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

 Thanks for joining us on this week in Health It. This is a solution showcase. My name is Bill Russell. . Former healthcare CIO for a 16 hospital system and the creator of this week in Health IT a channel dedicated to keeping health IT staff current and engaged. Alright, today we are joined by Marty Magen and we are returning guest Josh Peacock.

And I'm looking forward to this conversation. We're gonna talk. We're gonna talk the epic IRIS migration and, and what opportunities exist for health systems as they look at this. Welcome, welcome to the show, guys. Hey, thanks for having us. Yeah, so this should be good. I want you guys to feel free to, you know, really, uh, go at it and, and talk about this and explore this topic.

Marty and I actually just did a, a webinar on this. We got a lot of questions, so people are very interested in this topic. There's a lot of back and forth going on, so I, I'm looking forward to hearing, uh, what you guys have to say. Let's, let's start with just a, a broad question, which is, what is this?

What's, what's Epic doing? What's the opportunity? Why, what are health systems? Why, why should the health systems have this be top of mind right now? Marty, you wanna start with this and then, and then we will kick the next one to Josh. Yeah. Yeah. So historically, epic core database was always running on cachet.

That's been the core database engine for a very long time. InterSystems released the next generation, you know, database platform, which is. What Epic is doing, essentially helping customers migrate to

sooner. They're also giving customers the opportunity to change their platform licensing without a upfront capital investment. So historically, if I was running on power or if I was running on Linux and I needed to change platforms, there was always a, you know, license reissuing cost associated with it for the.

Healthcare systems, the opportunity to move to platform independent or platform specific licensing to essentially future proof, their upgrade and, you know, future platform roadmap for if they wanna run on power or if they wanna run on IBM and sorry, if they wanna run IBM power or if they wanna run on Linux and switch platforms, there won't be cost associated with it.

Wow. All right. So, so platform independence is, is one of the, one of the biggest things, and there's licensing benefits as well. So let's, let's go down this, it really does represent a, a, when you talk of platform independence for the first time, I'm not boxed in. So I'm not boxed in by a financial, weird financial model that doesn't allow me to go to a platform.

And I'm also not boxed in because now I can literally look at my data center and say what makes the most sense for me? So, Josh, as, as you, as you are working with clients on this and, and talking through it. What, what do you see as the, the biggest benefits and opportunities we have platform independence, we have licensing.

What, what are, how are health systems approaching this? Uh, well, I think because of the platform independence option and the ability to kind of take that, that switch without the penalty. I mean, we've been rec recommending that that's the direction people go. Just for flexibility now or in the, based on what we're.

E CCP used to almost be kind of a nasty thing to say. Enterprise cache protocols, multi-tier, kind of spread users out and also have utility servers for kind of spreading the load as these environments get so large. Well, it's been kind of recognized that that's gonna be a regular thing for most organizations that have any size going forward.

And the other thing is they move to IRIS. That's an inherently multi-server license, which allows ECP to be deployed without another additional charge incurred. So I think. Be able to switch to a different platform and that native ability to run multi-server without an additional cost has, has really opened up the doors for really any sized organization to go that direction without incurring additional costs associated with it.

Licensing wise, you know, one of, one of the questions we got in the webinar, which I think you know, is, is interesting because how ecp. You know, does it still require us to manually sync the CCPs files with the remote scripts? We got a lot of very technical questions in the webinar. I was really fascinated, but you know, how does this scale better than ecp?

Well, I, I think the unfortunate part of that is it's probably gonna be very similar. In the way that we've traditionally managed ECP and dealt with it, uh, some of the techniques and technologies that InterSystems has in from a IRIS perspective where, you know, whether or not that's being able to scale in a different way across more servers.

Those really aren't available to us when we're talking about an operational database for Epic. If you were developing your own thing or some other technologies that, you know, like an ensemble or something like that, they might be able to take advantage of that. But really the way the data structure sits in Epic, I think we'll see things being, you know, more of the same, at least for the near future.

