Today on This Week Health.

We actually call it clear box analytics because I think that's the other thing that got people in trouble a lot with data science in healthcare is people are trying to create these models or create these black box algorithms to say distrust this. And that doesn't work. Right. So we're really focused on how can we make it really flexible and easy to use? How can we make it really transparent. And how could we do it so that the average person, like you or I can actually create a predictive model and apply that into a setting.

Today, we have a solution showcase from ViVE 2022, and we're going to talk to Field, President of 📍 Clearsense and Chuck Podesta, the CIO for Reknown Health. And we're going to talk about data, essentially. Creating your data solution, your data platform and what Chuck has done. I'm actually gonna share some of the things I did as a CIO because Clearsense was a platform that I used across the board. And Chuck used it back in the day at UCI down the street from me and also now at Reknown. So I hope you enjoy this 📍 episode.

Here we from ViVE 2022 and we are joined by Chuck Podesta, CIO of Reknown Health. We used to be CIOs right down the street from each other.

Yeah, UCI. That's right.

I was at St. Joe's in Orange County. I could see your building every day when I drove by it, which was pretty awesome. And Kevin Field, Clearsense President. We're going to talk a little bit about data architecture and bringing all this disparate data together and cleaning it up. It's interesting cause this is one of the things that I can really weigh in on because we were a we were a client of Clearsense when I was at St. Joe's and you were an early client as well. Give us an idea of that origins story. Because you were around at the very beginning.

Yeah. So we, Charles Boicey is kind of the, one of the founders of Clearsense used to work for UCI. And then he left and he came back and at the time we were, we were looking at Clearsense. We were looking for a solution that wasn't SQL based, right. There was more a data lake, data lake house beyond what we were doing in SQL because we wanted to get into predictive analytics.

And so we started piloting with Charles and his team early on. Small team. And at the time we were taking physiological monitoring data into into Clearsense as part of the pilot. And we were trying to figure out what to do with this data, because we had five years worth of data we were taking every minute of the day for the last five years.

So we had a lot of, a lot of data almost in a streaming format. And so we met with some intensivists and they actually said, well, yeah, let's look at the data from the perspective of patients that crash, whether it's in the ICU or the ED. What happened to those patients 30 minutes before an hour before the day before, and see if we can come up with an alert algorithm around that.

So we mixed in some medications, some labs with the physiological monitoring and actually created an application called Code Blue. And separate from we had Epic as our EHR was a separate alert system and it actually did work. And we documented, I think for patients at the time that the alert fires. It doesn't tell you what to do or what's wrong. It just says, Hey, there's something going on.

This pattern looks like other paths that led to.

That led to 30 minutes, 90 minutes before a patient crashing. Now you can imagine catching a patient before they crash before a heart event or something like that.

That's the best of all worlds because now you're pre heart attack and the much better recovery process associated. And then we went back what we did, it was on those four patients. We went back to the intensivist, showed them Epic at the time. And what Epic was doing with these patients and said, would you have done anything different by looking at the medical record?

And they said no. It just wasn't. Even though you say try to work because it's real time, it's not real time enough. Right. Because you're not looking at streaming data.

And we're going to talk a lot about clean data, but here's what we're going to do. He's going to share his story. I'm going to share my story. My story's a little different. The Code Blue app's awesome because it's taking physiological data, which is clean. Very clean, right? No one else. No, one's typing it in, it's just all going to native

Natively as well. Yeah.

That's the beauty of it. We're we're seeing people start to look at that model and say, okay, what's the data that we can clean the fastest. Video data we can clean very fast. And the monitoring data is, is pretty clear. When we implemented at St. Joe's we had a little different problem set, right? So first of all, we had a lot of different data repositories, so we've had that going, but what happened almost immediately after we made the investment in Clearsense was we started doing our, our EMR migration.

And one of the things we've done in the past, I found all these applications that were still running out there. And I said, why, why is this still running? It's like, oh, well we have to get, we have to get archived. I'd be like, all right. So why don't we shut that off? It's like, well, we're still getting archived. It's like, it's been four years since we migrated. I mean, we're still paying a license on this and this. I thought the ROI on this was to get rid of this license and we're still paying

Right. Read only too right.

And it's interesting because that was not our intended first use case. Right. But it's such a natural use case. And we just, we started popping all that data in there and we were able to shut off our old BMRs within months instead of years.

