AH108 - Fixing Healthcare Interoperability and Modernizing Digital Workflows, with Brendan Keeler

Lightly edited for clarity.

‍

[00:22] Justin Venneri: Hello and thank you for listening to another episode of the Astonishing Healthcare Podcast. This is Justin Venneri, your host and Senior Director of Communications at Judi Health. Today I'm joined by Brendan Keeler, Interoperability Practice Lead at HTD Health and the author of Health API Guy, pretty popular, one of my favorite Substacks out there. That was kind of the genesis of this episode and our outreach to Brendan was because of a couple of things he wrote. Month after month, they just kept catching my eye, so I figured I'd reach out and see if we could get him on the show.

‍

There is a lot happening. CMS continues to push changes that affect how providers, payers, and tech platforms exchange data. Prior authorization remains a friction point for regulators, employers, clinicians, and patients alike. And across the healthcare ecosystem, there's a stubborn gap between what modern tech can do and what day-to-day workflows look like in practice.  

‍

Brendan, thanks for being here to talk about all that.

‍

[01:10] Brendan Keeler: Yeah, thanks for having me on. Excited to chat.

‍

[01:12] Justin Venneri: So tell us a little bit about yourself and your journey within healthcare.

‍

[01:15] Brendan Keeler: I come from a long line of providers. My mom's a nurse, my dad's a urologist, my granddad's a urologist. So logically, I looked at healthcare, and I was like, "Oh, maybe I'll do something there." It did not go well in terms of not really liking blood and things like that. I thought I'd go into other professions, and then boom, I ended up in Verona, Wisconsin, working for Epic Systems, a little old health technology company.

‍

[01:38] Justin Venneri: Small one, yeah.

‍

[01:39] Brendan Keeler: Out of college, I thought I was working for a gaming company. Turns out, no, it's EHR software. I've been in it ever since, working on integration, interoperability, health information exchange, both domestically and abroad. So, went to the Netherlands for most of my career with Epic and did fun things in Dutch over there. Freaky deaky things, Austin Powers.

‍

[02:02] Justin Venneri: And you are a doctor now because of LinkedIn, right?

‍

[02:04] Brendan Keeler: That's right. LinkedIn has anointed me a doctor, so that's pretty cool. Shout out to Monica Keeler and Luke Keeler. I finally did it. I've worked for Redox, Zus Health, and Flexpa, all different angles on how we digitize workflows between organizations, between patients. That's what interoperability is. I've had the chance to work on regulation, look a lot at litigation, and write way too much on the Substack.

‍

[02:28] Justin Venneri: So what in particular about this area -- I mean, Epic's obviously one of the 800-pound gorillas in terms of EHR, EMR software. I think everybody knows what they do. Interoperability in general, it's kind of a good buzzword for the last decade plus, right? Everything needs to be integrated, interoperable, longitudinal data, etc., so you can kind of work along the continuum of care efficiently.  

‍

What about that drew you to this, in terms of your writing and your content and just your affinity for this intersection of policy, tech, and workflows?

‍

[02:57] Brendan Keeler: Yeah, I mean, I randomly drew it from the deck when I went to Epic, is one thing. But after I did it, right, I did integration solutions within an organization, connect to Epic, and then I did interoperability. I did, how do we do data exchange between organizations, particularly for clinical data exchange or e-prescribing? Stood both of those up in the Netherlands, some of the first instances of that of any kind over there.  

‍

And I was like, man, this is pretty cool, working with the Dutch government as a young 20-something, which Epic was like, here, you get to go liaison with the Dutch government. That was all something that I'm like, oh, this is all interesting. And the fundamental problems of how do we move between systems of record, like Epic, how do we build ubiquitous digital networks, how do the regulatory mechanisms influence and accelerate that, that was just interesting academic problems, and just the career continued to compound in terms of different vantage points working on that.

‍

In general, as network-based businesses have the chance for some of the greatest impact, or are some of the hardest to build, right? Building network-based businesses, whether they're businesses or industry consortiums, that collaboration between parties with different incentives is really challenging. But when you do it correctly and get it to ubiquity, that has some of the greatest impact in terms of technology, right, in terms of reducing waste, in terms of making hard manual processes easy.  

