An Exclusive Interview with Scott Shreeve, Medsphere Co-Founder and Chief Medical Officer

The first thing you notice in talking to Scott Shreeve is his boyish energy and enthusiasm. The second, at least for us non-Left Coasters, is his use of surfing metaphors and decidedly un-corporate, SoCal expressions. He's the friendly Beach Boy who's impossible to dislike, other than for living what seems like an enviable surfer lifestyle that will keep him young forever. He's also a passionate physician and businessman who may be the perfect person to talk about disruptive technology and the open source movement in healthcare IT. Medsphere is the un-cola of the industry, the little company whose patch of beach is to bring a commercialized version of the VA's acclaimed VistA system to the hospital masses, offering a less-expensive alternative to that majority of hospitals unable to afford what the big boys are peddling. 


Tell me about Medsphere and your role in the company.

I’m one of the co-founders of Medsphere and its Chief Medical Officer. My brother Steve is the other co-founder and CTO. We always wanted to combine business and healthcare. I played football and basketball in high school and college and originally wanted to go into sports medicine. I took zoology and economics, and instead of running a zoo, went into healthcare and found it was similar (laughs.) As an economics student, I watched the industry and saw a sea of change happening around IT.

In my third and fourth year of medical school, my job was to gather information – rounding, tracking labs, picking up films, and bringing information back to the team to help them make decisions. I was amazed at how unproductive and inefficient that was, no matter where I was rotating. All these information management tools were available in other industries. We had paper.

Then, I rotated through a VA hospital. I was really impressed with their technology, which was then called CPRS. It let me sit down and interact with patient data, take care of consults, order tests, and do patient follow-up. Back then, other students and residents joked that the VA was a place where you “practiced” medicine, but I was blown away that they had developed this system of a caliber above those in the private sector hospitals I had worked in.

My brother Steve was finishing his first two years of medical school and decided to help them create their first combined MD-MBA program. When he was ready for his third and fourth year, I told him to check out the VA’s health record as an example of a system that was pretty good. He’d been an open source hacker since the early 90s and had expected to see really integrated applications to help physicians manage massive amounts of information. Until the VA, it wasn’t there. Having completed business school, he looked at the problem differently … why hadn’t someone done something to make this public domain software available to hospitals that needed it?

Linux was starting to take off and we were looking at this massive application that had been highly successful in the VA’s closed environment. We saw many of the successes that Linux was having. This could be the hospital's OS. We felt there was a market opportunity for a cost-effective, proven system that could be used by hospitals that couldn’t afford commercial products.

The last step was to find a license-free database. A company called Greystone Tech, one of the last remaining independent MUMPS shops, was acquired by Sanchez. GT.M wasn’t a core component for them, so they decided to release it as open source in the summer of 2001. That decision made the whole VistA application stack open source.

VistA had been in the public domain for 20 years. The Indian Health Service had created their own system from it, staying fairly close to the VistA platform in modifying it for their population. Ironically, the Department of Defense paid SAIC several billion dollars to customize a version of VistA they called CHCS. However, the VA had more funding, better vision, and a better development environment, so they developed the clinical pieces, such as the EMR, CPOE, bar coding, and imaging, all through the late 90s.

By 2000, they had fully implemented the system in all 170 hospitals and were beginning to be recognized as innovators in healthcare IT. The FDA asked them for bar coding guidance. They were consulted on how HIPAA could be implemented. Device companies starting going to the VA to have their products interfaced to VistA, with the count now up to 5,600 interfaced medical devices. About 85% of all physicians in the US rotate through the VA during their training, so the word was getting out.

In 2001, the IOM published a report called “Leadership by Example," where they stated the position that the government could best influence healthcare by setting standards for improving quality of care, including interoperability. Then, private industry should build and extend from there. They referenced the VA’s system as one of the best in the nation. The VA had unparalleled experience in widespread deployment and data was starting to come out of their large-scale implementation, allowing them to prove that their quality measures were outstanding.

