Display Date
November 17, 2022
Episode 18: Identifying the signal in the noise—How to make sense of 19 terabytes of data each year, with Chase Zaputil of Veda
The U.S. healthcare sector generates 19 terabytes of clinical data each year—and that’s doesn’t even include any financial, operational, patient, or systems data. Chase Zaputil, chief growth officer at Veda, joins Justin to talk about all that data: how to measure its quality and value, how to reduce inaccuracy, and why old data isn’t necessarily bad data. Justin and Chase explore the role of data in healthcare interoperability, the current state of pharmacy benefits and drug costs, and how data can help reduce inequity in healthcare.
Justin and Chase also attempt to answer some tricky questions facing the industry: Is the sheer volume of healthcare data actually muddling decision-making? Should the government intervene to promote interoperability? And can pharma manufacturers lower drug prices by getting involved with distribution?
We want to hear from you...
Have an idea for an episode? Got the inside scoop on a trending topic? Let our team know!
Justin Steinman:
Definitively Speaking is a Definitive Healthcare podcast series recorded and produced in Framingham, Massachusetts. To learn more about healthcare commercial intelligence, please visit us at definitivehc.com.
Hello and welcome to the latest episode of Definitively Speaking, the podcast where we have data-driven conversations on the current state of healthcare. I'm Justin Steinman, chief marketing officer at Definitive Healthcare and your host for this podcast. I'm joined today by Chase Zaputil, chief growth office at Veda Data. Like me, Chase has been all around the healthcare ecosystem over the past two decades. He's had a variety of roles at places like Wolters Kluwer Health, Mercer, and Wellframe before his current role at Veda. And unlike me, Chase has a clinical background. He's a trained and certified pharmacist, which is how he started his career. We'll cover Chase's experience as a pharmacist later in the podcast, but the reason that I asked Chase to join me today is because he is also an expert in healthcare data. For a podcast focused on data-driven conversations around healthcare, Chase is a bit of a home run guest, if you will. So Chase, welcome to Definitively Speaking.
Chase Zaputil:
Thank you, Justin. Very excited to be here.
Justin Steinman:
So Chase, data is a big topic. I mean, you can literally talk about anything. There's clinical data, operational data, financial data, affiliation data, reference data, patient data, provider data, system data. I could keep going, but for the sake of our listeners, I'll stop. How do you begin to make sense of all that data? Is any one form of it more important than the other in the healthcare ecosystem?
Chase Zaputil:
Yeah, it's a great question, and just to preface all this, you mentioned I'm a pharmacist but I'm not a data scientist, so just to level-set on this. Yeah, I mean there's so many types of data and it's all got its importance, and it all helps us hopefully make better decisions and drive better outcomes. So while it's vast, there's expertise in various areas, and I think that's what's critical to us being able to utilize data effectively.
Justin Steinman:
Is any one type of data more important in the healthcare ecosystem do you think about?
Chase Zaputil:
Yeah, I mean, from my perspective I don't know that it can be one or the other, right? I think that if you look at, let's say, financial data, that's important to the healthcare system. We have to make great financial decisions, right? If you look at outcomes data from community clinical studies and whatnot, there's a lot of great data there, right? It's important, too. So my thing is I don't think there's necessarily more important data, I think it's about making sure that we utilize the data we have in the most appropriate way, and that's ultimately what's going to drive, like I said, better decision-making.
Justin Steinman:
Yeah. I mean, I do think it's really hard to separate the financial data from the clinical data from the operational data, because it is so interlocked. But to give our listeners a frame of reference here, according to the website jdsupra.com, the healthcare sector generates more than 19 terabytes of clinical data alone each year, and that sum doesn't even begin to consider the other forms of healthcare data that I just listed. I mean, 19 terabytes. I can't even begin to fathom that, right? I've always found, and I know I'm far from unique in this insight, that there's a wide variety in the quality of data out there, right? The old garbage in, garbage out. How do you measure quality of data?
Chase Zaputil:
One of the things that's exciting about Veda and a little bit of my background is we're really focused on the idea of scientific method. When you think about that, that's really how do we systematically observe things, measure things, experiment, and ultimately modify our hypothesis that we're testing if we need to, right? And like I said, that does align with my background as a pharmacist. Obviously, a significant part of my education, as fun as it sounds, was how do you assess clinical information through statistical methods, and ultimately, hopefully, making the best decisions to drive sound clinical decisions in that regard. Yeah, so that's all well and good, but then we talk about just the sheer amount of data. An example I think that's interesting, I was just running across some stuff this weekend about clinical trials, and that's a place that, as my pharmacy background, spent a lot of time assessing clinical trials, like, "Is this a good observational study? Is it a good randomized, controlled trial? What is this?"
