This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.
Hey, podcast listener. Thanks for listening. If you're enjoying this week in Health it, we just wanna give you some information on how you can support these conversations to keep them going. This week In Health, it's goal is to keep you, your organization and your employees updated with the emerging thought and trends in the healthcare industry.
Through our conversations with healthcare and technology leaders. The best and easiest way you can show your support is to go over to this week in Health IT on iTunes and leave us a review. Also, you can subscribe on iTunes, Google Play, or Stitcher, or go over to our YouTube page and subscribe and hit the notification bell.
Again, we really appreciate you spending your valuable time listening to this podcast.
Welcome to this Weekend in Health it where we discuss the news, information and emerging thought with leaders from across the healthcare industry. This is episode number 17. It's Friday, May 4th. Today, we used the VA as a canvas for. What a health system might accomplish with the right platforms. This podcast is brought to you by health lyrics.
Uh, is your health system positioned to ride the wave of change? We've delivered agile, efficient, and cost effective it for healthcare? Get ahead of the wave. Visit health lyrics.com to schedule your free consultation. My name is Bill Russell, recovering Healthcare, c I o, writer and consultant with previously mentioned Health Lyric.
I'm always asking people, you know, who they would like, uh, for me to have on the show, who, who's interesting, who's doing, uh, fun things. And today's guest is actually the name which comes up the most. Uh, I had dinner with some of his, uh, compatriots last week in Arizona, and they told me not to tell him that, but, uh, it, it just flat, flat out.
Uh, people say, you know, we want to hear from, uh, from Dale Sanders. So today I'm joined by Dale Sanders, president of Technology for Health Catalyst. Uh, Thanks, bill. That's terrible pressure, by the way, friend. Now I don't know how I'd live up to that . Uh, well, uh, yeah, it is funny. I was having, having dinner with a couple of, uh, people at the Scottsdale Institute last week and, and, uh, it turns out they were from Health Catalyst.mountain Healthcare from, uh,: Warehousing Association from:
You served as a C I O for a national, for the National Health System of the Cayman Islands. So did you live in the Cayman Islands? Yeah, I did. What was that? What was that like? Uh, it was life changing. I mean, you know, of course it's beautiful for all the reasons that we all would appreciate, right. Uh, I mean, just fantastic people and culture and we, you know, I lived on the beach and I was, you know, I'd swim every morning and, uh, but what was really cool is it's a national health system, uh, you know, an economy that doesn't have personal income tax or corporate income tax.
Wow. Yeah. They provide guaranteed care for everybody on in the population. And there's also a private market there. So it was like this little health ecosystem in a laboratory and I got to work on, you know, everything from national policy with the Minister of Health and the premier down to, you know, configuring web browsers for physicians.
So it was, it. Just great. So the obvious question is why, why did you come back? Why ? It is an island. I mean, it's pretty remote. So, uh, I mean, is that the reason you came back to do something else? Well, I, my mom was in the later years of her life and, um, I just thought to myself, I wanna spend time with her while she's still healthy.
Yeah. So I put a hold on my career and moved back to my hometown in Durango, Colorado. And spent time with her. And as it turned out, it was awesome because she was in great health and two years after I moved back, she passed away in her sleep in her home that she'd lived in for 60 years. So it was a best decision ever to move from the Cayman Islands, even though it was a little hard to.
Yeah. And you know, I've, I've heard that story now. I mean, we're get, this is a tangent, but I've heard that story now from a handful of people that put their, uh, you know, made career decisions based on caring for their parents. And, um, in those latter years, and not a single one have I talked to yet, that has regretted I.
Decision and, uh, you're just another one on that list. So, yeah, I can't imagine. I, it, yeah, it just was wonderful. And, you know, uh, I encourage it if and every, you know, it's not everybody can do it, obviously, but here's just one last comment on this. I really didn't know how I was gonna make a living when I did that.
I just cut the cord and said, I'm gonna figure it out if I have to, you know, 10 bar and drive a forklift, I will do that. Spend time with her. And as it turned out, um, That's about the same time that Health Catalyst kind of got spun up and, but I had no clue that that was gonna happen at that time. So, yeah.
Um, yeah. Things have a way of working out. And then the last piece of information, which is relevant for today's conversation, c i o, on looking glass airborne command posts, uh, in the US US Air Force it support for the Reagan Gorbachev Summits, which Bob, you have so many stories I'd love to go off on, but We'll, we have to get to the show here.
