Intro: 00:00

Rich Palarea: Welcome to Healing the Hospital Podcast, I'm your host, Rich Palarea 

We would like to dedicate today's show to the victims of the fires in West Maui. As the relief and rebuilding efforts continue, we especially support the hospitals of that area: Mālama I Ke Ola Health Center in the heart of west Maui, the Maui Memorial Medical Center and the other Kaiser Permanente partners in Maui Lani and upcountry Kula. 

If you're tuning in via audio only to today's show, let me give you a visual… I'm wearing a floral aloha shirt that I bought on my last trip to West Maui in late June -- only a month before the major fires. I'm so grateful to have once again visited this treasure and to spend time in the Lahaina area. This last trip was special, as I was able to take my adult children along and I'm grateful to have shared this special area with them before it was destroyed. Personally, I have many memories of this area going back to my first trips in the early 1980s and throughout the years and such a fondness for the area's natural beauty, aquatic wildlife, its customs and culture and its people. Our thoughts and prayers are with you all for the loss you have endured and for your rebuilding efforts. 

If you'd like to donate to the relief efforts, you can visit directrelief.org or redcross.org. You can also call 1-800-RED-CROSS (800-733-2767). 

On today's show, we'll introduce you to a research professor on the cutting edge of healthcare economics who has researched some of the controversial and complex relationships between industry payments, hospital procurement and physician preference in the medical device area. I will introduce Matt to you in detail in a moment, but listen to the titles of some of his research completed alongside his capable partners and colleagues: 

• Lobbying Physicians: Payments from Industry and Hospital Procurement of Medical Devices
• Medical Device Firm Payments to Physicians Exceed What Drug Companies Pay Physicians, Target Surgical Specialists
• Transparency and Negotiated Prices: The Value of Information in Hospital-Supplier Bargaining, and;
• Price Discrimination and Bargaining: Empirical Evidence from Medical Devices, just to name a few

With that, let's move to the interview portion of the show. 

Rich: Okay, so welcome to the interview portion of the show. Today I'm really excited about this one. I've got, I'm going to say, an old friend of mine, just because we've been working together for a little while on something that he's very passionate about, that is the intersection of what we do here at Kermit. Dr. Matthew Grennan is an associate professor of economics and analysis and policy, and the faculty co-director of the Robinson Life Science Business and Entrepreneurship Program at the Haas School of the University of California, Berkeley. I knew I could get that out in one breath.

I just say he works in Berkeley. He was previously on the faculty at Wharton actually, and that's where I met him when he was working there. And he was with the Robbins School of Management in Toronto as well. Received his PhD from NYU Stern. He's also a faculty research fellow with the National Bureau of Economic Research and a senior fellow at the Leonard Davis Institute of Health Economics. And Matt's research studies various markets in healthcare. Specifically, around products and organizations using empirical and theoretical models from industrial organization economics. His recent work examines how complex incentives and imperfect information affect how health technologies are adopted, priced, and ultimately deliver value for society. Matt's research relates closely to his teaching in healthcare entrepreneurship, data analytics, and technology strategy. It also informs recent business and public policy debates.

Regarding price transparency, one of my favorite topics. Relationships between physicians and industry, another big one. Regulation of new products and antitrust concerns about market power in the healthcare sector. Now you understand why Matt's here today. Matt even authored a case study using our company Kermit and has taught the case at Berkeley and other universities, even in the Mideast. Matt has received teaching awards from Wharton, Rotman in Toronto, and Poets & Quants. His research has been published in the top.

General interest journals in economics, management, and policy, including American Economic Review, Management Science, and Health Affairs. Matt's research has been funded through leading institutions such as the National Science Foundation and the National Institute for Healthcare Management. Matt, that was a mouthful, but I thought it was important because I really wanted to set up the academic side of what you do. You're the first hospital healer we've had on the show. That comes from this side of the industry.

So welcome, Matt, it's great to have you here.

Matthew Grennan: Thanks so much, Rich. It's great to be here with you.

