Episode 28 : Harnessing Technology for Patient Safety:

SFTS-Ralph Johnson.mp3: Audio automatically transcribed by Sonix

SFTS-Ralph Johnson.mp3: this mp3 audio file was automatically transcribed by Sonix with the best speech-to-text algorithms. This transcript may contain errors.

Intro/Outro:
Welcome to Smart from the Start, presented by Care.ai, the Smart Care Facility, platform company and leader in AI and ambient intelligence for healthcare. Join Steve Lieber, former CEO of HIMSS, as he interviews the brightest minds in the health provider space on truly transformative technologies that are modernizing healthcare.

Steve Lieber:
Hello and welcome to Smart from the Start. I am your host, Steve Lieber, and it is my pleasure to bring to you a series of conversations with some of the sharpest minds in information technology. We'll discuss the smart directions healthcare companies and providers are pursuing to create smart care teams. Today I'm joined by Ralph Johnson. Ralph currently serves as the Vice President of Informatics and Technology for the Leapfrog Group. At Leapfrog, Ralph is responsible for the organization's technology needs, as well as managing public policy positions for Leapfrog. He joined the group after a 35-year career as a CIO at, in Maine at various hospitals. And as a hospital CEO, Ralph led many hospital and physician practice EMR implementations. He's also led major initiatives to improve patient safety. His career also includes past responsibilities for hospital quality programs, health information management, and risk management. Ralph is past president of the New England HIMSS chapter, which is where we first made our connections some number of years ago, and continues to volunteer for HIMSS at the local and national levels, currently serving on the HIMSS Public Policy Committee. So welcome, Ralph. Good to see you again.

Ralph Johnson:
Hey Steve, it's really great to reconnect with you. I'm looking forward to this discussion today.

Steve Lieber:
I am as well. As I mentioned in the introduction from my time at HIMSS, Ralph was volunteer and, obviously from the introduction, continues as a very active volunteer with HIMSS organization. And, you know, let's start out there. And in terms of talking about public policy, especially as it relates to AI. There's a lot of conversation, certainly in Washington and probably in state capitals all over the country. I was reading recently that Nikki Tripathy is working on an HHS policy for AI Congress, making, you know, all of the execs trot up before they're on the Hill and testify and say, you know, everybody's really focused on this. And so HIMSS has historically been very active in the public policy arena. My longtime good friend Tom Leary, I think, still runs the government affairs program there. Next time you see Tom, give him my best. So what's HIMSS talking about as it relates to AI and public policy? What should we as people in the field be thinking about that's likely to be coming down the path?

Ralph Johnson:
Yeah, it's a good question because I think the Public Policy Committee is just starting to embrace this as something they've got to really take some action on. Just in February, they formed a workgroup to start formulating a plan, and they've already come out with a first draft of principles. They're really focused on clinical applications, although some on the workgroup are questioning whether it should be expanded to include administrative and non-clinical workflows. But coming from the Leapfrog Group stage, personally, I'm really pleased to see an emphasis on patient care and patient safety. I see the workgroup recognizes a need for requirements to monitor introduction of bias and model drift post-deployment. You know, how do you monitor and stay on top to make sure that you don't perpetuate the biases that currently exist in healthcare?

Steve Lieber:
Clearly kind of see it going. You mentioned principles and all. And so bias and ensuring that the AI engine doesn't introduce particular biases and that sort of thing. Elaborate a little bit more on that, because I'd be willing to assume here, and you can help us, that in the field, people are already utilizing tools in certain ways. And so, in a sense, practice is ahead of policy here, which are oftentimes, as we know, is the case. So a little more elaboration on that.

Ralph Johnson:
Sure. So at Leapfrog, one of our major areas of study is maternity care. And we know that for example, black women have a 40% higher chance of having a serious error happen during prenatal and postnatal and childbirth. That's an inherent bias that's been built into our system. How do you use AI to remove that, so that you don't see that part of the AI actually showing up in the system?

Steve Lieber:
So on something like that help me understand. I want to kind of peek under the curtain at Leapfrog a little bit. How is the organization thinking about how it's going to bring this into its metrics? And so, kind of help us understand, okay, got you, got where you're going on public policy here. Okay. Now how's Leapfrog thinking they might translate that?

Ralph Johnson:
So there's a couple of areas that we're looking at that we hope it'll be helpful. But you've got like you said, peek under the covers. The way Leapfrog works, we're actually a pretty small organization with a lot of support. And part of that support is that we have expert panels on a number of different topics. And so it's really will turn to those expert panels to help us understand where AI is going to fold into the models.

Steve Lieber:
Excellent. Good. So as I think about Leapfrog, I mean, certainly quality is probably the thing I remember most. I also remember back in the day at HIMSS Analytics, we did some correlation studies between Leapfrog metrics and MREM stages and all, and found that there was a correlation there. What are you seeing in terms of what's going on more broadly, not just AI, but more broadly out in the field around quality and technology? And, you know, what are you folks talking about in terms of where Leapfrog is headed beyond AI specifically?

