David Lareau is CEO of Medicomp Systems of Chantilly, VA.
Tell me about yourself and the company.
Medicomp provides a clinical engine and tools to use it that provide physicians, nurses, and other clinicians what they need at the point of care so they can do their jobs. We give them what they need when they need it, stay out of their way, and let them focus on the patient. I discovered Medicomp in 1992, joined the company in 1995, and have been CEO for about 10 years. We are continuing to build our content and tools to support point-of-care use for better patient care.
How would you characterize the EHR industry and how it has changed?
Over the last few years, I would say the last 10 years, the focus has been on getting the money. Getting the $30 billion to $40 billion that the government gave out to promote and make electronic records ubiquitous in the industry. That’s been done, but I don’t know that it has moved the needle at all in providing better patient care.
Now the industry is being forced to turn away from the number of transactions processed or encounters documented to, how well are we taking care of these patients? How do we report that? How do we measure it? How do we make it happen? How do we prove that we did it? It’s a major change to try to make these systems usable for clinicians at the point of care.
Are EHR vendors committed and capable of making usability what users want it to be?
I’m not sure they are. As part of my evidence for that, people seem to be thinking they can rely on artificial intelligence, NLP, and machine learning to solve the problem of usability. Or the use of scribes. That indicates to me that they recognize that they can’t do what it takes to make these systems, the way they’re currently constructed, usable for physicians at the point of care to meet all the requirements that they need to. They need to be efficient, effective, they need to meet these quality metrics, and they need to do it without getting in the physician’s way or slowing them down.
They seem to be turning to, “OK, ambient AI. Say anything you want and the machine will figure it out.” The problem with that is that the machines are taught by programmers, not necessarily by clinicians. So I don’t see where the ball has been moved toward clinical usability much at all in the last 10 years.
Is it fair or unfair to say that EHRs cause burnout?
It’s fair to say that EHRs cause burnout, because EHRs, as currently constructed and implemented, weren’t designed with the patient or provider in mind. They were designed to maximize reimbursement and track transactions. Clinicians they have this reputation as being difficult to work with because they are the most highly trained knowledge users in the world and the systems they are using actually dumb them down. They get no benefit from it. They pour stuff in there and they get nothing out of it that helps them, because they already know what to do. Just get out of my way and let me do it.
What value can be added to a vendor’s EHR to make it more useful and satisfying to clinicians without having the EHR vendor themselves making changes?
Every EHR should have a problem list. If there’s something in the problem list, you ought to be able to just click on that problem and see everything in the patient’s chart that’s relevant for that problem. For diabetes, you ought to be able to see on one screen the lab results, medications, symptoms, relevant family history, relevant past procedures, et cetera.
We provide a huge clinical engine that has about 120,000,000 links between problems and the other clinical concepts related to that problem. That helps people get clean data in, use it, and see it at the point of care, rather than having to go to six different places in the EHR to look at labs, meds, and procedures. It needs to be pulled together. Physicians, because of their training and experience, already know what they need to see. They’re highly trained. They know it. Just give it to them.
We’ve been working with clinicians for 41 years to say, if you’re thinking about diabetes, what do you need to see? If you’re thinking about chronic renal failure, or thinking about this other issue, what do you need to see? They know it. Just give it to them and they can focus on the patient in front of them at that moment without having to navigate through the EHR to find every little piece of information.
How has private equity and other forms of investment changed healthcare lately?
You had a nice article on Jonathan Bush where he said, boy, once you let them in the tent, they own the tent. There’s really no room for anybody else. They have a three- to five-year time horizon to get in there, get things lean, flip it, and get out. Typically. Typically. Now, a lot of them say they’re long-term investors.
We’ve been doing this 41 years. We’ve decided to keep it, and we will continue to keep it, private and closely held so that we can continue to focus on doing what we’re doing and not be distracted by the flavor of the month. “Why aren’t you guys doing NLP? Why aren’t you guys doing AI?” We have a form of AI. Our engine was built using doctors to say what’s relevant, but we don’t talk about it as machine learning. It was actually taught by our clinicians.
If we accepted private equity money or outside investors at this point, we would lose our focus on the long-term vision, which is providing tools to let clinicians provide better care for patients at the point of care. That’s really what we’re trying to do here. Everybody has their own idea of the latest thing they should focus on. We stick to our knitting.
Is it hard not to get wrapped up in the AI buzzword that everyone suddenly claims to be using in their old products?
It’s tough. What is competition? Competition is anything that causes people to not engage with you at this point in time.
People have been telling me for three to five years that, “In three to five years, AI and machine learning will be able to do what you do.” That’s competition for us, in that it causes somebody to say, OK, we will wait and see what happens with that. People are now getting used to the fact that even if you use artificial intelligence and machine learning that’s programmatically-based rather than clinically-based, if you put garbage into these systems, you get garbage out. What error rate are you willing to accept?
We don’t try to compete with AI or machine learning. I don’t want to sound like a troglodyte. It’s valuable in identifying associations from large populations of data, saying, “We need to do more of this. We need to do more of that. This is happening in our population.” But for an individual patient, at any point in time at the point of care, I don’t think it’s going to be ready any time soon.
