In our video health data series, "Leaders in Leveraging Health Data", we chat with Jean Drouin, CEO of Clarify Health Solutions.
Improve patient outcomes by taking advantage of your own data.
Recently receiving Frost & Sullivan’s 2021 New Product Innovation award, Clarify is a company transforming the healthcare industry with advanced analytics and big data that empowers providers, insurers, and lives sciences companies to deliver better care.
Utilizing over 300 million individual patient health records, along with statistical modeling and machine learning, they provide insightful actions to take to impact performance, cost, quality and outcomes.
Transcript
Clarify Health with Jean Drouin: Better Care with Actionable Insights
Ryan Carlson: I'm Ryan Carlson with Healthjump. And we're here today with Jean Drouin from Clarify Health Solutions. So thanks for coming in and talking to us about what you're doing with health data.
Jean Drouin: Thanks for having me. Great pleasure to see you guys here today.
Ryan Carlson: Yeah, I would like to know who the heck are you, what do you do?
Jean Drouin: Terrific. So think of us as, essentially having put the equivalent of a Bloomberg terminal on top of one of the largest ever patient-level data sets in healthcare. So we now have 300 million American lives worth of patient-level data where we can bring together claims data, lab, data, prescription data, EMR data, and then social determinants of health. Not just at zipper block level, which is how many other people do it. But at the individual level, like a bank or Amazon would to understand a behavior. And with that combination of the technology stack, we've built in the data. We are able to deliver a set of business applications to our end customers.
Our end customers are hospitals, payers, life science companies and the business applications we sell to them answer fundamental questions about their business. So for payers, one of our most successful products is our network design analytics, where the payers are picking and choosing who are the top doctors to put in their plans.
And what we've helped solve for them is being precise enough that they can score those docs in the same way that baseball teams are applying money ball to score their, their players.
Ryan Carlson: Is it that you have specific information about a specific provider or is it a model that is informed through real world evidence?
Jean Drouin: Okay. Terrific questions. One of the initial premises or founding ideas of Clarify is that if you could far better turn massive data into insights about patient journeys. Okay. If you have individual patient journeys and can stack them up, you can then far more precisely answer questions about anything that's going on with the population.
So if a payer says, I want to know who the best doctors are. You can spin the cube if you will, around the doctor being the end point at the same time, though, if a life science company says, Hey, I want to understand how my drug does for certain types of people, you now spin it around the drug as the outcome.
So think of us as having a technology stack that is data. Then a processing ability that uses machine learning, AI and a grouper. Yep. And then ultimately you're able to surface into the workflow, the insight that answers the question at hand. So is this the best doctor? Is that the right drug? Et cetera.
Ryan Carlson: And my understanding of machine learning is it's always the amount of data you have access to and the quality of your data is almost an exponential relationship with the insights that you can generate, right? With 350 million records, that's amazing. Who is it that your primary customer would be?
Who is it that you serve with all of this cool stuff?
Jean Drouin: Yeah.
So think of it as maybe anthropologically. What we're trying to do is somewhat similar to what the Bloomberg terminal did in banking which is to enable the analyst at a provider, a payer or a life science company to self-service on demand, turn big data into actionable insight.
Whereas what happens today and the problem we're solving is often what will happen actually is a customer will say, oh, Clarify, you're the new kid on the block. And the industry is completely wired to say, what data do you have? I will do an overlap analysis of the data that I have with what you have in the sliver. That's different. Maybe I'll buy it and then I'm going to bring it into my data lake and then I'm going to have either some of my analysts. or last minute consulting support, go in sequel query for four to six weeks in a dark room and then come out with something that's still looks like a spreadsheet and might be prettied up in Tableau.
Ryan Carlson: With six to eight weeks of invoices.
Jean Drouin: Correct. Absolutely. Imagine a different world where somebody can log in and either automated or very easily through filter. Couple of seconds goes by and out pops the same analysis that would have taken six to eight weeks of time and invoices in a SQL world. And that's effectively what we've built is the capability or the intelligence to turn data into insights self-service on demand.
And so to use an analogy often used about doctors and nurses, we enable the analyst to now practice at the top of his or her license. And do away with the drudgery and the manual labor of ingesting the data, cleaning it, stitching it together, et cetera, because we've automated all of that. Now, why have we been able to do this?
I was enormously fortunate when I started Clarify six years ago to be introduced to the chief technology officer of what was then the leading cloud analytics business in financial services. And when that company was sold, Todd, my co-founder who was the chief technology officer came over with the CIO, the head of engineering, the three lead engineers and the two top product developers.
So in a way, Clarify is the translation of the power of financial services, analytics into healthcare. The bringing together of this massive data set and to your point, to be able to do machine learning and AI in a precise enough way you need that large data set. Yeah, absolutely. Yeah. And part of the reason that we love what you're doing at Healthjump is that for the category of data, that is EMR data you are, enabling your own customers to get value out of it in a way they didn't before. You're also making it possible though for that data to now be extracted and it would be possible to union it to the kind of data that we have, to enable some, even more powerful use cases. So it's actually a pleasure to be here with you today.