The number of clinical trials on an annual basis has doubled just in the last 10 years. I mean, that's pretty significant, and that's just that much more data, which hopefully is leading to better decisions on the care of patients and leading to better outcomes. But at the same time, that's just more data that's out there to also possibly make decisions more difficult, because you see this says one thing, one says the other. So sometimes that amount of data is good and sometimes it also could probably add to complexity of things, right? So yeah, that is amazing to see just the sheer amount of data that's out there. In general, the amount of data that's generated, while it can be overwhelming at times, there's times where it actually can be useful as well.
Data in general is two forms, right? It's signal, which is hopefully the things that are proving what you're trying to test, and then it's usually noise, which are things that are outside of what actually is proving your hypothesis, and that could be things as simple as somebody miskeyed something into a dataset and it looks outside the ram, it could be actually that your hypothesis is inaccurate and you need to adjust what you're trying to prove out. Now, that's when we talk about things like artificial intelligence, machine learning, especially the machine learning side of things, data, and more data can be fueled to help that machine learning improve its process and its modeling and what we're actually driving it. So I think those things are interesting. While there's a sheer amount of data, sometimes it can be overwhelming, and we can have it not be useful. There's times where it actually does drive better modeling as well.
Justin Steinman:
Is there a way to measure the quality of data?
Chase Zaputil:
Yeah, I mean I think that's the one thing that when we talk about, from a Veda perspective, accuracy of data is key to us, and high quality data, for example, is really important to us. And there are things. Obviously you want to validate your modeling is working how you would assume, and sometimes that actually requires you to do things like human intervention to validate the data. For instance, we're maybe validating a set of phone numbers. You can run models and go out and source a ton of data and try to make sure that those numbers are accurate, that they're actually can be called, but you may actually want somebody to call and make sure somebody picks up on that other end to validate that. So there are various ways that you could obviously show that there's accuracy in that data. That's not always the case, and sometimes you have to assume that your modeling is appropriate, and you're going to have to make ultimately decisions on where that goes.
Justin Steinman:
Why do you think there's so much inaccuracy in data?
Chase Zaputil:
I think it's around a number of sources. I think there's also the fact that in today's world data is produced at such a rapid pace. I think timeliness is also a really big factor in this. So while data may have been accurate at one point in time, as you progress to the future, it may no longer be accurate and it's still out there, right? And so I think that's one of the hardest things if we look at things like modeling is you have a certain time span of data, and data does evolve over time. It's not static. It's very dynamic, and so in that regard you really have to stay up with your modeling and constantly be processing that data to ensure you're picking up the appropriate signals, if you will.
Justin Steinman:
So is old data bad?
Chase Zaputil:
Well, no, not necessarily. So there's times where something could be accurate for a long period of time, so it's not always that old data is bad. It does get to the point where it evolves over time and it can lead to noise in what you're trying to do. So I don't think old data is necessarily going to mean bad data.
Justin Steinman:
So then how do you deal with the aging of data?
Chase Zaputil:
From what we see is the fact that when you're utilizing things like machine learning and building these models or artificial intelligence, that has to pick up on those nuances of how the data ages and what that means to the model and constantly being updated those models to assess that, right? I'll tell you a really good example. We focus a lot on provider demographic information, for example, and we've been talking a lot about something like rural healthcare. Rural healthcare is really difficult, because there's oftentimes a scarcity of providers in those areas. And so if you were looking at data and a simple question, "Where does Dr. Smith practice at," should be a fairly simple question. You're very aware of this, right?
That's not. It's a very difficult question, especially if Dr. Smith covers a large geographic area and has practice locations in a number of different, say, small rural towns. What if Dr. Smith only shows up in one of those towns once every two or three months? So you really have to assess what is that utilization pattern over a time period, right? So that's in the case that older data may be good data, it may actually help you think about what is the appropriate practice habits of that particular provider, for example.
Justin Steinman:
So the Dr. Smith question is actually really interesting one, because let's just say Dr. Smith is affiliated with a large IDN as a hospital employed physician, and Dr. Smith files his claims through the hospital. We're talking Texas and the claims for Dr. Smith go through Dallas, but Dr. Smith is out two hours east of Dallas, right? How does that begin to work and how do you think about that?
