Good morning and welcome to Health Care PittStop. This is Alan Pitt, and I’m welcoming Dr. Rick Vaughn. He is the chief medical information officer for SSM, a large health care system in the Midwest. Today I want to talk about the impact of technology on both providers and patients—the good and the bad. Rick, welcome to the show and thank you very much for being here.
RV: Thank you, Alan. Great to be here and looking forward to the conversation.
AP: Rick, you and I are about the same age. I remember practicing with large three-ring binders where I would have yellow sheets of paper in triplicate, and that would be my medical chart. And I know that electronic medical records were introduced as part of patient safety, but really also as a medical billing device. How has the electronic medical record changed the practice of healthcare?
RV: I remember those days as well. And if we recall some of those frustrations, it can also help lay out where we are and where we need to go.
I recall, especially in the office, how I would transcribe using only my notes and maybe some things that got penned in and would be loosely organized, from things that got faxed to me or sent in the mail. All of it would be delayed, especially if it was being mailed to me. If I prescribed anything, it would take several days for it to come back to us. So there was definitely a frustration on the part of patients that the file wasn’t complete, and frustration on my part as well as the staff’s.
The earliest promise of the electronic medical record was that it was going to connect everybody so we’d have a contemporaneous record that we were all sharing. And we’d have things that were more up to date, and we would not have to order things and redo things because the information would be right there. Have we delivered on that promise? I think at least partially we have. There’s a lot of work still to be done, but if you look at the incremental progress we’ve made over the last couple of years, especially in the inter-operability space, it’s promising.
It’s easy to get frustrated. It’s easy to be negative. But at the same time, we’re looking at where we have made successes. We have delivered a chart that is much more comprehensive and certainly one that gives a much better story about the patient than what we had when we were all working on isolated charts in our own offices.
On the hospital side, I remember having to assemble the story. We had to go find the cardex to see what medications were delivered at what time. We had to go find a bedside chart to find the vital signs. At one point in my career, I actually had to go to the lab to get the lab results when I was in the hospital, to assemble those. So if you look at, again, the basic promise of delivering a unified patient record even within the walls of the hospital, there are a lot of advantages to having that unified record available to you today.
If you went to physicians, they would still express a lot of frustration about EHR and rightfully so. But according to the numbers out there, something like 90 percent of physicians when reviewed say they would not want to go back to paper. It’s that connectivity and having that information in one location, knowing reliably if you opened that chart up you could find what you need.
But overall I think we’re definitely an order of magnitude better off, even with the systems we have today, than we were in the past.
AP: I definitely hear about being able to find that information, but at the same time, I hear a lot of my colleagues grumble because they have to chart into that EMR late at night. They can’t get their work done because they’re trying to figure out where to put that information into the chart. And it takes them longer to get that done. Do you see anything on the horizon that will improve the physicians’ workflow in terms of their ability to get that data into the EMR?
RV: We’ve tried. As I said before, I think these complaints are justified. So, it’s taking longer to get things into the chart. There are a couple reasons for that. One: I don’t have the stats behind me to back this up, but there is sort of a bell curve, if you will, of what physicians’ productivity looks like in the EHR across the specialty. Some docs can be highly efficient, get it done, and are not spending all their time in the EHR after hours. There are other docs that clearly have some challenges in those areas. So it’s not a simple uniform approach, nor is it a simple uniform problem. That’s the first thing to recognize.
The problem, again, is that technologies can only go down a certain number of ways. How do you get your ideas into a record? You either hand write them; you can speak them and rely on voice recognition; or you can have somebody else do the work for you. Right? So in terms of technologies there’s really only three ways until we get to being able to read your mind and get the information from your mind into the chart.
We’ve explored all those. I think voice recognition (VR) has kind of reached its peak. The accuracy of VR is very good compared to where it was in 1988 when I was using it. However, it’s still not perfect and probably never will be. So there’s a frustration even when we’re using voice recognition that you’re still doing a lot of work and correction and that’s inefficient.
