Dan:
Chuck, thanks for joining us.

Chuck:
Dan, it's a pleasure. Thanks for having me.

Dan:
You've held a lot of jobs since I've known you and I would love to know kind of your journey and what led you to Clarify Health.

Chuck:
I'd say if there was one of those theme in my career, it's been trying to have a larger and larger impact at scale and bring equitable healthcare to more and more people. That's a noble goal, right? And I don't know if I've been completely effective, but we certainly try. Began my career at Amerigroup Corporation, so managed Medicare and Medicaid under the Anthem umbrella, so working directly in managed care, first part of the leadership development program and then doing consulting with individual doctor's offices, doing transformation towards patient centered medical homes. Spent some time there and started to get that entrepreneurship itch and left to join a five person startup in Chicago building a dual eligible Medicaid/Medicare health plan for the city of Chicago. Basically, got to take that from the ground and that up from what we were as that initial five to upwards of 150 employees with over 40,000 full risk members. Having done that zero to one was beginning to really get that itch of the, how do I go more broadly? How do we use technology to have a bigger impact?

Chuck:
While I was at Next Level Health at the time, I did pretty much everything from government relations, to writing RFPs, to doing the product management for our care management software platform and went and joined Healthbox after Next Level was up and running. Healthbox is a services and innovation consultancy firm based in Chicago. It also ran a $35 million fund for Inter Mountain Healthcare. So got to spend some time there doing innovation consulting with some of the country and the world's most leading healthcare system, so Baylor Scott & White, Blue Shield, things like that and began to get that impact that I was looking to have of working directly with startup founders, entrepreneurs on helping them take their businesses to the next level, run them through Healthboxes, that accelerator, a studio program at the time and take those technologies and help them gain a foothold within these larger organizations to really help them determine, "Well, what does innovation mean? How do I actually grow that within my organization?"

Chuck:
The answer is that as a different flavor for whoever the end user is, whoever the customer is for those larger organizations because what innovation at one company is may not be at a different company. So getting to experience a bunch of different flavors of that was really impactful. Spent two years at Healthbox and then wanting to go back to business school. So I did that, went to Kellogg and coming out of Kellogg, joined Anthem as part of their innovation team. Anthem has a entire innovation department focused on spreading innovation throughout the entire organization and that looks at everything from optimizing internal processes using AI technology to establishing a blockchain ledger and creating a unified patient record so that all of Anthem's numbers can share across that. I was there for about a year and then most recently, moved to San Francisco. I've been with Clarify for quite a while now and focusing on helping us grow a large scalable software business.

Dan:
I'm interested just on a personal note, how did you choose insurance as a passion? Because not everyone does that, and so I'm interested in what intrigues you about dealing with the most complex part of the whole healthcare system?

Chuck:
As we know, people always get sick and you can always kind of make the joke that healthcare will always be broken. So I figured the best way to really have an impact on that was doing it from the inside. I grew up with a mom who was a nurse. She was from the nurse at my school to giving us the bandaids at home to then working with Kaiser for 15 plus years doing everything there and then you know a dad who is super into fitness. So healthcare was always part of our lives. I got the opportunity coming out of undergrad to join a leadership development program at Amerigroup, so kind of left with the opportunity to really get a head start in an industry that had sort of been slower moving and older if you will, as compared to other industries in terms of the demographic.

Chuck:
So wanted to be able to jump in and build my career there to really, like I said, have a larger impact and understand the ins and outs of it so that when going to smaller companies in the future, I would have that critical understanding of the challenges those types of businesses faced and also, what the interaction was like with our members and with our patients to be able to figure out what are the types of technologies, solutions, workflow, innovation that we can do to have an impact. I heard [inaudible 00:06:38] Jane talk one time, I think he said it best that innovation can be virtual technology or virtual reality for some people, whereas for a Medicare/Medicaid patient, it might just be taking the carpet out of their house so their walker doesn't trip them when they're trying to get around the house. Both of those can have immense cost savings and health benefits for the individual, but one is a much lower tech version of innovation. Taking a broad expanse like what that definition actually means was super helpful when I was starting my career.

Dan:
Yeah, that was a question I was asked yesterday. I was at a panel at UCSF and they asked, "Would you rather start in a startup or would you rather start in a large company and then go to a startup?" I struggled with the question because there's aspects of both I think that are valuable, but I think you put it well is if you get the large system view then it helps you narrow in and actually have the focus of what a startup is. Whereas if you start necessarily in a startup, you may be trying to figure out that system view while you're trying to figure out your niche at the same time and it may not work as well.