Chase Zaputil:
Yeah and, I mean, from our perspective, this is why deep expertise into what you are actually trying to model and measure is really important. So this is why I think it's really interesting when the idea of artificial intelligence or these systems that are just going to be computer brains, if you will, and learn all this. Well, they will maybe perhaps over time, but there does take a lot of deep knowledge of the industry or what you're looking at and understanding those nuances of practice locations and how somebody like Dr. Smith actually practices. And I think without that knowledge and helping understand that through your modeling, that in itself leads to inaccurate data. You could process those data and maybe you would just throw out all the fact that Dr. Smith filed a claim that was two hours east of Dallas and that wasn't accurate. Well, it was accurate because he goes out there once every two months to see a bunch of patients at that point in time.
Justin Steinman:
So I'm glad you brought up artificial intelligence, because let's dive in a little bit of that, right? Part of me thinks artificial intelligence is just like data science and it's algorithms, and then you go to the flip side, you've got your "Artificial Intelligence," and you've got "The Matrix" and "Terminator" and all these great science fiction movies and we're all working for our robot overlords, right? Is artificial intelligence really real when it comes to healthcare data? Are our computers getting smarter on their own?
Chase Zaputil:
I don't know. At some point maybe we will all be working for a computer. I don't know that. But I mean, my thing... It's funny, we have a tagline at Veda. It's "Technology helps people help people." I do feel human knowledge, deep expertise is really important to whatever we're utilizing. Maybe someday in the future, as we've all seen the sci-fi channels, it will evolve to something different. But right now I do feel the idea that those technologies are extremely important to help us process, as we just talked about, a large amount of data very quickly, right? We couldn't do that without those type of computer systems, those models that are out there. However, I believe if you do that without that expertise and ability to assess and validate that those models are working correctly, it can lead to a lot of inaccurate information, right? You know.
I mean, think about the bias of AI, those type of things. They happen if you leave those unchecked or if you program them in a specific way, they can have biases, and that's the same thing from a human validation perspective. We need to understand. Are there biases in our models? Are we really assessing this appropriate way? So I think it's a very interesting question, and hopefully we'll evolve into a time that'll lead us to an answer.
Justin Steinman:
So the great holy grail of healthcare data is interoperability, right? I've been in this industry for a long time. I've always been talking about interoperability. Back when I was at GE Healthcare way back when, I remember going to HIMSS, and every year at HIMSS the engineering team would pull all day the night before the start of the show at the interoperability lab to prove that we can connect system A is system B and they could share meaningful data, and that's meaningful use for all of our friends. A little bit of history in the industry here, right? I feel like we've been talking about interoperability in healthcare as long as I can remember, and I'm sure long before that, long before I got here. Are we any closer to true interoperability with healthcare data today?
Chase Zaputil:
Yeah, I mean, it's a great question. I mean, I think there's parts of the healthcare industry that are further along and mature in that side of things, and others that I think have more to go. It's interesting because, as you mentioned, I grew up through the pharmacy world, and actually when you think about pharmacy, I mean, go way back, I think some of the original HIPAA laws really standardized a lot of what we were saying during transaction processing. And pharmacy evolved pretty quickly in a standard transaction for realtime billing, for example. And that to me was a huge benefit for pharmacy in many regards, because you could do realtime billing, give responses back, you understood a copay, you understood all those things very quickly for the patient.
There are things, though, that I think probably allowed that to happen. One, there was an organization called NCPDP that was a standards organization within pharmacy and I think been around from the seventies I believe, and had really taken a leading role in that. The other piece was there had been quite a bit of consolidation and it happened through pharmacy, so you had fairly big entities, large pharmacy retail chains, that were big players in that who adopted those standards. And one of the reasons I think is for efficiency as well. So there was things that benefited that transaction being in place.
On the medical billing side, it's taken a lot longer. I think now we think to the High Tech Act, there was other things that... This is where the government and private sector coming together to really drive change is important, and it doesn't happen overnight. I think that's the other piece that's always really interesting is, especially in a large fragmented industry like healthcare, change is often slower than anybody would like. It usually evolves. It doesn't just happen, and I think that's important to always understand. So I think there's been tremendous progress over the last 20 or 30 years in interoperability and standardization of things like claims transactions. I think that leads to ultimately better data at the end of the day, and will continue to lead to better data.