Some docs would like to say, Hey, I don’t want to do the correction; I’m just going to say it’s the fault of the voice recognition system. Of course there are legal folks who will tell you that’s not an acceptable way to reduce your risk. You’re really responsible for whatever you put into the chart via whatever mechanism. When you sign it you’re agreeing that it’s accurate and up to date.
The other part of this is the challenge that we had with all the regulation, and again, rightfully so. Docs are here to take care of patients. That’s their passion. And that’s what we need to enable with our IT systems. But the IT team is not responsible for coding and is not responsible for federal programs for quality reporting. It’s not responsible for creating counter intuitive approaches to collecting information that then results in how you get paid—as we start talking about how you do HCC coding and how you do documentation that supports the level of complexity and describes the level of acuity of your patient. These systems had to be enabled and you could not have sold or designed an EHR in the last few decades that didn’t support that approach.
So I absolutely agree with you, Alan, that one of the tragedies as we look back on this period of billions and billions of dollars that were invested in creating a model of electronic health records [was that it] had to be based on making sure that the physician got paid for their work. And to make sure a physician gets paid for the work, it had to adhere to and be built by the rules that the federal government, state government, and insurers put in place that required excessive documentation—required physicians to do things that have nothing to do with taking care of the patient. But it was inevitable in the design because of the rules that existed in the marketplace.
AP: So you’re really calling out a problem, which is billing and coding, and the EMR is just the result of billing and coding. I’ve often said that, given this capitalistic influx on top of health care, the billing system is actually the root cause more than anything else. And there’s a lot of debate about single payer. There are some in health care that actually think that may make it better because you may not have as many rule sets on top of the clinical care of those patients.
My daughter is a fourth-year medical student and she’ll be graduating soon. Do you think in 10-15 years she’ll be entering in data similar to doctors today?
RV: I hope not. I’ve been a fan of the Gartner generations of EHR model that they put out a couple of decades ago. And one of their generations starts off with the EHR being a document. But at a certain point, it becomes a colleague. And then at a certain point, it becomes a mentor. Lawrence Weed said long ago that the human mind cannot hold all the facts in medicine. It would be similar to asking somebody to memorize all the airplane schedules across the entire United States or the world and being able to recall those. So EHR must evolve into an intelligence agent that’s going to work hand-in-hand with the provider because the true value of this relationship will be an AI on the EHR side.
Artificial intelligence processes millions of lines of data. Right? I can’t do that. And it provides to me a set of suggestions: 1) for things I’m not thinking about; and 2) things that I can’t find easily. So I’m really looking forward to the day that your daughter comes in and the EHR says, Hey, I’ve reviewed everything in this chart, I think this is the problem list, and it’s based on a set of heuristics that makes sense, that are exposed and transparent. And do you agree that this is the set of problems this patient has based on all the available information from every provider, every test, every consult, every EKG and X-ray that the patient has seen?
There’s a lot to be excited about, and if you give me that decade—or perhaps even less than that decade since we seem to be on the uptake of an exponential explosion in AI—I think there’s really something exciting about helping the provider so that they are looking forward to opening up that chart and saying, What can you tell me that’s going to help me take better care of my patient?
AP: I’ve often said that Dr. Watson as a patient should be your first consultation and your doctor should be your second. Artificial intelligence has the opportunity to talk to patients and definitely gather data. So once the patient finally meets with another human being, you have a very complete chart.
Let’s flip the lens just a little bit. I’ve written a blog that said your doctor’s not ignoring you, they’re just playing a video game, speaking to how a lot of physicians tend to look at the screen rather than look at the patient now. Do you have any advice to patients who are sitting in the room with the doctor, and the doctor is paying more attention to getting the data in than the encounter?
RV: I’m a huge fan of patient advocacy, and there are leading organizations around the country now that are learning from getting a patient advocate on a board or into their committees so that physicians and IT teams are really forced to deal with the patient’s thought on how this all ought to be set up.