Chuck:
Yeah, it's a perfect exemplar of I would say that the team that we have here at Clarify skews a little bit older in terms of how you would essentially look at started demographics. Because we do come from industry experience and got such a deep talent pool here, it's hard to come fresh out of college and have that type of expertise that you get through becoming seasoned in the industry for a while.

Dan:
Tell us a little bit about Clarify. What do you guys up to? From what you can see on the website, it's big data and all these buzzy words [crosstalk 00:08:14] that are kind of invading healthcare at the moment, but I know you guys are doing some really cool stuff with some really big clients. I would love to get an overview of what Clarify's up to.

Chuck:
Our goal is to have the most of the world's data to create actionable impact. We have data on over 300 million patient lives from across the country, many of those longitudinal, so having the full patient journey. We're also a qualified entity, meaning that we have all of Medicare's fee for service claims data in its raw state. We can then roll claims up to patients or members, members to providers and providers to larger health systems or contracting entities. What that has allowed us to do is taking a team that came from the financial industry space that was a platform built for hedge fund and dynamic trading and things like that and apply those same analytics to healthcare. So our founder, Jean Drouin, came from healthcare consulting and McKinsey for over 15 years and joined with our other co-founder, Todd Gottula,, who came from the banking industry and married those two technologies and that industry expertise together.

Chuck:
We have what we believe is the cleanest data set in healthcare and allows us to do prediction analytics with extreme precision across the payer sector and the provider sector in life sciences. The thing that we do differentiated than anyone else in the industry is creating what we consider to be our blue diamond predicted value. For any metric, for any provider in any geography in the country, we have a predictive value on how an outcome should have been for a patient on an individual metric and then use that to do all kinds of analytics from building new networks, to improving provider performance, to managing value based contracts, to enrolling in bundles programs as CMS or empowering your ACO. So that's sort of what our company does overall. Then we'd have it broken down by those different products. That's a high level and then we can get into any information further you'd like to talk about from the data stack and how that works or how we clean in tokenize those individual data points in provider or member lives.

Dan:
Yeah. We've talked about this before around how you can really get granular on an individual provider and their impact to outcomes, which I think is one of the holy grails of healthcare data. I think every health system in the world is trying to do that and they are in some varying degrees, but you guys have kind of knocked it down and are able to make some measurable impacts like 10% reduction in network referrals, 25% addressable savings. Those are all some big, big numbers there. What would you say your most successful impact is right now at Clarify?

Chuck:
You named a few there and I think to hit on one of the differentiators is that every provider will say, "Well, my patients are different," and we actually say, "Yes, that's true and we agree actually. Your patients are different. That's why your predictive value of Dr. Smith is different than the predictive value on the same metric for a very different set of patients that you serve, Dr. Jones," and we open up the black box and show what are all the factors that go into creating that predictive value. By identifying that and seeing where are the metrics that where you're performing above average or above predicted or below predicted, you can do that based on outcomes. Because we create that predictive value, any metric or score that we create, they become a quality metric.

Chuck:
By digging into that for some of our success stories or working with bundles companies, we were working with a provider, large provider here in the Northern California region that was I think, number 250 of 300 provider groups that enrolled in CMS's bundled care improvement program and we took them to the number two top performer within one year helping them identify what were the things that were driving core provider performance and giving them actionable insights. So telling them not just, "You're not doing well," but what is driving that? What do I do to improve? I know one of the themes that we're talking about here is automation.

Chuck:
So with that, it's figuring out how do we bring to the surface the most impactful things and the treasure map of where to start. Is it too much referrals to skilled nursing facilities? Is it too many days spent in ICCU when you're managing a patient post-surgery? Am I bringing those to the surface automatically and telling the frontline leadership about, "Here's where we need to go. Let's start addressing these problems. Here's the conversation that you need to have with this provider. Here's the process that we can put in place by using AI, data, science, automation to bring those insights directly to the surface."

Dan:
Some of the other guests around the automation theme that we're doing talk about this data not replacing any job or role necessarily, but actually making the clinician's superheroes, like giving them super powers so they have insights into things they've never had before to make better decisions. I think in the past, a lot of this work was either never spoken about because we didn't have the data to know about it or it was manually done and you got some spreadsheet every quarter about, "Well, you need to get more of this and that and just kind of work on it," or some of these kinds of global metrics that didn't really give actionable insights in the moment. How is that data fed back to the providers and to the network? Is is it in real time or is it in reports that come out monthly? How are they consuming the data that you're pushing out?