Rich: So, there's one question that I always ask our guests. I usually put it at the top because it kind of sets the tone for the rest of the conversation. And that is this, the name of the show is Healing the Hospital. And we feature people on the show that we call Hospital Healers. You are officially dubbed now that you've been here a hospital healer. So, you can proudly wear that title and see what doors it might open for you. I don't think too many. But the question is this, hospital healers are people in our minds that when they see a problem or they see a situation, they really aren't satisfied with just leaving alone. The status quo is not good enough. We see issues and we have to solve them for the betterment of everybody.

And so I'm wondering through everything that you've done and all the tremendous people you've worked with over the years and taught and been taught by, is there any one person that comes to mind for you who is a hospital healer, somebody who's maybe influenced you or really kind of changed the trajectory of your research and your professional career.

Matt: Yeah, I mean, this is an intimidating question to think about, right? I feel even a bit of a poser to be dubbed a hospital healer when there's so many amazing caregivers and administrators and others who have really gone above and beyond, especially these last couple of years through the pandemic. I don't know, you know, for me, one single person as much as, you know, maybe to give a little bit of background, I was an undergraduate biomedical engineering major who kind of was always interested in healthcare technology but then somehow stumbled into becoming a PhD economist who studied competition and markets.

And the way I've made sense of that over the years has been to study health care markets, and a lot, as you said, for medical devices and pharmaceuticals and the product side of things. And when I got into this sort of hospital purchasing and procurement world, and especially after I moved to Wharton and started working in the health care management group there, I just started being exposed to it.

So many people who were working in this space and really just trying to figure out, how do you deliver better care, but still do it in an affordable way? And so really to me, I'm always just impressed by the people who are trying to figure out how to solve that really thorny problem. And especially, this is clearly my biased view, but I think those who are trying to harness data and trying to figure out ways to use data to inform better decision making, to figure out how to make things more efficient and more affordable. And you know, so just many, many people who have kind of influenced and kind of continued to excite me about that over the years.

Rich: I can only imagine them.

There's probably a great number of people that you've come in contact with who all have that passion because, one thing that I recall about my time at university and even interacting with academia thereafter, it's all about passion. Nobody's getting paid enough to do this work really. It's because there's a love, there's a real love for it, there's a real love to getting to the truth. And that's one of the things I've really enjoyed about getting to know you, Matt.

I was first exposed to you through the work that you did with Ashley Swanson, that really kind of introduced me to these big ideas about transparency and all this. And I think it was transparency and negotiated prices was the title of that. And the subtitle was the value of information and hospital supplier bargaining. And that was back in 2016. So, we launched Kermit in 2011, and I felt like we were just kind of a lone voice out there trying to espouse that there was this problem, a situation where surgeons pick, hospitals pay, and nobody really understands the interaction going on.

There's this asymmetrical buying thing that's going on. And I was so thrilled when I ran across that paper that you and Ashley did because it was like, oh, there's somebody who speaks our language, somebody else gets it. And I said, we have to track Matt down at any cost and see if we can talk to him. And fortunately, we were able to do that pretty easily and found you and you were gracious enough to at least take a phone call. This whole journey for both of us which has been really cool. So that was at that time I remember too you had an association with the Harvard Business School as well.

Matt: Yeah, well, I mean, I remember you tracking me down very well because I remember that first conversation. I think we probably stretched a half an hour call into an hour, and it all went by in a blink of an eye. We had so much to talk about. Yeah, I mean, so that was the link to Harvard Business School ended up being through a former student of Ashley's and mine from Wharton, Kyle Myers, who is still at Harvard Business School, who helped write that case study on Kermit with us. And that was, I think, really exciting for me because it brought together, as you say, that research world, thinking about negotiated prices, thinking about information and how it can affect procurement, but also thinking about some of the work I do more on the teaching end and thinking about data analytics and entrepreneurial strategy.

You know, how you take a company like Kermit that was doing super interesting work, you know, bringing together new data, new analytics, but then doing the really tough job of figuring out how to sell that into kind of large legacy hospital systems. And, you know, and that's just a thorny business problem. And so, you know, that was just a great project to work on. And I still love teaching that case study.

You know, every year they teach it now in the data analytics core class actually that we have at Haas and executive education classes that I'm involved with there as well.

