What is the five-year trajectory for Epic?

One of the biggest and most successful software companies in health care remains privately-held and no one has a clue what it’s truly worth or how well it’s doing. That company, of course, is Epic Systems. Since it was founded in the late 1970s by Judy Faulkner, it has remained privately-held - and it’s unlikely that it will ever IPO, given its commitment to never going public. 

The best guess amongst my network is that the company’s market cap - if it were publicly-traded - would be somewhere around $40 to 60 billion, given it owns about 40% of the hospital market and likely generates about $6 or $7 billion per year at this point, or even more. I’d be basing that on Oracle’s $28 billion acquisition of Epic competitor Cerner, and some metrics from 2022 that have not been updated since. 

Epic is a tough company to write about for many reasons, some of which we’ll get into. For this piece, I tried to set aside my own experiences as a former journalist covering Epic. I talked to some super smart people, mostly a mix of operators, hospital CMIOs and CIOs. People have strong opinions about Epic, and all the EHRs really. But Epic in particular given that it seems to be the elephant in the room. As one chief digital officer of a large health system put it to me this week: “It may not be super obvious, but the bottom line is that if you look at total inpatient beds Epic is cleaning up…. They’re quietly dominating.” 

The reason that isn’t out there as much is that Epic doesn’t seem to be as focused on new clients, and more in expanding within existing via the “community connect” model. The largest systems on Epic - the AMCs and IDNs - are essentially resellers at this point to smaller systems, and have been for years. The idea is fewer clients, but more beds per client. This is the classic “land and expand” strategy that you see across software, with a bit of a unique twist. 

That has meant that some of the problems are hard to budge: The software remains extremely expensive; and there is not enough competition in this space to keep pricing down. What’s the overall sentiment and vibe?

People in the industry seem more resigned that this is the status quo, even as it comes with problems (the software isn’t nimble; the user experience; the price). 

In terms of new players on the block, no one seems to think there’s going to be another inpatient EHR to rule them all anytime soon. A phrase that a CIO used with me recently to describe Epic: “The nicest house in an ugly neighborhood.” In other words, it’s the best we’re going to get. One of my favorite quotes comes from Chris Longhurst from UC San Diego Health who once described Epic to Modern Healthcare as the “cream of the crap.”

To be fair to Epic: It’s a hard thing to build. As Brendan Keeler put it: “Inpatient EHR that does cardiology, connects to ER workflows, the ICU, and connects monitors… it’s just so much.” But there’s definitely still frustration amongst providers that the user experience isn’t better. Most people seem to think still that they better go with one of the big players, and the most popular with AMC’s still is Epic. Cerner, which is now Oracle, seems to be doing better with government contracts. Then there’s also Meditech and a few others, but overall it’s not as many large players as you might think.

Some Backstory

A lot of the frustration about the challenges with EHR software is rooted in the HITECH Act. The federal government subsidized these vendors back in the 2009 era to the tune of tens of billions of dollars in taxpayers’ money via the HITECH Act (remember “meaningful use?”). The goal was to move these digital record systems from paper to digital. That transition was far from seamless. We couldn’t even get basic information transferred from one hospital to another; or identify patients well by name. And many of these systems were built for billing more than anything else.

Hospital executives for a long time were pissed they had to spend hundreds of millions, even billions of dollars, on these software systems that seemed to be missing critical functionality. Some of that could be chalked up to old technology and misaligned incentives. But a lot of it was directly linked to regulatory constraints, compliance, and a fear of medical liability. All of that drove providers to a state of crisis, given the heavy documentation burden.

That’s a big part of where Epic is going next. Let’s dig in. 

The Forward Momentum

What we do know by following the signals is that there’s been some strong progress in a few areas:

1) Solving the documentation burnout crisis

2) New markets like the payer/pharma space: 

3) Large language models and AI

I’m intrigued by all three of these and I personally think the biggest near term opportunity is with health plans. This will also bring Epic into a position of competing with Cerner in some new ways, including in Life Sciences solutions space. The first and the third buckets, of course, are heavily linked as much of the early AI work is geared towards documentation burden. There’s a reason why Epic has extremely close ties with Microsoft, and I’m sure it will stay close with many of the large technology companies.

Let’s talk about pajama time

Studies have shown that doctors are spending hours a day on documentation within the EHR. Hours. This is a huge problem. No one signed up to become a doctor or nurse to do data entry, and that has become a big part of the job. It’s the reason why we sit in clinic and the person who’s supposed to be focused on our health care is glued to a screen.

Things are getting a tiny bit better from what I hear, albeit slowly. There’s plenty of groups moving that work forward, including patient advocacy groups and medical/hospital associations. There’s also a whole contingent of smart folks over at ONC, the federal entity tasked with overseeing health information technology, that are working to crack interoperability and make EHRs function better. From what I can tell, the group is knee deep in the rule-making phase, and our industry is bracing for some big changes up ahead (there’s a big election coming up). That work deserves its own separate post, and I’ll get to it I promise! 

So where does that leave us — both with solving key problems like interoperability and burnout? And what is Epic up to as it looks for growth outside of its core market: Hospitals and health systems? As someone who once wrote a three-part series on interoperability and considers themselves a health IT nerd, these questions continue to both fascinate and perplex me. 

The rise of DAX and Abridge

There’s a reason why Abridge and Nuance - two of the companies duking it out to turn clinical conversations into structured notes - are quickly gaining traction within hospitals. In my experience, some hospitals are team Nuance; some are team Abridge; and some are piloting both simultaneously. Nuance seems to benefit from its close relationship with Microsoft, but a lot of providers like the Abridge user experience. I hear good things about both.

