I learned the data science, I learned, wait a second, an open repair for some sub-ipulation of people, not laparoscopic, was probably the best value. I just didn't know who because we didn't have the data tools or the algorithms to see which sub-ipulation this patient fit into. And that was eye opening, obviously, a little bit of a humbling experience to realize I didn't know what the best mesh was or what the best technique was for any single patient. Yeah, and I think Bruce, that's true on both sides of the table here, right? I think I know in my experience as a data scientist, I came across a lot of clinicians that were really accepting of leveraging AI, of being a part of the process. But on the flip side, you also had a lot of clinicians that were very skeptical. And the same is true of data scientists, right? So you have a lot of prideful data scientists that they don't think they need input and that context we referenced earlier. They want to do it all themselves instead of really opening themselves up. And I think it's going to take everybody putting their pride aside, right, to really come together to say, we want to build better AI and we want to leverage it better. We want to see these better outcomes because of it. So I, a huge proponent of clinicians like yourself who are willing to understand that the purpose of AI isn't to necessarily override their instincts or their best practices, but to guide, which I know we've we mentioned earlier. Kudos to you. And I know that that's guided you on this journey that I'm very happy to be a part of. Yeah, it's a mindset shift, right? It's going from thinking we have to prove whether you're a hospital a CBO or a surgeon. The mindset shift is going away from thinking we got to prove the way we're doing things is right. And that this is good and it's good enough to thinking I can always improve. How can I improve? And that's the mindset shift. Let's look at what we're doing today and our outcomes today and let's all work together to make them better. And when you have that and mindset of improvement, it totally changes things. You don't look at things as good or bad. You look at where can we do better? How can we do better? And you take the shift away from trying to blame somebody or something that's a mesh is good or bad or a surgeon is good or bad. No, we're all human beings and products are just products. They're not good or bad, but they could be used better in better situation. Where they have value and how can we do better? That's the shift. Agreed. In the last few minutes, maybe things that you have seen either books or movies or resources that you've seen that have really impacted you or could help people in healthcare understand data science. I always use movies like Moneyball or the Book Moneyball, but are there other things that you've seen that you could recommend for people to help on a journey to learn and improve their understanding of data science and healthcare? Sure. So I'll reference a book that I always recommend to people who want to understand. I'll call it the sort of the good and the bad of AI rates. So hopefully most people have heard of the book, Weapons of Math Distruction. I want to make sure I say that so we don't get flagged or something, but that book is great. Now it's not healthcare specific, right? But it really starts to hone in on how AI systems are built and the potential they have to go astray in terms of the impact that they want to have versus the ultimate impact that they do have. And so I think that's a great book for anyone interested in really understanding how the potential of AI can be replaced by these farms. When we're talking healthcare specific, I really just encourage everyone to engage with some podcasts. I listen to a lot of data science podcasts specifically for healthcare to understand the trends that are coming up to really understand how this concept of equity and AI is starting to evolve both on a policy front and both on technology and application with plant because they think that those are going to go hand in hand. They're both really intended to build trust. And so we need to be embracing those at the same time. And so I would just encourage folks to just stay up to speed in terms of how AI is evolving. That's potential, that's pitfalls. It's hope and it's hype. That's awesome. Thank you for those resources. It's been a while, but I remember reading that book and it was awesome. It was really full of the examples of how we could use AI better. And both the good and the bad, like you said, it was I remember that really well. Any other questions or comments? Can you tell us a little bit about why you named your company Platypus Health? Great question. Everyone's familiar with the actual animal, the Platypus knows that it has a lot of different components from a lot of different animals. It has a million features and the duck fill has otter fur, a beaver tail, webtoon feet. So it has all of these different very unique components. All to create something very new and very interesting. We can view that's the purpose of Platypus, right? It's to bring all of these very diverse parts of different corpses of data together to build something unique, something that can stand alone and be a part of its own kind of calls in AI. And we thought that the Platypus really resembles that. That's perfect. And then this next question I'm going to try to phrase. So you guys were talking about ensemble modeling and how layering the algorithms and networking them led to reduction in biases. And so I want to dig a little on that because when Bruce, she gave the example of the football and card values and how some of them were 70% or 70% saw that there was bias or bias and 30% weren't. So how does layering those algorithms or networking