This Week Health
Today: Epic's AI Integration - Transforming EHR Workflows & Clinical Documentation

Subscribe to This Week Health

Share this episode

Kate Gamble and Sarah Richardson discuss Epic's initiative to integrate AI agents into its EHR system, exploring the impact on healthcare organizations. They examine key benefits including AI-powered clinical documentation through Nuance's Dragon Ambient Experience, patient self-service through MyChart, predictive analytics for early intervention, and enhanced interoperability.

Subscribe: This Week Health 

Twitter: This Week Health 

LinkedIn: This Week Health 

Donate: Alex’s Lemonade Stand: Foundation for Childhood Cancer 

Transcript

  This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

 Today in Health IT, we're discussing integrating AI agents into EPIC's EHR system. My name is Kate Gamble, and I'm Managing Editor at This Week Health, where we host a set of channels and events dedicated to transforming healthcare, one connection at a time. And I'm joined by Sarah Richardson, President of Community Development here at This Week Health.

Sarah, welcome to the show. Good morning, Kate. Let's dive right into Epic Systems initiative to integrate artificial intelligence agents into its EHR system. This move is generating a lot of buzz, especially among health IT leaders. It absolutely is, because Epic's AI integration could reshape the way clinicians and healthcare organizations interact with their EHRs.

But beyond the excitement, what does it actually mean for healthcare leaders? And we're going to break down a few key highlights, looking at why this matters and how leaders can approach these changes in a way that benefits their teams and patients. So let's start with one of the most immediate benefits, AI powered clinical documentation.

Epic has embedded Nuance's Dragon Ambient experience co pilot into its EHR, which of course automates the creation of clinical notes from patient visits. It's a game changer for reducing clinician burnout because we know that physicians spend nearly two hours on documentation for every one hour of direct patient care.

So by automating notes, AI can help free up clinicians to focus on what matters most, which is back to patient interaction without the computer being at the center of the conversation. And if you're a CIO or even CMIOs and CNIOs, you need to think about how this aligns with the existing documentation workflows.

For example, if your hospital is already using voice recognition, how will AI enhance or replace that process? And how do clinicians verify that AI generated notes are accurate so if the system's generating incorrect or misleading information, it could create liability risks. And we always have to think about trust.

As Jennifer Owen, Senior AI Programming Administrator at Cleveland Clinic, said during a recent keynote, all the statistical and clinical validation in the world doesn't matter if your end users don't trust

That's a great statement. Next, let's look at how Epic is rolling out AI assistants within the HR to help patients through MyChart, answering questions and guiding them through self care recommendations.

You have to think about how patients call health care providers with simple questions about medications, test results, even scheduling. If AI can handle, let's say, 30 to 40 percent of those inquiries, that's a huge Operational efficiency boost and something we hear about often is that whole in basket and task management component of the relationship with the patient.

And then, of course, you have to think about when AI should be handed off to a human. So if a patient has an urgent need or a nuanced concern, how do we ensure AI doesn't create friction or delays in care? A really good way to test this is by running a pilot program. So let's just say your organization's considering AI driven MyChart messaging.

Start with the non critical interactions like appointment reminders and see how patients respond. And what I would say about this initiative to Kate is that regardless of. What type of pilot you are putting out there, make sure the users know it's a pilot. So you're testing it out. It may or may not stick around.

It may or may not end up in its current version. And you may or may not use what is embedded within the EHR, ambient listening tools and other tools that are available for these types of workflows. Do you integrate effectively with Epic? And so your organization may choose to use a point solution that integrates effectively.

So really looking at all of the aspects of what this could be for your organization is going to be critical in that governance process. Yeah, that's a good point you made too about the pilots because we have heard cases where the users say after it's over wait, we didn't want that to go away. We wanted that.

So really important to communicate that from the get go. So another area where AI is making an impact is predictive analytics for patient deterioration. So EPIC's AI models can analyze real time patient data to identify early warning signs of complications. This is huge for things like sepsis detection, fall risk, and post surgical complications.

The earlier we can catch warning signs, the faster clinicians can intervene. I remember you and I did an interview probably almost 12 years ago where sepsis alerting was something I had put in place with Cerner EMR in a facility, and we then moved quickly from the ability to do the detection of certain deterioration criteria.

To alarm fatigue and how we also had to balance the alarm fatigue that came with that. So same thing with these new integrated capabilities within the EMR is understanding if it is going to give you early warning signs for early intervention, how are people being informed about what is occurring with the well being of that patient?

Oh, so glad you brought that up. We might have to revisit that one day. We'll see how different we are. We'd have to find it all the way out there in the ether somewhere. It is. But yeah, very good point there. And another thing that you have to consider is that predictive AI is only as good as the data it's trained on.

If an EHR lacks clean, comprehensive data, the predictions can be unreliable. One, if your hospital system has interoperability gaps where some departments don't share data in real time, it limits how effective your predictive analysis is going to be. So you need to ask, do we have the right data infrastructure to support predictive AI?

And if not, that's where your investment needs to start. So along those lines of data, another critical aspect is AI driven interoperability, helping Epic EHRs exchange data more effectively with other systems. If we go back to interoperability has been a gap or a pain point, which we just discussed in healthcare systems for decades.

But even today, a lot of hospitals are struggling with fragmented data across different systems. Having data governance in place and the ability to understand And I think this is a really valuable tool to be able to understand where your sources of truth are coming from, how clean they are, who's maintaining them, who has access to them.