Ryan Carlson: Thank you. So if I had, a bunch of data trapped inside of my electronic health record system. Are you saying that I could then take my data? Put it up in the cloud. Let's say how we use Healthjump to if, to pull it out and put a, into a standardized format that could then be married up would it be that I could then run an analysis against your data set to go help me compare. Where am I strong? Where am I weak? Cause you, you said you're not licensing your data, but it's leveraging it to help train or gain insight from my own data.
Jean Drouin: So here's one thing we could do. You all use Datavant tokens as we do so we could theoretically link those two datasets in a secure cloud and then absolutely either run our or your models on top and derive a set of insights that would now be the union of the insight that's in the EMR data with the insight that's in all of the claims, lab, prescription, and social determinants data that we have.
Ryan Carlson: Yeah.
So at no point, am I getting your data, I'm getting the benefit of it?
Jean Drouin: Correct.
Ryan Carlson: So I had a conversation with Cliff, our CTO about this, trying to understand this idea of training, like machine learning models, and how do you leverage data you have to create insight, but without ever giving access to that data. And here's what I came away. Are you familiar with the dog whisper, Cesar Milan.
Jean Drouin: I'm not actually.
Ryan Carlson: So yeah, he comes in and he helps people with these troubled dogs. They usually have a sorted past or have been treated wrong, whichever, and he rehabilitates them.
And the way it works is anyone who's seen the show is going to oh yeah, I totally know this. He brings this dog, the owner comes up to his facility and he's got this big fenced in acreage and he's got hundred dogs that have all been rehabilitated and trained, that's his pack. Yep. Okay. So they all have the skills and knowledge of how to be a good dog, how to be a good member of the pack.
Jean Drouin: Yeah.
Ryan Carlson: And so he goes out and leaves the owner on the curb going, bye! and then they always do the reveal, the extreme dog make-over they bring that dog in and they learn how to be a good dog from his pack.
Jean Drouin: Sure.
Ryan Carlson: And then they get a walk back out and he hands them over the new keys to the brand new, with the new behavior, the dog doesn't bring the pack with them. They stay inside that walled lot. The dogs are never in motion other than in the right. I think it's a kind of a cool extrapolation to think of, like, how do you not bring a hundred dogs home with you, but you still have a better data model.
Jean Drouin: 100% on the data models and then the other, analogy you might use is it's possible in almost a Snapchat, like fashion to start with a source dataset. Yup. You do a set of manipulations and calculations. And you end up with a derived insight or a derived work. And the process by which that happened disappears. And so it's impossible to re identify any individual information that led to that, derived work.
So I love those two analogies because it's part of the explanation for why today we're able to have this discussion. In a way that if we'd been here five years ago, people would have said, what are you guys talking about?
Ryan Carlson: Yeah.
Yeah. It's fascinating. Tell me, there's the technology challenges and then there's business model challenges that a lot of businesses are addressing. What is it about what you're doing at Clarify that uniquely, allows you to address this problem? Is it that you've got a different business model, pricing model, or is it something unique in your technology stack that really allows your solution to stand out amongst anything else out there?
Jean Drouin: I would say it's a couple of things. Number one for us is the team. The fact that the team of engineers and data scientists came out of financial services and had built that platform at scale for hundreds of banks and hedge funds gave us an ability to approach healthcare in a different way.
And just a little nugget on that one. When we described social determinants of health to the team, they said, okay. And they went off and they came back and we said, wait a sec, how did you pull off the stunt of doing it at individual level instead of zip code level? And they said, oh, but that's what we did in banking to automate who gets a credit card and who gets a mortgage.
Okay, so team is number one. Number two then is you're absolutely right. Is the technology that, that team then translated and innovated into healthcare, if you will, it's
Ryan Carlson: know-how and a lack of ability to reproduce, right? It's ...that
Jean Drouin: Absolutely. And one thing that's often missed and it's a technicality, but there's the power and speed of the platform, but there's also having a grouper in there.
A grouper in healthcare is like a ledger for blockchain.
Ryan Carlson: Oh, okay. I am knew. I thought it was a fish.
Jean Drouin: And it allows you to more rapidly organize into your patient journey, my patient journey, his patient journey. And there's only three other companies that have groupers, right? So Optum, there's Promethius that was bought, and 3M.
That's a massive difference, ultimately in our ability to very quickly personalize the insights that, we're generating.
Ryan Carlson: And to Clarify, it's not the evil Skynet AI takes over the world because your data set doesn't allow it to sell to you and me. It doesn't allow a marketing company to go, we want to know how to sell to Jean.
Jean Drouin: 100%.
Ryan Carlson: People like Jean are, are prone to this type of behavior activity, depending on where you are. Is that more accurate?
Jean Drouin: That's right. And this is one whereby I believe that HIPAA is actually philosophically in the right place, which is HIPAA says, for the purpose of improving care, making sure the billing was okay or research.