So, uh, nuclear Threat Assessment for the National Security Agency and Star Treaty Chief Architect. For Intel Corps Integrated Logistics, data warehouse and co-founder of Information Technology, uh, international and obviously, uh, president for Health Catalyst. Uh, one of the things we like to do is to ask each of our, uh, co-hosts, uh, to give us an idea of what they're currently working on or what they're excited about.
The, uh, well, you know, it's Health Catalyst centric right now for sure. Um, in the last two years, we've basically rebuilt the technology at Health Catalyst. Uh, for the most part, it almost bulldozed everything we were doing prior to two years ago to the ground and, and, uh, started over. So, uh, this thing that we now call the data operating system, which, um, may sound like a bit of a buzz phrase, but it's, it's the, the term is genuine and it's purposely, um, intended to convey something different about analytics and workflow application development.
So we're having a lot of fun with that, and we're building applications on top of the data operating system. And uh, I'm, you know, what I'm trying to be in this vendor space now is kind of the golden rule vendor, right? I'm trying to be the vendor back to all of my colleagues and friends in the industry that I wish we all would've had, right?
And so that's, it's fun trying to live up to that. It's not easy, but that's, that's what we're doing. Right. And, uh, you know, the thing we hear o over and over again from health systems is we need to make the data actionable and a data operating system. Uh, from the presentation I saw, emerin gave me a presentation on this and a couple others.
Um, it that is, its purpose is to make it actionable, to build applications on it, to be able to move it, to be able to, uh, uh, really make it, uh, something that can be put to work to improve health and health outcomes. Yeah. Very, uh, well, that's, that's the extent of the, what we allow for in, in terms of a, a, a commercial.
We give everybody sort of. Uh, so, you know, our format is, uh, typically we go in the news, then we do some sound bites, then we do, uh, social media close. Um, instead of, uh, doing two stories and going back and forth, I wanted to take this opportunity, uh, to really delve into platforms. It's where you've lived.
It's where I've lived. It's what Anish Chopper calls. Uh, we're supply siders, right? We're the mm-hmm. . We're the people who build out the infrastructure that makes these things possible. And, uh, I'm gonna use. Uh, the, the VA situation as a muse and it, the, it's, it's the same intro I gave, uh, two weeks ago when I was talking to, uh, ed Marks with the Cleveland Clinic.
But Ed, I sort of veered towards, uh, you know, what you would do politically and what you would do, you know, as a c I O to sort of, you know, in your first couple of days at the, uh, at the va. We're gonna take a little different approach of, um, we're not gonna worry about the politics, we're not gonna worry about the culture, we're just gonna talk about.
If we were building this thing from the ground up, If we were, or rebuilding it from where it's at, this is what we would do. So let me give a little background for our listeners and then we'll jump into it. Um, uh, so yeah, you have the d o d, the Coast Guard and the va. They all have E H R stories. Um mm-hmm.
you know, the patients, uh, are obviously really important. I mean, these are the people who risk their lives for us, uh, for the freedoms that we enjoy. These are, uh, these are people that we would all love to serve. Uh, to just give back to them. Uh, the situation in each of these is very public, uh, very political.
So the d o d has had, uh, the, one of the worst EHRs as measured by surveys. Uh, and so they responded to that and they launched a new r, uh, which they called the, uh, m h s Genesis platform. It's built on Cerner Millennium product. And from what we can tell so far, it's uh, it's going pretty well. I mean, I mean, from time to time you'll hear a senator come out, but, uh, we, we understand it's going really well.
Uh, the Coast Guard has, uh, had a fail. The HR implementation on Epic for whatever reason could be change management, you know, whatever, whatever the cause is. They hit the reset button. They chose to latch onto the d o D project, uh, because they believe that the requirements are the same. So the va, this is where we're at now.
So the VA has Vista built on top of C P R S. They recently did a no bid contract that was given to Cerner Millennium to replace Vista. Very public, very controversial. Um, one more piece of information is, um, you know, the VA actually was a pioneer in this data record space. I, people don't remember that, but Yeah.Um, you know, back in:
They sort of us into this. For whatever reason now it's uh, you know, it's aged a little bit. It doesn't plug and play real well with the rest of the health, whatever the rationale is. Mm-hmm. , they decide to replace it. So let's get started on this. Uh, again, we're gonna, so we're gonna take the VA's a backdrop and, um, You know, let's, let's start with our thinking.