Justin Steinman:
But you could argue that there are certain business inhibitors to interoperability, right? Let's talk about the concept of leakage, right? So I'm working in a healthcare network and I'm part of the Beth Israel network here in the greater Boston area, and I want to keep all my patients and my system. I really don't want to share information with the hospital system in New York, even if they have the best specialty in the world, because I can treat them there, and so I have, frankly, economic disincentives to promote interoperability. Can we ever overcome that?
Chase Zaputil:
Yeah, I mean, I always like to say there's perverse incentives everywhere a little bit. There all are a number of things that are in this healthcare industry that are not ideal. There's no doubt. And I would agree. I mean, there's times where it doesn't make sense for one entity to want to share information with another entity. Again, I think that it's also a matter of as we think about things, like talk about quality, it's a very difficult thing. What is high quality, right? And when we're talking about quality, obviously quality of data, but the quality of the provider for example, and what leads us to assess that quality of that provider in a way that's different from quality of provider A versus B?
But I think if we can get to a point, to your point, where we get better standardization, actually analyze data and show that quality matters and drives outcomes and ultimately reduces things like the cost of care or improves the health of patients, hopefully those do lead to better decisions at the patient level to drive to the best outcomes for that patient. If it was easy, I think we would've solved it. And I do think in our industry, as you know, the fragmentation does sometimes impede progress. I think the overarching, though, sentiment of the healthcare industry is ultimately you should go back and what's best for the patient, and hopefully we continuously do that. I'm not saying that always happens, but I think that's what we have to focus on is how do we drive the best outcomes for the patient.
Justin Steinman:
Yeah. I'm going to take a quick aside here for our listeners and mention this is the second podcast in a row we've talked about perverse incentives. Dr. Andrew Norton talked about perverse incentives in oncology at our last podcast. Amazing. Twice in a row here. But I got a follow up question for you here, Chase. I want to know is there a role for the government to legislate interoperability? We've tried multiple, multiple times. Is that just never going to happen? Should it ever happen?
Chase Zaputil:
It's a great question. Honestly, I'm not a policy expert, but-
Justin Steinman:
Neither am I.
Chase Zaputil:
I do think there's things that have been put into policy that have driven ultimately changes, right? If you think about HIPAA, that really did drive a lot of standardization through the X12 stuff, we talked about the NCPDP transactions, there were things that happened in there that did come out of that policy. Or even in the high tech side, that really did launch into what is now electronic medical records and utiliz... I think before that you're talking less than 10% of providers utilized electronic medical records. I think today it's well over what? 80, 90% plus. So it's like-
Justin Steinman:
Mm-hm. Over 90%.
Chase Zaputil:
Yeah. So those are great opportunities. That said, policy has [inaudible 00:18:08] chained. I mean, I think that there's also a lot of discussion on things like does a single payer system solve all our issues, right? Those are really larger questions than I'll ever be able to probably answer in my lifetime. I think there's going to be tradeoffs no matter what we do. I also think the US healthcare system in general has also led to a lot of innovation and a lot of great things that have come out at the system. So again, I think change does not happen overnight. I think it evolves over time. There may be a place where policy would be able to standardize and allow more interoperability and allow more sharing of data more freely, and I think that will continue to be, I'm sure, pushed and pulled on for years to come.
Justin Steinman:
So you're not moving to Canada?
Chase Zaputil:
No. I'm all right.
Justin Steinman:
All right. Just checking. Just checking. Let's pivot a bit. Talk a little bit about your background as a pharmacist, and, frankly, why did you make a transition from being a pharmacist into the data and technology industry?
Chase Zaputil:
Yeah. Came up as a pharmacist and worked in community pharmacy, so that's within the retail pharmacy world. A lot of it on the large chain side, I actually did a residency with large retail pharmacy chain right after school and happened to get involved in the technology group at that particular moment. And they were building a new pharmacy dispensing system so that got me in the technology side. I was working with developers every day really thinking about how do we use technology to drive the efficiency, effectiveness of pharmacists in clinical practice. And so that was my start in the career. I always said I never thought I would evolve from being a pharmacist to being in technology and software development, but I did, and then had the opportunity to go over and work with a growth organization at Wolters Kluwer Health, as you mentioned. So that was my first foray into growth and business development. So yeah, never thought that the career path would happen the way it did, but it's been really exciting, and a lot of great opportunities as I advance in my career.
