Healthcare systems today are incredibly complex beasts.
From EHRs and lab results to doctors notes and imaging, they hold vast datasets that are structured, managed and processed in different ways across disparate departments, creating major challenges in integration and interoperability.
“It's been at least a half a century since it was as simple as, hey, my doctor knows everything about me,” explains Jitin Asnaani, Chief Product Officer at Rhapsody. “From the moment we had specialization in the U.S., data started getting fragmented.”
In healthcare, this fragmentation is the source of several major challenges.
Not only can it hinder patient care, limiting the ability of providers to create a complete and holistic view of a patient’s health that can, in turn, potentially impact diagnoses and treatment decisions. Additionally, siloed and inconsistent data can also stifle healthcare innovation.
Today, this is one of the leading headaches facing healthcare companies: How can they connect the dots between each patient’s disparate records from multiple different specialist departments to create a clear, holistic picture of each patient’s needs?
In this exclusive interview, we sit down with Asnaani to explore how modern interoperability solutions are transforming the ability of healthcare companies to build effective enterprise-wide master patient indexes (EMPIs) to maintain accurate medical data across their various departments.
Chris Hayden:
Hi, everybody. My name is Chris Hayden. I'm a producer here at Fierce Healthcare and today I'm speaking with Jitin Asnaani, the chief product Officer at Rhapsody. Jitin. Before we get started, can you tell us a little bit about Rhapsody?
Jitin Asnaani:
Absolutely. Nice to be here, Chris. Thanks for inviting me. Rhapsody is a leader in digital health enablement and has built a platform to enable data integration and transformation. We have 1, 700-ish customers all across the globe and have been pioneers and innovators in interoperability for over 20 years.
Chris Hayden:
That's great. That's great. I didn't know you were across the globe. That's some unique challenges, I'm sure.
Jitin Asnaani:
Unique and really interesting challenges as we've tackled interoperability all across the globe and continue to do so. Yeah, it's a fun time.
Chris Hayden:
All right. Today we're going to be talking about Rhapsody's identity solutions. First question is what are some of the key challenges CIOs and product leaders who work with Rhapsody Face?
Jitin Asnaani:
I mentioned we have these 1,700 customers across so many different segments. If you think about the world distilled into two big segment buckets, there are the folks who are providing clinical care, the clinical provider organizations, and for them it's all about just do more with less. Hey, Mr. CIO, do not get in the way of the innovation. We as a healthcare organization are trying to do the innovations we're trying to bring in, and by the way, we need you to enable this innovation as we bring it in for less money than in your budget than you had last year or the year before. So, it's all about do more with less. That's on the healthcare provider side of the story. That's their challenge. That's what they face. On the other side, it's all these digital health innovators, these large digital health companies, these health technology companies, just this amazing market of innovation and transformation and technology that comes to bear.
For them it's really about let's get these innovations into the market, let's get them faster. It's enable interoperability to enable the innovation. As they are in this space, they have to overcome things like cybersecurity concerns and cybersecurity scrutiny have to deal with all across the board. We've had this as an emerging challenge across the industry in all of the world, for example, over the last few years, and it's becoming more and more of a pain point for them. How do we overcome that and how do we overcome resource challenges? That's another aspect that they have to deal with and that's very specific to healthcare. If you think about it, I mentioned the dwindling IT budgets if you're on the provider side of the house, and that's oftentimes who the health tech company and digital health company are trying to sell to. But there's also the issues of physician burnout and the concomitant issue of nursing shortages.
So, it's about how do you get this to the market when you have these hurdles to get over and you critically require the data for your innovation to actually work. This is what they're working through. And then remembering again that this is healthcare. Whichever part of the world you're in on both sides of the equation, both those large segments, which themselves have so many segments, everybody has to deal with the fact that there's multiple sources of data. The data is fragmented. Your ability to utilize and actually innovate, whether you're building the innovation or buying the innovation or renting the innovation, if the data is inaccessible or unreliable or fragmented, you're not going to be able to do the job that you are set out to do. And there, it's all about data quality, that's the name of the game, in order to actually create that business outcome, that clinical outcome that you trying to accomplish.