So how do you think about this? What, what, what should a health system prioritize in terms of the, uh, Infrastructure, both data, infrastructure, you name it. Uh, and, and how should they think about their E H R? Uh, you know, before we make any e H R decision, how should they be thinking? Should they be thinking about, uh, you know, what things like agility, security, what, what things are we trying to accomplish?
Well, I, I mean, to me, I'd be thinking about agility and personalization. You know, one of the concerns I've always had about EHRs is that they, you know, it's a monolithic, I mean, frankly, all EHRs, this included, are built on pretty old technology that we're never envisioned to be as agile as the modern software engineering, data engineering platforms that we have today.
So they're, they're hard to change, they're hard to modify, they're certainly hard to extend, and, uh, And, and so that's one bias I've always had, and I, I think that bias is shared by a lot of folks, right? I don't think many people would argue against that. But the other thing that's always bothered me, and you know, clinicians routinely reminded me of this in the, in the trenches of care, is that the, the, it's too generic, right?
Most EHRs tend to reflect sort of a primary care, general internal medicine kind of paradigm, and then we twist it and turn it and put baling wire and duct tape on it. For specialties and other more personalized uses in including patients and, uh, you know, the, at the, the same time then the rest of the world.
And I always hold up, you know, my, my iPhone is on a platform. I have somewhere in the neighborhood of 150 to 160 applications on this. Some of them overlap a little bit, but no single vendor dominates the way I interact with the platform. The platform's dominated by Apple for sure, but the apps give me the freedom to choose what I want and interact as I want, and that's the paradigm of society in general that we have to drive healthcare towards.
So a platform of data and infrastructure, but flexible, agile, personalized applications on top of that data. So if you are a cardiologist, you're, if you are on a platform, a cardiologist would see one thing in primary care, not, not necessarily different data, but a different representation of the data based on what you're doing.
Absolutely. What, what makes a platform a platform? Uh, that's a good question. I like that. Um, for me and the way that I've approached the data operating system, and by the way, I'm not plugging Health Catalyst today here, friends. I absolutely, it makes me awkward even. Pretend or fake that I'm plugging Health Catalyst.
What I'm advocating today is things that I think everybody should be doing, including the traditional E H R vendors. They need to have something that looks more like this architecture that we're calling the data operating system. And so for me, if you look at the, the traditional technology stack that starts down in hardware operating system and the application layer.
The thing that we've overlooked, and especially in healthcare, is the data layer between the operating system and the application layer. It's still very, very difficult to work with the data layer in healthcare especially, and so you've got software engineering that has progressed on the platform of the public cloud, so Azure, I mean, what we can spin up now in the public cloud is incredible.
It's been commoditized and it's even better than commoditized. It's capabilities that, you know, you and I never had in our data centers as C I O. Right? And we can spin that up in an afternoon. Well now we've got these incredible software engineering tools and applications on top for agility. But we can spin up applications in Angular and D three at rates, you know, 10 to a hundred times faster than anything I could do earlier in my career.
Yep. But that, that layer between the OS and the. Application is still, that data layer is still very hard to work with in healthcare. So what we need to do is build out something that looks like this data operating system where we take advantage of the public cloud and all that great hybrid architecture that exists now between SQL and NoSQL and, and, and big data, uh, uh, data lake concepts.
We take advantage of the software engineering tools that exist on top of it, but the platform now is the data on top of the infrastructure of the public cloud. And of course, APIs on top of that data so that these, all these brilliant application developers have a much easier environment to work in. So to me, that's what a platform is.
It's take advantage of the public cloud, move it up to the data layer. Curate the data in healthcare, then put APIs on top of that data layer. And then, and then the, the, the, the possibilities are almost limitless about the applications you can write on top of the data. Yep. And you know when, when, uh, when I'm asked that question, 'cause I talk about platforms a lot, I talk about platforms to, uh, to, uh, my clients and, and others, but, but what we talk about is the distinction of a platform is, first of all, it's intentionality.
It was designed for others to build on top of it. There you go. Yeah. And so that's, that's the intentionality of it. And then the other is the layers of abstraction, right? Yeah. So the, there is a data data layer, there's a business logic layer, there's a presentation layer, uh, there's a security framework which envelops that whole thing.