them result in a better outcome? Or do you kind of result or can it veer us, right? You network it improperly and you end up with a worse result. Can you talk me through that a little bit? First of all, I'll explain a little bit in more detail what ensemble modeling is. So let's say you have 100 data points. If I were to build a single model on top of those 100 data points, I'm likely going to be taking the averages across those 100. But if I split those 100 into 10 different buckets, now I have 10 buckets of 10 different data points. And I build a model on top of each of those. I'm still doing averages, but now I have a very random sample, right? So those averages are going to change. And therefore my model changes and its ability to predict changes. And so that's the beauty of an ensemble model is you're taking samples of data versus in each sample is diverse. It has its own distribution of data. So that's one example and really the most common example of how you build ensemble modeling. You take all the best insights from these models and put them together. What Rooza's referencing is now you have the same data set. You know, let's say you still have all those 100, but you give it to 29 different diverse teams and they all take a different approach being maybe they maybe they created different predictors from their data, right? So maybe they bucketed race differently or they bucketed age differently. So they did all these different things and they all arrive at a different result. That doesn't mean that one of them is better or worse than the other. It just means they all came up with something different. And I really think it speaks to not only the diversity in how we statistically approach our problem, maybe by creating random samples and building different models, but also how the lived experience or the diversity of a data science team and how they actually apply those algorithms and those processing and all of the kind of technical components can really yield different results. And I think that we is the importance of that is to note that there is no one size fits all that the more knowledge we can start to pull together, really starts helping find more applicable and impactful AI. Yeah, that's exactly what the data science says. And it's so important. That's where we get back to. We all need each other and we we can decentralize the data and that's a much more valuable way to do data science than aggregated, which just does averages. The other thing I would just add as we learn this as the clinical team, we learn each local environments going to generate new ideas and unique learnings that are valuable to others. And then a different local environment will generate something different. So the example I use is one of most for me, impactful things we learned by doing data science with our hernia program was how important the neurocognitive component in the patient was to outcomes. And we had very complex group of patients and many of them were in chronic pain. And we learned to measure the neurocognitive state of the patient in terms of chronic stress state. We called it emotional complexity at first. So we put that into our data algorithm, because we learned that by doing data science for our own patients, a different pernietine might figure out that a genetic component is really important. And they need to look at the genetic component and put that into their algorithm when we combine those algorithms or network those algorithms in an ensemble model. And another will be a different sub population of the minority in Africa. When you combine those differences, those algorithms get better and better predicting for all people. That's my understanding and how important that decentralization and that ensemble model is to reflect all of the learning and all of the human beings on our planet. Does that make sense? It was beautiful. I think that's probably a great way to end because I talk about this again. It's back to the science. It's so hopeful that we have a science that includes everybody and requires everybody to participate and benefit from each other's human beings. I really believe as we get this going in healthcare, I'm looking forward to working with both of you for many years, because it's going to take unfortunate to kind of a long-term process. But I think with it, if we can change healthcare and help healthcare become the industry that learns how to learn together, we're going to have a much more sustainable, better healthcare system, but I also think it can lead to a better world for all of us. Thank you, Lee, thank you, Anian, and I really appreciate your time. Thanks for joining us for another episode of Data Nerds in the OR, a surgeon's journey toward value-based care. You'll find links in the show notes to any resources mentioned in today's show. If you're enjoying our podcast, please subscribe so you never miss an episode. And if you want more content like this, you can always sign up for our smart surgery blog via the link in the show notes. Or, if you want to ask Dr. Remshaw or the team a question directly, please send an email to
[email protected]. This episode is brought to you by Care Syntax, the leading vendor neutral surgical intelligence platform. The Care Syntax platform can ingest and analyze data throughout the surgical workflow, including clinical, operational, financial, and outcomes data to capture a complete picture of the surgical pathway. That data can produce actionable insights to enhance efficiency and innovation for surgical teams, hospital administrators, and MedTech developers, driving innovation for the future of healthcare delivery. Care Syntax impacts over 3,000 ORs and more than 3 million annual procedures across the globe. Learn more at www.caresyntax.com.
PodcastTranscript AI