And it took that huge ROI very quickly. But the other thing we started to do was we looked at claims in almost that predictive way that, that you were talking about. We were looking at our claims and the stuff we were submitting, and we're looking at patterns and say when you submit claims that look like this, we were we were able to predict which ones were bouncing.

Now we've never finished that product. I don't know if you guys ever finished that. It was an interesting model to look at it and go look, I think within a very high accuracy rate we can predict if the claim that this, this entry person just put in is going to be denied or not. Right. And be proactive to potentially get more information before we even send it advocate rejected, come back and go through another process. But that's the stuff you can do with a data platform per se. So I'm going to have you tie all this together.

Absolutely. Yeah. So I think you both giving really good use cases, right? These are end points that benefit from having good, clean, trusted data. When we started to really look at this and look at code patients. And we started looking at septic patients and other things and looking at the breadcrumbs, we realized that those analytics are only as good as the data that you're feeding into them.

Right. We started looking at the enterprise of these systems. There's EMR data out there. There's operational data. There's all these different patient monitors. There's also environmental data. The list is endless. It's sort of looking at all that. If you could really focus on getting all that data together and trusted and started to build some transparency and governance around it.

You can solve for a lot of problems. And that's really been our focus is being able to figure out how can we get that data centrally located, organized, intelligent and applicable. So it can actually solve problems. So whether it's claims or whether it's streaming analytics for a patient event, you can do any of those things.

So, so you guys are going through a growth spurt right now. So we just had somebody come by the booth and say, well, how long have you guys been around? It's almost like the best kept secret out there. But, I mean, you were, you were talking about use cases back in

2016.

2016. Same thing, 2016. So you guys have been around for a while. The technology is baked and I think I ran into four Clearsense people that have joined the team in the last six months. Absolutely. So an awful lot of records, a lot of systems signed up a lot of records being handled at this point.

That's right. We are. We're growing just exponentially. I think a lot of organizations are starting to think about data differently. I think there's been a lot of misconception that people had access to data in ways that they really needed it and wanted to drive their organizations where there's a lot of people that were trying to become data-driven or starting to save this digital transformation or these kind of buzzy type things. But they didn't really know how to do that. And we've been focusing on this for a long time now. Will they be in data first and data-driven.

So Chuck. And you're now a Clearsense client even now. Where are you? Where are you going to take this? Where are you thinking?

So when you're looking at what COVID has done with telehealth right, and it's really kicked down the digital front door. Right? And so you need a tool like this to collect data from all these, there's much more disparate systems out there than ever have been. Right. Because now you've got patient monitoring in the home. You've got EICUs. You've got virtual telemetry. You got the video visits that you're doing. So you're starting to build a patient persona, so from a patient experience perspective, you want to bring people in a digital way and then decide whether they stay in a digital way or they need to have an analog experience, which would be face-to-face with a provider. So when you start looking at all of that, you've got to have a data, like you got to have an ecosystem that can take all this information from all these disparate systems. Because it's not just Epic in the middle anymore. All these different products out there, whether it's a Zeit or an And or an Amwell. I mean, they're all getting into creating platforms for that virtual journey. Right. And Epic is certainly a big part of that or Cerner or any EHR but it's not the end all anymore.

So you've got to be able to pull that data and you've gotta be able to pull it in natively. You don't, you can't have a hundred people mapping it to SQL. That's impossible. Right. And so it's just been, you can't get into virtual care unless you have a product like Clearsense. It's inevitable now. Just like you, 10 years ago, you couldn't treat patients unless you had an EHR.

Now we're heading in that direction of a, of an analytic platform. So yeah. Yeah.

The data, I mean, data becomes the lifeblood of a lot of these applications. Talk about AI and AI applications on top of this, I know have AI capabilities on top of this engine, not only, not only on the backend to do some really interesting things, but also on the front end to do the, the ingestion and the cleaning I guess.

Yeah. Unfortunately when all of these applications were built in healthcare, there was no master plan, right. There was no standard way that we were gonna store this data, the way that we're going to identify this data, the way we're going to profile that data. So fortunately, there's been a lot of advancements in AI and ML.

And what we're able to do with that as, as we started to pull data into the platform, we can intelligently look at that data and start to profile it and get closer and closer. So we look at data in what we call domains. Domains are things like contracts or providers or clusters of different critical data elements.

