HIStalk Interviews Marc Willard, CEO, Certify Data Systems

Marc Willard is founder and CEO of Certify Data Systems of San Jose, CA.

Let’s start off with a brief description of yourself and the company.

I’m from England. I’ve been here for 12+ years. I’m one of those serial entrepreneurs. I’ve been in technology for most of my career. 

Certify was founded by myself in 2004. We had a vision, very early back then, of connecting physicians with hospitals or health systems. We’ve been doing that ever since. We’re in the enterprise health information exchange market.

Who would you consider to be your main competitors?

It’s changing rapidly. I would say for sure we would see Medicity. Sometimes IT units within health systems developing their own products, but that’s not really very common any more. Maybe a company like a MobileMD as well.

The market is in two segments now — state or public HIEs and enterprise. In enterprise, there aren’t too many companies at the moment. There’s a lot in the state-based, though.

Describe how you see the market shaking out and the difference between the enterprise ones and the public ones.

The public ones are normally driven by public funds or grants. They tend to try and encompass a whole state or a whole county. Their goal is to try to create a common medical record. The challenge with the public ones is that they’re driven unfortunately by politics. I think in the past we’ve seen CHINs and RHIOS all try to do a similar sort of thing.

The enterprise market is something that I’d say in the last two or three years has become very interesting. It’s probably is the fastest-growing segment now. That is where a health system is trying to enhance its relationships and exchange data with its physician community. They protect and increase their revenues for all members involved. It’s a much more sustainable business model because it doesn’t rely on  grant funding. It tends to have a much stronger ROI.

The public organizations had a challenge getting providers to sign up. Are enterprise ones more successful, and what reasons are causing providers to either sign up or decline to?

It definitely has more success. Unfortunately, it’s politics. When you try and bring everyone together in a public HIE, everyone has a different agenda. England is the best example of how a free HIE just doesn’t work.

The reason the enterprise does work is that healthcare is local. Most of the time we’re within 20 miles or 15 miles of our healthcare systems. It’s very rare that we’re even 50 miles away. Physicians feel very compelled to help in their community. It just makes a lot of sense to receive information electronically from the hospitals who they refer with. They do not feel there’s any hidden agenda. 

I think today with Meaningful Use coming on board, that’s helped as well. With some of the things going in healthcare reform, in medical home, I think the emphasis is shifting where the physicians feel a lot more comfortable.

Your model also may have helped with that since you have the federated model, where you’re not insisting that all the demographics be pulled into a third-party system that practices can’t control, instead placing the HealthDock server inside the practice’s firewall. Are customers aware of that as an advantage and are any of your competitors following that lead?

We call it a network approach, and you’re exactly right. By not asking all the providers to centralize their patient information — they feel threatened by that — but the ability for them to control it within their environment and not only share and offer up the information they want. Some offer everything. Some, if they are split between two health systems, a little bit. It’s definitely appeased their issues. 

We are at the moment about the only vendor around that’s got this true hybrid edge server model that will go down to a one- or a two-doctor office. I mean, 75% of the physicians today are less than five docs in a practice, and unless you can bring those primary care guys in, the small practices, you don’t really have a true health information exchange. You’re not really looking up the complete medical records.

Yes, absolutely it’s definitely helped. I spent between ’04 and ‘07 90% of my time in very small physician offices. We had focused user group meetings where we’d understand their requirements, their concerns. This is the way Certify has been designed — to meet that challenge. It definitely helps an awful lot.

Do you think centralized data made it attractive for other companies to buy up most of your competitors?

Yes, I do. There’s nothing wrong with a centralized model. I just think we all just need to understand the kind of dynamics that happen within an HIE. 

For example, even with us we’re a hybrid, we will bring information into the middle if you want to run analytics on it. And yes, definitely I think there are many companies today looking at companies like mine and Medicity and Axolotl that see the value of having access to that data.

The key is to make sure that the owners of that data are happy to share it. With the ACO structures being formed and now the medical home plans, a lot of the information is able to be shared. There are many, many companies out there that see value in it.