‍

Well, it's cool to build a database with a nice little UI, but to facilitate digital exchange at a national level, you suddenly can really drastically improve things for patients, or providers, or really any other stakeholder.

‍

[04:33] Justin Venneri: Yeah, it's pretty powerful. And one of the stronger threads or themes in the discussion, I think, today, and just in your content recently, is current CMS rulemaking and that environment, especially around prior authorizations and related transaction standards. Can you walk us through the most important rules or proposals people should be paying attention to right now? What do those CMS actions -- what are they aiming to change, and why should everyone care?

‍

[04:54] Brendan Keeler: Yeah, I mean, like you said, interoperability is a buzzword. Why do we even give a damn about it? Well, I think it's because the chance to improve really manual processes of phone calls, of faxes, that take minutes or hours for people on the payer, or provider, or PBM side, suddenly there's a push to make those digital. And we've had pushes before for different workflows, right?  

‍

E-prescribing in the late 2000s and early 2010s, the government pushed it along with various different regulatory programs. And all of a sudden, we have ubiquitous e-prescribing. Yay! That's the outcome that we can get to with well-designed, timely regulation, is to push things over the hump to an atomic network and then to ubiquity.  

‍

And we've had other pushes along those lines for different problem sets. And the one du jour, the problem set of this administration, of the last administration, and of Trump one, we've really been on a journey to this point, is prior authorization, which today is not digitized as a subset of interoperability. It is decidedly un-interoperable. It is phone calls, it is faxes, it is, if you're lucky, a portal. And it is just burden for providers, burden for payers.  

‍

And so there's a push via the CMS 0057 rule, the prior authorization rule. There's a push via the 006 rule, which is the prior authorization for drugs, right, the drugs on the medical benefit rule, to digitize those, and not just digitize them, but make them ubiquitous standards for all CMS-regulated plans -- so Medicare, Medicaid, CHIP, and ACA. And then on the provider side, via what's called HTI-4, a rule for provider software to support the same transactions. That's pushing us from not digitized at all to, in the next couple of years, we're going to see electronic prior authorization rapidly mature across those stakeholders and beyond.  

‍

And so for PBMs, even, you start to talk about drugs on the medical benefit, there's impact for those stakeholders, too. And so what will it do? Well, it will drastically reduce burden, right? If we digitize these things, the incremental cost of grading and processing prior authorizations will go down. Will it mean that there's more or less prior authorizations? Well, there's a bit of a paradox here, in that the ease of the transaction means that overall the volume will go up, right? The number of times that we submit prior authorizations -- it took me 10 days to do a prior authorization, and then if there was a failure, I might just do it once and then say, too bad. But if I can do it instantly, well, I might do it once, and then twice, and then three times for different drugs or different procedures. And so the total volume will go up, but the burden will likely be lower.  

‍

And so, overall, interoperability doesn't change some of the social problems we have, some of the incentive misalignment we have between stakeholders, but it at least has the promise of removing some of the burden on people as they go through those convoluted processes. But tech can't solve regulatory and social and policy problems that are more upstream in terms of, do we have prior authorization or not? When do we have prior authorization? The complexity that we've instilled at a process level.

‍

[08:03] Justin Venneri: Yeah, no, that makes sense. We see that a lot, of course, on the pharmacy side and the prior authorizations, and especially for things like GLP-1s and others where, you know, which prescriber is writing the script or suggesting the patient get it? And how many times are they submitting it, and is it part of the plan design?

‍

[08:18] Brendan Keeler: CMS has the GLP-1 for Medicare CMMI program. I forgot the name of it, but that's a fascinating one because, all of a sudden, they're like, hey, maybe we allow for GLP-1s for traditional Medicare beneficiaries, but first they have to go and exhaust any other coverage they might have.  

‍

So all of a sudden, for PBMs and payers and everyone in that mix, you're having this really convoluted check across different insurers. And how do you communicate back and forth that, yep, I checked these other benefits beforehand? So what would be interesting and great to see, this transformative drug be appropriately accessible for Medicare beneficiaries, possibly -- GLP-1s in the right context can be very powerful medications. But at the same time, the layered complexity falls to the PBMs, to the payers, to the providers to deal with, from the processes, the policy measures we put upstream, rather than the tech itself.