And we can actually use the AML to get things into those different domain structures faster. And when we do that allows us to have less people. Armies and armies of people coming in to do ETL exercises anymore. Or if we can actually get that data really into a form that we can use it. And it's flexible enough that it can feed into all sorts of different systems.

It's not plug and play, but it's getting pretty close to that with, with certain systems.

Absolutely. Absolutely. The really nice part is once you learn a system, but once you bring a data set in that benefits everybody, right? So it just becomes more and more intelligent over time.

So on the backend. Are we going to start to see AI applications against the data? I would assume we are.

Yeah, we already are. We're already starting that now. So we have something that we call it approachable AI. We know that healthcare organizations are certainly to lean more and more into creating a data science initiative. Right. They want to be able to get into predictive analytics and you were doing this in 2015, but people are starting to really catch up.

But with that, we have to make it a lot more approachable. We have to be able to have business units that are trying to solve their problems, have access to tools that actually work for them to do so. So that's really where we're focusing as well as how can we make these tools really flexible and easy to use for the, the business person the person actually needs to consume and use that data. So we do have a tool at the end of that, that platform once we get the data trusted, have a govern ready to go, that we can actually apply it into actually looking forward.

It's an interesting data set. So I don't need to have my frontline staff know our pipeline and whatnot.

That's right. That's absolutely right. And we actually call it clear box analytics too, because I think that's the other thing that got people in trouble a lot with data science in healthcare is people are trying to create these models or create these these black box algorithms to say distrust this. And that doesn't work. Right. So we're really focused on how can we make it really flexible and easy to use? How can we make it really transparent. And how could we do it so that the average person, like you or I can actually create a predictive model and apply that into a setting.

So one of the things, again, if people came in to go, Hey, what series are you on? And that kind of stuff, but I'm not sure. I'm not sure my listeners are going to care what series you're on. What they're going to want to know is who else is using it? Are they using it? How many records do you have? This is not like, Hey, we just signed our first health system. You guys are

We're growing pretty rapidly. We're actually working with five of the top 10 largest IDNs in the country right now which is fantastic. With some of our customers we're talking about hundreds of applications worth the data within our environment, which equates to millions and millions of patient records. So lots of volume, lots of variety. Like you said, we've had been the, the quiet player in this space but that's because we had tried to get the data right for such a long time.

So we, we talked about some applications. Are there other applications that these health systems are using that, that maybe Chuck and I didn't touch on at this point?

Yeah. There's a few other ones. So one, first and foremost, we're completely agnostic as a platform, right? So data doesn't come and live and die in Clearsense. So we can pull all that data together, get it prepared and feed it out to any other third party application. So being able to feed that clean data to trust the data to other sources, I think is a big, big differentiator for us. We also do have tools for researchers as well, and I think that's a really big one

Big for us. We just had an affiliation last July with the university of Nevada Reno medical school. And there's a big research aspect of that as well, that we're starting. Clearsense is going to be front and center on that research side. We also have a regional transfer operational center we call Our Talk. It's basically an FAA landing the planes instead of planes it's patients.

We take them through the inpatient journey all the way to discharge and we actually doing remote patient monitoring as well. We just started a pilot with a company called Bio IntelliSense. It's a little thing called a bio button. Slap it on and you can monitor patients. You can monitor their gate.

You can see whether the lying down right side left side what their gate is walking wise, whether they fell. I mean, amazing besides just the vital signs. So now you're talking about getting patients out of the hospital per surgery, a lot faster a day or two sooner average length of stay goes down. Readmission rates go down as well because you're monitoring 24 7. You don't have to have family members come all the way across the country to take care of them because we can give them a report on a daily basis of all the monitoring that's going on and if they everything's green they know their loved one is being taken care of and that's, and that's where we're going.

And so we're actually expanding the, our talk into an ambulatory side from the inpatient side as well and utilizing Clearsense cause we're, we're just starting to gather more and more data and we've got a stored somewhere and then be able to put the reports, whether it's operations, whether it's for research or whether it's just to take care of the patient.

Yeah that's right., I mean, we're really trying to build it so that pulling all that data together, giving people the ability to apply data governance on top of it, we're actually putting that in the hands of the people that understand the data. So we're making decisions about data. It goes out to it or decentralized a group of data stewards.

Right? So they can make decisions around it that builds the trust within the data for that enterprise. And then we can feed it into these other tools where we can do data discovery for researchers, where they could look for different cohorts, relationships with different patients, take it from that and apply it directly into more of a visualization tool. And then they can ultimately apply it into those data science approaches as well.