Chuck:
Yeah, to your first point was absolutely spot on, right? Is that it's about making the provider the superhero or helping them operate with the top of their licensure. No provider wants to be a, "bad doctor," and nobody desires to do that, but if you don't have the analytics to tell me what am I doing well and what I need to improve, I might not know that my older diabetic patients have a much higher use of the emergency department than my young healthy patients. That may be obvious, but if I didn't know that was a problem, I can't do anything about it.

Chuck:
So we serve as insights either directly through our platform, which is a web based, high trusted, secure platform, but we also serve as that through reports that are extracted from our platform and delivered directly to the providers. Sort of depends on who the messenger is. For our provider customers, they may access the platform directly. For our pair customers, they may take those reports and then help their providers who are enrolled in value based contracting to give them basically the answer of how they can perform and improve in their sharing agreements and where exactly the areas to focus on are.

Dan:
What's kind of the Next Level for Clarify? Is there a day where as the physician, nurse practitioner or whatever the is in the future is going through their decision making, popping up insights and references and recommendations for them to follow? Or is it being able to match the right performing physician to the right patient population? Where's that future look like for Clarify?

Chuck:
We have a lot of different roadmaps depending on our different products. Where we don't sit is directly, right now at the point of care. We're not embedded necessarily into an EMR to get those types of clinical decision support types of analytics, but on the roadmap really is that individual patient to provider mapping that says, "For the patient sitting in front of me who has these types of patient features, who has this types of social and behavioral determinant needs," one thing I forgot to mention is not only with that claims data do we link all of the individual risk factors and any claim points data, we also marry that with social and behavioral determinants of data so we can actually understand what is this patient's access to transportation? Do they live near where the providers are? Do they live with another person?

Chuck:
Our data science team has proven that you would think that living with another person would actually make you at a lower risk for a negative health outcome because you have somebody to take care of or somebody to take care of you. But what we learn is that if the person living with you is below 18 or over 65, you're actually a higher risk because you put your needs secondary to the person that you're taking care of.

Dan:
Wow.

Chuck:
So when building in those types of features into our models, we can get really, really precise on for providers who have seen patients like this who live in this area and are part of this insurance program, who's the absolute best doctor either to match me as a PCP or for the specific type of surgery that I'm getting. That's all the types of things that are on our roadmap. Then there's also one of the products that I lead is our networking referral optimization solution, our Clarify network which looks at building newer networks or refining current networks. The biggest problem here is that providers and payers haven't had that lens into their performance outside of the data that they currently have.

Chuck:
Our outside in analytics and where we've married our Medicare data with our commercial data help me define who are the top performers in an area. By automating the process of network building and moving it into software, using our analytics to define who are top providers, we take a process that normally takes about two years of trying to go and figure out, "Okay, what geography do I want to enter? Okay, who are the providers in this market that I don't have any data on, so I'm kind of guessing? Okay, now I need to go contract with them. All right. Well, now my network is stood up and I've gone live and now I have to wait another year before I actually have any performance data on these providers," and sped that up so day one, when you enter a new market, you can look at the provider performance, you can understand where they are, you can make sure your network is both accessible and adequate to where your patient population is and then stand up that network in a matter days rather than a number of years.

Dan:
Yeah, I think that real time data set and assessment per patient even gives you just amazing speed to market. Whereas before you needed to get some big and number to be able to have any insights into anything. I think that's a really cool differentiator. I'd love to jump into some of the data side of things. We were talking about automation in this series. We've talked to some experts in bots and that kind of stuff. What kind of infrastructure do you have to go through these massive datasets and clean it and make sure that you're getting valid insights to give back to the clients?

Chuck:
There's probably better people on the team to speak about this than me, but we have a huge data science team that basically take in all of that data, clean it, and then tokenize it and then marry it across datasets. Because we have both the Medicare fee for service data and commercial data and Medicaid data, you tokenize each individual claim up at the member level. So if we see a member who has gone from a commercial plan and then has enrolled into Medicare and then maybe becomes too eligible, we can link all the claims together in that tokenized fashion and then marry that with the social and behavioral determinants of data. With that requires is a really strong tech backbone so that we can keep that data clean because then that is the data that we have to extract from to do all of our analytics.