Yeah. And, but all of 'em are accessible programmatically. Yeah, so, so now you hand this to, well, actually you walk into any college now and you can put together a, uh, you know, a hackathon or whatever they're, what they're called today. And literally you can say, all right, here's, here's the APIs. Here's how you access this data and, you know, build it on whatever you want.
Build a web app, build a, uh, a mobile app. Uh, use whatever development platform, uh, you want to develop, but because of it, It, it's specific intentionality to enable an ecosystem of developers. And, uh, you know, in the case of Amazon, it's enable an eco, it's a platform 'cause it enables an ecosystem of, uh, businesses to resell on it.
Uh, iTune. Is another platform. Um, you know, and I, I just, Salesforce is probably the, the classic example, right? Yeah, yeah, exactly. And, uh, and the thing we've experienced, and the thing we're talking about here is the, the level of agility a health system is able to come up with. So, You, you and I have both experienced this.
When you have an E H R and you go, alright, we want to do this new thing in the E H R and, and we did, we did, you know, about 20 of these a year where we sat down with the E H R provider and we said we want this custom. We, we always call 'em customs and yeah, we sit down, each one cost us, I don't know, 250,000 to a million dollars to do.
Wow. 'cause they're, they're complex and they're, yep. And it's across 16 hospitals. So it's, it's not a, it's not a simple thing. Um, whereas if it's built on a platform, uh, the, the integrity of the data is handled by the data layer, the integrity of the business logic is, and you can start to split those things apart and build.
Very personalized applications, uh, for that. Yeah, totally. Is any, is any e h r provider heading in this direction that you know of? Uh, not that I know of. R I mean, you know, there are bits and pieces and of course, you know, fire is emerging and, uh, we can talk about fire later if you want to, but I just don't see it.
And I, and, and, uh, of course, That might sound like I'm, and we are, we compete against E H R vendors, but I, I can tell you that opinion would come from me whether I was competing or not. If I were a C I O in healthcare right now, I would be really disturbed by the lack of progress that the E H R providers have made in this.
But I would also say this, those platforms, the core of those platforms that dominate the market right now started decades ago. Right. And I say all the time, there's a reason we don't build code in VB 32. And, and, you know, and COBOL and FORTRAN anymore. At some point all tech companies have to bulldoze themselves and re-engineer, especially in the old days.
I mean, nowadays you can actually, um, evolve a little more gracefully. But I don't see any evidence that the dominant e h r vendors are bulldozing taking all that high tech money and all those profits. They should be bulldozing that old technology and they should be building next generation EHRs around modern Silicon Valley kinds of concepts.
And I don't see it happening. So let's, uh, I, I, I think I read an article, Jonathan Bush talked about, uh, rebuilding based on microservices and those kind of things. Um, so let's talk about what, what are the core technologies. If you were redesigned, so let's go back to the va. VA has Vista and they have a choice of going to Millennium.
Um, I think it's, by the way, I think it's an irrelevant choice. Me too. Totally. And you know, 15, if you give me $15 billion, my, my response would be, um, I'd, I'd rather stay on Vista. I'm sure operational operationally it does what it does, and then you can build out a platform. But what things would we, you know, I, I threw out microservices and services.
What kinds of things would we look for in a modern e h R from that would define a, a technology platform that, that you think would work well? Um,
I would say, let's give the attributes of the platform and then you can spin off whatever you want to, including an e h R from a platform. Right. Once you have the data, then you've got the platform. There's like unlimited use cases with it, literally. Right? Right. And, and so you, you have to have sort of the abstract layer and the reusable content and logic, the curated data that's facilitated with APIs.
That's fundamental. In healthcare so that the application doesn't have to constantly do that. So registries, you know, core things like registries, um, uh, metrics, value sets, embedded machine learning so that it's not something that you do as an afterthought, but you can call an a p i and you can bind your data to a machine learning AI model without doing it externally to the application.
It's a natural part of the . Of the data first application, um, it certainly has to support real time. It has to support batch analytics, right? Yep. It has to be microservices based. Um, and there's a lot of debate about what does microservices really mean. We can talk about that, but it, it, what it means is continuous delivery, right?