If you haven’t heard of Abridge or Nuance, I won’t use jargon to explain what they do due to my jargon allergy. Think of it as this tool that “listens” to an encounter between a patient and a doctor. All of that gets transcribed, and if the doctor is worried about a potential fabrication, they can highlight a specific part of the text and hear the audio. There’s also a way to turn that clinical summary into a set of “next steps,” and then make it easy to bill with the right codes. Patients can view the summary in the patient portal for one of the EHRs they use, assuming they bother to log in (most of us forget because we have busy lives, but it’s nice to know it’s there). Note: I’ve tested Abridge and you can watch the video here.

What I’m hearing so far is that these products essentially do the thing they advertise. One CIO told me that he thinks the documentation time is going to be cut down by 50% as his system continues to roll out DAX. The big questions right now seem to revolve around price, as it’s expensive, and how many clinicians within the system will actively use this new technology (many of the health systems I’ve been speaking to want to roll out specifically to the clinicians who are going to use it to save cost). The name of the game right now is about identifying who the early adopters are going to be, and making sure they have the right number of seats.

Both companies have a close relationship with Epic by the way. Abridge is a “Pal,” so the information it collects can sit within Epic; and Nuance is fully integrated into the workflow. 

Build or buy?

Back to Epic. The big question everyone is asking is whether Epic will buy a company like Abridge, or a competitor. I’m sure the VCs will back a few more of them, and I’m already seeing evidence of that.

Truly, in my mind, it could go either way. I’ve spoken with a few folks about this who say it’s unlikely, because this isn’t Epic’s core competency and others are solving it. And yet, this is the core use-case that needs to be solved right now. Abridge and Nuance are going after the real pain-point in health care right now on the provider side. There are literally dozens of studies on this. Health systems are paying for these types of tools for a reason: The documentation is causing massive burnout. My bet - and this is based on zero insider knowledge to be clear - is Epic does try to get into the game through an acquisition or they leverage the close relationship with Microsoft to build something together. Otherwise they’d be leaving money on the table. But again, that’s entirely speculation. 

What about Cerner?

The press has not been kind of late to the Oracle/Cerner combined entity. The Business Insider piece led to a very strongly worded response from Oracle, one that I was frankly surprised to see given that most tech companies tend to be very conservative with their comms strategies. Bloomberg wrote something more measured, which hinted at problems that we’ve all been hearing about for a while.

So TBD. It’s been about 18 months since the acquisition and most of us in the industry are watching closely to see what Oracle will do with Cerner. Will it go after Epic in its core market, or focus on the government work? How about AI, given Oracle/Cerner does seem to have made some announcements here also? It’s clearly a major focus. There’s going to be new battlegrounds in the next few years, so I’ll be sitting back with my popcorn. 

Where else is Epic going next?

Well it’s clear to me that Epic is sitting on a very valuable data-set but I do not believe that it’s a data company or data broker given the contracts it has with health systems. I do not see it competing with, say, an IQVIA.

We’ve touched on some of the most exciting areas for the business in terms of growth. In the near-term, AI will continue to be a big area of focus for EVP of R&D Sumit Rana. From talking to clinicians and bio-informaticists, it seems like there’s an opportunity around Epic Haiku, where there’s already technology in-built for ambient listening. That tool is designed to streamline physician note taking, per Stat. Don’t underestimate Epic when it comes to new AI. One thing that is worth pointing out about Epic’s decision to stay private is that the company can spend on R&D without any intrusion from Wall Street. 

Because of AI, Rana said one site reported a savings of “five-and-a-half hours” per week. (Also, after you click on the link above, can we talk about Sumit’s outfit? It’s giving Jurassic Park, but I don’t hate it.)

Where I’m most interested in the next steps for Epic are in two areas: Payer and Life Sciences. There’s already a lot going on in the health plan space - check out what’s already publicly available to peruse here on its “payer platform” site. A lot of folks I spoke to said the opportunity here is a “slam dunk,” because of CMS mandates and the need for more solutions to connect payers and providers (essentially to bring together both claims and clinical information). This is going to be a huge focus in the next five years, and there’s already been a lot of traction here. 

The other area is Life Sciences. In my mind, this is where the growth will be (see “Cosmos”). Keeler, who’s been following this space for years, talked about how the value prop will be for pharma companies to identify specific patients for their trials with hospitals that have opted in (Truveta is likely a potential competitor). It won’t be like a blunt instrument - taking a huge amount of anonymized data and sending it over to a pharma. It’ll be very specific and fully bi-directional, Keeler notes.

What’s going on with Particle?

I don’t think we know all the details yet, although this coverage seemed to be more balanced and is worth reading to get up to speed. A lot of the issue seems to be that there’s a very loose definition around what types of data can be accessed via some of these networks, like Carequality, in particular the definition of the term “Treatment.” I suspect we’ll learn more from both companies in the coming months. I will stay tuned rather than comment on it for now.

Where are the opportunities for startups?

I wish I could say that now is the moment to raise a bunch of capital and build the next great inpatient EHR. But it probably isn’t for a variety of reasons. There’s already been billions of dollars invested in software, and the switching costs are high. That said, there has been a rise of startups that are geared to more virtual providers. Canvas and Healthie are venture-backed examples. AthenaHealth is also still around, focused on the ambulatory space, and has new leadership on the executive team. There’s also a bunch of leaner, smaller EMRs out there that focus on a subspecialty. 

So that’s all for this week! If you hadn’t noticed, this post is coming to you via Beehiiv, my new newsletter platform. I hope you didn’t experience any kinks with the migration - but if you did, don’t hesitate to reach out and I’ll do my best to sort you out.