Rich: So, you're reminding me of stuff that you explained to me in the early understanding of how all this works. There's a lot that goes into these papers and the research that you do. And that wasn't the only piece that you published with the one that you did with Ashley Swanson from 2016. There have been many others. You talk a little bit more about some of the other works you've created, maybe a little bit about how you come up with the ideas, maybe how you gain funding, how you go about the research process.

Matt: Sure. Yeah, I mean, you know, so maybe as I was saying before, I'm an economist who studies competition and when it came time to think about where I wanted to study that, you know, I'd always been a big fan of medical technology and it turned out there wasn't a lot of people studying medical technology markets, as many as you would hope, I guess, given what I think is the importance of the area.

But also there turned out to be some really incredible data sources, like the data I used in my dissertation from Millennium Research Group, which is now a part of Decision Resources Group, and then ECRI Institute, who ended up supplying the data for the paper you mentioned that Ashley and I wrote. And that really, you know, from the kind of economist, competition economist aspect of things offered this really detailed data on negotiated prices.

So, this situation where prices were negotiated, hadn't really been available before. And so, economists really hadn't thought through how to model that, how to bring data and theory together. And so that ended up being a really cool piece of that whole research journey, is that it helped contribute to this really nascent study of modeling negotiated prices and others have contributed to that as well. And really that area has grown and matured to the point where models that we worked on 10 years ago now are kind of the status quo models that the FTC uses to analyze hospital mergers and negotiations between hospital systems and insurers, for example. So that was kind of one side of that, probably the part that's least, you know, least correlated with the healing of the hospital world, right?

But I think what's kept me so engaged in it is that it also, you know, gave a chance to kind of add some value to thinking about, you know, public policy and business strategy around, you know, hospital purchasing and this kind of, you know, complicated area where you have, you know, these new technologies being developed.

Physicians are critical to developing those new technologies a lot of the time. They're critical, obviously, to learning to deliver these new technologies. And as a result, end up having close relationships with the manufacturers who typically actually commercialize these technologies. But at the same time, we're the kind of flip side of those relationships, if you read articles in the popular press, and academic medical centers who are being very restrictive on those sorts of interactions because they're worried about conflicts of interest. And so, there's been kind of this agenda that I've embarked on with a lot of co-authors, including Ashley, who you mentioned, and Kyle, who wrote the case study with us.

Others like Alon Bergman who's at Wharton now, just trying to unpack and think through some of these issues and some of the trade-offs and really figure out how do we get policy and managerial incentives right in this world where we really need these close relationships to create value and create good care, but we want to get the good stuff without getting the bad stuff.

Rich: Yep. It's a bit of a tight rope, I guess, that we all walk who are decided, who have decided to participate in this category. Um, and I, and I'd agree with you too, like, you know, back in 2014, you were doing some research, 2016, the paper comes out, you know, and you've done other things too, the, the data models that you use with ECRI, ECRI is a, you know, a hundred plus year old organization. They've got a lot of data. This stuff was.

I don't want to say taboo, but nobody really ventured in there to dig around in these relationships. It was kind of like, you know, the surgeon is there to provide patient care, and we don't really want to mix patient outcomes with the cost of delivering that. You know, they should have carte blanche, maybe somebody would call it, to be able to do whatever they need to do for the patient. It wasn't really until the present era that we're in, especially the post-COVID era that we find ourselves in and where costs have become highly scrutinized and so many hospitals are posting operating losses that we really need to look at this data.

And I kind of look at you and your cohort of researchers as journeymen who have just ventured into this area doing the pioneering work and clearing the weeds for everybody, but it was very important foundational work that you've put together that now really has a lot of merit because we're all talking about this. And the Center for Medicare is very interested in this category too, with bundle payments and things. So, you had just touched on it a little bit, but I'd love to hear a little bit more about it, you mentioned you're a biomed engineer. That's where you kind of cut your teeth. And then that led to economics and then research in the empirical study. But how did you make that transition? And do you miss the Biomed stuff, I guess you're still doing a little bit of it in the research you're doing, right?