‍

[09:08] Justin Venneri: Makes sense. Yeah, using AI as well to help process these fax documents and images and wonder how much more streamlined it could be with a fully digital system kind of running. I mean, prescriptions generally process pretty quickly and efficiently when it's digital. So, I mean, it's a milliseconds transaction.

‍

[09:24] Brendan Keeler: So, yeah, what's interesting is people obsess over standards, like do we encode in FHIR is one standard, or NCPDP is another standard, or HL7v2. But at the end of the day, that actually, for interoperability purposes, isn't the most important facet or measure. It's actually ubiquity. It's can I send and receive between all the counterparties that are out there?  

‍

Because if you have a digital standard and it only works for two counterparties or a small region, but you're national in scope, the utilization of that pathway is kind of a hindrance, kind of a barrier to you, because you have to remember, okay, this is digital, but then I have to go phone and fax over here. And you find yourself falling back to the ways you can do things ubiquitously. And so ubiquity of, is everyone on the network, can I send and receive this transaction digitally, really is the measure of success for these initiatives and not how cool have we designed the technology or the standard that underpins it.

‍

[10:21] Justin Venneri: That's a good lead – like, my next question was along the lines of, where do you think the next most meaningful gains will come from? I mean, you have this kind of regulatory push in the back, and you kind of see the potential to follow a similar path to e-prescribing. What else is going on out there between payers, providers, and vendors trying to reduce friction in the real workflows that everyone's engaged in, taking care of patients and trying to get paid for taking care of patients?

‍

[10:44] Brendan Keeler: Yeah, I mean, I would just think about it like, when you see fax, or PDFs, or paper, or email, or portals, you can just look and you're like, what am I trying to do? Because interoperability is a superset of 60 different problems and jobs to be done, workflows in a trench coat, right? You unpack them and you're like, okay, this one's fully digitized, right? Like e-prescribing. And you might look at another, like clinical data exchange for treatment, and you're like, wow, that's pretty ubiquitous. That's semi-interoperable. Then you look at electronic prior auth and it's totally unsolved.  

‍

And so realistically, people are like, why isn't interoperability solved yet? That's a silly question. There's infinite workflows between parties and counterparties and organizations of different types. So, for instance, I can send the e-prescription, but if I want to transfer my script to another pharmacy -- they send it to CVS, I want to pick up at Safeway -- that's not ubiquitous. That's a phone call or fax between those pharmacies. They take a day, it might take hours, like whatever, but it's painfully manual. And so if you look at just that piece of interoperability, you go, wow, there's no interoperability. And so I decompose it typically into different workflows, and then you can measure and be like, how far along are we towards ubiquity? And so I think, where are we going to see meaningful gains? Electronic prior authorization very clearly is a focus, a bipartisan focus.  

‍

So we're going to see gains there. Clinical data exchange continues to get closer and closer to ubiquity. We see all the major EHRs participating in a treatment exchange of, if you're at your PCP and you go get hit by a bus and you go to some random acute care, probably they can pull your clinical records, is where we're at, like 80% of providers, all the major EHRs. But then some of the long-tail specialty ones -- behavioral health, PT, oncology, specific EHRs, maybe not.  

‍

And so they're coming on slowly, and we're seeing pushes there. But that's one where it's more market-driven adoption with light incentives via the Trusted Exchange Framework and Common Agreement as a voluntary network, the CMS-aligned networks as a voluntary agreement. These programs, they help encourage people, help raise awareness. But to drive all the way to ubiquity, maybe you need some incentive changes for people, that pure market incentives don't work financially.

‍

[12:54] Justin Venneri: As systems become more open and can connect, and information can be shared, you know, securely and in a normalized fashion, when you look at the market, what do you see really changing in how healthcare organizations think about their infrastructure? Because we spend a lot of time talking about, like, hey, you can come up with the best idea in the world -- AJ says this all the time -- the best idea in the world, but good luck implementing it on archaic infrastructure. And, you know, we try to pride ourselves on being open, secure, scalable.  