Tom Scully at CMS started pushing the industry to meet quality metrics, telling them if that they didn’t get their act together on standards, the government would use the VA experience to create them. He even told a roomful of healthcare IT executives that maybe the government should just promote VistA or even send out free disks to providers and tell them, "This is what Medicare supports." This was VistA’s coming out party in 2003. Until then, people had the false impression that the VA sucked and that the government could produce only crap software.

What's your business plan?

We started Medsphere in 2002. Our first contract in 2003 was for seven long-term care facilities, the first VistA deployment outside of the VA. We created a commercially viable product and provided professional implementation services and support. Working with partners, we could provide an end-to-end solution. For example, for those Oklahoma LTC facilities, we worked with HP for hardware and project management experience. Those seven facilities standardized around our application - order sets, templates, notes, and business processes

We see four stages of our company. The first is "proof of vision." Is this is a good idea, can we build a team, and can we raise money? We spent one and a half years telling the story. No one believed us that this system was available and as good as it was. This first phase ended in 2003 with our Series A funding and the Oklahoma contract.

The second stage is "proof of concept." We had the money, the team, and the customer, but could we really do this? We developed processes around implementation, training, documentation, and support. We were building the company.

We’re in the third stage now, the "go-to-market" stage. We got the second round of money, brought in additional management expertise, built up sales and services, and created a company around this new paradigm. Midland Memorial in Texas, Laura Bush’s hometown, was our first true commercial customer. It’s a classic community hospital, with 425 beds over three campuses. Their funding is limited, but they have the same information management needs as everyone else. They were in sticker shock after being quoted $18-25 million for a full EHR system. Their CIO was a Linux aficionado. He saw the opportunity to be involved with a disruptive technology and found Medsphere while poking around open source apps. He went to a VA facility to see VistA work, listened to our story, and was satisfied at the level of risk involved with working with a young company compared to the value and functionality. We’re implementing them now. They’ll be live with CPOE in December.

The last stage is "grow the market." The next three to five customers will prove that we can scale and support them. Then, we’re ready for rapid expansion and penetration. The timing is incredible. The interest in VistA is phenomenal, this multi-billion dollar, highly successful federal government investment that can be commercialized.

VistA Office has caught everyone’s attention. We think the federal government’s role is to set standards, like they did for railroad gauges and the Internet protocol. That will let the private sector really begin to innovate and get rapid adoption. The government can then get out of the way and let the private sector go. Maybe CMS is slightly overstepping their bounds by offering software, but we understand why they’re doing it. Their goal is to create an EHR infrastructure that 75% of ambulatory-based physicians can be a part of.

Their real objective is to gather information to improve the quality of care. This can happen only when you can measure, report, and monitor the quality of care. The pervasive platform lets them rewire physicians for quality care through a pay-for-performance approach. VistA office is a low-cost, standards-based product that’s publicly available and serves as the lowest common denominator to meet the CMS goals for quality.

Many people think that VistA Office is a new product. It is not a product. It’s an adaptation of VistA that lets it operate standalone and to work with foreign systems. VA runs VistA in 1300 clinics, but because it’s tied to the VistA mother ship in the medical center, it doesn’t know where to get information otherwise. VistA Office begins to address some of those limitations.

How many employees does Medsphere have?

50.

What’s the relationship between VA Software and Medsphere?

There is no relationship, other than VA Software was the former company of Larry Augustin, our CEO.

What do you think of the VistA Software Alliance?

We were a founding member and original author of the concept. It was born because the VA was unable to work individually with companies like Medsphere. They have stringent rules about vendor favoritism, so while they were very interested in what we were doing, we couldn’t interact with them. They requested that a foundation or buffer organization be created so they could interact with a neutral body, true to the open source spirit. We formed a consortium of complimentary partners. We felt that this trade organization could facilitate increased adoption of VistA. We were also confident that the market would recognize our technology, product, and service differentiators as more members joined.

Since the announcements about VistA Office, there’s a ton of consulting shops looking to get at the table. But, there’s a reason that it took 20 years to get VistA commercialized. Its like SAP … successful implementations require detailed, complex knowledge that takes years to acquire. Would you build a nuclear submarine just because someone offered you free blueprints? We’ve developed a portfolio of configurators, rapid implementation tools, patch management, system monitoring, and other tools that make us different.

