00:02.31 Ratnadeep Bhattacharjee Hi, everyone, and welcome to another episode of Leaders Perspective, where we explore the evolving frontiers of health care and technology through conversations with visionary leaders driving change. 00:13.68 Ratnadeep Bhattacharjee I'm your host, Ratnadeep, co-founder of Tech Variable, where we build compliant, secure, interoperable, and scalable solutions for the health care ecosystem. Today's episode is titled Interoperability in Action, and I couldn't have asked for a better guest to walk us through it. 00:29.61 Ratnadeep Bhattacharjee My guest today is Keith Deutsch, a very experienced advisor to early stage companies in the health tech, IoT and AI space. As a fractional CTO, Keith supports companies with market entry strategies for disruptive products and guides both product and technology development. 00:45.81 Ratnadeep Bhattacharjee In health tech especially, he has been deeply involved with the growing role of remote and wearable devices in clinical settings. Keith also brings a wealth of hands-on experience in interoperability and beyond, right? And has worked across both device integration, system level architecture in care delivery and diagnostics and whatnot. 01:07.68 Ratnadeep Bhattacharjee Keith, welcome to Leaders Perspective. It's a pleasure to have you here. 01:10.58 Keith Deutsch Thank you. Thank you. I'm very glad to be here 01:15.14 Ratnadeep Bhattacharjee So Keith, let's kick off things by grounding our listeners a bit. Keith, you have had quite a journey across different areas, not just in tech, but healthcare healthcare care and health tech in general, right? 01:25.19 Keith Deutsch Mm-hmm. 01:29.49 Ratnadeep Bhattacharjee can you Can you walk us through your background and how you found your yourself at the intersection of all these things? 01:37.48 Keith Deutsch Well, so I've been doing this for quite a while. and And for several decades, I've been working with early stage companies. And I've had the unique experience of working with a very large number of very early stage companies. And sometimes for early stage companies, I'm talking about two or three people, not necessarily in a garage, although I am in Silicon Valley, where that's sort of the and the mode of the preferred mode of starting a company, but two or three people with an idea. 02:04.09 Keith Deutsch um So we're talking very early stage, up to you know a few hundred people and and you know a first product. So very much a zero to one kind of guy ah for a very long time. long time um I would say in the last decade or so, I have moved into sort of the IoT world and an industrial world as a focus. 02:28.17 Keith Deutsch And interestingly enough, that turned out to be an interesting ah segue into the health tech world because particularly in the last four or five years of the emergence and the acceptance, they've been around for a while. Wearable devices have been around for a while, smartwatches have been collecting data, but If you've ever had the experience in the US s in particular of going to a doctor and saying, look at all my Fitbit data, um you know, you got exactly nowhere. They looked at you blankly and said, OK, that's fine. 02:56.08 Keith Deutsch Be happy, but let's talk seriously. But that started that started to change in the last five or six years. um And a lot of it was driven by the experience of the pandemic. I wrote an article back in, I guess, 2018 predicting this just before the the pandemic really kicked off, talking about you you know the the difference that wearables could make in the health tech space um and and the change that big data, which is not usually available in the health tech world, ah could make. 03:27.19 Keith Deutsch um And so that that was sort of one of the, and and I've done some other things in the health tech world. I've done things like what you guys do around, ah you know, quality metrics. um But a lot of my focus has been this crossover of the IoT world with the health tech world through the medium, so to speak, of ah remote devices and and the potential they offer. 03:49.99 Keith Deutsch And that leads us to where we are today. 03:53.93 Ratnadeep Bhattacharjee Okay, okay. So some of the things that you mentioned, Keith, revolved around, ah you know, medical devices, and big data, health tech. 04:06.64 Ratnadeep Bhattacharjee So ah interoperability, I guess, plays a very, very important role in all of this, does it? 04:13.61 Keith Deutsch Well, you have to integrate these devices into the clinical world, and that is a non-trivial operation exercise in interoperability and and integration. 04:25.02 Keith Deutsch um And I'd say more integration and interoperability. These devices are not necessarily working directly with each other, but therere but but the insight that they bring has got to be brought together. 