No, more of the, I mean, what, remember back to the days when you had to upgrade an E H R. Uh, it was, um, it was a thousands of hours initiative that took months and months to plan and execute. And even then it was painful, right? As opposed to the microservices continuous release cycle that we see now. But, so lemme comment on that, by the way.
So this, the data operating system we have right now is microservices based, and we're able to push out updates to our apps and the platform now faster than the cultural. Ability of health systems is to adapt to it because the culture, you know, our IT shops are accustomed to very rigorous configuration control and release schedules.
At best, it's like once a month, generally speaking, it's certainly not daily. So that's, that's an interesting thing that's evolving back to the attributes of the platform. Um, it, it, the platform has to support the integration of text discreet and image data. You have to be able to support that and make that a natural part of the data ecosystem.
Um, I think I mentioned real-time streaming. It has to do that. Batch analytics. By the way, the, if folks wanna study this a little more, it's an easy study. Kappa and Lambda architectures in Silicon Valley are, um, the role model that we should all be following. Those design patterns are what is what we should all be following.
Yeah. No, it's, uh, wow. , there's so many directions to go, go from, from here. So, um, but I'm gonna, I'm gonna bring it back a little bit because, uh, we just, you know, we're, we're plumbers, right? So we're talking, we're talking really down into the weeds. Yeah. So let's pull it, let's pull it back a little bit. Um, So we were talking about the va, so the VA's gonna stay with Vista, and Vista is going to be a core, let's call it a transactional platform, right?
Yeah. So it's a transactional thing, and we're gonna build around it. So, uh, the first thing is, you know, cloud DevOps, let's, let's get that sort of stood up and then let's keep that transactional layer there. Um, but then there's probably another layer. So for our health systems that are watching this saying, well, this is a little bit out there.
It's not really out there. What we're talking about is. Using the traditional I, what they would think of as a traditional enterprise data warehouse and making that into, um, yeah. And, and really enabling that platform and then putting a set of APIs and, and, and architecture around that, that can tap into machine learning, tap into AI totally, uh, into the cloud and those kind things.
So this is very practical for. For every health system that's saying, how do we get to the next level? And the E H R isn't moving and it can't move fast enough. No. It's the, the, the architecture of it will not allow it to, to to, to move the quickly. They almost have to do the same thing we're talking about, which is take all the data, move it out, put a new architecture on top of it, and that's what they're gonna build on.
So what does the. What does the traditional c i o, what does the architect of a current health system, how should they be thinking about, let's, let's go in this direction. How should they be thinking about tapping into AI and machine learning? Everyone's talking about it and they're saying, uh, okay, how do I do this?
Do I just move all my stuff to Azure and that's how I'm gonna do it? Well, the, the challenge that you have as a C I O right now is that you're, uh, unless you're in a forward thinking organization, you the, the loyalty and the commitment to the EHRs is significant right now. Right? You've just spent hundreds of millions, if not billions of dollars on an E H R implementation.
Yep. And the notion of doing anything else other than that is, is not very appealing right now. Everybody's a little worn out by it financially and culturally. So what I, what, what I see in the market right now is, um, the hope that the e h R vendors will evolve towards the platform and that they will provide AI and machine learning capability.
I see a lot of hope about that. The other thing that I do see occasionally is, um, uh, a, a very unique, um, niche based AI platform being installed in an organization that extracts, it extracts just e h r data out, and you might run predictive models for, um, cardiovascular events or oncology or something like that.
Uh, and then I see maybe a third tier where, or it's kind of a fascinating, uh, Thing that's happening. There are some organizations that are saying, we recognize that the EHRs on the current trajectory not gonna meet our platform needs, so we're going to build our own. Which is interesting. It's because, I mean, I would, it's attractive because the public cloud has made building your own feel more.
Within Grasp than ever before, right? In the old days of building a data warehouse out, a data warehouse platform, that was a significant infrastructure investment. Very unique skills, hard to do. Now that's been commoditized, but I also draw a parallel between that and the old white box pc. You know, build your own pc.
Um, uh, you know, 20 years ago when we all thought, you know, why should we buy from I B M and HP when I can go buy all the bits and pieces and build my own PC for a fraction of the cost? Right. Nobody does that anymore. Right? And that's the same thing that's gonna happen. It feels like you can build your own platform now, uh, and sustain it.
