This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

This episode is brought to you by Rhapsody. You want to accelerate your efforts on interoperability? Have a chat with Rhapsody. If you want a platform with seamless data integration, one supporting all the protocols for improved operational efficiency, Talk to Rhapsody. They simplify data exchange, helping you drive digital transformation, and deliver better patient care outcomes.

Revolutionize your approach to interoperability. Visit ThisWeekHealth. com slash Rhapsody today and unlock your full potential.

  📍 📍

My name is Bill Russell. I'm a former CIO for a 16 hospital system and creator of This Week Health, where we are dedicated to transforming healthcare one connection at a time. Our keynote show is designed to share conference level value with you every week.

Now, let's jump right into the episode.

(Main)  All right. Today we have an executive interview and I'm talking to the CEO of Rhapsody Health, which is Sagnik Bhattacharya. How did I do?

You did wonderful.

Now that we're past that, we can talk about something that is near and dear to my heart, which is interoperability and the.

movement of patient data across the healthcare ecosystem. Hey, welcome to the show, by the way.

thank

you for

having

me. I'm looking forward to it. This is a anyone who's listened to the show for any period of time knows that this is a topic that I can sometimes get pretty passionate about.

so first of all, tell us a little bit about Rhapsody and your place. I know. Rhapsody has a history of being an integration engine per se, but I'm sure you're much more than that now.

Yeah at Rhapsody, we started out as in the core of being an integration engine. The way we think about our job and the way we help our customers, is really to help accelerate digital health adoption.

So there's lots of innovative technologies out there outside of just the core workflow systems of EHRs or revenue cycle systems. And for most of these innovations, in order for them to get adopted, they need to be integrated into the workflow. And for them to scale, they need to be able to do that integration at a much higher level of scale than they do at the first one or two deployments.

So that's where we are there to help. And the way we do that is through a suite of products, first of which is integration and interoperability. which is around how do you connect up all the systems that need to exchange data to have that seamless workflow for end users. And then on top of that, we do work to improve and enhance data quality.

So how do clinicians get usable data and not just the, someone threw a whole bunch of data at them. So we do that through semantic interoperability so that we can create a usable picture of the patient. We do that through patient identity management and since, Bill, you've been in healthcare, patient identity management is this vexing problem that we have to solve in healthcare, which probably not, is not as big of a problem in other parts of the economy, but unfortunately is in healthcare, and we've been doing that for a number of years.

It does all start with a single patient record. And I remember when I got into St. Joe's and I said, I don't understand the problem. Why can't we get to a single record? And they said 14 percent of the people who showed up at the ED in Orange County did not present a social security number because They didn't have one and they're like, they don't want to be tracked.

Therefore they give us different information every time they show up. and they're like, however, there are some things we can do to try to mitigate that problem. The quality of data is something I'm hearing people talk a lot about now, especially with the advent of AI and other things talk a little bit about patient quality and how.

Rhapsody is moving that forward.. data quality. Sorry.

That's right. Patient data quality has been a bit of a journey for an over a long period of time. In the, let's say 10, 15 years ago, where EHRs were just being implemented, there was most of the data creation, if you will, was very localized, right?

You have your EHR, That physicians and clinicians document in and that generates patient data. And yes, you had some of those challenges around folks without a clear identifier and such, but that's where the origin of patient identity really happened, which is how do you try to use the different pieces of information you can get about the patient to identify the right 📍 person.

What's changed over the last 10 years is, now there is a lot more exchange of data happening between systems. Started off with Commonwealth, Care Everywhere, Care Equality, and what have you. So now there is a lot more data that physicians can see. However, that data sometimes is not commingled with what they already have in the database.

And you're getting the same lab, the same diagnosis, or slightly different labs, et cetera, showing up all at once. And the clinician has to process that by just by looking through reams and reams of data. So I think that's one area where we look for patterns, look for matches and how we can bring all of that data together.

The second part is around just knowing who the patient is. So today, if you're building an AI data set, You're trying to gather data from lots of different locations and knowing how to connect the same patient's information together. to build your AI models or to build these multi tenant SaaS type of digital front doors that people are building, that has become a lot more important than it used to be 10 years ago.

Because the, ultimately, the quality of the solutions of your AI models or what have you really is only as good as the data that's going into it. And specifically within healthcare you can get to a low 90s type of match rate through some probabilistic mechanisms. That's not good enough.