04:38.13 Keith Deutsch And, you know, the the electronic health record system, the EHR EMR, depending on what which way like to talk about it, is the place that tends to be where these things have to come together. 04:49.53 Keith Deutsch But making that happen has been a challenge ah for a few reasons. but But yes, interoperability, integration, without that And clinicians have no way of making use of this data. 05:02.50 Keith Deutsch Again, if you bring it to your doctor on on ah on ah on a memory stick or something, they'll have no idea what to do with it. So it has to be brought to them in a form that is accessible in the normal practice of medicine. 05:20.43 Ratnadeep Bhattacharjee Interesting. You know, let's dive ah more into it, Keith. When people ah talk about interoperability, right, as you have been, ah as you touched just touched up upon the point, it often gets reduced to just APIs and data exchange, right? 05:35.30 Ratnadeep Bhattacharjee But you have seen probably firsthand how interoperability plays out in actual care environments, right? How would you define interoperability in a way that kind of reflects real world complexity? 05:50.16 Keith Deutsch So you're right that from at some level, at a technical level, you know you are talking APIs, yes. I have a a hosted platform out in the cloud somewhere, and it needs to talk to your EHR, and there's a there's something there's an API for that, right? Yes. 06:06.42 Keith Deutsch But that doesn't that's not the entire story, because there is a workflow. There is a practice sequence that you have to integrate with. If you don't get inside the way that The medical community operates um and the prescription, the way that things are prescribed, you end up getting involved in integrating with the payer scenario in the U.S., which is very complicated. 06:30.57 Keith Deutsch Again, in other places in the world, it's not as hard. But the U.S., insurance companies, the payers are central components and you can't ignore them. And you have to figure out how to weave all the way. I mean, so it's more than just an API. It's this whole system of operation that you've got to get inside of. 06:49.04 Keith Deutsch um And half of it is just human interactions and how they work. Right. um So, ah you know, how does a doctor become aware that this is even an option for them? 07:02.01 Keith Deutsch I'm working with one company that wants to ah change the way remote cardiac care happens with a device. um It's great device. ah it it It is every bit as good as a an ICU resident electrocardiogram, and it can monitor a a patient continuously wherever they are in the world. 07:22.30 Keith Deutsch Great. How does a doctor... you know How does that become available to a doctor, even as an option, and and what do they do with it? So there's a whole flow of, okay, the doctor's got to be aware that they can prescribe this thing you know for for a patient. 07:38.23 Keith Deutsch They've got them aware of that. They have to have a means to do it, which again involves presenting it to the payer for preauthorization, um getting the device attached to somebody because, a you know, patient needs instruction on how to do that, ah registering the device with the system, and then sending the kind of the patient home, but then looking at the data, getting to the data. How do I get to where that data is? Because it's not just dumped into the EHR. EHR is not very good at that kind of large volumes of data. 08:09.06 Keith Deutsch So need another place for it. So all of that is part of what you have to solve. It's not just a matter of saying, OK, there's a FHIR API. i'll just dump my data through the fire atl well nobody wants to do that anyway because they get into the business side of things ah no device manufacturer wants to be just a device manufacturer right they want to be a service provider and that means they need to own the data and control the data and they also need to control what's done with the data because that's where their visible value add comes in but that's that's an integration problem again 08:44.64 Keith Deutsch It's not just a purely technical problem. There is a technical aspect to it. Yes, i mean, you have to understand how these FHIR standards work if you're using FHIR and what the sequences are, some of which are a bit arcane as to how to launch something. 08:57.53 Keith Deutsch But you also have to understand how to fit your device into this entire flow, this entire ecosystem of payers, providers, technicians, and patients. 09:10.71 Keith Deutsch and and And, you know, a bigger picture than just APIs, as you say. 09:18.66 Ratnadeep Bhattacharjee okay I hear a lot of things from you Keith. One of the important keywords you use is FHIR, FHIR. 09:30.01 Ratnadeep Bhattacharjee There is also a very common ah data format called HL7V2. 