Matt: Yeah, I mean, I don't know if I can really call myself ever having been a biomedical engineer. I was an undergraduate who studied biomedical engineering and worked in a few labs, a bio ceramics lab once upon a time. And I think it was a combination of I was an undergraduate and probably not disciplined enough to be a real scientist, to be honest, sitting in a lab and doing the groundwork that it takes to do some of that lab work.

As well as, I was just really interested in kind of business and policy problems and searching for a way to take some of the analytical thinking that you learn as an engineer and think harder about problems that were kind of, you know, more in the social world, I guess, than the physical world. And that's what led me to economics.

And so, I guess I've just been kind of cheating ever since then in doing the stuff I wanted to do in economics, but still double dipping and getting to do it in this medical technology space that I think is so interesting and important. So, I really just feel like I'm just fortunate to be doing the economics that I really do love to do in this field where I get to kind of be a bit of an observer along the way in teaching classes in med tech strategy and healthcare entrepreneurship and healthcare commercialization.

Rich: Yeah, it's really cool that you find yourself at the intersection of your passion, your gifting, basically, that you get to participate in that every single day. It's, you said fortunate. I think the same way. It's rare that people actually get to experience that. So, however long it lasts, I hope you get to stay there and you're thriving.

You talked a little bit about the previous work. What new work, what are you working on now that maybe that's kind of under wraps, but you can at least tell us a little bit about to kind of get us excited about what is coming next for you, guys.

Matt: Yeah, I mean, so some work that's, I guess, not under wraps, but still in the trying to kind of cross all the T's and dot all the I's really at this intersection of kind of relationships and purchasing. So, you know, when you, you might read in, you know, an article and then say the New York Times or something like that, that's, you know, an expose on a surgeon behaving badly or something, right?

Putting pacemakers in everybody who walks through the door or something. And then they might mention at some point in that article, and they have some consulting agreement with one of the device makers or something, and either directly or indirectly insinuating that there might be some relationship there. I'd read articles like that, and I just think, like, how common is this, right? Or another example in the pharmaceutical space that we've read about over the last decade would be opioids, right?

So, some companies who were promoting opioids, it's gone through court cases and people are in jail for having things like sham speaker programs where they were paying doctors a lot of money for consulting and speaking fees, but the doctors weren't doing anything besides prescribing opioids, right?

And so, you read really discouraging cases like that, and you wonder, how representative is this of what's going on out there? These are the sort of problems that make academic medical centers want to shut the door to manufacturers. In some states, pass restrictive policies, et cetera. But we know there are a lot of legitimate reasons for this.

Like you have to have some way to pass on, you know, new information about technologies to busy docs or docs who want to be trained and so on. And you know, who is better, you know, both qualified and incentivized to do that than the manufacturers of those products. And so, kind of set out on this, you know, let's figure out what's actually going on out there. And this was in the days even before, you know, open payments. 

You know, as a part of the Affordable Care Act, they passed this transparency law that now ever since 2013, you can go on openpayments.cms.gov and you can look up any sort of compensation or even just, not even compensation really, just anything of value that's passed on from a pharmaceutical or medical device manufacturer to a health care provider, right?

So, if you're going to get a knee implant, you could look up your doctor and see, you know, has Zimmer paid them any money in the last, you know, any time since 2013. And, you know, then the question when you, you know, we look that up is, wow, I see my physician is, you know, getting some consulting contracts from Zimmer, which is not very common by the way, so they probably aren't getting anything other than maybe a lunch a year or something like that. But then the question in your mind is, well, should I be really excited about that because that means my physician is so talented and so amazing that Zimmer values their time and wants to hear what they have to say? Or should I be worried about that because of the conflicts that we've been talking about? 

And so, we have a series of studies, both in pharmaceuticals and in medical devices, where we're trying to kind of start to unpack and answer that question. And it's such a complex question that by necessity, the studies are kind of incremental, and case studies in different segments. But so far, the results have been pretty encouraging, actually. I didn't really know what we were going to find, because you hear both sides of the story out there. But for example, in a paper where we studied statin prescribing and meals to physicians.