Justin Steinman:
So I love asking people questions that they may or may not be qualified to answer, but hey, you're a pharmacist, so I'll ask you anyways. What's your perspective on the state of pharmacy benefits today?
Chase Zaputil:
Yeah, it's really interesting. Pharmacy benefit management, it's an interesting one. I think there's a lot of opportunities for disruption, honestly. I mean if you look at the fact that you have really three large pharmacy benefit managers that control 80% plus of the overall prescription volume is pretty significant, right? And all three of those are aligned with a health insurance company, so there's vertical integration there, and one of them has a large retail pharmacy chain attached to it. So those are big entities. Those are large entities that are controlling a large portion of the drug spend in the United States. And I think there's opportunities to think about that differently. If you think about even the concept of things like discount cards, those are really bypassing that process of the pharmacy benefit management industry, and those have been around for a number of years. They've made inroads, and it's interesting to think that those actually even exist, right? And if you look at who's using those, I think some data I saw was 50% of those discount cards are utilized by people who don't have insurance.
But that population in the United States for the uninsured has continued to shrink. I think we're down to maybe less than 10% or around 10% of people who don't have healthcare benefits today, so that's shrunk. And I think I saw like 5% of insured people are using a discount card in some way, shape, or form. So that's pretty significant amount of utilization, and there's a reason. They can get just as good a pricing going through a discount car program as through a pharmacy benefit management company at this point. Now, PBMs do a lot of good as well. I mean, there are things that they are driving from overall trying to look at cost savings and how we rationalize the spend of our resources, and they have a tremendous amount of buying power, those things that are happening. So there are things that happen within the PBMs that is positive, there's no doubt about it. But I do think there's potential [inaudible 00:22:19] to really rethink that a little bit differently, if you will.
Justin Steinman:
Some of our listeners, and maybe you if you had heard of Mark Cuban's Cost Plus Drugs, right? Pretty much just a straight 15% markup, and it's pretty transparent right on the front page of the website, and then they're like, "Hey, it's 15%. We add a little labor cost on top of it and that's your price." Is he onto something?
Chase Zaputil:
Yeah, I mean it's really interesting. The thing I do love about it, to your point, it's transparent and it's a direct to consumer model, right? And my career has spanned, as I talked about, retail pharmacist. I spent a little time in the PDM world way back in the day as well, and I remember us at that time really talking about, at that time I think it was called consumer directed healthcare or the idea that we're going to have the consumerization of healthcare and allow the consumer to make more choices and actually make decisions on their own and have more transparency in how they're making those decisions.
And I think that's really interesting about this is that it really is trying to think about that model differently, and even in my experience in being in the retail pharmacy world, there was times where it could have made a lot more sense to just say, "We'll sell the drugs at cost, and just add a dispensing fee and make sure that we're making a margin on that," because there is a lot of really interesting things that happen in the pharmacy world from a pricing... How you buy drugs, ultimately, how you negotiate contracts, all those things. And it's not always transparent. I think more models are moving that direction, and it's exciting to see that because I do think transparency in things... And this goes beyond pharmacy by the way. I think the same thing happens on the medical side as well. This is a little bit on the No Surprises Act, the whole idea, how do you get these surprise bills? How can you not understand what you're going to pay for a medical service?
And I think that all of that, in healthcare in general, I think if there's one thing we can do as an industry, it's bring more transparency to the industry.
Justin Steinman:
Man, we've hit all the big topics here. We've hit interoperability, consumerization of healthcare, transparency. I mean, this is a who's who of the unobtainable in healthcare, right? Going back to this Cost Plus thing though, to some degree, isn't Cuban just moving costs around the supply chain? I mean, at the very end of day, the huge drug costs are being driven by the drug manufacturer, not the middleman or the PBM. Can we tackle the high costs of drugs? Do we need to go back and do something different at the manufacturing level?
Chase Zaputil:
There's costs that you could be looked at across the board, I think. But I do think that particular what the Cost Plus Drugs is looking at is really interesting. I do think as they progress into things, I think I've seen some releases on where they're going in their vision. The idea of things like a vertical integration where possibly the manufacturer of medications and being able to completely go direct to consumer with drugs at a price that is discounted is really interesting, because that's really something new. We talked a little bit about vertical integration that exists in the existing models, right? That's health insurance, you got a PBM, maybe even have a retail pharmacy attached, but manufacturer's always been outside of that. And so things like generics are really an interesting spot because I think today 85 plus maybe close to 90% of the drugs that are spent are generics, so that's a huge piece of the overall landscape of what prescriptions people are taking.