And again, healthcare, that last 1 percent matters because it's all about, patient safety, quality of care, and what have you. So that last mile of getting from the low 90s to the high 90s of patient matching is so critically important. And that's what we have been innovating on over the past several years, including trying, oftentimes people don't do that last mile because it takes a lot of work and a lot of human labor to actually go through the potential matches and mismatches.

So we have introduced some AI enabled automation so that we can reduce the human labor involved in getting it from, say, a 92% to a 98%.

When a health tech startup approaches you, typically, I would imagine it's to. Is to access certain data and as you said earlier, to get into that workflow.

Is that a common thing for a health tech startup come to you? Or is that usually facilitated by a health care provider?

It's usually it's both. And so today we serve about, about a thousand provider organizations on one side. So we act as their integration platform and partner.

So if they need to integrate a digital health technology we do that for them through our platform. We also serve about 350 health tech customers. They range from the startups, like you said, a lot of the up and coming innovations in the industry. As well as many enterprise level players, folks such as Philips or Roche, like when they're deploying their new devices into a healthcare ecosystem or their digital health solutions.

The problem is The same for whether you're a large healthcare technology conglomerate or a small startup, which is your solution, whether it's hardware, software, or a mix thereof, unless they can exchange data seamlessly within the clinician workflow, they're not going to get adopted and they're not going to scale.

So it's, we work on behalf of those health tech companies, big and small, to take on the work of integration for them. And we will deploy that across all of their provider customers as need be.

That's interesting. So you have a background of working at Epic and were you a developer back then?

I started my career as a developer and maybe I can share a very short story to sure, that'd be great. Where I'm coming from. So I started as a software developer at Epic about two years into my career. I had to take a friend of mine to the ER. And at the when we went to the ER, a triage nurse took his vitals and then they took him back.

And thankfully he was okay. It all turned out fine. And then the floor nurse took another set of vitals for him. And in that moment I was really thrilled because I was the developer who had programmed that feature in the system that allowed nurses to take multiple sets of vitals. Now, it's such a small little thing, but, I was really proud and that was a big moment for me.

And that's really been my calling since then, which is. How do you use technology to really solve healthcare problems and impact patient care? And so through that journey, I was at Epic for 16 years. I went from being a software developer to being having product leadership roles.

And then having general management type roles. I used to lead the outpatient EHR and value based care side of Epic. So that's really and it's been amazing to see how the world has evolved over all that time. It was the early days of the early adopters of EHRs. And today, everyone has an EHR.

And so the problems of integration are different, but the core of the clinician experience, the core of getting better patient outcomes still remains the same.

Does that background of knowing the EPIC ecosystem that well, does that serve you from time to time? I would assume it does.

It does, and there have been many trial by Fire

I have spent thousands of hours shadowing clinicians over my career, and it's it's one thing when you're building product, but it generates another level of empathy when you're actually seeing your products used in the wild and just seeing how that operates and really understanding what clinicians are going through.

It really matters. And it definitely influences a lot of my thinking and how I approach things. And through the course of that journey, I've become friends with several CIOs and CMIOs because we've been in the trenches fighting the fires together.

Does that change how you build your culture at Rhapsody Health, how you hire and how you develop programmers at Rhapsody?

Yeah. So at Rhapsody, we are definitely a technology first company. We try to solve problems through technology and partly I have a technical background and education. So that definitely influences that. I think the other thing that I've carried throughout my career and having learned it early on is the, really the importance of customer partnering with customers to get them the outcome that they need because the balancing act is, sometimes on engineering, we get really excited about building cool stuff.

But then on the other side, that making sure that works really well in the wild and we provide excellent customer support. That is key. We take a lot of pride in being best in class. Rhapsody has been ranked best in class in our vertical for the last 14 years running. And to me, that is a real testament of culture.

So technology and customer focus.

When I was CIO, the technology companies that would actually come out and round, that would actually spend time with the doctors they did so much for me as a CIO because they were seen as an extension of the IT organization.

And they were taking the time to really understand how things were being done. And I appreciate companies that. come at the problem with that.

Talk about your healthcare provider clients. I assume the use cases are all over the board. There's a lot of different things. You need to build out a clinically integrated network, you need to cross different EHRs, you need to pull in data from all those disparate systems.

I'm curious, what are the most common use cases for healthcare providers?

Yeah at Rhapsody, we serve probably about 600 odd health systems and then a few hundred clinics, lab organizations, radiology providers, and so on. And we do that across 30 countries. So that's actually a key insight that we have, which is there's a U.