But the reality is it's a fool's errand because you can't, the platform's been commoditized. The data is the important part, and the curation and the management of the data and the logic is the hard part to scale. Uh, so yeah, that I, uh, I, I have this theory that, um, in probably three to five years, I think the dissatisfaction with the current EHRs will reach a pain point.
I. At which the market demands something else and, and an alternative will start to emerge. And the loyalty to those current E H R vendors is going to be diluted. And I think we're gonna see a bit of a renaissance. I think the second phase of healthcare it renaissance, the real phase is gonna happen in about three to five years.
That's my theory. Yeah, I've, uh, so my, I have a similar theory, and this will actually be a segue. So, uh, my theory is that, uh, c m s continues to step up their game in terms of interoperability. Yeah. Moving that data around. Once we, uh, are able to, I, I have this mantra, free the data, share the data, apply the data, and I think that's what leads to transformation within healthcare.
And if we can, if we can, uh, And if they continue to push. And c m s is Medicare and Medicare's the largest payer, um, and, and e r providers are sort of forced into this, opening up that data set and moving it out. I think you're gonna see in probably that same timeframe, three to five years, uh, really an emergence, uh, a renaissance really around the, the experience around outcomes because, uh, these.
These innovators have been sitting on the sidelines saying, I can't get the data. And they, they, they go to these various ways of getting it. Uh, and then the other thing is it's really expensive to get it. So yeah, fire becomes the other way that you can get it without paying the exorbitant, uh, cost. So where, where do you think we're at on the fire lifecycle?
Obviously, It's an important, uh, step in interoperability. Um, and it's, and it's a huge undertaking. Um, you know, what's the maturity level of the fire a p i? What's the, uh, in terms of working with it? What are you finding? We like it. We, and, and, um, you know, I've been notoriously critical of HL seven. Message oriented architectures.
They're so fragile, you know, outdated as you know, from your background service, you know, ignoring services oriented architectures. Um, so the fact that that fire kind of emerged out of the re the rebels in HL seven has been awesome. And it's a, it's a very solid approach to this problem, and we embrace it.
We like it. It's not moving as quickly, of course, as we'd like it to because it is consortium based, but it's moving fast enough. Um, You know, the e h R vendors are now starting to catch on a little bit. The, the leading E H R vendors for a long time are dragging their feet. Because I had a conversation like this back at Northwestern, I approached one of those leading E H R vendors with concepts around services oriented architectures and APIs.
I described the concept to this leader of the that E H R company and their response, I will never forget, they said, well, Dale, that's very interesting, but we see ourselves as more than a database vendor. And so what they, their mindset was, if we open our APIs, we diminish the value of our product. Which is completely opposite of reality.
So the, those cultures have changed some, now we see, you know, the, the, um, the application, the app stores, you know, coming out of all the leading HR vendors, and there's some progress being made. But the reality is, and I did this in the old days, you can't wrap modern APIs around old architectures and expect a miracle.
Right. That's the bottom line. We did this at Intermountain, right when we had help all written on tandem system and, and tackle and you know, very proprietary languages and we wrote Java based APIs around that environment. I. And so that the, the writing, those Java based APIs around that environment was interesting and kind of helpful, but it didn't do what we needed it to do and it didn't survive.
So you can't just write web services and APIs around old architecture and expect it to be the miracle that people expect. Yeah. Well that's interesting. I, you know, I'd love to go on in this conversation, but, uh, it, it is a 30 minute podcast, so I'm, I'm gonna jump to, um, and actually we, we really could talk about this for, for, oh yeah, it would be fun.
Yeah, it really would be. Uh, I, we're gonna move to our soundbite section. I have five questions. It's just rapid fire. I ask the question, you know, two minute answer on your side. And I know that some of these questions are pretty big, so, You know, give it your best shot. So, first question. Yeah. Okay. Alright, here we go.
First question. How will precision medicine change healthcare in the next three to five years? Uh, probably if I had to guess in pharmacogenetics mostly. Um, yeah, I, I mean I have every time, I have a lot of faith in genetic precision me and I think precision medicine, I'm always thinking about genetics.
But, um, then along comes epigenetics and the microbiome and we think we understand things and suddenly we don't. So, Uh, I think probably pharmacogenetics is the most likely place that it will be. Um, high impact. Yeah. I went down to the human longevity, had my, uh, genome mapped and yeah, they came back with a report and they said, Hey, these medications will work for you.