Chris Hayden:
Why do healthcare and healthcare tech organizations need an enterprise master patient index?
Jitin Asnaani:
For a healthcare organization, if you think about the world they live in, they have patients' data stored on a wide variety of systems, numerous systems. It's fragmented across all of these systems. And if they're not able to pull that data together and actually create a holistic view of the patient, then it affects multiple aspects of their operations. You can think of them as two gross buckets of things that it affects, clinical operations, actually taking care of the patient, knowing what the patient needs, ensuring that they have the right diagnosis and the right care both here and downstream when they transition out of your facility. And by the way, all the business associated with that as well. You could pretend that healthcare is all just about the clinical, but there's also, hey, let me make sure I take care of this patient and I get appropriately paid for it and so on downstream as well. So, all the aspects of the clinical operations that have to be accomplished.
There's also this aspect and this emerging aspect of the patient's engagement. And to the extent that you don't know about your patient to the extent that you'd have only partial information on your patient, it creates increased credibility issue with you between you and the patient and your patient's engagement with your system. And to the extent that you do get it right, it's psychologically fulfilling and a better moment for the patient as well. They engage in a very positive way with the healthcare system or your healthcare whatever, it's innovation clinic, outpatient home health, whatever that setting is. It means more when the patient feels like you actually know them and how do you know them if you can't actually connect the dots on their care. That's a healthcare side of the house.
On the technology provider side of the house, if you think about it, they're building really interesting innovations, analytics. Today, it's more and more AI. If you look at any of these innovations, data is a critical input. As you just said a few minutes ago, it's all about GIGO. That's the name of the game here. If you're giving me data that's poor or fragmented, it doesn't actually give me the quality experience that I need, then my downstream algorithms, whatever they are, are going to either have to really work hard to overcome them, which is a little possible, but actually fairly difficult even with AI and certainly with traditional innovation mechanisms and analytics mechanisms, it's virtually, totally impossible and hard to detect. So, you really need to make sure that you've got the data quality right up front and pull the right data in so that as it feeds through your innovation, what you have on the other side is a product that actually is valuable.
Chris Hayden:
Speaking of which, you just mentioned two things that we'll touch on in this next question, AI and product. I understand you're launching a new AI piece, so how does your upcoming AI launch change the game and can you tell us a little bit about your approach to AI?
Jitin Asnaani:
As I mentioned, if you are a provider organization, in this case, let's even take a technology organization, let's say all of healthcare. You have different data sources you need to connect. In the past, over the last 20, 25, 30 years, it's always been a challenge. It's been at least a half a century since it was as simple as, hey, my doctor knows everything about me and he follows me no matter where I go. It's been a long time since we had that. From the moment we had specialization in the U.S., data started getting fragmented. And the reality is it's pretty hard and it's always been hard to connect up all those internal systems, but you've always had this fallback. If the algorithms that were created, even as it got sophisticated over time, if they did fall down or they just couldn't make a match. Imagine a hospital system in a metropolitan area that has 100 patients named Jay Smith. Without an EMPI, duplicate records or mismatched data could cause major safety point issues.
Even if it's not able to make that match, at the end of the day a human can. A person can look at it and say, yep, based on this, what I know here over here, what I know about this patient, I can make that match and say, this is a match of these two patient records and this is not a match of these two patient records. But now think about where we are today. That was all fine and good for 20, 30 years. Now we have so many more data sources. We have consolidating healthcare systems, we have more connected healthcare systems, which is a good thing. We have multiple sites of care getting connected, ambulatory, post-acute, acute, they're all getting further and further connected, connecting not only your internal systems, but you're connecting to the outside because of government regulation, because of your needs for value-based care, because of the needs of your patients as they move around.
Now you have this dramatically increasing surface area where you need to connect the dots. Human intelligence, as excellent as it may be, just doesn't scale for it. How many people can you hire to do nothing but try to connect these dots as the data itself gets more and more fragmented. By the way, each system you add, each system you connect to has its own vagaries in terms of what data they actually capture and don't capture. So, the problems are becoming really, really hard. Human intelligence, in some ways, is becoming more necessary than ever, except it doesn't scale. That brings us to the answer to your question here, Chris. What we decided is we took, in this particular case, we have a whole toolkit of AI tools as you can imagine as people have seen out there in the world. In this case, what we decided to do is take machine learning as a tool of choice because what we can do with machine learning and through training our algorithms is really learn what the humans are... what choices they make, through a model we call neural networking because it mimics thought processes in the brain.