I saw some examples of things that hospitals might choose to pull in from those connected EMRs of the practices that they’re affiliated with. What are hospitals doing with that analytic capability?

Quality measures are a great example. We have a very nice health system that’s built an ACO and really believes it’s the better kind of environment. They’re pulling information in for quality measures.

Analytics to me is broken down into two segments. One is a rules-based engine — quality measures — and then the other is population management, which is more predictive analytics. I would say the rules-based stuff today, especially in rev cycle management, is pretty popular out there.

But as health systems connect more and more and more physicians in the community and really start to see that the data from the inspection of care … when I walk into my primary care office with a cough and they can have access to that information, predictive analytics become something that is very, very real and doable. I expect in the next couple of years that will be a really nice product line for Certify in the marketplace.

How does that work when you have a hospital attached practices using a bunch of different EMRs? What’s the technology involved in trying to pull all that data from all these different systems into a single database for analytics that takes into account differences in the way their data is used, stored, and defined?

That’s a big question. You’ve got two types of feeds at the moment. You’ve got an HL7 feed, and now you’ve got some of the popular XML feeds, like the Continuity of Care Document.

We spent eight years working with EMR vendors and finding ways of allowing for easy connections and trying not to make every single connection from every single health system a custom integration. That is the kind of power what our product does. Once you can achieve those connections, then we can pull out patient summaries, scheduling information, ADT, admit /discharge / transfer information, patient summaries. 

Once we have that information on our platform, we can then dice and slice it, and in some cases maybe we’ll ship an XML file to an analytics engine, and in other cases maybe we’ll ship a couple of Continuity of Care Documents to a central repository that the health system has. Once you’re in there and connected it, it’s actually fairly easy for us to manage and pull up data.

Of course, then as you start to run analytics, you’ll get into things like a vocabulary server to make sure that a blood lab test doesn’t have five different ontologies. You need to go do mapping, and that gets a little bit trickier.

Is there any potential for a standard from ONC or NIST that will eliminate the need to dig into the data to understand everything about it before you can actually have systems talk to each other?

If everyone just jumped onto LOINC and SNOMED and ICD-10, then life would be real simple, but we know it’s not that way. I think maybe 10 years down the road possibly, but at the moment not really. You’re always going to need to have some sort of vocabulary server in there. But the IP is out there. We’ve got access to great technology to do that. It’s all very solvable.

The government licensed SNOMED for everybody.

Yes, you’re right. The problem is not everybody uses SNOMED.

So that wasn’t enough encouragement? Or do EMR vendors have no incentive to use it?

It’s not really the EMR solution at the edge. It’s the human interaction. 

The lab is the easiest example. Quest or LabCorp back in the day would use different terminologies for the same thing. Then the health system would say, LOINC is the standard, and we would have to map for LOINC. The technology already exists. It’s just getting humans to adopt it and to agree to it.

I guess we’re kind of back to the age-old problem of asking people to do more work or spend more money for someone else’s benefit.

Absolutely. Absolutely. Today I would say that most of the health systems would just like to connect with their physicians. Just for the things that you and I are talking about, I see that some health systems could be three to five years out.

But the majority of health systems today would just like to connect with their physicians. They would just like to push out a clinical summary. Just like to be able to do a query for a patient record if the patient unfortunately is in the ER. All of the analytics and everything else for them is probably two or three years down the road.

But we IT companies have to prepare for the future. The market today is in a different place than maybe we’ll see at HIMSS, but I think it’s going to get there pretty quickly. It’s going to change pretty quickly.

Do you think ONC is putting the carrot out there through the Meaningful Use requirements?

I do. I think they’ve softened it, which is good. They’ve realized it’s a carrot and a stick. I think the carrot was too small and the stick was too big, so they’ve changed it a little bit now.

A number of our health systems are doing it for Meaningful Use, but most of them are doing it because it’s the right thing to do — increase quality of care. I think the energy around forming ACOs — I think that created more enthusiasm to pull HIEs together than even Meaningful Use.