I think the big takeaway that we found there is really a lot of these meals seem to be going to physicians who otherwise wouldn't have been prescribing as many statins classic, you know, a classic tale for under prescribing for a variety of reasons. And so, it seems like at least in that case, you know, a lot of these meals were really helping to encourage patients getting statins that they should have been getting that they otherwise weren't, right? You know, and in the study that's kind of more in progress where we've been looking at devices, you know, so far, at least on average, we aren't seeing any evidence that.

These relationships seem to be leading to, you know, more devices being implanted or something like that, right? It really seems like the effects to the extent that they're there have to do with kind of what manufacturer you're choosing versus other manufacturers. And, you know, just figuring out the extent to which that's, you know, a good or a bad thing is pretty, pretty thorny. And so, I think what we've kind of found is that we can, at least in some of the cases we've studied, there doesn't seem to be a lot of evidence there for some of the bad stuff you might be worried about being really there systematically. And kind of whether or not things are net good or bad is I think still TBD maybe, and maybe case dependent as well.

Rich: Yeah, and far less empirical than what you're doing, but I'll give you an anecdote also from the work that we do. Okay, so can we just coin that tag phrase and put it under your picture? We'll just, we'll say that you said it's okay. I think, you know, so first of all, any time you dabble in anything that has implications like this, especially where large dollar amounts are attached to it, there's going to be bias. We start with a bias at Kermit that hospitals are paying too much. And so, we're championing bringing the price down to something that makes sense, that's already out in the market and leveling that price for maybe the underdog hospital doesn't understand how to negotiate as well as some others are not sophisticated enough to do that. And so, we help them with that. But it's been the same journey for us here too.

Gosh, it's really only one or two that are doing it, but they're getting the spotlight on them. And that's the story that's being told, and that's what's being noticed. And the social media environment and the 24-hour news cycle that we live in today, if a story gets picked up, then it gets circulated very quickly, and everybody says it must be true. And while there probably is some truth to some of these stories, they're sensational, they're just the tip of the iceberg, the rest of the community is playing by the rules, and the rest of the community values patient outcome and want to do the right thing.

And I'd say by and large, every single surgeon that we've met, except for just a scant few, and this is 12 years, so a lot of surgeons are interested in being good stewards of the hospital's funds and want the best for the patients. So yes, it's the same kind of thing we're finding. And we came up with this idea, having been former med device reps, that everybody is operating this way and come to find out, we can shut down the ones that are operating that way pretty quickly and suss it out. 

But there's a lot of value we can bring too on the empirical side to say, hey, it's not everybody. And we can take this data and do really good things with it too. So that's been what we've also seen here. Now you mentioned, I'm glad you explained open payments for the audience because I think some people understand that really well and others are amazed that they can actually take just the standard Chrome web browser and pop open the URL that you gave and put a surgeon's name in or a doctor's name in and see all those relationships, even if it was, as you mentioned, just a meal, up to large amounts and royalties that are being paid for inventions and speaking engagements and other things.

And so that's an interesting thing if you're going to be receiving care from a provider to just pop in there and do that. I would say maybe if you could just, you know, tie your research with the average person who might do that and give the cautionary tale of don't try to interpret those results on your own because there's a lot of other things that are going on in those relationships. It's hard to say to make that claim that you were saying before, should I be excited about this and their relationship that they're that revered and respected to be called by the manufacturer to do this work or should I be concerned about it? I mean, how would the average patient really navigate this data if they were to go in and not really looking for advice, but just off the top of your head, what kinds of things would you tell, say, your mom if she were poking around in this data?

Matt: Yeah, that's a good question. I mean, I think it's just as you said, the data tells you if a relationship's there, it doesn't really tell you what that represents, whether it's something you should be excited, worried about, or both. I think the first thing I would tell my mom when she was going to get a procedure was just find out how many of this procedure this person does and how they seem to be turning out. I wouldn't even go to look at any sort of payments.

But building off of that, I think there are a number of reasons why some positions might not have payments because of their employer, whether it's a big hospital system or an academic medical center that forbids it. So just because they don't have a consulting agreement doesn't mean that they're not a world expert, for example.

