Why MultiCare Chose to Invest in an Open Source Data Analytics Platform !

Tacoma, Wash.-based MultiCare Health System’s accountable care organization has partnered with an open source data analytics platform company called Tuva, and MultiCare’s venture arm has invested in the company. Anna Taylor, associate vice president of population health and value-based care at MultiCare Connected Care (MCC), and Tuva CEO Aaron Neiderhiser recently spoke with Healthcare Innovation about the opportunities the open source framework opens up.

Salt Lake City-based Tuva Health says its goal is to establish the open standard for healthcare data transformation and unlock the true potential of data to transform health and healthcare for every organization.

MCC is a wholly owned subsidiary of MultiCare Health System that operates as an independent entity. MCC has established a clinically integrated network comprised of doctors and other healthcare providers, as well as hospitals, clinics and other healthcare services, such as imaging, labs and pharmacies.

Neiderhiser is a former Health Catalyst employee, and co-founder Coco Zuloaga previously worked at Strive Health, which focuses on chronic kidney disease with a value-based care approach. The two are squash players and discussed forming the new company between games of squash, Neiderhiser said.

HCI: Aaron, could you tell the story behind the foundation of Tuva and the problem you and your co-founder were trying to solve?

Neiderhiser: Coco was leading the data team at Strive and I was leading a team at Health Catalyst that was bringing in clinical and claims data from across the entire customer base into a single repository. It was one of the largest clinical and claims data sets in the world, and we were using that data to do benchmarking, to train machine learning models to generate evidence for pharma from a real-world evidence standpoint.

The more we talked, we realized our teams were building the exact same things. We need a common data model to standardize clinical and claims data. We need all these terminology sets. We need data quality testing of the clinical and claims data. We need these higher level concepts built into the data — like how do you define different therapies or conditions or healthcare services?

The more we chatted, the more we thought we're completely reinventing the wheel on this stuff. It took longer than this, but that's ultimately what became Tuva. Everybody who's dealing with population-scale healthcare data, whether you're doing value-based care or whether you're doing real-world evidence from a pharma standpoint, you're dealing with the same problems, and there are no good tools out there. As an industry, we just keep reinventing the wheel, solving these problems over and over again. So the idea behind Tuva is what if we open-source all this stuff? What if we give these tools to the people in the teams that need them? We could move past these foundational problems and actually start spending more time analyzing the data to get interesting insights out of it.

HCI: What are some of the implications from a business model perspective of it being open source?

Neiderhiser: We went down the open source path for two reasons. One is we imagined ourselves working at other companies that discovered Tuva, and we imagined our stuff being behind a paywall. If we built all this stuff and we couldn't use it, we would just, like, kill ourselves. So we said OK, we can't do that.

The other thing is that the healthcare analytics space is a very crowded industry. There are a few very big companies, and there are lots of smaller companies. There's also a long tail of consultants doing this stuff. Whenever you're doing anything in business, first and foremost, you have to have a very clear idea of how you're different. I think that's even more important than the business model. We knew with open source that it would be different. The bet is OK, it does make it harder to build the company at first, because you're giving away all this technology that you’re spending money to develop, and the early business model can just be services, right? But now we're getting to the point where we say let's open-source all this foundational stuff, and then we can build technology to solve harder problems that arise. That's the stage that we're getting into.

HCI: Anna, could you talk about some of the things the team at MultiCare was perhaps dissatisfied about with their previous data analytics infrastructure, and why you were open to looking at something taking a new approach?

Taylor: All of our foundations are built on the economic model of fee for service, and we are trying to perform in both fee for service and value. We needed an infrastructure that serves our ability to have a P&L for both models, so that when we're running volume through the ED, we know how it impacts our risk-based lives, and that is a different data infrastructure than we have today. We knew we had to transform to survive. We are a not-for-profit health system in Washington state, and we want to continue to be independent. To be successful, we needed to be able to run both financial models.

Tuva was an answer for us to clearly understand what the architecture was underneath. It was visible, transparent to us, and it was a low-cost option. We have contracts that we can run through other services that afford them. We might have a fully capitated product, like our employee health plan, where we we own the bottom line, that we run through a platform like Innovaccer, let's say. But for the contracts that may not afford us that capability, we needed a solution where we could house all this data and put agents on top of it so that I'm plugging and playing across the data infrastructure and ecosystem. We wanted a center of the universe that did that for any type of contract that we would have in place, both fee for service and risk-based contracts.

HCI: Did I see you quoted as saying that you actually considered building something like this internally before you found Tuva?

Taylor: Yes, that’s right. We said, OK, there's nothing out there that you can buy that is going to give you this transparency. It's a black box. We wanted to build our own infrastructure, because there's nothing that was going to serve both worlds in this sophisticated way and and enable us to put it on something modern, like Fabric or AWS, so we can take advantage of those services, too. So we were going to build it ourselves, but then our actuaries heard about Tuva, and our data scientists took a look at it, and it was the perfect match for our problem.

HCI: Could the open source nature of this enable things developed at one health system to be taken advantage of by other health system partners without them having to reinvent the wheel?

Taylor: Deep in my heart and written into the values of MultiCare is the fact that we don't want to compete on this. What we want to compete on is how much care we're providing the community. As Aaron described, health systems are solving this 100 times over. We don't need to do that anymore. We can just have this semantic, shared infrastructure that we have the ability to customize to our enterprise culture, and that is what's going to give us that edge, because whatever customization we do is to lead to better service, better health. But the basics should be shared, because we we shouldn't be competing on that in the marketplace.

HCI: Anything else you want to add?

Taylor: We’re all trying to solve this really hard problem with a lot fewer resources than we had before the pandemic because we're all still in deep recovery mode. It's incredibly energizing to find a place that has an answer that is not a million dollars, because that seems to be the price tag for every agent that we're trying to solve healthcare with: a million dollars.

We're hoping to have some great results by the end of the year. So far, we deployed the data warehouse in five weeks. We were in production, and we ran contracts through there in three weeks and had them in QA, and we're doing data analysis out of there. So in in a matter of eight weeks, we had an enterprise data warehouse, which is amazing.

International Research Data Analysis Excellence Awards

Theme: Exploring Recent Research and Advancements in Research Data Analysis

The International Research Data Analysis Excellence Awards celebrate groundbreaking contributions in the field of research data analysis. This year’s theme, "Exploring Recent Research and Advancements in Research Data Analysis," highlights the latest innovations, methodologies, and transformative applications that drive scientific discovery and practical solutions.

By recognizing outstanding researchers, teams, and organizations, these awards aim to:

Honor Excellence – Acknowledge remarkable achievements in data-driven research and innovation.


Visit Our Website : researchdataanalysis.com
Nomination Link : researchdataanalysis.com/award-nomination
Registration Link : researchdataanalysis.com/award-registration
member link : researchdataanalysis.com/conference-abstract-submission
Awards-Winners : researchdataanalysis.com/awards-winners
Contact us : contact@researchdataanalysis.com

Get Connected Here:
==================
Facebook : www.facebook.com/profile.php?id=61550609841317
Twitter : twitter.com/Dataanalys57236
Pinterest : in.pinterest.com/dataanalysisconference
Blog : dataanalysisconference.blogspot.com
Instagram : www.instagram.com/eleen_marissa