Since Medsphere was first, we had the opportunity to gather an incredible team of ex-VA and VistA luminaries. George Timson, our Senior Architect, is viewed as the rock star of the Underground Railroad, as they called it back then. He was one of the original Hardhats, the programmers who started working on VistA in 1979. There are great stories of the intrigue and political wrangling that forced this renegade group of programmers to go underground to collaborate before the VA cried uncle in 1982 and agreed to adapt VistA as their platform of choice. George created Fileman, an incredible file management tool that’s been running 20 years and still runs great today. He wants to get this incredible legacy he created out there and loves working with us to bring it to the broader market. Lots of people want to work with him. He’s one of the most intelligent and witty people I’ve every met in my life.

Larry Augustin was a paper billionaire before the dot-com bust.

In 1993, Larry was an electrical engineering post-grad student at Stanford. He was doing number crunching, needed hardware that his student-sized budget could afford, and started playing around with Linux. He formed a company that shipped preconfigured Linux boxes. He was in the eye of the storm in the Silicon Valley, riding the wave. Complete insanity and they all knew it, but what could they do? We do a lot of surfing down here in SoCal and Larry caught the gnarly mother of all waves and rode it in style. He even managed the downturn well. Like any good surfer, he got up after some incredible wipeouts.

Larry is bringing a more measured approach to growth, built around repeatable metrics. He’s instilling a sense of vision, methodology, and structure to Medsphere. When Steve and I started the company, we did whatever we had to, working ridiculous hours doing everything. I’m an ER doc, so I never know what time of day it is anyway, but we were putting in mondo crazy hours. We’re transitioning into something more sustainable, with a clear sense of purpose and mission. Larry’s looking for measured growth and an emphasis on doing the right things at the right time.

What hospitals are live on OpenVista and with which modules?

The seven Oklahoma facilities are totally live. Midland’s implementation starts in the fall.

Can you build a business around free software?

Absolutely. We spend a lot of time explaining the concept. The software’s not free. Midlands had a quote for $18 million from a big vendor and we came in at around $7 million for everything – hardware, software, and services. We have to charge significant dollars because of the massive size of the application and the expertise needed to implement it. Midlands demonstrated that we can do all of this for a lot less than what others offered.

Clearly, the core technology from VistA is publicly available. We’re taking that multi-billion dollar investment and adding functionality, features, and integration to make it commercially viable. People are more than happy to pay for that because it’s a great value for them. That’s the power and advantage of open source. We also leverage open source at the hardware, OS, database, and tools level. We’re not OS bigots – we’re OS practical. We’ll lower your cost, but if you want Cache’ on Windows instead OS alternatives, then go for it. That’s fine with us. OS is about freedom to choose and to save or spend money wherever you fall out on the risk-to-cost profile.

Is healthcare, particularly in the US, ready for open-source, mission-critical software?

We think there’s a strong corollary. Healthcare is cost-sensitive, all about quality, peer review, standards, and interoperability. When you look at open source, those are its qualities as well. Many people think open source is of higher quality because many people have input into the code and are checking each other in a form of peer review. Open source is powerful in its ability to commoditize functionality, so when there’s a commodity market, open source does very well.

Functionality was once a differentiator in healthcare, but vendors offer mostly the same things these days. Everybody has order entry and results reporting and most have CPOE. The feature/function has become a commodity. It pushes you up the value chain as a vendor. You have to be better in quality, training, or something that’s not as quantifiable.

We think that open source is a disruptive innovation for healthcare. Systems are way over-engineered, overly complex, and overly expensive. If you can provide a “good enough” solution that meets the need at the right price point, you’ll blow the market away. We offer a cost-effective, comprehensive, proven system that can penetrate the places that have never had the opportunity to get these kinds of systems.

We don’t even try to compete with Cerner, Epic, McKesson, Siemens or other big vendors with their marketing and sales. However, 85% of hospitals still don’t have an EHR. We’ll go after that other 85%. We’d be crazy to pitch OpenVista to Kaiser Permanente at this stage of our company. The first thing they’d want to see is our last 15 implementations or 10 years of financials, and we don’t have those. We'll go to the Midlands of the world. They don’t want crappy software and their needs are the same as anyone else's, but they’re ecstatic about the price and the value we offer. They’re confident in the platform and in us as a company. They evaluated the risk and went with us after meeting our team.