Chuck:
So we get rid of all of the sequel queries in the time that business analysts are spent doing that either at your providers system and move that all into our platform that gives these lightning fast insights in real time. There of course, is a little bit of data lag that comes with that because that's the nature of the industry. But to be able to bring all of that into one platform, our Clarify platform basically allows you to think of any question that you might have from a business or clinical perspective and put in the right inputs, ask the platform the right question to get at the insights that you're trying to manage.

Dan:
I think there's value just in cleaning data for healthcare. On our side, we work in the staffing area and we went to one hospital, they had 237 spreadsheets that documented all of all of their staffing information for a year. You can just imagine the amount of spreadsheets. Microsoft's making a killing. But yeah, I think there's this an opportunity that data cleaning itself is just something that sets a baseline that's so important for healthcare to actually do their work better and actually get some insights.

Chuck:
You have to be able to determine what's good data and what's bad data and extract that noise from a lot of the data that you get. That means also we have to de-dupe every client and make sure that we're not double counting any patients or that we're pulling in the correct mounts that are being paid and understanding how those links as part of the patient journey when a patient leaves the inpatient setting and goes to a post acute setting and making sure that all of that ties back into that one patient journey or that one episode of care, and being very clear around the caregivers that we use to define what should be part of the episode of care that we're looking at.

Dan:
That's so important because we've talked about it before, but put insights from poor data, you're going to just break the system even further.

Chuck:
Yeah.

Dan:
That's interesting work. It'd be interesting to see behind the scenes how that all happens. One of the things that's really important as we look at automation is you can't forget the human touch piece and actually, the end user of this thing. So you've talked about physicians, you've talked about kind of network administrators from a health plan standpoint. How do you involve those people in the design of the product and ensure that it's really meeting their needs and not just adding extra noise to them?

Chuck:
There's a number of factors of this. I think to the latter part of your question that's hard about being a good product organization is making sure that everything we do is informed by the needs of our end users and our customers. What we found over time is that clinicians actually become our biggest advocates because we actually tell them the things that they can do to get better and provide better healthcare to their patients or to their members. Because we open up that black box and we help them understand that you have a unique predicted value for every single outcome, and here are the factors that are driving that up. It's because you see more patients that need joint or extremity replacements or more of your members are Medicaid eligible or over 65. Here are the reasons that we've arrived at this value and now here's the things you can do to actually improve.

Chuck:
So bringing that human element into it, it's not about automating the things to be done to get better. It's about bringing to the surface the steps that need to be taken so that the provider can provide better care. That's what our care product does is it's a right dashboard that says to a provider, "Here's the things you're doing really well at. Keep those up. Here's where you have areas for improvement and an underlying cause of that. Here's how the trends have changed over time," so you can be empowered to actually go make those changes. The same thing with payers who want to kick the best doctors and be part of their networks and make sure that they are enrolling the right providers in value based contracts and doing risk sharing with them what do they need to do to actually improve.

Chuck:
So it's more of a month operating at the top of the licensure then trying to remove any staff or basically, automate away any of the care that's so critical. You and I've probably both heard a lot about virtual chat bots or getting rid of radiologists. I'm sure that is a trend we'll begin to see more of as AI does become better and better, but I think we both know, and nursing's such a great example of this, so we'd love to hear your thoughts on this as I turn the question back on you, but the human touch is so important in healthcare, whether it's delivering a message, whether it's changing a behavior pattern or being able to effectively communicate with the patient. It's not just about telling you to eat healthier and exercise more., but specifically understand my condition, converse with me as a human, help me understand what I need to do to be healthier and live a better life.

Dan:
The stat that I keep quoting is about 36% of nurses time is spent finding people and stuff. So that's four hours out of every 12 hour shift that's being used in non-value-added activities for a very smart, well-trained clinical expert. The more we can automate the workflows around the mundane data entry, finding people, getting equipment, programming equipment, that kind of stuff, the more time we can give back to those interactions that actually change lives and help people feel connected and kind of better the whole planet ultimately.

Dan:
So I'm a 100% for taking away all those tasks because they drag down the care team. The same thing from physicians. They feel like their data entry monsters now, like they've lost that passion. It's leading to burnout. They've tried all kinds of different solutions from scribes to Google glass recording their sessions to everything. I think there's a lot of promise in the fact that we can consume data through machines and give insights that allow them to go back and spend more time in the thing that they trained for and the thing that they love, which is treating disease and interacting with people. So yeah, it's really interesting. We'll see how the future unfolds with that.