These won't work for you, uh, kind of thing. And I thought, yeah, that's the future. That's what it looks like. It's, it's very specific, um, you know, uh, dosages and medication, uh, based on our, on our makeup. Question number two. What do you think, what do you think innovation in patient engagement is going to look like?
Well, I think we're going to knock the, um, expectations down about patient engagement. We're already starting to see that in the industry where, um, there's probably two thirds of patients that don't really want to be engaged. That is true. Now that's the reality. Right? That's true. And I'm, I'm kind of one of them.
Don't bother me. Really? Don't engage me. I'm okay. Just kind of leave me alone. Yeah. Don't call me. I'll call you. Yeah. But when I want, when I have a condition, when I wanna be treated, let's make it safe. Let's make it efficient, let's make it personalized. So I think we need to shift this notion that we've gotta constantly engage with patients.
Expect them to engage. Some folks don't wanna be engaged. Some folks can't engage for various reasons. Um, and we just need to refocus and say, let's provide really good, personalized, safe, efficient care. I, I love that answer. That's, that's really interesting. I, I had not taken that tack before. I will put that in my, uh, I'll file that away.
Mm-hmm. . Um, what's the, uh, what's the biggest barrier to innovation in healthcare today? Well, uh, two things. No, don't, don't, don't get yourself in trouble. Just . I'm, I'm not good at that. Actually, getting myself in trouble is what I excel at. Uh, it's certainly the economic environment, the behavioral economics of healthcare and mess, um, the administrative overhead.
Is a mess. That whole, you know, because we have so much administrative inefficiencies and so many rent seeker economic layers, um, the resources we could use for innovation are constantly on the treadmill of overhead. Right. So we've gotta drive towards more efficient economic models. We've gotta disintermediate those inefficient layers in the economic environment to free up money and time for innovation.
That's one thing. Um, reduce the number of measures, right? We have to stop over measuring clinicians. That's one thing that's, you know, that that's not the fault of the ehr, is it makes 'em so unusable. It's we're, we're over measuring clinicians. We need to stop that. It's dis it's, uh, disheartening to them.
But then the other thing that's holding innovation back, frankly, is the software. All other industries are differentiating themselves through software and we can't do it in healthcare. I will agree. Uh, what will it take for machine learning AI to take hold within healthcare? Well, it's starting to take hold, but here's the thing that is a little bit of, um, a fallacy.
It, AI and machine learning require breadth and depth of data, right? My background includes time with N Ss, a pioneers in, in AI, machine learning, unlimited budgets. And what I learned then is that you can't just have rows of data. You need to have lots of facts about those rows of data as breadth and depth.
If you think about it in healthcare, On average, we only see a physician or a hospital three times per year. That's equivalent to our digital sampling rate. We have a very, very thin digital sample about us as patients that the healthcare system sees. So when we talk about big data in healthcare, we're not big data.
We collect a hundred megabytes of data per year per patient. That's nothing. So AI and machine learning will never be fully realized in healthcare until we round out the full digitization of the patient and we shift the focus of data from the limited data set we have in the E H R to a, a patient that's fully censored and collecting data on a, you know, seven by 24 basis of some kind.
So we're doing some cool things and interesting things with AI and healthcare, with the data that we have, but the reality is you can't, you can't believe the results. Too much because the data is so thin. You have to be very careful about believing AI and, and machine learning in healthcare right now.
'cause the data is so thin. That's interesting. Uh, you know, our experience was that the, uh, machines at the bedside provided the best opportunity Yeah. As, as a place to start, um, in, you know, in monitoring and analyzing that data and predicting, uh, uh, you know, events before they were going to happen because you were getting that steady stream of data stream, right?
Yeah. Yep. It's, it's what I call instrumentation of the payload, which comes from my satellite days in the Air Force. It's instrumentation of the patient. Yep. Yep. Uh, what you're huge in the E D W, what's the future of the E D W within healthcare? Well, old school batch oriented, you know, read-only data warehouses are already outdated.
So, um, if you have one of those, you gotta figure out how to get off of it. Um, it's, what it amounts to is that it has to be a platform that looks like these Lambda and kappa architectures. Which is a single stream of data feeding both workflow and batch analytics. Um, and you can run real time applications and analytics off of that platform, but old school batch oriented EDWs are over.
