The First Medicare Data Vendor, the Nintendo Wii of healthcare data

CareSet Systems is the first Medicare Data Vendor. Soon, we will release our first analysis from Medicaid and also be the first Medicare/Medicaid data vendor. We think this will make CareSet into the “innovator’s dilemma” for the healthcare data industry. Traditional data vendors like IMS, Truven, Optum and First Databank will likely continue to do very well in healthcare, but slowly the market will reward CareSet and other Medicare Data Vendors that inevitably follow us, with every data task that we are able to perform.

And there will be other Medicare Data Vendors. There are now two programs under which CMS gives the rights to private vendors to sell Medicare and Medicaid data. The first is the Virtual Research Data Center (VRDC) program, which is how CareSet (through its sister company, DocGraph) acquires Medicare data. The second is the Qualified Entity (or QE) program, which is a way for vendors who already have commercial claims data to gain access to Medicare data. Both programs allow private participants to resell data ultimately sourced from Medicare or Medicaid claims.

Of course, there are limitations. Strict rules prevent the release of Medicare data that can easily be connected with patient identities. Medicare data vendors are limited to describing what happens to groups of Medicare beneficiaries. These data release standards are a very conservative approach to protecting patient privacy. This conservative approach is embodied in the required cohort size. When data is being released to the public, using either VRDC or QE systems, cohorts smaller than eleven patients are forbidden.

In comparison, the equivalent NHS policy defines the smallest cohort of patients that can be released in public data sets as six. About 65 Million UK residents are covered by the NHS, vs about 55 Million people covered by Medicare in the US. That means that the NHS has a 20% greater sample size, but a cohort size requirement that is half the one chosen by CMS.

So Medicare is being very careful as it allows the release and sale of aggregated data sets. Every data set that we acquire from CMS is carefully reviewed by CMS to ensure that we are following the CMS standards. The commercial claims data that is routinely sold by companies like Truven and IMS have no such limitation. The data sets sold by traditional healthcare data vendors include the specific, but deidentified, claims history of individual patients. In these data sets, you can see specific information about the path that a single patient took as they moved through the healthcare system. There are many use cases where researchers need to have this type of granularity and many problems that cannot be solved without these level of access.

But also, there is a large number of use cases that can easily be solved with aggregated data sets. In fact, there are many use cases where the most important feature of the data set is its size, and even with the increasing popularity of Medicare Advantage (Medicare Part C) plans, Medicare fee for service remains the largest single source of payment for healthcare in the United States.

All of that qualifies CareSet as an “Innovator’s Dilemma” to the rest of the healthcare data industry. Most of the time, we can answer questions about the healthcare system at a fraction of the cost of traditional healthcare vendors. This is not to say that we are a “cheap” source of data, far from it. We are in fact a premium vendor of Medicare data. There are at least two reasons for us to charge at a premium rate:

First, our years of experience mining currently available public use files (PUFs), and wrangling data out of HHS using Freedom of Information Act (FOIA) Requests allow us to get much more value out of VRDC.

Second, we have years of experience leveraging machine learning technologies on top of these datasets. Especially leveraging graph-based learning systems to allow for deeper examination of patterns in the healthcare system. This allows us to get more clinical insights out of aggregated clinical data sets than other vendors.

We charge a substantial premium for our Medicare/Medicaid data services. But ultimately, that is a premium on top of our costs for data acquisition, which are orders of magnitude cheaper than the data acquisition costs for IMS or Truven. In the end, we are a premium vendor and still much cheaper than current proprietary healthcare data vendors.

If you work in the tech industry, you would probably have to be living under a rock for the last decade not to have heard about the book “The Innovator’s Dilemma”. The author of the book, Clayton Christensen, argues that the primary threat for technology companies is not competitors who have equivalent offerings. Instead, threats to a current technology are new technologies that are not as capable, yet have specific appeal in some way. To quote straight from the book:

Generally, disruptive technologies underperform established products in mainstream markets. But they have other features that a few fringe (and generally new) customers value.

Christensen, Clayton (2015–12–15). The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail (Management of Innovation and Change) (Kindle Locations 265–266). Harvard Business Review Press. Kindle Edition.

There are countless examples of the innovator’s dilemma in the book, including the steel industry (large scale steel to minimills), the computing industry (mainframes -> PCs -> Laptops -> Tablets -> Cell Phones) and the hard drive industry. But my personal favorite example is lessons learned in the Seventh Generation of Video Game Consoles.

In fact, in those Lean Canvas marketing exercises, where you sort out your pithy way of explaining the role you play in an industry (i.e. like “We are the Netflix of the taco industry”, or “We are the Beats headphones of the frisbee industry”), I like to argue that CareSet is the “Nintendo of the healthcare data industry”. But not just any version of Nintendo (which is by far the longest running video game console company), specifically the Nintendo that released the Nintendo Wii, during the Seventh Generation of the Video Game console wars.