The kinds of things that we're talking about are fair game. It says though, if you're about to go and market, then now we're on the other side. We have spent a lot of time building into the platform, the safeguards that ensure that we're on the side of the use cases that are absolutely HIPAA compliant.
For that reason we have not focused on use cases that are more about marketing. Yeah.
Ryan Carlson: And ultimately you get to control, which use case. Absolutely. So you have your own, is there an ethics board or an ethics committee or someone that helps provide that level of oversight on where you can say, we say no. To all projects that look like x, Y, Z.
Jean Drouin: I love that you asked the question, we've just set one up and we have good mix of, internal clinicians and then some, external folks. And, yeah, like it's, it makes a ton of sense because, it's an important, true north.
Ryan Carlson: So what does getting started look like, for, your average ...
Jean Drouin: Customer?
Ryan Carlson: Yeah. What does that journey look like?
Jean Drouin: It's actually relatively quick, when you think of the typical implementation for software as a service. Okay. Usually, let's say it's a hospital that has brought in our referral optimization analytics. Typically there will be analysts in the strategy or finance team, and we would come in now, the pandemic taught us how to do this remotely.
But train them on literally how to use the software in response to the kinds of business problems that they have and, which they may not have identified had they not been forced into this situation. absolutely.
Ryan Carlson: I think it's a 10 year timeline of like technology roadmap compressed down into a couple of years. Just how many times. Have you tried to explain what you do pre pandemic when it was still acceptable to keep like files and paper forms, or like you just assumed that you could always go back to some of the manual processes. Sure. We need to go digital, but we'll get there. Yep. And then it was wake up call time. And so that pain that people feel is close enough to the pain that you solve.
Jean Drouin: Yes. And I'm so very much so the ability, I would even say the believability of what we do has gone way up, partly, and one interesting thing too, is we used to have this notion that too. For a customer to gain enough trust in us that they would want to bring us in that it had to be a face-to-face interaction. And yeah, and I think the pandemic proved to us that in certain circumstances it's possible to do just in the two dimensions of zoom. Yeah.
Ryan Carlson: What's that we can trust humans to not take advantage of that, like being at home in their pajamas and just turning their camera off and sorry, I'm still scarred from remote learning from kids right now.
Jean Drouin: Having said that it is really nice to be able to interact like this again. I'm hopeful that we're going to come out on the other side and that this will open up, a better work-life balance for folks because there'll be more flexibility. And, we're seeing that with our customers and with ourselves.
Ryan Carlson: Yeah. I'm realizing that I jumped right in. So someone identifies the problem. They bring you in. Are they, is there like a class or?
Jean Drouin: Sure, absolutely. We have Clarify academy, which is both a direct right now, online and then virtual where you can go asynchronously at your own pace. And, you can acquire, qualifications or levels of proficiency.
Ryan Carlson: Cool.
Jean Drouin: And, typically it's a two month onboarding. And then, as any technology company, we then have, a customer success customer care department and people can call and say, Hey, I'm working on this thing, et cetera.
And then obviously there's a set of new releases, of additional functionality. The quarters go by.
Ryan Carlson: So probably about the second quarter, they're now making their own queries and getting their own insights. And then if they got a problem, they get on the phone and you've got a customer success, exact person that's walking you through it.
Absolutely. It's almost as if it's a model that has been proven to be successful.
Jean Drouin: Ah, so it's interesting because in healthcare we often get told, rather than asked. You have to have an army of MBAs with clipboards now providing services to ensure people achieve the impact. And this is actually one of the things we were seeking to prove, which is that, no, you don't have to have an army of folks with clipboards because you can enable your own team.
This is this concept of practicing at the top of their license to, deliver the insights into the workflows they need to.
Ryan Carlson: I love about what you're talking about is a theme that I've been sensing, which is, you're not just buying a product, you're buying the ability to create a new capability that you didn't have to have.
Jean Drouin: Absolutely. Yes.
Ryan Carlson: And, and it's that human in the loop data analytics thing that I see is just mission critical, right? Data turns into information turns. Insight turns into knowledge, right? Somewhere in there, a human has to be involved very much, whether it's a data scientist or providing the context behind things, but, AI is awesome at turning data into information and information into that, like that gap into there's variances here that may require us to think why is this happening?
Yes. Which I love. Anyway, I really think that, this is. The way things are moving. Yeah. And I'd say the success of using data and real world evidence like it is the thing that we're moving to. And I see it's inevitable. So thank you so much.
How would people learn more about clarify health solutions? If they wanted to learn more.
Jean Drouin: Probably the easiest thing is to come to our website, www.clarifyhealth.com some really interesting videos on there and, even better would be to reach out directly. Interestingly, what we found is, it's when people see the actual software that we have, that they light up and they say that is possible!?
Seeing is believing ultimately.
Ryan Carlson: Oh, that's great. Yeah. You heard it. Her first seeing is believing you should check it out. So thank you so much, Jean for coming in, the chief executive officer of Clarify Health Solutions for sharing about this whole new world of our great data overlords, insight generators.
Thank you so much.
Jean Drouin: Thank you for the opportunity really appreciate it.
.svg)