S. healthcare delivery system and then how it's delivered outside of the U. S. We see some variations there, and which is why we are extremely flexible to meet our customers where they are. Some of the key use cases first and foremost is integration of the core system. So you have your EHR typically on one side, but for a radiology provider, it could be their RIS system.

For a lab provider, it's their LIS. And there's a revenue cycle system. Then they have a number of other systems on the surround that they need to connect into the workflow. And it is really inefficient to try to do point to point connections. As an example we have one of our customers, a really large health system in the Midwest.

They have an enterprise EHR, so they're an enterprise Epic shop. But even for that, they have about 200 different systems that they connect together. And I think , that is the fundamental problem that we solve. And the second part of that is, how do you do that scalably? Now Bill, you've been a CIO and I'm sure you have many CIO friends.

The core theme that I hear from CIOs these days is number one, do more with less. Number two do more with ai, right? Or do innovation of the day. Most CIOs that I know, they want to do more of the innovation projects that they come in, that's why they went into information tech.

But oftentimes. They have a backlog of just day to day stuff that is, really hard for them to handle and the budgets are under pressure. So for us, some of the technology innovations that we have been doing is to really figure out how to reduce the cost of integration. And, instead of having a really large team, that is a lot of a huge backlog.

How can we make it faster to integrate? How can we reduce the amount of monitoring that we need to do? There's some. Really interesting technologies that we have invested in around edge computing, which is, you don't want your PHI to live your firewall, but you want to have remote monitoring support.

So there are some technology advancements that have made that possible, which was not possible 10 years ago. So those are some of the primary use cases, like really. Do the integration of tens, if not hundreds of systems in an extremely scalable and extremely efficient manner and do so with extremely high reliability.

And as the integration platform goes down, that health system is having almost like a black swan event. So it is so important to get it right.

AI hasn't been mentioned much so far. I'm going to close with this. Question. Again, not a current sitting CIO, but if I were, and I really thought about it, there's a lot of opportunity here just on the integration side, like to overlay machine learning, and we've already seen some of this machine learning on top of images and whatnot to really move things forward.

But in terms of the interoperability, just, cleaning the data, moving the data, making it available are we applying AI at that fundamental level to really clean up the data and make it more accessible?

Absolutely. There's, in fact, on our website, you'll find some case studies for some of our customers.

One of them is building AI models for medication management. And they have really smart data scientists who really want to spend time building their models. And their challenge was, a lot of their time was just spent in cleansing the data that they acquire from different sources.

So what we do is we help them spend significantly less time in putting that data together into a nicely normalized data set so that they can build the AI models on top of it. Same thing applies for even, I know a lot of large health systems, they have data science teams on staff for themselves too.

, the dirty secret of data science is you do data cleanup more than data science often. So we can reduce that cleanup time significantly. So that's part one. And then part two is and I cannot leave how many conversations I've had around this. People are innovating a lot. There are like, for example, you mentioned radiology.

There are 80 different FDA algorithms that are approved today for different modalities and different body parts. Hardly any of them are getting used because they are, that last mile challenge that they have, which is, How do I fit it into the radiologist's workflow? How do I integrate with the RIS system so that in their work list you have like light bulbs go off to say there is an insight here and not have something that they have to jump out to another system for?

So that last mile workflow piece is so important. And I think most people are realizing that, and that's where we'd see our place, which is how do we bring those innovations into the workflow and help them accelerate?

I was talking to a imaging professional and I said, X rays.

He goes, look, AI can do an X ray as good as any human being. It's just a basic, you're looking at it, you see it, it's very one dimensional. \ it's not really complex. He goes, now, if you're talking, cardiology and some of the other, it just gets more complex as you go along.

But he goes, we can work up that ladder. Pretty quickly. But one of the things you mentioned was exactly what you said. It's like it, none of it matters unless it makes me more effective, unless it gets into my workflow, unless it essentially if I can do 10 reads instead of eight reads then I'm going to adopt the technology and it's going to move forward.

Yeah. That's the ROI for them.

Yep. Absolutely. Hey I want to thank you for your time and I'm looking forward connecting in person sometime soon.

I hope so, and I'm sure we'll have lots of notes to compare from your experience and mine.

Absolutely.  

Thanks for listening to this week's keynote. If you found value, share it with a peer. It's a great chance to discuss and in some cases start a mentoring relationship. One way you can support the show is to subscribe and leave us a rating. it if you could do that. Thanks for listening. That's all for now..