We believe there are five more Midlands-type hospitals out there right now and we would invite your readers to help us identify those. It’s time to bring this cost-effective and proven product that can improve patient safety and clinical efficiency to a customer base desperate for alternatives. OpenVista is healthcare IT for the masses.

Cerner and many other companies are doing very well selling proprietary IT systems. What are they doing wrong that will drive customers to seek you out?

We have the utmost respect for the large vendors There’s a reason they’ve been successful: they have great products and impressive customer experience. However, what they’ve built comes at a price.

Epic has earned their reputation, but because they’re expensive, they don’t have to talk to small hospitals. If only Jaguars and Cadillacs were offered, a lot of people wouldn’t be driving cars.

MEDITECH has a great app and business model. Anyone not envious of their numbers is crazy. We often see them as similar to what we’re looking to do. We hope to replicate their success. George Timson says he was MEDITECH’s first employee. That was an amazing era. MUMPS was absolutely the right language at the right time, given the hardware limitations. Epic, MEDITECH, IDX, QuadraMed, and others have been highly successful leveraging that architecture.

We’re offering a high-value Toyota or Honda that works. Our software is expensive, costing billions to write and hundreds of millions a year to update. The beauty of it is that we don’t have to pay for it … or, maybe I should say that all of us pay for it (laughs.) Medsphere has the biggest R&D department of any healthcare IT vendor.

The challenge is in stacking VistA modules head-to-head against the big products. It would get blown out of the water in some areas. However, when you look at the comprehensive whole, the overall cost and value received for it, you start to see the value of a “good enough” solution. VistA starts to look rock solid. Medsphere OpenVista can meet and exceed features and functionality expectations.

As a physician, how do you think clinical systems fit into the Medsphere strategy?

The VA never worried about billing software since they highly focused on making sure the providers had the best tools for taking care of veterans. Most other vendors started with financials and back-filled with clinicals. VA was ahead of everyone else – CPOE, GUI, physician chart – that’s why it became a standard. They have 200,000 people a day banging on the system and those residents and doctors are making lots of suggestions to product development.

The VA could say, “This is what we’re doing, this is the VA way, this is how we take care of our nation’s heroes. If you don’t like it, don’t work here.” What were the docs supposed to say? The developers were responsive and the applications worked, so eventually the physicians starting using and actually liking the system to the point where 95% of their orders are entered electronically. That is amazing.

The VA wants to rewrite VistA, presumably to eliminate MUMPS and Cache’. How would that affect Medsphere’s product strategy?

If you talk to the IT staff within VA, they’re flabbergasted. They have this totally awesome system and they’re trying to replace it in a way that is high risk. The VA is thinking ahead and trying to modernize, but their approach has baffled their own people. There’s politics, bureaucracy, and money involved, not a good combination.

CoreFLS is a good example. A couple of people decided to grab a little Java, a little Oracle, and hire some consultants build a revised system without user input, appropriate oversight, or an understanding of what VistA already had available. The results speak for themselves. It was a spectacular failure at the wrong time. They got called out by people questioning everything about VA’s software projects. There’s a big shadow cast on everything they’ve done, and that’s unfortunate. HealtheVet was reviewed by Carnegie Mellon at the VA’s request and they rattled off a lot of problems. It was good plan for the VA to improve around. The message was to look at the methods and means before moving ahead, and they’re doing that.

We have tremendous respect for what the VA has done. We’re continually impressed with the thoughtfulness of design, its scalability, and how they rolled it out. We also respect the environment they have to work in. Still, we recognize that the VA is not a software shop. They get called out by Congress for developing software and there’s a lot of pressure on those guys. We think we can help with their effort.