Chuck:
Yeah. What we try to do is is we call our smarts AI partner Claire and Claire surfaces the insights for you, right? So rather than making a provider dig or a payer dig to have to find the insight, we surface the highest value impact insights directly through the software. A provider can get the insights to know, "Okay, who is the patient that I'm about to go in and see? What would I need to know about them so that I can have a better care experience with them and not be there digging for information and do the wrong things that might not be right for this patient sitting in front of me?"

Dan:
There's a lot of change management I'm sure in your business. You have to teach people that their data is not the best data, that you need to clean it and that you might have different insights. What is some advice you have for healthcare leaders who are thinking like, "Oh my gosh, maybe I'm making decisions on really broken or bad data and maybe I need to talk to someone, like a Clarify." What's that change to management or what are the things that leaders need to start considering if they want to start that conversation?

Chuck:
Yeah, I think a couple of things that come to mind are understanding where your data's coming from and being aware of the types of things that could corrupt it. Since we start with the claim level, we first have to clean everything and make sure everything's in the right formats and variable and accessible before we can even build it up into data science pipelines and create these types of analytics. The other part would be knowing that your data is only part of the picture. So for payers, for example, they may have an understanding of how medical costs are trending for their population.

Chuck:
But if you don't have a global perspective on how healthcare is trending or healthcare costs are trending overall or in a region or for the patients or members that your providers aren't seeing, the rest of the geography, then you don't really have a full picture of what's going on. So you're only getting the perspective of inside, inside where we bring that sort of outside in perspective that say, "Here's how a provider is doing across all of their patients," and they may be actually much better than your data to say they are and they might actually be much lower performing.

Chuck:
Then what are the drivers of that? I'm sure our data science team can get a much better answer about what constitutes good data versus bad data, but I look at it as from what are the types of insights that I can take from this and are my insights based on processed focused metrics? We get a lot of stuff in the EMR that tells me did something happen, but that you're really measuring outcomes. Are you really taking into account how many people should have had an outcome based on the data set that we have, rather than how many of a certain type of screening. That I could do that doesn't help me understand if my patient at the end of the day in that patient journey actually came out healthier. It just tells me that this doctor did a bunch of screenings for a certain thing and checked the number of boxes in the EMR.

Dan:
Great. It really needs to come down to that outcome level, not the process data. Well, should link to the outcome data, but the process [crosstalk 00:28:03].

Chuck:
They should hopefully correlate.

Dan:
Yeah, they should correlate. Right. It's not just how many diabetics did you prescribe insulin to? It's what did that end up doing for the A1C and where'd they end up over time, and did they go to the emergency?

Chuck:
And did they stay on their medication? Right. [crosstalk 00:28:18] What'd the readmission rate look like and what drove that?

Dan:
Yeah. I love that. My PhD was on complexity science, so I love all the connections and it sounds like you guys are making those connections across the system. So I love it. Just kind of wrapping this up, you know, this is The Handoff podcast. We like to handoff essential information to healthcare leaders. What's the number one thing that you would handoff to listeners about Clarify and automation and insights in data?

Chuck:
For Clarify, I would say that it really starts with having the right data in a clean way to ask the right business questions and surface the right insights. We're all well-meaning, we all want healthcare to be better, but without the right understanding of what's going on, what are the real outcomes that providers are having and that my members of my patients are receiving, then it's hard to really get there and really hard to make improvements.

Chuck:
For the listeners in general overall, I think my advice comes back to what we talked about earlier is that the human touch in providers are really the crux of allowing this automation to really become tangible. So I can service you the right insights [inaudible 00:29:26] do anything about it, then all the software in the world won't help solve the problem. Having the right data and having the right people to go and make that data reality, make those insights into action is, is really how we make progress.

Dan:
Awesome. I love. Chuck, where can we find Clarify and where can we find you if people are interested in learning more?

Chuck:
We're at ClarifyHealth.com and if you want to contact any of us, it's just usually first name @ClarifyHealth.com. So I'm Chuck@ClarifyHealth and then all social channels is just @ChuckFeerick, Just my name.

Dan:
Chuck, thanks so much for being a part of the podcast today. It's really interesting to see the outcomes you're achieving with clean data and automated insights and those types of things. So really appreciate the time and look forward to more innovation from you and your company.

Chuck:
This is great, Dan. I appreciate the great questions and I'm hoping we get to catch up in person again sometime soon.

Dan:
Thank you so much for tuning in to The Handoff. If you like what you heard today, please consider writing us a review on iTunes or wherever you listen to podcasts. This is Doctor Nurse Dan. See you next time.