09:35.44 Keith Deutsch the 09:35.87 Ratnadeep Bhattacharjee ah oh you know there is a ah constant discussions there's a constant discussion happening in the industry of FHIR you know coming in and really you know saving a lot of ah I mean it it is really helping these data engineers and you know analysts to churn out data as well as you know use those data effectively for different you know yeah you know whatever clinical decision making support or whatever you call it various use cases right now 09:49.14 Keith Deutsch Mm-hmm. 10:04.23 Ratnadeep Bhattacharjee in in in In real scenarios, right, where each of these has ah has still a strong role to play, both, right, whether it's a V2 version or an FHIR version, 10:16.21 Keith Deutsch Well, and so there's even more than that, because one of the things that's happened is, know, fire first showed up, you know, maybe it was 2017, 2018 when it first really there was an earlier version, but when it first, and at that point, it seemed like it was just going to take over. 10:32.41 Keith Deutsch ah But it hasn't. and And some of this has been resistance from EHR, the big EHR providers like Epic and Cerner, who have resisted to some extent simply embracing the open standard. 10:46.34 Keith Deutsch And that's made things complicated. HL7 v2 and now v3 is still around. ah It's been used even before FHIR as a data form. remember What the HL7 format is, is a is a data format. FHIR is a complete interoperability scheme, an open you know interoperability scheme ah for EHRs. 11:10.05 Keith Deutsch um So at some level, there is a data model inside of FHIR. One of the differences and one of the reasons I like the FHIR data model better, but that's not necessarily... 11:21.96 Keith Deutsch What I like doesn't necessarily matter, right? Because you have to deal with what's there. And you'll find that internally, there's a lot of the HL7 format as the way things are stored. And I don't think anybody has quite completely figured out how to merge the two because they're they're not completely consistent. 11:38.39 Keith Deutsch But if you're on the far side of a FHIR API, what you're going to see is FHIR objects. And so the FHIR server's job will be to convert whatever is that internal format is. So if you're in the if you're in the world of building FHIR servers or building you know a FHIR server on top of any EHR, that conversion is important. 11:57.81 Keith Deutsch If you're outside, and again, most of the world that I've dealt with is we're outside of the EHR, we need to get into that flow. You don't necessarily see the HL7v2 format. 12:10.30 Keith Deutsch What you'll see is FHIR objects, which typically manifest as ad JSON format um and you know more modern than and the kind of h l seven the older HL7 approach. 12:22.71 Keith Deutsch um So again, from the outside, you're probably not going to, I would guess you're going see less of that. But there again, you know nobody has fully adopted even the FHIR model. 12:34.75 Keith Deutsch everybody has their own flavor of it so it's not fire it's epic fire plus plus or cerner fire minus minus however you want to view that um that was not a slight on either epic or cerner by the way love them both but they can be painful because they have their own business objectives to work with um so that's kind of i mean that does that answer your question 13:02.79 Ratnadeep Bhattacharjee yeah Yes, yes, yes. I mean, ah it does, Keith, honestly, more than what I expected. 13:04.71 Keith Deutsch Good. Mm-hmm. 13:10.49 Ratnadeep Bhattacharjee One of the things that we have been ah talking about while you were explaining FHIR and HL7 and integration, you were talking about EHRs, right? And how we important it is for EHR data to be integrated for any of your downstream use cases, right? 13:29.72 Ratnadeep Bhattacharjee so even in ah your experience or real world experience for that matter, right? What are some of the common pitfalls that you have seen organizations running into when trying to implement data integration, especially with EHRs and remote devices that you, that you actively work, right? 