The VA is fascinated by what we’re doing. They don’t know where this is going, but they want to stay involved. The VA is the author and the source of VistA. In the future, we think they will be a contributor to the product, but not the only one. That’s the endgame of open source, to have a broad audience contributing to an ever-improving application

These systems can save lives. It’s a shame that Charity Hospital in New Orleans doesn’t have CPOE. It’s available and patients are dying because they don’t have it. A product like OpenVista could have a dramatic impact on their quality of care.

You’ve got a big-name board chair in Ken Kizer. What attracted him to Medsphere?

Ken is one of my personal heroes and mentors. As we learned about VistA and the VA, it became apparent that he was the visionary and thought leader behind their intelligent use of healthcare IT as the enabler of massive quality improvement initiatives. Ken led the largest reorganization of any federal agency in the history of the US. He got everyone behind his vision and implemented an EHR in the largest healthcare organization in the country in four years. We called him out of the blue, two kids with laptops and a theme song out in California, and told him we’d love to get him involved. He’d left the VA by then and was a believer in IT as an enabler of quality. He wanted to get involved with us and we found a way for him to contribute.

Frank Pecaitis had a following at QuadraMed. What will be doing?

Spending time with his family – I saw your comments (laughs.) He’s phenomenal. We were so pleased that he was interested in our opportunity and our vision. It didn’t hurt that we’re 15 minutes from his house, either (laughs.) Frank has the opportunity to be one of the most influential hires in our company’s history, to build a sales and marketing organization to make this thing happen. He’s on the phone constantly, talking with people all over the industry to see what he’s doing and jazzed about the Medsphere business model.

Marlene McCurdy is another industry all-star who will play a key role in our deployment, training, and support organizations as we enter the "go to market" phase. We’re 100% impressed with both of them.

Does bringing on well-known industry talent signal an increased visibility of Medsphere?

You didn’t hear much from us in the year and a half when we were implementing Oklahoma. We kept quiet until we had something to show other than plans. We had planned to come out at the same time as VistA Office hype begins to build and support those media efforts through the fall with the Midland implementation. As part of the natural progression and staging of our company, we feel comfortable talking more and I’m happy to be talking to you.

It’s funny to be speaking to someone who writes a blog, but we recognize that it is a highly influential medium. We know CEOs and other influential people read HIStalk and we’d like our story to get out there. We also look forward to HIMSS this year. It is the big show and we clearly want to be building momentum towards that. As appropriate and as we have something to talk about, we hope to introduce the new paradigm around our business model. We’re getting interesting calls from RHIOs and medical associations that want to make big things happen. Surf’s up and we plan to rude that wave, too.

There’s an interesting lag in adoption. The organizations who can afford current healthcare IT offerings have already caught the first few waves. However, a lot of healthcare shops are sitting on their boards and waiting for a wave they can ride. We believe that Medsphere and OpenVista can serve as a diffusion catalyst in the widespread adoption of healthcare IT. Once there’s a solid EHR infrastructure in place, with the appropriate information being collected and exchanged, things will get really interesting. Now you’re talking about metadata analysis, population data, interoperability, genomics. That’s the transition from information management to knowledge management. That’s the future state.

Will the high cost of add-on specialty platforms like propriety decision support systems limit your ability to provide a less expensive product?

The wave of open source started around hardware and infrastructure. Then the operating system. Then the database. Then middleware. Now we’re pushing up the food chain to applications. We’ll see over time that open source will provide an alternative to high-end systems, such as clinical decision support. Somebody will write that “good enough” solution that will provide choice. There will be options to allow us to stay competitive.

What happens in the next 1-2 years with Medsphere?

Our main focus is to implement the next five reference sites. Then, continue to define a clear roadmap for the product as we commercialize it by adding features, functions, and performance enhancements to meet market demands. We hope that clearly defining our product roadmap, implementing those next 3-5 reference sites, and executing on our business model will have a dramatic impact on widespread adoption of OpenVista.

Any final comments?

I’d like to give props to all the Underground Railroad pioneers who believed and persevered. Any success that Medsphere enjoys is a result of their original efforts to build an application that has succeeded for over two decades. None of what we do would be possible without our OpenVista Team, those incredible people who share a common passion and just flat out perform. Finally, we honor the VA and their success in utilizing information technology to provide the best healthcare to our nation’s veterans.