Are there value-based care applications for this as well?

It's something that we're capable of building. Not today. There's nothing out of the box for it, but the way that we're really designed for the, the ecosystem is to have flexible tools sets so whatever the problem statement is, as long as it's related to data, we can apply it.

You're the data platform.

We're the data platform.

So if I have a clinically integrated network need to build a dashboard across. When people hear that they go, well, why don't you just do it across the HR? Well, in California, our clinically integrated network was 75 EHRs.

Yeah, no, it's the same. We're moving in Nevada. We've got 13 critical access hospitals, right? They have CPSI. They have MEDITECH. And the three or four years ago, if you want to move into remote patient monitoring, and those areas are monitoring their patients in any way, you'd have to put them on your EHR, right?

You'd have to implement Epic in a connect way. With these virtual care platforms, you can plug these CPS sighs and Meditech in, take the data natively into your product Clearsense product, and you're done.

And it's like what was really the driver for doing that. Right. And the more you start to talk to people about it, they don't want to go through this huge change management exercises of having to go through and pull for organization to a new EMR every time they're doing an acquisition, right. They want to do it for the analytics, the data, the ability for them to look at their organization and approve. And if we can do that centrally in a data platform, then there's no need to go through a major conversion work like that.

Well, Rknown's probably not a huge system if I thought about it. And one of the things that St Joseph's. Every year that the data team would come to me and want to double in size. It's like, you're asking for too many things, too many reports, too many, too many integration points, too many new physician practices coming on. We need to double the team. And it was, it was like clockwork. And that's, that's the demands that are on the data. So that's where the ingestion the ETL making that, that aspect easier plugging into. How many? If you were to ask me right now, how many of those different ingestion points do you think you guys can handle this?

Well, we have what, we've hundreds of different connectors, so we can talk about anything from any of the public cloud environments. And we're trying to pull data in from those, all the connectors are built in for those types of things. We've worked with many of the major EMR systems. So all of that, work's already in place. A lot of the ERP systems we're talking about hundreds and hundreds, and as far as volume of how much you want to pull into it, it's really limitless. You can pull as much.

And so if he dreams up something is more than likely it's going to be standard space. You guys are going to be able to pull that data.

That's right. And if it puts out a signal, we probably have a way to pull it in.

Yeah. That's why it's an ecosystem. You're building this leaving. It's a living, breathing kind of environment. So whatever you're doing next year, the year after by investing in Clearsense, you've invested in the future. You're not going to run into a SQL dead end that we have today. And that's why we bought it, was from the standpoint of we view it as we're just going to grow with it for years and years to come.

That's exactly it. I mean, the idea with having a platform is that you don't have to continue to go out and buy another third party application and another application and another application to solve for problems. If there's a platform in place that's really flexible. It can really just build it and grow it into whatever problem state when you haven't solved for it.

One of the questions people are gonna ask and I realize it's irrelevant. I'm going to ask it anyway, which is, they're going to say, Hey, what cloud are you running on it? But it really doesn't matter does it?

It doesn't matter. So we actually started in our own private cloud. We actually started our private cloud because security, right. There was just that there was SOC2, high trust, all those types of things that public clouds just weren't ready for that. Now that said, we know that the market's starting to migrate people are getting more comfortable with this cloud migration. So we actually are standing up in public cloud environments as well with a Reknown, for example, we'll be in Azure.

Okay. So you could be in AWS.

We are designed specifically to be multi-cloud, hybrid cloud, cloud agnostic.

What other technical aspects do I need to know if I'm a CIO or CTO?

Other technical aspects. I think that there are the big things that kind of

Because the President should know all the technical aspects.

Every single, every single ounce of it. I think the I think the big thing to know is that everything is being designed in such a way that it's a, a single pane of glass for people to look into. And on the backend, there's a lot of complexities are happening, right? There's a lot of different connectors. There's a lot of different technologies that go into this. There's a lot of moving parts, but we spend such a long time working to build this together and make it work seamlessly that my technical buyers are probably just going to be able to sleep well at night, knowing that we're handling everything behind the scenes.

What about your clincial buyers? What do they, I mean, I understand. Is Code Blue. Part of the solution set potentially?