13:48.91 Keith Deutsch Well, so there's a few different ways I can answer that. I see this one major challenge in terms of remote devices, which I'll get to in a second. But on a more practical, technical level, you know, again, one of the obstacles is that Epic, Cerner, you know, PointClickCare, all of these individual yeah EHRs have their own flavors, and they're not... For all the fact that FHIR has been around now for seven or eight years as this well-established standard, it has not... There's been... 14:18.68 Keith Deutsch If you'd asked me six or seven years ago, I'd have said, it's all going be fire. Obviously, it's such an obvious solution, but it isn't. It hasn't. It hasn't worked that way. It's still not been fully adopted, and I'm not sure it ever will be. 14:31.64 Keith Deutsch Just at least in the U.S., again, that's the market I'm familiar with. I don't know how this plays out elsewhere in the world, ah but at least in the U.S., the business is so complicated, and the the issues driving decisions that these companies big EHR companies make around what they'll support is driven by so many other motivations than the purely obvious technical needs that I don't know when that will ever happen. 14:58.07 Keith Deutsch But from the standpoint of um you know devices and remote devices, and there are so many of them that are emerging, And each one, some of them are, some of them are some some folks are taking the approach of, well, we will want to integrate a bunch of ambient devices. We don't want to have to go through this prescription thing. 15:18.24 Keith Deutsch We just want to take devices that are out there, pull in that data, and do something useful with them. And ambient devices can be, you know wearable devices. They can be cameras. 15:29.15 Keith Deutsch ah In one case, they were smart mirrors. So in assisted living facilities, they were putting mirrors in with cameras and intelligence that would do a ah facial analysis and determine stress levels and things like that. and It's hard to, i mean, this is not traditional clinical information, but it can be used that way and there's some openness to doing that. 15:51.64 Keith Deutsch So got all these different data streams coming in. And the problem is that each of these companies wants to present themselves as a service. And the problem that creates is not just the technical, you know, how do I, as one of these device defenders, build my service and then integrate across all these EHRs. The problem becomes, I'm a hospital or a hospital chain. So like, you know, one of the big hospital chains in my area is Sutter or Kaiser is another one. 16:19.99 Keith Deutsch um How many different integrations is a Sutter or a Kaiser going to be willing to do? Do they need to integrate individually with every one of these device platforms and then help their train their physicians on every one of these device platforms? 16:35.79 Keith Deutsch So there's an integration needed. One of the things that I've been a big fan of, but I haven't seen a merge quite yet. I'm working with some companies who are trying to do this, is this notion of let's create some kind of a hub, a single point of interaction, single point of integration. 16:51.60 Keith Deutsch Then the challenge is, how do I go to these device vendors and say, you're not giving up anything by using our hub other than the pain of individually integrating, right? So we're we're going to provide you with a way for you to add your device and your data, keep it separate, and your special source, your analysis, your clinical decision support mechanisms on top of our platform, but it'll all become visible in a single unified way. 17:16.97 Keith Deutsch If somebody can pull that off, I think there's a huge opportunity, but it's hard, right? Because everybody has to trust you and there'll be a couple of them if they emerge. my I would guess it would have to come from a sizable company because they have to have the mass to be a sense of gravity. 17:33.40 Keith Deutsch But that to me is one of the big things that hasn't happened that needs to happen. And so any of you folks out there who are listening to this and thinking about going that route, you've got my support. 17:45.14 Keith Deutsch yeah and and And I wish you well because somebody needs to do that. I think that's the next big breakthrough that has to happen to enable this wave of devices to really become present. 17:59.70 Ratnadeep Bhattacharjee Yeah, I mean, I can see kids that, you know, number one, you're really passionate about this stuff. And secondly, it's very evident that you have been doing a lot of work around remote and wearable devices within clinical workflows, right? 18:16.72 Ratnadeep Bhattacharjee I wanted to further further understand, right? 18:17.19 Keith Deutsch Mm-hmm. 18:19.51 Ratnadeep Bhattacharjee What are some of the promising use cases? I think you've already touched upon one of those use cases, but what are some of 18:25.91 Keith Deutsch Right. 18:26.95 Ratnadeep Bhattacharjee the other promising use cases you have seen in remote patient monitoring or monitoring as a whole, where the integration of device data has genuinely improved, you know, outcomes or reduced clinical burdens. 