What are the secrets of working with EHR vendors instead of trying to compete against them?
You have to provide something of value to them and to their users. We provide a clinical data capability that they don’t have otherwise.
One of the secrets is when they say, “We have to have certain technologies. We need Angular. We need React. You have to do Docker containers. You have to host it in the cloud. It must be able to be web-based. We don’t want to use your UI, we just want to call out to it and have it linked,” you have to make sure that whatever tools you create allow them to stay in place to do all the things that they do well. Patient registration — we don’t do that. Storing of data — we have data services that pass it back and forth.
You’ve got to be willing to not enforce your vision of what their application should be. You’ve got to make your tool customizable enough and flexible enough and you’ve got to constantly redevelop your technology so that it meets the latest requirements for integration with these systems.
With these systems, the concrete is poured. They’ve got a bridge in and the concrete is poured. They want to improve their roadway? Yes, we can put down parts of a roadway, but we can’t rebuild the bridge, and nobody wants us to. So you’ve got to be willing to be part of their implementation. And in our case, our clinical stuff becomes a core piece of what they do, but it doesn’t look like it is to their customers. It just does what it needs to do and sits there. You’ve got to make it work in their environment.
Health IT vendors are making splashy announcements about embracing Amazon Web Services, Google Cloud, or Microsoft Azure and using their tools for AI and speech recognition. How will that change your business?
We’ve already begun part of that. We already have people using AWS and Microsoft Azure to host our stuff, or their applications with our stuff in it.
We will probably be asked very shortly to provide some sort of a clinical relevancy service to some of these people who are making announcements so they can find anything they want with just one or two words. For clinicians, finding anything you want means finding the things that are relevant to that, given the fact that somebody has asthma.
Over time, probably in the next two or three years, we will probably have to split our stuff into separate consumable services, one of which we’re already doing with our HCC and Medicare Advantage service, so that people who aren’t even using our concepts, engine, or terminology can do risk adjustment reviews if you just give a problem list based on SNOMED or ICD-10.
Microservices was a big buzzword a few years ago. We’re probably going to have to be willing to work with not just vendors, but suppliers of technology to those vendors, by allowing people to consume services from our engine, but not necessarily the whole thing.
Google has expressed interest in creating an EHR search engine, but it seems like it would find “patient denies chest pain” just as readily as “patient complains of chest pain.” How important is the contextual element of the search?
There is no question that natural language processing, based on noun phrases alone, is not going to work in medicine. You need context, you need to know relevancy. Is pain relevant for somebody with asthma? Yes, chest pain. What about wheezing? Did it start suddenly or not? The more that people drill down into this, the more they realize that you really need the clinical context within which the phrase you’re looking for exists.That’s a big part of what we provide.
What do you expect to see at HIMSS20?
A lot of the vendors complain that the CIOs, CMOs, and CMIOs don’t go to HIMSS. I think that’s true more and more. If they already have a platform — Epic, Cerner, Meditech, Allscripts, some of the big guys — there’s nothing they can do about it. They’ve got it. They might go to HIMSS looking for the ancillary vendors to add on certain products. We’re still seeing those people come through.
Ten years ago, health systems would send like 15 people, and say to the 15 people, “Fan out. You check this out, you check this out, you check this out.” You’d see them meeting at breakfast and planning their day because they were going to switch vendors. They were looking for a vendor. You don’t see that much any more. Because we’re not there to replace vendors, we’re there to have our stuff in as additive value, the fact that those people aren’t there doesn’t affect us so much any more.
Going into HIMSS last year, I was thinking about downsizing the booth. But by being there and by making a stronger statement about what we do – “We fix EHRs” — people said, “Finally, somebody says what they do. No buzzwords.” We met two major new accounts and opportunities that we have license agreements with. I decided for the first time to splurge and get a booth on the main aisle, because if you’ve got a good message and you know who your target market is, it’s still worthwhile for people like us. But boy, there’s a lot of noise.
Half the team you bring is there just to figure out, “Somebody just stopped by the booth. Are they serious, or are they not?” We’ve been going to HIMSS since 1996, so we’ve gotten pretty good at that.
It’s worthwhile for us still because we’re solving a problem that everybody has, and they know they have, which is clinical usability. We’ve managed to hone our message on that in the last few years, so it’s effective for us. But if I was going in there for the first time, it would be like a deer getting caught in the headlights.
How do you see the future of the company?
We’ll continue to do what we do. We’ve been doing it since 1978, 41 years. We’re really starting now to benefit from people realizing, “We’re not just tracking transactions any more. We’ve got to manage the patients better. It takes really good data to do that. We’ve got to make it usable at the point of care. What are we going to do? What are we going to do?”
This, I think, is our time. We’ve got to stay focused on what we do. Going after outside investors, changing ownership, changing leadership would just distract us from our mission. One of my main challenges is identifying and nurturing the next generation of leadership here, because we’re going to continue to do what we’re doing. I look at the senior people at companies to see, how long have they been there? How many generations of those people have they been through in the last 10 years? If it’s more than two, that’s a bad sign for continuity. The only way we can continue to do what we’re doing is by continuing to do what we’re doing.