‍

Our Judi platform -- this is not a rah-rah Capital Rx/Judi Health podcast, but it's important. You know, I don't understand how everyone's just like, oh well, it still works, so it's okay. But the system hasn't been updated in 10 years.

‍

[13:36] Brendan Keeler: This is the systems of record problem. Like every business has its system, the last system you pull out, right? The operational heartbeat that typically has a financial or clinical component to it. So that could be an EHR for a payer; it could be a core administrative processing system for the claims processing. PBMs have it too. And what's interesting about this type of software is that it's cement, right? If you look across any industry, school information systems, or dealer management systems, payers, PBMs, pharmacy information management systems for outpatient pharmacy, these vendors are not new, by and large, except for Capital Rx/Judi Health, as a new person coming in.  

‍

But once you put it in place and you get successful, the challenge of changing that system of record is not linear with your size, it's actually exponential, because you start to build data gravity, the weight of data is exponential compared to your organization size and history; workflow gravity, the people just learn how to use it and work around it and build their workflows around it, and integration gravity, right? You build a web of systems around it. And given your business, the lifeblood of it is running through that system; the investment to change, your switching costs, are astronomical.  

‍

And so, as a result, you're a CTO or CIO. You're someone who gets in charge and you're like, we've got to change and modernize. And you look and go, but if we mess this up, I'm going to get fired. And so people lack the confidence, the intestinal fortitude, the risk tolerance in particular, to take that leap, even though there's outsized gains to be had. It's easier to say, well, let's go change how this department's organized, or let's go slap some AI on top of it, right? That's an easier, more palatable equation for most leaders.  

‍

And so it favors the disruptor, the insurgents, someone starting de novo, Greenfield, because you can say, how do we build really sustainably using the best available technology at this time? The drawback being that you should think ahead. What's it going to be like in 5 years or 10 years when you're in the same situation? This is for every organization, for people with their EHR, people with their ERP. Any system of record has this switching-cost issue and dependency that is hard to break away from.

‍

[15:46] Justin Venneri: All right. And I mentioned your Health API Guy Substack, and it's always got great memes in it. What was your favorite one from the April monthly review, and why? Tell us a little story about what. And I'll link it of course.

‍

[15:58] Brendan Keeler: Yeah, so Monthly Review just looks back on all the posts from the month. You know, topical to what we're talking about. It's this girl, it says, "Should we use NCPDP or FHIR for prior auth?" Two different standards. And that's a picture of CMS and says, "Why don't we have both?" And they throw up their shoulders. So anyway, a lot of good memes in there. Some making fun of Epic, if that's, you know, people's alley, but really a diverse selection.

‍

[16:21] Justin Venneri: Got it. You know, the need to leverage your technology and processes to help with, like, basically taking complex issues and processes and adapt them to the environment and the workflows, have them reach people outside the quote-unquote expert circle. I'd love your thoughts on just why that's so important in healthcare.  

‍

I know at one level it's like you're taking care of patients, you want to make it easy, and you want people to be able to explain things. But maybe tell me a little bit about your take on, you know, areas like drug pricing, regulation, or benefit design, even if you have thoughts there, I'd love to hear.

‍

[16:52] Brendan Keeler: Yeah. Why is it important? Because people want to solve problems, and their ability to solve problems is predicated on empathy. And most people, unless they've worked in healthcare, but with any industry, it's like you don't have empathy unless you've done it before or you've been through it before.  

‍

And so most of our experiences are limited to being a patient. And that's a very limited vantage point into, like, how do provider organizations work, how do PBMs work, how do payers work? You know, what are the softwares they use, what are the problems they solve day to day, what are the constraints that they have?  

‍

And as a regulated industry, it means that you just layer in complexities, a lot of times unintuitive complexity, that we can't have empathy for until you understand it. And so I think there's been a big push in the past couple years in healthcare for great content to bring people in to help solve problems around, like, why does benefit design work the way it does? Why does prior authorization exist? Like, what are these technical standards? If you don't bring it to people's level, then they're dissuaded, and they're just going to go build a cool little gaming app or consumer app, because the empathy gap is so much less there.

‍

So if we want people to work on and solve big, meaty, important societal problems, then you have to come and meet them in the middle and guide them to understanding, to speed-run them through the empathy gap.