18:39.81 Keith Deutsch Well, there's strong evidence that they could genuinely improve outcomes. and The problem has been that all of these things are still nascent. There's kind of two universes that I'd split this up into. 18:51.33 Keith Deutsch And again, this goes to the kinds of technologies. There's a set of naturally remote patients out there that, you know the the assisted living communities and this sort of place, you know, the the nursing homes, these are not hospital settings. 19:07.18 Keith Deutsch but they need observation, and by definition, definition they're remote. There's a certain set of patients that are always going to be remote because they're they're they're ambulatory. But but but this is long-term, and it's not necessarily, you know, it's not like, so the the example I gave you of the cardiac, that's an acute situation. That's not something where you're going to strap that on somebody for for a year for for the next 10 years, right, and you're monitoring them forever. 19:36.21 Keith Deutsch So the ambient devices tend to be in that in that first set of groups, you know, where you're trying to remotely monitor and give useful information to to clinicians in a hospital from a group of patients that are not in a hospital. They're, again, they're in nursing homes, assisted living facilities, and so forth. 19:53.83 Keith Deutsch That's one big area of of where this stuff is becoming critical. And, you know, some of it falls out of the so-called surveillance society. You see this elsewhere, but this is a clinical flavor of that, right? 20:07.20 Keith Deutsch um These are people who you're not surveilling them because they're criminal. They might be criminals. You're surveilling them because you need to know if they're okay, right? And and it's to their benefit. But then there's the more clinical set of use cases. 20:21.14 Keith Deutsch And these are devices like the one I described. you know, you don't want to keep a a cardiac patient in the hospital longer than they need to be. But oftentimes, they need to be observed. They need clinical quality observation, um you know, and how do I do that? 20:41.45 Keith Deutsch So you have these more clinic, and there the problem for the device manufacturers is that they need more like FDA approval kinds of things. They need to meet a much higher bar. So they have an additional issue beyond, um you know, just integration, right? they need They need to meet the clinical bar. they have to demonstrate that the use case exists, they're valid, and that they are reliable. 21:04.68 Keith Deutsch um So, you know, cardiac is is one thing. Other things are a little more invasive. There are all of these glucose monitoring devices for insulin patients, you know, that that need to be, not insulin, diabetic patients um that need to be monitored. 21:23.57 Keith Deutsch um There are other blood-borne diseases, other disease conditions that require a little more invasion, but also can be monitored. um There are, you know, not necessarily ECG, but other kinds of... 21:40.88 Keith Deutsch non-invasive ah wearable monitors um and sometimes you want to combine those with ambient data and that's an interesting problem as well i'm kind of wandering in various tangents here because this is a rich a rich area that you've talked about um so i'll try not to take up the entire hour on that but um you know um These clinical devices have the, you know, they they might be more invasive. They're harder to attach to the patient. You know, the ECG patch, for example, has to be attached correctly. You can't just slap it on your chest. 22:11.61 Keith Deutsch It needs to be positioned reasonably well, right? um You know, if you just if you just let the patient say, stick it on you somewhere, they'll get it wrong, right? So, you know, um and and typically they also need help setting up an application on their phone to to send the data up to the cloud and so forth. 22:30.60 Keith Deutsch So that needs support as well. So there's there's additional factors there. But again, if you look at those two broad areas, there are the more clinical devices of which there are some of them are variants of like the the Aura ring. 22:46.98 Keith Deutsch and Suppose it is like your smartwatch, but It reputedly, because if it's more specialized and and in um you know it's a ring, so it has a ah better bond, it supposedly can produce much better data. 23:00.06 Keith Deutsch And so it might wrap it's trying to ramp itself up to the clinical world from the sort of more casual world of observation. And these devices are are there. And in some sense, i it's that I've been more involved with those um because of the more technical issues and the more 23:16.76 Ratnadeep Bhattacharjee Thank you. 23:20.32 Keith Deutsch the deeper way they have to get into the workflows of of the physician. But both of those use cases are out there, and they're both important. Yeah, 23:29.45 Ratnadeep Bhattacharjee Wow, Keith, I can already see so many things that you must have done throughout your career, right? And these are some only some of the recent ones, I believe. And and I don't even, I mean, as you said, right, if if if you can keep on going, you know, with all the different use cases that you you have come across. 