There’s a guy that I follow on Twitter, a physician named Jimmie Vanagon, who has, very simply, with existing technology, tried to conquer this problem by saying, Look, I’m just going to put a big screen on the wall that we’re both going to look at. There’s so much data now and there’s so much data-driven discussion that you and I need to have shared decision-making. Why don’t we both, side-by-side, look at this screen at the same time? Right? So if it’s your child, I’m showing the growth chart that we can both look at, showing existing technology. He’s a big advocate of it. Check him out on Twitter. And he’s written a lot of stuff around this and it’s been very successful, I think, in advocating for that approach. And the beautiful thing of course is, you can go on to Amazon—or if you’re actually old school you can go down to Best Buy— and buy a big screen TV and stick it on the wall. It’s not a huge investment.
As we were talking earlier with AI, that’s what I’d like to be able to do: Walk into the room with the patient sitting right there and say, Hi Siri, let’s talk about what’s going on with the patient; let’s review the chart. And then, as you were saying earlier, see what other things we can do to help navigate and document. Having that intelligent agent like Siri in the room is going to be very valuable—to have that screen on the wall, navigate by voice through the chart with a lot of anticipatory work done by the AI in the backside and anticipate what I’m going to want to see, have it loaded and ready to go.
AP: I love that shared experience.
RV: Again, with existing technology, not too hard to do. We’re not talking tens of thousands of dollars to put it in place.
AP: As the chief medical information officer, I’m sure you’re responsible for the road map of your system. Are there two or three things you think are really pressing that you need to address? That you’d like to help industry with?
RV: The first one is, again, the purpose of this conversation we’re having. Although I don’t agree that EHR is the sole cause of burnout, it’s certainly a contributor. And we have to figure out a couple of things, some of which are in our hands and some of which are in the hands of some of our EHR vendors, in terms of how we simplify the interface.
How do we relieve physicians of the extraordinary cognitive energy they have to waste essentially on navigating these systems that aren’t anticipating their needs? How do we move away from a training model that says physicians, to become efficient, must customize? In other words, we are telling physicians in each of these EHRs that, Hey, we couldn’t figure out the design for you, so you have to spend an additional hundred hours of your personal time to customize your order sets and customize your office visit and your documentation template. That’s really not a sustainable model, and it’s very wasteful if you consider what a physician’s time is worth in the number of hours. So we need some help there. The topic of this podcast? Absolutely on point.
I think the second part of this that we’re really going to need some help with is interoperability. We still need to keep the pressure on. As we take on more risk as health care organizations, we’ve got to figure out how we make it seamless to incorporate information from an outside source. That’s going to take, as it has in the past history of this, a kind of public-private consortium that’s going to continue to develop strong standards based on a clinical care document (CCD) approach; figure out the network of networks.
My goal here is that “interoperability” becomes an unknown word. People don’t even think about the inter-operability. It’s just there; it’s not a topic any longer. It’s anticipated and it’s actually delivered. When I open up the EHR, everything I need to know about this patient is there reliably. I’m not worried about gaps or missing information. And more importantly, it doesn’t require me as a physician to do the reconciliation, which is extraordinarily time consuming and, again, a waste of cognitive energy.
Brett James of Intermountain Healthcare said many years ago the most expensive resource we have is the physician’s trained mind, and if we used it for what we needed it for, we wouldn’t have a physician shortage. We’d have plenty of physicians. But we’re asking them to do a lot of silly things that don’t make sense. And that’s where we need the vendors to stand up and make this more efficient.
AP: I know your EMR is EPIC. Do you expect EPIC to fix this for you, or do you expect a third party to come in and help you with the interface for the physicians, the inter-operability? How do you see that playing out?
RV: I think parts of it have to be together. Suddenly you’re paying extraordinary amounts of money for a proprietary system that is the daily workflow engine for the provider. And I don’t think third-party tools even with SMArt on FHIR and FHIR (fast health interoperability resources allowing relatively easy integrations) APIs are going to solve this problem in the next 5 to 10 years. If we tell physicians they have to wait five years to have better usability, they’re not going to be very happy about that. If we feel like we, as clinic leaders, have to address the frustrations that are leading to burn out and see some progress on the EHR side, I think there’s a — even moral —responsibility on the EHR vendors to figure out how we improve our interfaces. How do we reduce the amount of time people are spending in EPIC? How do we make this more of an enjoyable exercise for providers?