Ralph Johnson:
I'm glad you asked that, because that's actually a new emphasis that we have here. We have had a long-standing interest in computerized physician order entry. But that's about as far as we've gone, actually, on the electronic side of things with the patient safety and preventing medical errors. And we're really starting to study now how we can, especially with the explosion of AI coming out, how we can start embracing more of the electronic medical record technology into preventing medical errors for many years. And I hate to say it, Steve, going to HIMSS, I would see, what are you doing? What, how are you using decision support in your systems? And it was primarily to increase revenue or increase physician or provider productivity, which equals revenue. And I was always discouraged at the emphasis on looking at revenue as the end-all and be-all, when really the patient should be at the center of that. And I was pleasantly surprised, I think in Orlando, recently at HIMSS 24, I saw much more evidence of companies presenting technology just for the purpose of improving patient safety and patient care.

Steve Lieber:
That's great to hear. I didn't make it to HIMSS this year, but I have been reading more and sort of picking up on what you just identified there in a trend towards, you know, we've talked patient is at the center of care, but we didn't really practice it that way. And so I'm encouraged to hear you say that at the HIMSS conference, you're seeing that sort of trend. Ralph, just taking a short digression here, Ralph sits on a national advisory panel that Care.ai has sponsored for the building of a smart hospital maturity model. And in the feedback we're getting from that group so far, it is consistently: are you focused on how smart technology is impacting the patient's experience? Make sure the model includes a component that measures patient engagement and patient activation and this sort of thing. And you're right. You know, over the years we've seen a different focus on some of these technologies and all. So how does patient experience and such, factor into Leapfrog. Is there a metric that exists already or is that also a new direction?

Ralph Johnson:
No. You know, part of our model for years has been to incorporate the Hcap scores from CMS into our methodology and how we do the scoring. My boss, though, you know, on picking up on what you said, my boss has a really good saying that there is no good price for bad care. It's hard to gauge what's value, but if you have to look at that clinical quality and preventing medical errors right up there at the top.

Steve Lieber:
Excellent. Good. So let's stay with the HIMSS conference this year, in terms of: what are other takeaways? What can you share with the audience here of things that you are seeing and hearing in terms of various activities? I know, I think they still do or used to had a day focused on public policy, starting with public policy breakfast and government speakers and that sort of thing. So, you know, just give us some insights and some takeaways of what you saw this year.

Ralph Johnson:
First thing I noted was that especially among my former CIO colleagues, there is a strong recognition that they can't sit back and watch how this unfolds. Everybody's got to jump in and embrace this. How you do that, you know, how do you do it without really opening up new risks? But finding opportunities is the challenge that a lot of them are facing. So it'd be good to see how that plays out. I think the second big thing I noted, and I actually shared this with our CEO yesterday, is that a lot of organizations are focused on the governance of AI and setting up a governance model right up front and making sure that it's all-encompassing with the right stakeholders, because you've got to have this. And what that plays into is another hat that I wore previously that you noted is risk management. I actually sat with some people who are in the risk arena and they're concerned about how do you, not only govern this, but how do you monitor and track it? You know, they're used to things like the security logs and the EMR to go back and defend when something bad happened in the hospital. What if something bad happens because of a poorly written AI model? How are you going to mitigate that risk? And make sure that it's trackable. I could foresee at some time you could use AI to actually trigger nurse protocol orders, and one of those orders was inappropriate. How do you go back and find out where does the responsibility lie in that bad order?

Steve Lieber:
Yeah, we're certainly going to go through a period of significant transition from no machine interaction to, okay, what is the right balance of when human intervention is required versus the points at which machine can drive some thinking at all? This is, I'm going to carry it out a little farther here. Is anybody talking about, from a risk standpoint, the risk of not doing AI? In other words, is there an argument out there, not yet necessarily, but thinking ahead that not having a technology that would have predicted the direction of patient is going or something and not utilizing that or whatever? Am I going down a bad path here in terms of risk management or something here? But I just, you know, not using technology is a question as well as using technology.

Ralph Johnson:
Actually, I think you're looking into the crystal ball a little bit further out than what I was hearing, but I can see it steamrolling right to that point quickly, Steve. I think that's recognized in what I said earlier about the fact that to a person, the CIOs I spoke with recognize that they have to embrace this. They can't let it. They can't sit back and watch and see how it unfolds.

Steve Lieber:
Yeah, in my understanding, in terms of some of the readings I've done and chatting with some folks, your two major EMR vendors also are heading down this path in terms of embedding it into their products. I remember reading back last summer that Epic had said, Well, we're going to start out with administrative tasks and we're actually going to embed. So did you get a sense of that in terms of the technologies? And more and more are, and I'm trying to go beyond the hype part, because I would be willing to bet that almost every booth at HIMSS this year had AI on their backdrop, but that aside, in reality, is industry moving along at a rapid pace and bringing this day that you say I may be looking at in the crystal ball out into the future, but it may come faster than we think.