23:46.64 Keith Deutsch yeah well. It's a very rich and it's and it's all relatively recent because it's before the pandemic, there was a lot more resistance to this kind of stuff. And you wondered if this was ever going to happen, even though it seemed like it should. 24:01.06 Keith Deutsch um Now it seems like it's going to happen, but there are some obstacles that have to be overcome. 24:08.18 Ratnadeep Bhattacharjee I completely understand. So it it's an interesting segue because ah so a i mean some of our audience here would be you know founders or product teams or tech leaders in healthcare. care right Probably some are even trying to build with interoperability in mind. right as As a fractional CTO for several startups over the years. 24:31.75 Ratnadeep Bhattacharjee ah what's your advice for them as they try to scale systems that talk to multiple stakeholders, right? These may be your EHRs, this may be your peers, patients, devices and whatnot. 24:47.22 Keith Deutsch Well, before you scale, you got to get into the market. And for most of these companies, getting into the market is so market entry strategy for a disruptive and all of these things are disruptive. 24:58.05 Keith Deutsch You know, new technology is a needle threading problem. um and And most folks don't have a lot of experience. And this is something little plug for myself it's something i do a lot of as a as a fractional cto it's not all technology it's also market entry strategy because you want you have to understand the technology deeply in order to figure out how do i get this into the market but the problem you have in the market when you disrupt things is you're going to disrupt things um and people have you know things may not work well and this is not just this is not just in the healthcare universe this is universal 25:33.55 Keith Deutsch Every startup that has I've involved been involved in that has had a disruptive product has faced this. You may have a solution to to a problem. there There may be folks out there trying to do something and it really doesn't work the way they're doing it today. 25:48.00 Keith Deutsch It's broken. But that brokenness has become a feature of how they work because they've built in Ways of working around that, you know ways of distributing blame, for example, ways of distributing ownership so that when things fail, they fail in a way that works for them. 26:06.99 Keith Deutsch um The bad news is that they fail, but they lose sight of that because the good news is, okay, I'm okay with a failure now and again because I how to distribute the problem. I'm going to come in with this brand new technology that's going to say, well, um excuse me, but you don't need to distribute the blame anymore because we'll just'll just handle it. 26:25.45 Keith Deutsch They're not going to trust you. And by the way, your technology probably doesn't work reliably enough at that phase for them to trust you, right? i mean, that's real. When you're introducing your first version of a product, it's buggy. It doesn't have all the issues out of the way. 26:39.25 Keith Deutsch you know And you're gonna you're going to say, trust me, and you're going to break these elaborate things schemes that everybody has has learned to live with that's a problem so you have to figure out how to get past that what can i introduce into the market that will be useful that won't first of all make them overly reliant on my being completely accurate because i might not be right but it adds value gets them to understand what i'm doing gets them to start trusting me and gets me some chance to make my product a little bit more robust. 27:15.31 Keith Deutsch why Where can I fit that you know in this scheme? And I've seen lots of different, it's usually you take some aspect of what you're doing, you twist it a little bit to the left and you fit it in in a place where you're extending the reach of the of whoever your customer is rather than changing the way they do things. 27:34.77 Keith Deutsch And this is, you know, I can give you dozens of examples in all kinds of domains, but that's the trick. And it's a needle threading problem. It's not a, you know, gross, can it's not a selling problem, so to speak. you You'll never convince people just by pure force of pitching and sales. 