You can have your AI governance, your data governance, and to a degree, your project governance all in one space tends to get muddy when those things come together, unless you've been very thoughtful about how the decisions are being made across the continuum in those different environments. So that would be my recommendation.

And at least if you have to keep them separate, have the same

figure out which parameters by which decision making is being done in your organization. It takes a very long time to get there and it's still a conversation we have in every single city tour dinner and every single summit that we have.

Yeah, exactly. As much as we hear about there being so many systems on Epic, there are so many systems that are not. And, we need to think about how to facilitate that more seamless data exchange. Among those different organizations. And of course, anytime we're talking about stuff like this, you've got to bring in the CISOs too, because you need to be hypervigilant about security risks.

I always involved the CISO or whatever version of CISOs we've had throughout, let's be honest, the last couple of decades, because AI's ability to extract and transfer data at scale could also expose new cybersecurity vulnerabilities if not managed properly. Please have a team that is hyper focused on any kind of data transfer, inclusive of your APIs and any other way that you are sending information across.

Different systems, because. You might retire a system and somebody forgot to turn off that interface. It might still be broadcasting data externally and needs to have the right updates that are in place. So make sure that is an element of all of these conversations as well. Yeah. And sometimes it may seem like it's a footnote in the conversation, but as you said, cybersecurity should always be at the forefront, even if it's not, the first thing we talk about.

So let's go ahead and take a step back though. Okay. If you're the CIO, you're the CISO, you're the CMIO, you're the CTO, we've covered bases for all of them. And you're listening to this. What's the bigger picture at hand?

So taking a very top level look, and AI is obviously not a tech trend, it's an operational shift. So the key takeaway is that AI is going to fundamentally change how clinicians and patients interact with EHRs. And as a health care leader, I'd be asking these three questions. Where can AI add the most value to our organization?

Is it documentation, predictive analytics, patient engagement? I'd also ask the next question of what are the security and compliance risks? How do we ensure AI aligns with HIPAA and other regulations? That's going to be security, compliance, anything that's federal regulated, etc. And let's be honest, AI is not a space that's heavily governed.

From a higher level at this point. So put those frameworks in place and there's some great resources available. And then how are we actually measuring success? Are we tracking the right metrics to see if AI adoption is actually improving outcomes? And here's the thing, if it's not, it's okay to pull back a pilot, like you and I were discussing, even if it's AI, because you have to make sure it's the right product at the right time with the right fit with all the other things we've been discussing today.

Yes, you want to be using some of these newer technologies. You also need to make sure they're being used correctly. At the same time, don't let too much bureaucracy get in the way of doing the right thing. That is a very good statement. And yeah, all those three questions are really important and have to be addressed.

Along with this, how does AI affect healthcare workforce planning? If AI handles more administrative tasks, does that shift the roles of medical scribes or administrative staff? And if so, how? It's an interesting angle, and I do believe that it does. So some jobs are going to evolve, not disappear. AI is going to help free up resources to focus on higher value work, rather than placing employees for years, we've said, putting in technology so that XYZ can practice to their highest level of either capability or licensure.

This just adds one more element and ability to do that in a very thoughtful and productive way. Agreed. And let's not forget about the patient. If AI is handling more aspects of care, do patients trust it? Are they comfortable with an AI summarizing their medical history or providing recommendations?

It's something every health care executive should be thinking about. I did an interview with one of our partners earlier this week who developed a tool who's a physician specific to some of the workflow aspects within the EMR. And I said, how do you educate both patients and providers on how you are using AI?

And he says, it starts with that baseline conversation because the patient trusts their PCP. Even if they're going to subspecialists, even if they're going to other forms of care, they're gonna ask their PCP about it. So just know that about your own systems, the PCPs, the nurse practitioners, even the front desk, wherever that patient receives their primary care, that's where they're gonna go for these questions.

So make sure that staff is very well.

really important to focus first on what needs to be shared appropriately with the patient, inclusive of any state or regulatory laws, because different states have different laws about how you can use certain aspects of A. I. ambient listening, et cetera.

The person that they have the most access to is going to get asked these questions so important to, make sure that they know how to answer them. In closing, if you're having a conversation with your board, your leadership team, or even frontline staff about AI and EPIC's EHR, here's some of the things you might say.

AI is here, Epic is embedding it directly into the workflows, which, goes without saying, but it has the potential to reduce administrative burden, improve patient outcomes, and enhance interoperability, but it comes with challenges. Security, accuracy, and change management must be addressed,

the leadership teams need to proactively test and measure AI's impact before scaling it.

It's also an ongoing transformation. It's not a one time implementation. So organizations that strategically plan their AI adoption will be the ones that benefit the most. And that goes back to whether they're embedded in certain aspects of the technology that you have, or they are pouring solutions being integrated appropriately.

All of those conversations need to be happening. In real time, and then adapting those pilots in those environments to ensure you have the best possible outcomes for the patient, for the clinician, and for anybody who is providing care and dependent upon positive outcomes, which is the intent of the usage of the technology in the first

that's it for today's episode. Join us tomorrow when we discuss whether AI is providing actual value. Remember to share this podcast with a friend or colleague. Use it as a foundation for daily or weekly 📍 discussions on the topics that are relevant to you in the industry. They can subscribe wherever you listen to podcasts.

Sarah, thank you for joining and thank you all for listening. That's a wrap.

Thank You to Our Show Partners

Our Shows

Related Content

1 2 3 316
Healthcare Transformation Powered by Community

© Copyright 2024 Health Lyrics All rights reserved