That was one of the problems with Meaningful Use. It wasn’t a huge incentive, but it got everybody’s attention and they missed the whole Affordable Care Act, where maybe they should have been putting some energy into looking at ACOs instead of chasing what wasn’t much money comparatively.

You’re right. It’s what — a $40,000 reimbursement to a physician? But if they have no EMR, they’ve got to build an EMR. 

The healthcare reform stuff – the ACOs and medical homes — that one is very interesting. You create an organization where everybody can win. If we can all focus on wellness and not illness, then suddenly we’ll win. That’s a really clean example for the physicians, for the payers, the hospitals to all get on board. 

That to me is probably one of the most exciting things that’s happening. I really hope that it stays true and it stays on its course and more and more health systems create ACOs and there’s a good balance between the payer and the health system and ultimately we’re going to solve it.

Companies like Certify will end up empowering that network. Just be the veins underneath, where the information is flowing clean, and also cherry picking information off all these quality measures and so forth. But to me, that’s the exciting times over the next couple of years I’m going to personally watch.

I don’t think I asked you the question when I asked you about the company. How many customers do you have and what are they doing with your products?

Today we have, I believe, just over 70 health systems that have taken our products on board. All of them are health systems. They’re using it for exchanging clinical data in their communities. Some of them are using it to build out ACOs. But everybody’s marching down the same path. We’ve seen tremendous growth in the last two years. I mean, it’s just been phenomenal.

You have a relationship with Cerner that I don’t really understand. How does that work?

Every small company either needs to raise a fair amount of capital or they need to find a very good strategic partner or do both. We decided back in ’09 that wouldn’t it be great if we could sign up a strategic partner that could just introduce us to a large client base? We met with Cerner and our visions were aligned, and now Cerner has a relationship with Certify where they sell our products and services into their client base.

It’s been a great relationship. It still is a very good relationship. Certify now has a direct sales force and marketing team that will actually go out and sell to the rest of the world, which is the Epic, Meditech, McKesson, that kind of stuff. Most people think that we’re a Cerner company and we’re not. We just decided — and I think it was very clever for us to do it — to use Cerner as a channel to get it out to the market.

Do you have a way to share data other than just in one direction, so if you have a bunch of practices and hospital or two all connected, can any of them update things like allergies and insurance information and share that?

They definitely could. But the way our platform is designed is health systems can connect to health systems, physicians to physicians. You can have a healthcare community all aggregating up. They can all share information around. It depends on how transparent they want to be.

We have some scenarios where the health system wants the ADT data in from the practice to populate their own systems. Other health systems won’t,  and vice versa. We have controls. We have consent and data controls everywhere, but basically it’s, “OK, how comfortable are you with sharing information?” and setting the product to the conditions that you feel comfortable with. But ultimately, they could share everything with anybody. Obviously all according to HIPAA and it’s all encrypted — I don’t know want to make it like it’s a Yahoo Mail program.

We have the apparently declining RHIO model, the enterprise HIE, and some providers connecting to each other via their EMR vendor’s closed network. How do you see that playing out for the patient’s benefit in five years?

As I mentioned at the very beginning, healthcare is definitely local. I think it would be absolutely awesome for a patient to travel within their county or its state and have peace of mind that if something happened, duplicate tests won’t be performed, they’ve got basic information about who they are and what’s happened to them. I think personally if we get there in the next five years, then we’ve already created something very powerful.

It’s ultimately all about patient care and trying to reduce the cost around it. With healthcare being incredibly expensive, I think the faster we can there, then ultimately the better it’s going to be.

To do that, we also have to make sure that all of us vendors play well together. I’m a big advocate of that. We can’t create these silos. We all have to work well together. I think things like these IHE standards are very important. I think ONC’s driving stuff is very important. But I also think the healthcare vendors need to make sure they perform their part as well.

Any concluding thoughts?

We’ve spent a number of years flying underneath the radar screens and decided last year that we’re not going to do that any more. I think what you guys do is very exciting as well, giving a lot of people a voice. I appreciate your taking the time to get to know us.