27:51.35 Keith Deutsch you have to but You have to understand their pain and the way they work to a deep enough level that you can fit your first product into their world without breaking their world. 28:04.42 Keith Deutsch Subsequently, you can get them to trust you. And over a period of years in your second and as you scale, you can begin to expand what you do. And, you know, five years later, everybody may say, well, why were we even thinking about any other way of doing things? 28:18.14 Keith Deutsch Right. But you got to get there. And I will tell you that most product managers, and I apologize any product managers that are listening, are not trained in that. 28:29.49 Keith Deutsch They're trained to figure out the next feature of an existing product. They're not trained to say, how do i get past all the resistance that I'm going to face and thread this new product into the market? 28:43.09 Keith Deutsch Right. And, um you know, you see that with in the health care place in spades, because there is so much of this kind of thing I've talked about where everybody's got somebody else to blame. 28:55.75 Keith Deutsch They have to because that's because there's so much blame to go around. And there's this careful, delicate balance. Is it the payer's fault? Is it the doctor's fault? Is it the technician's fault? um you know What do I have to do to document all of this so that I'm okay? you know What codes do I use that make it seem like it's all cut and dried? 29:18.72 Keith Deutsch um and so forth. And you're going to come in with something that disrupts all that. You can't. You have to be very careful, you know, how you get yourself in there and think that through. 29:31.38 Keith Deutsch um And I've seen lots of folks thinking through like, okay, I'm going to describe this hub strategy. Well, who am I the hub for? Who's going to drive that? Well, one folks, some set of folks think, well, maybe the payers They're the ones who are the one who don't want to pay as much, and and ultimately, they yeah they this is to their benefit, so maybe they're the right place. 29:53.04 Keith Deutsch Other folks say, well, no, they don't want to do that because the doctors will never trust them, so we have to come in a different route. I don't necessarily know what the right answer is right now yet, but that is part of what a lot of these folks need to figure out is how do I introduce this? What's the right point of entry? 30:10.19 Keith Deutsch Even if I'm producing this hub that will integrate all these devices in wonderful ways, what's the right point of entry? and And so that's for early for folks who are thinking about starting something. 30:21.50 Keith Deutsch Be aware of that. It's not the case that if you build a better mousetrap, the world will be i mean ah a path to your door. That that that story is is a fantasy. 30:32.75 Keith Deutsch It doesn't work that way. If you'll build a better mousetrap and the world will say, i can't use that because I don't trust it enough to replace this this horrible way of managing the mice now. 30:35.61 Ratnadeep Bhattacharjee Wow, wow. 30:48.51 Ratnadeep Bhattacharjee No, I get this is so, so, ah you know, enriching. I'm pretty sure some of our listeners must be taking notes and, you know, writing down important, you know, information. Once the information, these are all knowledge flowing ah through your words. Right. I think I'm I'm pretty sure they must be taking notes. 31:09.07 Ratnadeep Bhattacharjee oh it Let's talk a little bit about the future now. like What really excites you most about what's coming in the world of interoperability? 31:21.64 Ratnadeep Bhattacharjee you know oh art isio Are the new FHIR paradigms, things like smart on FHIR or even LLMs, large language models leveraging clinical data areas? 31:33.83 Keith Deutsch Well, but so I would like to get past the interoperability phase and assume that that happens and then say, what do what are the opportunities past that? 31:35.82 Ratnadeep Bhattacharjee ah Do you think 31:43.71 Keith Deutsch Because to some extent, this' this conversation around fire and so forth, in my view, it should have happened years ago, right? i It's amazingly frustrating that it's still stumbling along all these years later. 31:59.47 Keith Deutsch So I would prefer and not get excited about that and just wish that was just done. And it's not. It's not. But it's it's going to take a while. and And to some extent, some of this is going to have to be forced. But You know, but it's looking past that. 32:14.66 Keith Deutsch and And the opportunity, the big shift that these device creates, aside from the fact that it gets people out of the hospital and allows the the the reach to folks who are elsewhere, who are who are in the world, is the nature of the data. 32:30.26 Keith Deutsch that that that this stuff produces because today the entire healthcare universe is built around very small windows of data that are collected in very expensive ways. 32:42.89 Keith Deutsch Little snapshots of extremely detailed data. You know, you take it an EMR, right? aye and excuse me, an MRI. And great, you've got this snapshot ah of of incredibly detailed data, and it costs vast sums of money to do that, right? 