That being said, the vendors can’t solve the problem, as we said before, around what’s required for building, coding, regulatory, quality reporting—all these other pieces that have complicated not only the provider’s life but the vendors’ lives. I’m stuck by one of the vendors saying that they spent 60000 hours of development time to get ready for Meaningful Use version 1. That’s 60000 hours of time that could have been spent on improving [other areas.]
The third thing I want to see solved is, again, analytics. We’ve got to get better at not only understanding descriptive analytics and predictive analytics but also pre-scripted analytics. And that includes understanding our patients well enough to know, for the risk match patients that have the same disease, which ones are going to be open and willing and benefit from expensive care management services.
So I think there’s a lot of room to work on analytics, and that’s probably one of the most pressing issues across that entire descriptive / predictive / prescriptive set of analytics. That’s going to be interesting because right now that’s playing out in the marketplace between the third-party vendors like Health Catalyst and others versus the EHR vendors who have said, You know what? Maybe we were sleeping a little bit at the wheel here and maybe we need to catch up. So we’re seeing the EHR vendors really putting a lot of resources in this area. It’s going to be interesting to watch this segment of the market play out, but it’s a big need right now.
AP: I think the same can be said for telehealth. Whether that’s going to be a third-party vendor—just like you mentioned with the analytics company—or it’s going to be the EMRs. Ultimately I see the EMRs as an operating system that needs applications to help move things forward.
RV: It’s always hard to look at historical precedent, and sometimes you twist the facts to meet these precedents. But watching Microsoft in its dominance for the last 20 or 30 years in the marketplace is interesting because it kind of went both ways. It was good and bad. Obviously they bought some vendors and incorporated those functionalities into their operating systems. Their strong adoption of the Office suite into their OS was absolutely killer for them, and we’re kind of watching this to a similar extent play out in the EHR market. So which things are they going to incorporate and put in the base EHR to make that proprietary system extraordinarily strong? What are they going to incorporate? And then again—going back to what you said earlier—what will they let go and try to create just the operating system part of it to connect to a third party application? So far they have opened up a crack, but I don’t think either one of us is seeing incredible amounts of third-party stuff that’s easily off the shelf, interoperable with the base EHR stuff.
AP: It’s very difficult. We have to figure out how to iterate, how to change faster. And I don’t think that the EMR vendors have the bandwidth—the number of people—necessary to integrate things like analytics and telehealth and all of this stuff. It’ll be a very interesting three to five years in healthcare.
I really liked your point about the shared experience around data between providers and patients. Hopefully we’ll see more of that because I think that may be part of the solution moving forward.
RV: Well, it’s easy to be overly reductionist, but if you’re looking for some of the big levers, what I have to be able to do and what we haven’t been responsible for as much in the past is to change the patient’s attitude and how they approach life. I’ve got to get them to change their diet, eat better, as well as be adherent to the medical regimen that I’m going to place in front of them. So it’s going to be more emphasis on patient engagement and adoption in a shared decision-making model than we’ve ever had before. Whatever we can do in that discussion while we’re in that room, that’s going to help patients see what I see and help me communicate, which is a huge issue. I’m particularly bad at this and I’ve been told this many times by our health literacy people that I’m not allowed to create patient level documents. But I think all physicians have this great difficulty in translating to patients what is important, how to understand it, and what they should be doing. We fall back into our bad habit of using technical jargon and shortcuts and acronyms that only we understand. And the patients are very respectful. They don’t stop and say, Doc, I don’t know what you’re saying. They just nod their heads and then hope that they can figure it out later. So I think health literacy and engagement and getting people to change is absolutely critical.
AP: Rick, I want to thank you for your time today. I hope you’ll consider doing this again. Maybe we can talk more about your role in patient engagement. For those out there listening, Dr. Vaughn is a very active tweeter. I recommend following his Twitter feed as well. Rick, Thank you very much.
Thanks, Alan. Great talking to you.