Code Blue could be. Absolutely. So we do a lot of work in Code Blue. We do things like renal, failure, heart failure studies, all sorts of things like that. I think for the clinical users, what's actually been something that's gotten us a long way is that we build a lot of trust in that data.

So we always preserve full lineage the whole way back. If you're looking at an output, we can follow every single data element the whole way back to where it came from. So maintaining that transparency, that lineage and data, the governance around data, we know where decisions were made and who made them allows us to start to slowly but surely build trusted data.

It's a single source of truth and you want to create an environment of what I call a 1-800-GET-DATA, right? And that's where you can centralize this, certainly plug and play in a decentralized fashion, but the data centralized in a governance way where it is 1-800-GET-DATA.

And so we at St. Joe's we had 800 applications, 1600 instances of those applications and each one had a piece of the record that created that whole person profile. And so when people are like, oh, well you can just use Epic and their whatever or Cerner and their solution around that. But at the end of the day where you really want to pull in is all these disparate pieces. And that's what you give us. You give us a snapshot of really as many of those systems, we want to bring the data in.

We call that building golden records. Right? So we know that sometimes if you're looking at a provider, the specialty might be a one time. And maybe they're demographic informations in another system. Or maybe their hours are in another system. Right. And there's all these different sources of truth that have critical data elements that are more truthful than others. So when you're looking at creating the ideal provider record, you're not going to go to one system, right. It's really going to be an aggregate of all those different systems pulled together to create a trusted golden record. So we do that for any type of data across the enterprise.

Delivering data into the workflow. That's the other thing, clinicians would ask for.

Absolutely. Yep. So you have to, if you can't get it back to that final mile nobody's going to use it. That's a, that's a classic thing. That's why we say nothing comes to live or die. So if there's any sort of ability to feed information back out through APIs, through integrations, through interfaces and what have you.

And it's imperative today with a nursing shortage, you can't find enough nurses, you can't find enough clinicians. I can't tell you how many hundreds of open positions we have that we can't fill. We're using travelers. Everybody's using travelers on the clinical side now. And so it's imperative to automate that workflow, their work as much as possible using these types of tools. We're not looking to replace them. We're looking to get them more productive because we don't have as many of them anymore. And it doesn't look like that's coming back anytime soon. Right? Yeah.

So let's talk about last thing. Last but not least the CFO's perspective ROI. Right? So that's the magic word. Absolutely. We saw almost immediate ROI with the, and I know you don't want to get pigeonholed here cause it's so easy to get pigeonholed into the application rationalization space, but there's an awful lot of ROI there.

Absolutely. I think it's always a good place to operate into it. It's always going to be core to our business, to do that work for the exact reasons you're talking about. We can preserve the data and we can help reduce costs. And at one of our clients, we've actually saved them a $20 million in recurring costs to date in about 18 months.

That's real money. That's real money. And actually it was a kind of a moving story quite honestly. One of our last meetings with them, we'd go and do some quarterly meetings, they had, one of their executives came in the room and said, this initiative is reducing our costs by this much, it's ahead of target and we're able to actually use those dollars to hire more nurses, hire more doctors and actually put it back into clinical care. So whether it's directly or indirectly that, that savings and cost reduction, it really improves the organization.

Yeah. Our business sponsor was the CFO for this initiative and she was on, she was great. She was on more meetings with Clearsense than I was. Yeah. She was really into that. And we also bought the archiving component cause we've got a lot of read only type systems, same as you had paying the software maintenance fees. So there's a huge ROI there. But beyond that, she recognized that this was the future not just in, in, in research and other things, but also in operations. In her operations using AI and some things that she could benefit from so I can see. Yeah, she was the final signature on it, so yeah. That's absolutely right.

Wow. Technical buyer. Clinical buyer. CFO. Sounds like you've, you've covered all the things I need.

Yeah. Fantastic. Thank you for your time. Appreciate it. Thanks so much. Really appreciate 📍 it.

What a great conversation with Kevin Field and Chuck Podesta. I loved exchanging the conversation with Chuck around what we've done. This is one of those technologies that I used as a CIO and I'm really happy that they're a sponsor of This Week Health so that I could share this information with you. So we appreciate Clearsense making this conversation possible. If you're looking for some more conversations like this, this is the conference channel. We have another YouTube channel. It's called This Week health newsroom and all the 📍 interviews that I've done from this week are actually on the newsroom channel. So head on over there and check those out. 📍