33:03.69 Keith Deutsch You take you know some of these other measuring techniques, and you get these very dense snapshots. What they don't have is continuous running large data from multiple sources over an extended period of time in the world. 33:19.02 Keith Deutsch and all of a sudden if you have that kind of data you can do all sorts and and some of these things you discover is so for just you come back to the ECG one and I'm excited about that because that we work with a company that's doing it turns out that and this will never happen in a hospital setting because in a hospital setting they attach the ECG to you and they take five seconds worth of data and that's it and then they unattach it and they have the data if you have continuous data of that electrical nature, it turns out there are all kinds of ways that you, there are all kinds of additional things. You can, you can extract blood pressure. 33:54.07 Keith Deutsch You can extract information about sepsis, about shock, um all kinds of, things can be figured out when you have a large volume of this electrical, because essentially, stop thinking about an ECG, it's just electrical measurements you know of your body, mostly around your heart, but nevertheless, there's a wealth of things that are buried in there um that with large amounts of data and the right kinds of models, you know and this isn't even LLMs, this is just machine learning, but you can start applying that stuff and you can extract 34:31.38 Keith Deutsch all of this secondary, you know, other other signals other from from this data that were invisible to you before. um And you can do this while the cut while the while the patient is at home or in transport. 34:45.88 Keith Deutsch um You know, um wow, that changes things tremendously. you And that's what gets me excited as I look into the future. The introduction of these big data techniques the things we can start to understand that that are going on in someone's body, that current techniques just, first of all, they cost too much and they don't give you. 35:10.48 Ratnadeep Bhattacharjee Mm-hmm. 35:12.98 Keith Deutsch i mean, I can go on some. mean just Just as an example, and one of the problems you have is that current analytic techniques are so expensive to collect the data that they're only collected in, you have to say, i think this is going on. 35:26.50 Keith Deutsch right you can't just sort of ambiently monitor for a wide range of things you know which is often the case i mean you somebody's got a heart issue but it turns out that there's a stroke problem emerging as ah as a secondary effect but you're not looking for that in the hospital because you don't have because ah because you'd have to have a whole different measuring paradigm right now you can look at it all at once over time And it's it's cost effective to look at everything. And if if it turns out there's a secondary effect going on, you can flag it. 36:00.58 Keith Deutsch Lives are saved. 36:02.65 Ratnadeep Bhattacharjee Yeah, definitely interesting. Interesting. 36:06.30 Keith Deutsch i hope I haven't overwhelmed your your your listeners with all of that. But, you know, this is to me where the excitement comes in. The opportunity for doing those kinds of things is where the excitement. But at first, you've got to get through the interoperability and the integration phase. 36:20.44 Keith Deutsch you know, the nasty details that happened happen to make this possible. 36:28.66 Ratnadeep Bhattacharjee Wow, this has been such a rich conversation, you know, ah thank you so much for joining us on leaders perspective or your experience across standards, you know, product strategy and real world interoperability challenges really adds a lot of depth to this topic. 36:34.11 Keith Deutsch Good. My pleasure. 36:46.26 Ratnadeep Bhattacharjee And I'm sure our listeners will walk away with valuable takeaways. 36:50.76 Keith Deutsch like 36:52.33 Ratnadeep Bhattacharjee To everyone tuning in, thank you for listening. If you enjoyed this episode, please subscribe our channel, share it with your colleagues and stay tuned for more conversation with thought leaders like Keith shaping the future of healthcare care tech. 37:04.72 Ratnadeep Bhattacharjee Until next time, I'm Ratna Deeb, signing off.
