Data-Driven Healthcare


  • The Data-Driven Healthcare session at Techonomy Health. From left, Daniela Hernandez of the Wall Street Journal, Andrew Thompson of Proteus Digital Health, and Oscar CEO Mario Schlosser.


Mario Schlosser
Cofounder and CEO, Oscar

Andrew Thompson
Cofounder and CEO, Proteus Digital Health, Inc.


Daniela Hernandez
Digital Science Editor, Wall Street Journal

Session Description: Data is coming from all overincluding inside our bodies, the systems of hospitals, and health insurance. How much will we have, and what will we do with it? How will it change the quality of care, how we fund it, and everything else?  

Below is an excerpt of the Data-Driven Healthcare panel. The full transcript can be accessed here.

Daniela Hernandez, Moderator: I have here Andrew Thompson, from Proteus Digital Health and Mario Schlosser from Oscar, and so I’ll let them introduce themselves before we get into the questions.

Andrew Thompson: My name is Andrew Thompson. I’m the chief executive of Proteus Digital Health. You could tell from my accent that I grew up in Texas but I’ve spent the last 30 years in Silicon Valley starting and building technology based healthcare companies. I’ll talk about Proteus where I’m the founder and CEO. Proteus makes digital medicines. These are drugs that, when you swallow them, talk to your cell phone. What we tell you is did you take it, and did it work? Those are the two most fundamental questions in all of chronic disease management, and in particular, ambulatory care with a drug.

Mario Schlosser: I’m Mario Schlosser, the CEO and cofounder of Oscar. I also have a European accent. I love the fact that we have two Europeans telling the American audience how broken the US healthcare system is. Apologies.

I am a computer scientist originally. I run a company called Oscar. Now it’s a health insurance company, at its core, a technology and health insurance company. What we do quite differently, I believe, than the other health insurer out there, is that we built the product we have on three different pillars. Number one is consumer engagements. We engage members in a very different, deeper kind of way.

One nice statistic illustrating that is the fact that in Oscar’s case, about 25% of membership last year has used telemedicine to treat some illness that they have. Usually, if you look across US health insurers, about 2–3% or so have membership that have used telemedicine.

The second point is that we build all the technology of the operations in house. We have a full technology stack in house. We do everything from claims management to clinical outreach, to network construction, etcetera. We use data—hence the panel here as well—for fueling of these activities, we’ll talk about that in a second. We’re watching right now in the Oscar office in New York here, at Lafayette and Houston and on the wall you would have, for example, a real time manifestation of the current medical loss ratio and the components that go into it. It’s one of the kind of advantages we have in really making sure data flows through our systems in a different kind of way.

And the third thing is we build different kinds of networks of physicians and hospitals. We try to connect in a more deeper fashion with the providers we have, doctors, and hospitals as an example. Doctors under the Oscar network can log into provider apps that give them all the data that we have, but members make it available for them, including conditions we suspect that the members might have, based on lab tests we’ve seen, tests we’ve seen them do, drugs that they are on, and things like that. We founded the company with an eye towards really being at the center, at the nexus of the various data flows, which tend to be fairly broken and un-unified in healthcare. The insurer is a natural nexus of all these data flows and we thought we could make them more powerful and more meaningful.

Hernandez: Why isn’t healthcare already data-driven? And, what is it right now? And, how can we get to this data-driven vision?

Schlosser: There’s one very simple answer and I think it’s hidden in the incentives that are built into the system in the US healthcare system at the moment. It’s a fact that almost this entire system tends to be a cost-plus system, if you will. Almost everyone in healthcare right now, if you really uncover the way it works, gets paid a constant percent margin on top of rising cost trends. And if you’re in that industry, you’re quite happy if, you know, the costs sort of go up over time. If you get paid fee-for-service as a provider, as a hospital, for example, you don’t necessarily use data to keep people out of the hospital. You’re better off financially if they come back into the hospital.

Hernandez: That’s changing, right?

Schlosser: It’s changing. In that change lies the power in the shift towards more data. In the end, I think, what we see from the insurance company perspective is that much of the data flows in the current healthcare system tended to be transactional. If you’re an insurance company, you had to pay claims. That’s what you did. You’re a transaction engine. So, your data is only towards that. If you are a hospital, you don’t necessarily need to have a clean so called roster file of doctors working for your hospital in where they practice what they practice. You need to get the claims paid. That is the emphasis on the type of data that you have. If you want to start using the data that’s in the various systems in healthcare that tend to be very fragmented towards better patient care and management, you’d interpret it differently and incentivize the use of data very differently.

Hernandez: Andrew, how do you see [that] in terms of what you do?

 Thompson: We do have a data-driven healthcare system. It’s just that it’s driven on very little data, right? So, if you look at how drugs are approved, it’s on relatively small panels of patients and you do some statistics. So, it is data-driven, but it’s small amounts of data. And then let’s look at some of the points that were made about how we then use data inside the system. So, the first thing to say is we don’t have a healthcare system. We have a sickcare system. Right? And it’s important to note that it was built in the last century to do a very important job, which was to deal with acute disease and trauma. And it was built using the best technologies we had in the last century. So it’s buildings where you plug into electricity, people with knowledge in their heads, and products that were designed to be safe in everybody, and work in somebody. So, you didn’t need much data, right? Mass standardization, great achievement of the industrial era.

Today, we have very different challenges, 75% or 85% of what we need to deal with is chronic disease that’s dealt with in community settings, not in hospitals. So, we need to supplement, and in many ways magnify the power of this magnificent sickcare system with a healthcare system, and we need to build it using the best technologies we have in our hands today. So, a building where you plug in is going to be magnified by the incredible power of the mobile device where you log on. And people, with knowledge in their heads, are going to be massively augmented and supplemented by software and servers with intelligence in the cloud. And products that were designed to be safe in everybody, and work in somebody, are going to become services that are tailored to you, your genes, your lifestyle, your behavior, delivered where you live, work, pray, and play, in ways that you can see, measure, and understand. That is digital health.

Now, let’s talk about data. What most insurance companies do, and what most data aggregators do, is to take data out the sickcare system. It’s all money flow and work flow. It has nothing to do with life flow and health flow, and that’s what we’re about. As a company, what we do is to capture information about what’s going on in people’s daily lives, in the settings where they live, using the drugs that support them, and prevent them from having strokes and heart attacks. That’s life flow and health flow and that’s the core of how we’ll build a data-driven healthcare system, not sickcare system.



Tags: , , , , , ,