On the other side, you know, you could, you know, be getting care from, from someone who, as you said, like, is a, has a royalty payment because they were an inventor on a device. Um, you know, I think those are cases where, you know, it's probably useful to have a conversation just to know. Um, if the, you know, a lot of times I think in those cases, uh, a physician will, will actually have a general policy of letting a patient know. So, for example, you know, if they, you know, like to use the device that they get a royalty from, I think the typical thing would be that they don't actually take royalties from the devices they use, right? Or it gets all donated to some worthy cause or something like that instead. And so, you would hope that if your physician is one of those people, that they would be willing to have a conversation and be pretty transparent about any potential conflicts that they would be facing, right?

I think the, you know the tricky kind of cases that you can see in that data are things like sometimes there are these physician-owned distributorships and things like this where physicians were making quite a bit of money off of some ownership stake in the devices they were implanting. And I think I don't want to paint those with a broad brush because every case has its own particular details, but you would hope at least, you know, a candid and transparent conversation wouldn't be an uncomfortable one for a physician if they're doing everything on the up and up.

Rich: Yeah, you raise very good points there.

Just I have a question actually, because it jogged my thinking on open payments. I think for all the grief we give our federal government, it's amazing that what Centers for Medicare has been able to do with the data sets they have, and they don't hoard the data. They're one agency that really exposes the data freely to anybody who wants to either look something up like what they're doing with open payments, where they've put a web front end on a database, and we as patients or consumers can search that information or

We're just offering large data sets to companies like Kermit and other companies, other technology companies who want to do some kind of cross sorting with that data or enrich the products that we have out on the market for providers or for patients. So, when I think about the cross section of large data sets, especially around things like large language data models and things. We have this whole emergence that everybody has heard about with chat GPT and artificial intelligence. Are you doing any work?

Are you interested in doing any work around those datasets? And also, kind of where do you see, I know it's very kind of nascent time right now for all this stuff, we're just trying to figure it out, but do you see a lot of this stuff playing a role and maybe how?

Matt: Yeah, I think that's a great kind of set of questions around that space. So maybe two thoughts that have to do with two different hats that I've been wearing recently with respect to kind of large language models and maybe just like AI and data usage in general. So, one is I've been kind of working with a kind of federation, I guess, of some healthcare data scientists, many in academia and many more in the consulting space, providing some advice and some data science development to healthcare clients who are trying to organize their data and to make use of it. A lot of what we've been focused on is really thinking about modeling things in a way that might be via large language models or might be other types of models. But the idea that healthcare can start to look a little bit more like some of the industries, like car manufacturing, like airplanes and things like, airlines have been revolutionized by the fact that every airplane has a zillion sensors on it and they're collecting data on all these aspects in 

that's telling them, you know, if you're getting these different readings, you might be starting to enter a risky zone for breakdown. And so, you should do some preventive maintenance now, right? And so, like the airline safety and maintenance world has just been transformed by that type of thinking and modeling.

And part of our hope is that hospital systems can start thinking in the same way, right? Like you can have a model of yours, you know any unit in the hospital, right, labor and delivery unit or something like that, that kind of tells you when, you know, maybe you need an extra, you're likely to need an extra, you know, nurse on board, you know, tomorrow or something like that, right? And so, you know, you can think through, I think, a lot of these kind of concepts, like the digital twins is what they call these models of airplane engines and so on, and porting some of these over to the healthcare sector because there's so much data collected in healthcare and so little of it is currently used. So, I think that's kind of one, maybe, touch point on this phenomenon you brought up. The other is, I think what's become clear to me is just like a kind of interested user of chat GPT kind of playing with it here and there in these large language models is I think it's bringing us much more quickly close to this world where we all kind of need to be data scientists a little bit, right? Because as I said, aside from the healthcare commercialization work, another kind of part of what I do at Berkeley is I'm involved with the data science and data analytics courses that we teach there 

And I think our mantra for the last few years was really thinking about all these organizations who want to bring data to bear and helping them think through the challenges of, OK, you need to be bringing on data scientists. But as managers, you need to be able to manage data scientists and help them connect what they're doing to business questions that you want to answer, being an intelligent user and consumer of the sort of work that they're going to do.