So if you think about the idea of how do you manufacture those generics, how do you actually get those drugs to the appropriate consumer, that's a really interesting model, and I do think what they're doing is pretty unique in that regard. They have obviously the direct to consumer experience, they've announced I think some stuff with Capital Blue Cross. They're doing some interesting, I think, programs, if you will, that I think could really change the way that people ultimately get benefits of a prescription drug program.
Justin Steinman:
So it's clear that there's definitely a lot of opportunity for change and innovation in the PBM space.
Chase Zaputil:
For sure. I think there's so much opportunity there. It's really an interesting one. And it always stayed near and dear to my heart because of growing up in the pharmacy space. And I think at the end of the day, making sure people have access to appropriate medications at the most reasonable cost is the most important outcome, and I will say that's the cause I think that something like Cost Plus Drug is really after, which is a noble effort for sure, which is exciting.
Justin Steinman:
Well, I'm glad you brought up health equity there, because that's the last question I wanted to get to you today before I let you go. I want to link together everything we've talked about. We talked about data quality, we talked about EMR, talked about transparency, pharmacy benefits, consumerization interoperability. And there's no question that one of the biggest challenges facing our country today is health equity and access, right? How do we use data to improve health equity?
Chase Zaputil:
Honestly, it's one of the most important questions I think any of us can ultimately think about. And I mean there are so many opportunities for us to use data ultimately to improve equity. I think the big thing is thinking about where inequities exist and how do we ultimately deliver a solution around that. And if I think about even when we were talking earlier about things like provider data, just the simple ability for a member of a health plan or anybody who's getting the benefits to go on and find a quality provider that actually is, for instance, in network that's close to them, that they can ultimately go and see that's accepting patients. Honestly, that that process is not easy today. I mean, it's very difficult. And I think those things, we'd assume that we could solve those very simplistically, it's not always the case.
And I think this is where though data and our ability to ultimately process data and prove out what we think are the appropriate hypotheses is really important. And I think that if we all work together on those types of solutions, I think that's obviously going to make that much quicker. The other thing I would say is you mentioned earlier too the ability to partner within ultimately our ecosystems and within the healthcare industry is really important, because one entity is not going to solve these issues alone, right? I think there's large organizations that are doing a lot of great stuff. Sometimes they can't solve these issues by themselves and they need maybe a younger company with a little more drive to do something different too, right? So that's okay, and I think that how we partner together to solve these really big challenging situations is the most critical. So yeah, it's interesting to think about that. I think it's always something that is, from our perspective at Veda, I'm sure for your perspective at Definitive, the idea of equity and healthcare continues to be one of the most important issues that we're all tackling together.
Justin Steinman:
Yeah, it is. And it's one that's sad that I don't think we'll be able to really fix any time soon. You just got to take a baby step every single day to a little bit. It ain't a lot.
Chase Zaputil:
Yeah, agree. We talked about it. I think that's the one thing that's frustrating about healthcare, but I think it's also one of the opportunities we have is that we are not going to probably solve anything overnight, but I think that all of us working together can improve things on an iterative basis, and hopefully small changes lead up to big changes. And I think we'll hopefully continue to see that.
Justin Steinman:
I love that, and I love ending there on a positive note. Small changes leading up to big changes. Chase, thanks again for coming on Definitively Speaking today.
Chase Zaputil:
Yeah, thanks so much for having me, Justin. Appreciate it.
Justin Steinman:
For all listeners out there, thanks as always for joining us on Definitively Speaking, a Definitive Healthcare podcast. Please join me next time for a conversation with Othman Laraki, CEO of Color. Color provides the technology and infrastructure for large scale health initiatives, for everything from a population genomics program to high throughput COVID-19 testing. Othman and I are going to have a wide ranging conversation around health equity and what's fundamentally broken in our healthcare ecosystem, and Othman's thoughts on how to fix it. You won't want to miss this. If you like what you've heard today, please remember to rate, review, and subscribe to the show on Apple Podcasts, Google Podcasts, Spotify, or wherever you get your podcasts. To learn more about how healthcare commercial intelligence can support your business, please follow us on Twitter at definitivehc, or visit us at definitivehc.com. Until next time, take care, please stay healthy, and remember that small changes lead to big ones.