To set the stage for the Seventh Generation, Microsoft had just bought its way into the game console industry with the release of the original Xbox platform, which leveraged everything that Microsoft had learned supporting decades of PC video games. Sony had entered the game console industry in the previous generation with the release of its hugely successful Playstation line of consoles. The new console from Microsoft and the massive improvements available in Sony Playstation 2 had been so substantial that Sega decided to entirely retire its console offering. Its sixth generation Dreamcast would be its last device.

Sony and Xbox were throwing tremendous amounts of computational power at winning the console markets. Both companies were offering their platforms at a huge loss on the hardware, sometimes losing hundreds of dollars on every device sold. Sony and Microsoft would still make money on their game platform, but only after the profits from two or three separate video game titles had been sold. The software (the games) were so profitable, that they were basically giving away hardware under cost in order to capture the valuable game marketplace. During this era, some of the world’s largest supercomputers were built by connecting thousands of these game consoles together. Clever supercomputing researchers would use Sony’s generous console hardware prices to fund their supercomputing grids. The capital requirements for this kind of “platform play” were staggering, and Sega simply could not take the heat. Now Sega just makes software for other companies game consoles.

What should Nintendo do? How should it compete against these huge investments? How could it afford to compete with the massive capital investments from the likes of Microsoft and Sony?

Their answer was to become the innovator’s dilemma for the video game console industry with the release of the Wii. The Wii did not boast tremendous hardware specs. The graphics were not as good as the other members of the seventh generation, both the Microsoft Xbox 360 and the Sony Playstation 3 were far more powerful. But instead, they did something different at the lower end of the console price market. At the end of 2005, the Xbox 360 launched at a $400 price tag, a year later Sony would launch the Playstation 3 at $500. The Nintendo Wii also launched in 2006, at a price of $250.

Nintendo sales of the Wii beat both Microsoft Xbox 360 and Sony Playstation 3 combined. But more importantly, Nintendo made a profit on every Wii sold, while Microsoft and Sony lost money on the sale of each device.

How did Nintendo do that? They made a new way to control video games, with the Wii controller. Rather than focusing on capabilities, they focused on the underlying purpose of a video game (which is to have fun) and revolutionized the video game industry. They offered a less expensive product that was not as capable as the competition but still had a significant compelling feature. The release of this device changed the calculus of video games forever.

As one of the technical founders of CareSet (with all founders being technical to some degree) that is what I believe CareSet is doing for healthcare data. I believe that inexpensive CMS data that summarizes what is happening in the healthcare industry at a fraction of the cost of commercial data and with improved overall patient privacy, is our “Wii”. Our “Wii controller” is going to be the specialized machine learning that we use to make that data nearly as valuable as claim-line-level unaggregated claims data. CareSet is a member of a new class of company, an Extract, Transform, Load and Learn company (or ETLL), that takes raw data in one side, and solves tough information problems on the other side. In our case, we intend to address as many of the Gordian Knots in healthcare as we can.

Currently, the only people who are solving problems with healthcare data are paying a tremendous price for it. Through our partnership with DocGraph (our sister company that holds the CMS VRDC seat) and CMS itself, we believe that CareSet can be part of a movement to dramatically improve what data is available to solve healthcare issues. To do that, we made a very conscious decision to build a new type of business, one designed from the ground-floor to be an innovative distribution in our industry.

Of all the impressive things that Nintendo accomplished with the Wii, probably the most important, was to democratize gaming. Nintendo’s prices and family-oriented style of fun encouraged people who would have never chosen to buy a console before to buy the video game and to enjoy it with friends and family. Nintendo ensured that lots of people “had fun”, which is the end goal of the industry, by making video games much cheaper and much more accessible.

At CareSet we believe that the purpose of the healthcare data industry is to improve the delivery of healthcare to patients, by using data science to sort out which treatments and practice patterns help the most. We also believe that our pricing and innovations will serve to ensure that more data consumers across healthcare can afford to “check the data” more frequently and with greater depth. We believe that the end result is that patients will have better and faster access to the right medications and other treatments to help ensure that they can maintain their health and happiness as long as possible.

I know it sounds silly, but CareSet is working to be the “Nintendo Wii of healthcare data”.

Thanks for reading,

– Fred Trotter, CTO CareSet Systems

Fred Trotter

Fred shapes our software development and data gathering strategies, which doesn't stop him from getting elbow-deep in the code on a regular basis. He is co-author of the first Health IT O’Reilly book Hacking Healthcare, and co-creator of the DIRECT protocol mandated in Meaningful Use. Fred’s technical commentary and data journalism work has been featured in several online and print journals including Wired, Forbes, U.S. News, NPR, Government Health IT, and Modern Healthcare.