Ralph Johnson:
Yeah, absolutely. I spoke with people at both of those major vendors, had good conversations, and they're clearly down this track of embracing AI and how are we going to incorporate it into the EMR and staying ahead of it rather than trying to. It's got to be a competitive advantage to them.

Steve Lieber:
Sure. And I would expect that when we look at data analytics tools, I mean, that's going to be another place, you mentioned physician computerized physician order entry, there are a lot of places where the opportunity for machine learning is obvious. Now, how it's done, how it's governed, the principles upon which it's based, and that sort of thing are all there. But to me, this has the appearance of one of the more major transformational developments that we've seen, and kind of take it back to the early 2000 time period, we both remember in terms of the digitization of health records getting away from paper. And, you know, that was a transformational moment in time. I'm kind of sensing it here. Do you sense that as well?

Ralph Johnson:
Oh, absolutely. But it's funny you bring that up because as you pointed out, we've both been around long enough to know that AI is this year's buzzword, right? It's actually been something we've been utilizing for over ten years. The difference is the computing power is exponentially stronger than it was. CPOE is a great example. We were deploying that a long time ago; alerting providers when they were prescribing or ordering a medication that there's a potential adverse event, you know, and trying to prevent those errors. Leapfrog recognized the importance of that technology a long time ago. We actually have developed a tool that ONC recognizes as a great tool to test your CPOE system or its safety aspects. And we've had that in place for a long time. You know, almost 50% of preventable medical errors that happen in hospitals are medication-related. Now, if we can apply AI to that, again, with that exponentially faster computing power we have, think of how many medication errors we could prevent and really reduce that number.

Steve Lieber:
You know, I really like the way you gave that somewhat look-back perspective and tied it into today, because we do, I think at times get a little anxious. We'll use that word in terms of where this is going, but in reality, it is a journey that healthcare has been on for a decade-plus, going on two decades now in terms of utilizing technology to inform, alert, even give direction in terms of what we should do and how we should react to patient conditions and that sort of thing. And certainly, the expanded capabilities of what we are able to do creates a different world. But it is, I think, is, you know, paraphrasing what you said, it is continuing on a path that we've been. This is not totally new territory. Yes, some of the tools and they're certainly their capabilities are new, but this is something that we ought to be somewhat familiar with.

Ralph Johnson:
15 years ago, would we think that speech recognition would be where it is now? No. You know, it was the bane of providers. They wanted to embrace speech recognition, to get away from the transcription and everything else. But then they'd get frustrated with all the editing they'd have to do because the speech recognition was so terrible. But now here we are. Now we're looking at large language models using that and ambient listening opportunities, because we've got such better computing power and ability to tackle that.

Steve Lieber:
Absolutely. The ambient listening and ambient monitoring is a big piece of sort of the smart technology and smart care team movement that's moving us much more to capture of information and patient activity and that sort of thing. Put it into a platform and give us some analysis, and then from that action can occur both whether by machine or human interaction and all. It's all part of a big package. That's a great insight there, Ralph, I appreciate it. So to wrap up here, we're at the end of our time, sort of you come from the world that our listeners are a part of: CIOs and CMIOs and Chief Digital Officers and such. So what's your takeaway from where you sit today in terms of what you would say to your former colleagues out there in the field as the big piece that you want to leave with them?

Ralph Johnson:
I would say it's heat, my earlier remarks about having good governance as you deploy it. Make sure you engage all the right stakeholders in that process, you know, the physicians, mid-levels, nurses, even radiology and lab representatives. I mean, there's AI opportunities there as well. And make sure you include the quality leaders as well as if, you know, larger organizations have Ephesus; engage those Ephesus in the process as well.

Steve Lieber:
Ralph, this has been a great catch-up. I really appreciate it; the insights you have coming out of the provider world, as well as the work you're doing at Leapfrog. Really do appreciate the opportunity you've given us today to have this conversation.

Ralph Johnson:
Well, thank you, Steve. I really appreciate catching up again. It's a really important topic, and I was happy to give you Leapfrog's perspective on how we embrace this.

Steve Lieber:
Excellent. And to our listeners, thank you for joining us. I hope this series helps you make healthcare smarter and move at the speed of tech. Be well.

Intro/Outro:
Thanks for listening to Smart from the Start. For best practices in AI, in ambient intelligence, and ways your organization can help lead the era of smart hospitals, visit us at SmartHospital.ai, and for information on the leading Smart Care Facility platform, visit Care.ai.

Sonix is the world’s most advanced automated transcription, translation, and subtitling platform. Fast, accurate, and affordable.

Automatically convert your mp3 files to text (txt file), Microsoft Word (docx file), and SubRip Subtitle (srt file) in minutes.

Sonix has many features that you'd love including upload many different filetypes, automated subtitles, automated translation, world-class support, and easily transcribe your Zoom meetings. Try Sonix for free today.