Wow. Well, um, yeah, so I, I know that this, this half hour goes really fast. Uh, let's, let's close out with social media posts. I, I usually try to close out with, uh, you know, something funny this week. I, I, I don't have something funny. I just, something that I find interesting. Uh, David Miller, chief Executive Officer for H C C I O Consulting shared, uh, a, uh, Gartner, uh, Study that said only 1% of CIOs are adopting blockchain for the enterprise.
I, I have theories as to why that is, but uh, that number does not surprise me all that much. Yeah, me either. It's, it's, it's overhyped like Hadoop was, you know, 10 years ago. Um, it's interesting. I mean, we've looked into it very deeply and. Uh, it's fascinating. It's, and I, somebody asked me the other day, shouldn't we be doing more in blockchain?
And I, and the, and my response was, it's almost too disrupt. Yeah, it's almost too disruptive. I mean, technically it has some challenges. It won't work quite as well as people think it will in, in healthcare, but it, it's almost too disruptive to adopt in healthcare right now. I think it will be pretty important maybe five to eight years from now.
Yeah, that's what, that's it. It's, I only have one client who has me working with them on it, and essentially, What I'm saying is it feels like five years out and yeah, there's so many, so many obstacles, but I don't wanna go down that path. Um, so, uh, so what's your, what's your post for the week? Uh, my favorite post of the week came, it's actually a Harvard Business Review case study that came out of the University of Utah, right down the street here in Salt Lake City.
Um, And, and in that the summary of that story was their use of, um, cost data and outcomes data to truly understand the healthcare value equation. So if the healthcare value equation is quality of care over cost of care, That's the value equation. All your care over cost of care. U of u's done a pretty good job with their own cost accounting system, and in particular in surgery, they've done very well.
So they know their costs quite precisely in the surgical area of the U of U. But what they've also done is very formally measured outcomes. Associated with the patients and everything that contributed to that particular outcome. So now they bring all their surgeons together with that analytics and they can explore the health value equation for, you know, both individual patients as well as aggregates of patients.
That's where we, that's fundamentally, that's where we need to go with all forms of decision support in healthcare, right, friend. It's the, it's the, um, it's what I call the outcome or cost per unit of outcome achieved. Yep. And we gotta get away from readmissions as a proxy for outcome. Right? It has to be based on promise and promise.
16 and functional status and that kind of thing. Yeah. And that's the, uh, to enlist physicians in reducing costs, show them the cost article in the Harvard Business Review. And that's interesting. I think we always assume that people want the highest quality of care regardless of cost. And that is, uh, that's just not the case.
Right. And. More and more as we become consumers of, uh, of healthcare, um, you know, quality of life, mental health. I mean, there's so many things that go into it. I don't wanna spend all my money on healthcare, you know, anyway. Right. Regardless. Right. Yeah. That's a good, so that article, uh, goes into the details.
I highly recommend everybody read that article. That's in general what we need to prop, propagate in all areas beyond surgery. Well, that's awesome. Hey, thanks, thanks for coming on the show. Um, is, is there, how can people follow you? Do you, uh, do you have a Twitter handle? Do you, do you out on LinkedIn? Yeah, I'm sort of up and down, you know, on that kind thing.
But I'm on my Twitter handle is d r Sanders, everything that's Dr. Sanders is, that's just my middle initial d r Sanders. And, uh, and then I'm, I'm LinkedIn too, and I post out there every once in a while. . Dr. Said, does that get you in trouble from time to time? Well, that doesn't really, you know, as long as I don't practice medicine, it keeps me safe, but I have to explain to people a lot of times I'm not a doc.
That's funny. Um, awesome. So you could follow me on Twitter at the patient c i o, my writing on the health lyrics website. And, uh, I have a new article out on health system c i o this morning on, uh, data brokers and, um, uh, and that this intermediation of, uh, data brokers due to the, uh, C M s announced our C M SS proposed rule.
We'll see where that, that, see where that goes. Um, don't forget to follow show at this week in h I t and check out a new website this week in health it.com. And, uh, if you like the show, please take a few seconds and, uh, give us a review on iTunes and Google Play. And as always, we are now up over a hundred videos on our YouTube channel this week in health it.com/video.
You can check those out and, uh, please come back every Friday for more news commentary and, uh, information from industry influencer. That's, uh, that's all for now.