The interesting thing about these decisions is that it depends on the data you're getting and the business objectives you're trying to account for. So, if you can learn from that organization what it is they're trying to accomplish, you can actually tailor your algorithms, your outputs to be what that human with their intelligence would have articulated as the match that you need to make. We use machine learning there. It allows us to more consistently and at dramatic scale be able to make those human decisions and compliment humans as they... Again, there will always be something that doesn't get matched, and so you always need to be on the forefront of that. But where today traditional EMPI mechanisms, even those which have been refined over dozens of years as ours have, as we've continued being pioneers in this space, even as those try to keep up with what's coming from the outside, being able to actually bring intelligence into the process through the guise of machine learning allows us to actually enable our clients to have that quality data experience even as the surface area underneath them increases.
Chris Hayden:
And you touched on this a little bit in one of your earlier statements, you mentioned just the psychological benefit that patients have when talking about some of the benefits of good data. When you're sitting in a doctor's office and they call up, oh, okay, you saw somebody else and they said XY, it's got to be rewarding for a patient. Can you talk about some of the other key benefits that your customers see?
Jitin Asnaani:
I think about the benefits of our product in two different dimensions. One is if you could give me high quality data by connecting the dots on this patient's records, what would I be able to do with it? That's one dimension, and so let me speak to that first and then I'll give you a slightly different dimension. In that dimension, it's all the aspects of the triple or quadruple aim. I think now it's at the quadruple, I don't know how many more parts of the aim we'll add over time, but in any case, it's helped me be able to take better care of this patient. The better the picture I have, the better care I can take. I can articulate the gaps in care better. I will know for sure that this patient has or has not had certain procedures and I can enable that picture to get better and richer.
Help me to keep costs down by making sure that I'm actually tackling the things that this patient needs, that I'm getting this patient to go get the care they readily require, that I more in sync with both with this patient and the ecosystem around me, the peer, the other providers around who are part of the care of this patient, like it or not. I can help this patient be more satisfied. And this is where you started. Is this patient having a good experience? Well, to the extent that he feels like you really know him or her, then you are in a position to provide a better experience for that patient and then better health for all. That's the opportunity you have by connecting the dots at both the patient level so that you're taking care of communities and at the larger level as you take care of whole communities and panels of patients and so on. As we go down those paths... That's a benefit again.
There's a different dimension though, and here's something that's not always appreciated, particularly from organizations who are now getting enough data coming to them that they've realized maybe I need to do something about making my data more useful than it has. For the longest time, the big problem was just get me the data. If you have a good integration engine like Rhapsody or Corepoint, those are two of our integration engines or you have some mechanism for connecting out there and bring the data or you have access to your vendor systems or you have all of these things. More likely than not the data is finally moving good news and now you're thinking about the data quality. Where we have specifically taken a stand is we've said, look, as you embrace data quality, you might want your quality mechanism to sit within your four walls. You might want it literally on your premises.
In healthcare, the infrastructure modes are like everything from I have my own servers and my own hospital and my old data center to I'm using a public cloud to, hey I'd rather just you do everything and we just want to host and subscribe to what we want, but you host it and maintain it and worry about cybersecurity and all that jazz and we'll just hold you to a contract that tells us that we know we are getting the service level you want. And the thing is, we have deliberately taken a stand of we're going to meet the patient... I'm sorry, not the patient. We're going to meet the healthcare or health tech organization wherever they are on that journey. That's an important dimension because in this industry you can't just take for granted that everybody wants a cloud-based service.