‍

[18:14] Justin Venneri: Makes sense. And so I have a couple more questions for you. Thanks, Brendan, for spending the time with us today. One is, I'm sure listeners are probably wondering, like, why haven't you asked about this yet? But, you know, if you saw our announcement that we partnered with CLEAR as a preferred identity verification partner, you know, I know identity verification is a topic you're interested in, you write about, worked on. Wondering for your take on just kind of leveraging new technology like that in partnership to kind of streamline things for patients when they walk into a hospital.

‍

[18:43] Brendan Keeler: It's foundational stuff to enable any type of interoperability. So commend you guys, commend the health systems adopting digital identity. It's foundational in the sense that appropriately knowing that “I am Brendan Keeler” unlocks the ability to much more easily have trust between counterparties and share things that could be sensitive or that you wouldn't put publicly.  

‍

If we had the ability to put everything public, then we would do so, but we can't. And so strong identity of patients, of providers, of users and their organization gives us the ability to understand the benefits of different plans or who's gold-carding or prior authorization patterns or clinical data access. And so it's awesome to see the adoption across the ecosystem by payers, by providers, by PBMs, by technology vendors, because it's going to unlock the next layer of not sort of guessing like, oh, is that Brendan Keeler, or is that Brendan Keeler pretending to be Lil Wayne? Like, you know, really you can start to do really trusted stuff when you can assess and know at a baseline who's who.

‍

[19:47] Justin Venneri: Yeah, we're excited about it. Okay, and last two questions for you. One, crystal ball. If you look out over the next, call it 12 months, maybe 24 months, what should healthcare and health tech leaders be watching most closely? What are you watching most closely?

‍

[19:59] Brendan Keeler: I'm watching the ONC, the Office of the National Coordinator, because they will finalize a rule, HTI-5, which is going to change the contours. It's going to deregulate the EHR certification program, but also change information blocking, which is an important statute. So maybe more burden than there is deregulation, we'll see. But then they're going to re-regulate back up with HTI-6, or at least they've alluded to that, and that's going to change.  

‍

What will that require? What next level of interoperability network will that push? Remains to be seen, but I'm excited for it because to get to build those next networks beyond prior authorization, so the next problem set, the government is uniquely positioned to push everyone along.

‍

And so I'm excited for that. I'm excited to see that CMS continues to cook, and they will not stop cooking – maybe midterms changes that. But my feeling is they'll continue to come up with new things and put more shots on goal. And again, there's opportunity to make things better in those changes.

‍

[20:55] Justin Venneri: Last question I ask everybody, and you've got to throw a compliance hat on for me. What's the most astonishing you've seen, a good story you can share related to our discussion today that you think the audience would find interesting? Could be funny, could be just shocking. Tell us a good story to send us off.

‍

[21:09] Brendan Keeler: I think - I'll tell you an interoperability one. So one use case that sucked that's now getting better is patient access to data, right? We've always had the right to our data via HIPAA, which is a great right that is unique and not every country has. But it's manual, it's very burdensome. Go hospital by hospital, or covered entity by covered entity.  

‍

Well, now we're pushing forward and seeing not ubiquity, but signs of future ubiquity in terms of you using digital identity, using CLEAR, or ID.me, or Persona, proving who you are and pulling your clinical data to use in a personal health record, to pull and underwrite life insurance, to go and use with OpenAI, to do all the things that you might want, in the same way that providers are able to exchange data pretty frictionlessly today.  

‍

That's awesome. If we can meet patients where they are instead of having to do convoluted processes, that can happen. You know, all these, the healthy middle of the bell curve, something can be activated to be healthier. Do the easy thing, because when it's a hard thing, they'll say, well, you know what, I'm going to go watch Netflix instead. But when it's baked in and it's part of their workflow and frictionless, you know, all of a sudden we can nudge things towards better outcomes. I think that's pretty cool.

‍

[22:18] Justin Venneri: That's very cool. And that's what it's all about, right? Better outcomes at a lower cost, if possible.  

‍

Brendan, thank you very much for spending time with us in the studio today. It's great chatting with you, and I hope you have a great rest of your day. Love to have you back on.

‍

[22:30] Brendan Keeler: Hey, thanks a lot.

‍

‍