That always seemed like, depending on the organization, a problem that was going to play out kind of slowly and expensively over time. Now, when you can have a bunch of unstructured data and ask some chat GPT or some other large language model, what's the relationship between patients having to return to the hospital and number of, how long we let them stay after their surgery or something like that. All of a sudden, the manager who can type in that question, they're data scientists, right? Like you don't need, your kind of, to some degree, I think we'll see how far it can be pushed, but this might be either allowing these data science questions to be asked by people who don't have coding and programming skills, or it might be allowing those people with coding and programming skills to do a lot more work as one individual than they might have done before. So, I think this pace at which managers need to get comfortable with these concepts is potentially accelerating quite a bit.

Rich: Yeah, for sure. Um, yes, I mean, this, this all feels a little bit like the early days of cryptocurrency when everybody thought, you know, this is just a fad, it's going to blow away. We really don't have to pay attention to this. I don't need to slow down and learn something new, really, do I? Because we all use the dollar. And yeah, it might take a little bit of time. But make no mistake, I think what we're seeing now is what you're talking about. We have unstructured data that's all over the place, much of which is, you know, we think about data that's being collected in hospitals that's electronic. And yes, it is. But what about all the data that goes back? Historically, it's locked on paper.

I mean, I think you had told me a story once about a project at the U.S. Patent Office where they digitized old index cards. That's fascinating to me because that is the genesis of why we launched the Kermit application, was to take this paper billing process in the operating room and somehow structure that data in a way that it's useful to multiple parties. So, I love those kinds of work. So, it's so easy that we can overlook all this other data that's locked away in file cabinets, but it is very valuable, isn't it?

Matt: Absolutely. No, it reminded me, I was sitting in the seminar and hearing the story about how these researchers found these old boxes of index cards about patent value assignments in the USPTO basement somewhere. And the way this would have ever become digital data if at all in the past would have been to hire a bunch of people to painstakingly go through them and just type stuff into a computer, which would have taken a long time and potentially been cost prohibitive.

But what they were able to do actually was just scan these all in and with a series of continuing prompts of a large language model, get that model to figure out how to kind of extract information and put it into a more structured database in a way that just really would have been difficult even if you were just trying to code that yourself to do. And so yeah, it immediately made me start thinking. My mind went to Kermit and the paper bills of the device firm associated with the surgery and kind of the exact same problem that you all have been dealing with.

been dealing with and cracking over the years. And so I think there's just still a ton of that out there in terms of data that's kind of on paper that can be brought into the digital world in a more meaningful way, or data that's digital but not super structured, but maybe you can extract meaning from it kind of more quickly than you otherwise would with some of these new models.

Rich: Yeah, it's a fascinating time to be in this space, is it not? I mean, so much is happening so quickly, right?

Matt: It's really exciting. And I anticipate bigger things than cryptocurrency from the data analytics in some of these models. But we'll see. Maybe I'm just undercounting cryptocurrencies.

Rich: We'll see. There's some time still that has yet to reveal what that's all about, I think. But I'm really excited, Matt, about staying, kind of watching you from afar and seeing the stuff that you're releasing. And I'm a big consumer of that information, especially when it has your name on it, because I know there's credible research behind it. So, I really appreciate the work you and your colleagues do. There's people who listen to the show from lots of different walks of life who I'm sure would love to Also watch you as you release this stuff and learn more about what it is that you're putting out there What's the best way for them to kind of find that information?

Matt: Sure, you know the crazy thing about professors is that all you really need to do is Google us and you can find us like seven different places on the internet. So, you know I have a website at Berkeley, I have my own personal website that links to as well and the MBR and you know all the places you mentioned. So, if they're interested, should be easy to find, I hope.

Rich: Yep, and for those who are listening from a car or from a device, it's Matthew Grennan, PhD, and that's what you should Google to find the work that Matt has been doing published previously and also the stuff he's working on today. So Matt, thanks so much for coming to the show. I'm geeking out about having you here and it's great to reconnect with you. And again, really looking forward to the stuff that you're going to be able to help healthcare with as you publish that information.

Matt: Thanks, Rich. Yeah, it's always great talking with you and thanks so much for having me on the show.