That feels like a very modern thing, but in reality, there's all sorts of great reasons, operational reasons, regulatory reasons. We operate in dozens of countries around the world. Some places, you can't do a cloud-based service for patient data without having to crawl through some very gnarly tunnels, and so it's just easier to deliver it there. This opportunity to allow you to get what you want is really important. There's another dimension here as well within this notion of we'll make it easy for me to actually adopt data quality and which is again, I mentioned... Actually, I mentioned this a little bit earlier. I mentioned that you're getting all these data sources. The data sources your organization connects to the business outcomes your organization wants might be different from the organization right down the street, supposedly in the same business. But if the patient matching algorithm is completely standardized, then how do you know you're actually getting the best outcome for your organization?
We already enable mass customization and we are bringing this combination of AI and mass customization into the product so that you as an organization can really get this EMPI to do the job that you want it to do. So, you can really tune it to get the results that you want. And that's an incredibly important part of the entire convenience picture here. By the way, we're talking about AI and that's what we're launching right now, so we're super excited about it. We brought in other types of really important innovations into the product over time too. Our algorithm has so many different aspects to it that have evolved over time. There's this thing called referential matching, which we brought in a couple of years ago as well, which really allows you to ensure that that patient is who you think they are based on not just healthcare data but outside healthcare references where permissible. We're just trying to solve a hard problem. That's really all we're trying to do.
Chris Hayden:
What makes Rhapsody superior to other EMPI solutions that are available out there?
Jitin Asnaani:
There's probably four or five aspects where I'd say we've really invested, and I think at this point we can fairly clearly say that we lead the industry on these dimensions. First of all, more than three decades of experience in tackling identity. We obviously have folks here who are brand new, bring a lot of the neatest and greatest innovative thinking into the company and into our EMPI product. We also have people here who have been here three decades and they've forgotten more about identity than most experts ever knew. And we have everything in between. We have an amazing team tackling identity and just really knows the space inside and out. We have a proven tool. The EMPI product we have, even as we have relaunched and added incredible features and functionality to it every few years to keep it fresh and to keep it powerful and at the front of the industry, it's a tool that's been proven to work at scale, tens and tens and millions of records at a time in a single instance, no problem.
It is absolutely a scale product that works, that meets all sorts of security and compliance requirements across the world. It works off of the cloud, works on-prem, it's incredibly high-tech. It's just an amazing stack underneath that product. And then the product itself builds on all the experience that we have. The third thing I'd say, and this is a little bit of both, we have so many customers who love what we do. They love the team of people. They know they're going to get the ace team of people on identity if they have to talk to anybody because that's what we have in this company. They love our supports, our professional services, everything we do for them during implementation and thereafter. And they know they have a product that is just going to keep their lights on and help them to get the job done and innovate with them, which is where, particularly recently, we've spent a lot of time. We have this just robust basis.
I'd say there are two other things which we kind of take for granted but actually put us in this really interesting position in the U.S. and globally, and which is we are really a healthcare company in the sense that we focus on healthcare. At least as far as we know, nobody is using our product in the automotive industry or financial services industry where there's great problems, great business, all of that jazz. Actually, my first job was in financial services. I know that world, well. It's been 23 years, but I did know that world well. But we at Rhapsody, we are all about healthcare. We know everything about it, we're focused on it. We're focused on the problems that are here and now and the problems that are coming as we think about what we're going to build.
Finally, and it's a little bit of full circle, Chris, so thank you for teeing this up because this is kind of where I started all the way back in the intro. We keep a full picture. We're trying to do more than moving the data around or even just transforming the data. We have built a digital health enablement platform that starts with everything from let's get this data unstuck and get it to move to let's make sure that we can get the highest quality data for our customers through identity, through semantic mapping, through taking the content and meaning of the data and actually surface it in a way that can be incredibly useful and valuable to the healthcare organization, the healthcare tech organization that's trying to do more with it. So, when you look at the whole picture that Rhapsody brings to the table with the EMPI being a market leader, but a number of other best of breed products coming together, I feel really good about our ability to say, yeah, this is the suite that can help you as a company to really transform your organization.
Chris Hayden:
Well, thank you so much for your time today, Jitin, we really appreciate taking the half hour or so out of your day to talk to us.
Jitin Asnaani:
Absolutely. My pleasure. Thank you so much for the time, Chris.