An Epic vision for using AI – but will it work in the real world?
The company plans to infuse artificial intelligence throughout its applications, but larger questions remain for customers and the wider industry.
It’s not hard to get caught up in the moment during the annual Epic Users Group Meeting (UGM). There are thousands of attendees, thousands of Epic staff doing everything to help the masses, lots of cheering and applause, and a pretty slick set of presentations that tell what’s been accomplished and what lies ahead.
As a member of the throng at Epic’s UGM this week, I could really get a glimpse of where the company is going, as Judy Faulkner and other company leaders sketched out a vision for what artificial intelligence could accomplish within the framework of the company’s applications.
In her executive address, Faulkner noted that Epic has been working with generative AI “quietly,” and has 160 to 200 AI projects going. “Some are completed or in process. Some will be permeating every area of our software,” she explained.
Epic’s AI plans are intended to help it build off the knowledge in its Cosmos database and other applications used by healthcare organizations. At UGM, Epic cast a vision for assisting human users by combining their intelligence with that of generative AI.
Faulkner and other Epic execs cast a vision for three monikers for AI use within its systems. The company calls it “Art” for the clinicians, “Emmie” for patients and “Penny” for use by administrative staff. The goal is to make all types of information that’s germane to each segment easily accessible when and where it’s needed, easing the burden and frustration of clinicians, patients and administrators.
The bulk of the morning session was built around morning demo-skits that fleshed out how this vision would work in the idealized real world (if your concept of the real world is a spaceship where a knee replacement is being considered, despite the risks of zero gravity…work with me, here). It was all very slick and polished, and to be honest, I so so so want something like this to work in healthcare. It shows the potential value of all the work that’s been done to bring computerization and data to healthcare, bringing value to the use of technology and downstream benefits.
I am neither a cheerleader for Epic nor a hater. But I did have to reflect on innumerable similar simulations I’ve seen in the past at HIMSS conferences or other industry get-togethers. And I feel like we need to add perspective on what still needs to happen in areas where the industry is near the vision cast during UGM, and also where it’s still a long way off.
As Epic presented these simulations, it noted that some of the tools are still in development, and it projected future delivery dates. The company revels in its positive track record of delivering on projects within its time targets, and it certainly has the resources and information repositories to draw upon. Still, there’s a lot of work that needs to happen over the next 12 months to achieve the ambitious targets that Epic has set. Tools developed in isolation don’t always play together well, and revamping existing products to incorporate AI when and where it’s needed could be challenging. Basing capabilities on still ongoing development adds questions to the eventual capabilities.
Then, after you develop this source of water for the industry, the question becomes whether you can lead the horses to the source and have them take a drink. Will the industry, broadly, be ready to adopt into it?
Epic touts those of its users that have fully implemented its systems and delivered measurable, exciting results. Some healthcare organizations have extensive IT staff and are completely bought in to the multiple components that Epic contains. And a fair number are already dabbling in projects such as agentic AI, which facilitates notetaking and records creation. But that’s not everybody. Many organizations have not turned on all the components in their Epic systems. The company consistently urges them to do so, but it can be a heavy lift for some.
And the use of AI-infused products by consumers may be a work in progress. Not all consumer-users are tech savvy, and they may not be able to fully use AI in MyChart. If the AI can be brought in so that it’s intuitive and essentially invisible, great. If it’s confusing and hampers the use of consumer-facing tools, that impact could fall back on healthcare organizations, which may take on an increasing burden for user support for patients and families.
And Epic angles its products for larger organizations, systems, regions or states. While that makes sense from a business perspective, there are whole swaths of providers that don’t have the size or resources to be considered for an Epic implementation. Granted, the company is trying different approaches to make it applicable for smaller medical organizations, but there’s also a business/cost component to that kind of extension.
Certainly, the future that Epic paints for the potential use of AI within its products is important. It’s where ALL providers need to get to, not just tech-savvy Epic customers. The societal question is, how do we best accomplish that when revenue appears to be shrinking, and yet with so much benefit just within reach?
Finally, making the shift to an AI-infused environment will be scary for some organizations, which, after all, are people who have day jobs, like caring for sick people, managing an overwhelming workload and keeping the lights on in organizations. Some handholding, and some fear-easing will be needed (not just from Epic, but from every other company that’s going to an AI-augmented product line).
Yes, Epic has the experience here, in spades. I met Ryan and Gwen from Epic’s staff there, who demonstrated kindness and compassion in helping this lost soul find a room for a meeting. That’s a great beginning.
I so want this to work – not just on a fictional spaceship, but on planet Earth. Not just for those getting care in Epic organizations, but for everyone.
Fred Bazzoli is the Editor in Chief at Health Data Management