You know, I, I think the biggest advantage is essentially what is Epic gonna do with IRIS, because it's still very new with the upgrade and with a virtualized data center. When you, if you move to an X 86 platform, it's additional virtual servers, right? You're treating it as a virtual server just like everything else, and it's throwing more memory and CPU, depending on how much you need to scale and when you need to scale for your standard database server and your ECP servers.

The other thing I think that it opens up to, and this kind of goes to the multi-platform, not that on AIX, we can't use some of the tools that are out there, but there's some really mainstream tools. Uh, one for example is Ansible that you can use to manage an environment, keep everything very concise and, and the same.

And that is one thing we've deployed at a couple clients where we've been able to essentially throw down a base image of Red Hat. The environment. Environment. And then from there, Ansible will take care of rolling out all the operating system changes to support Epic's requirements along with the organization's, you know, active directly, integration, user, authentic, you know, the way that the users access menus and all those sorts of things.

And you can make that super consistent across all of your servers. And then whenever you just need to add some ECP servers into the mix, it just becomes another one of the pool that you're managing in the same, same manner. We'll get back to our show in just a minute. We have a lot of good stuff after the break.

InterSystems, IRIS. As the next generation data platform, IRIS excels at delivering a modern, consistent, and reliable experience to clinicians and patients. However, at Sirius, we know that migrating or upgrading to a new platform presents its own set of challenges such as managing current and future costs.

Switching from SMP to ECP when migrating platforms or determining the best services or manufacturer for your future platform. When it comes to managing costs, serious healthcare experts can help you build out a modern infrastructure that will include a flexible upgrade cycle, along with a predictable cost model to ensure that you're aware of and prepared for any future upgrade costs.

We also know that switching from an SMP to an ECP architecture can be overwhelming. The Sirius Healthcare team is fully capable of offering guidance and support while building out your new architecture, minimizing the impact on your IT team and clinical users. When exploring disaster recovery options for new platform, we know you want the best technology available while controlling costs.

By utilizing a hybrid cloud metal, our experts can help you discover the perfect mix of services and technologies to do just that. A significant benefit of the IRIS platform is its compatibility with many intel manufacturers, such as Cisco, Dell, hp, and Lenovo. So if you're switching from power to Linux, the Sirius healthcare team can help plan a migration with a manufacturer you're most familiar with while developing a roadmap to manage your new platform for the long haul.

You guys are gonna be CTOs. And what we're gonna first talk about is how I should be thinking about this, right? What, what are some aspects I should be looking at? So I, I've got, I have too much variability in my data center. I still have too many systems, too much software. I have some challenges I need to deal with.

I need to deal with 21st Century Cures, API management. Those kinds of things. I've gotta think through, uh, architecture. Am I gonna be us utilizing cloud or not utilizing cloud? I got capital expenses, operating expenses. If I'm making you dizzy, yet, I've gotta be able to leverage, uh, all these new technologies.

I gotta, you know, NLP AI machine Learning. I've, I've gotta make better connections to my consumers and be able to build platforms and those kind of things. All right? So those are all the things from ACIO perspective that I'm thinking about that that's only half of what I'm thinking about during a pandemic.

But, but those are the things I'm thinking about. This is an opportunity, right? You're giving me platform independence, you're giving me flexibility. I can now, what consolidate am I gonna be able to consolidate? Platforms, am I gonna be able to reduce costs? What, what am I gonna be able to do? I'm the CIO asking you two, my cto.

So you could both formulate an answer to that of how should I be doing this and what, what is the opportunity for us? Honestly, I, I think it's looking at what is your technical team, sfa, right? What is my technical team skillset? What platform they manage better and what is my current data center hardware.

So you're

are.

But, but the hardware itself, I can really go in any direction at this point. Can I? You can, you can. I think it also, part of the hardware is what's my compute, what's my storage, what's my network? You know, have I standardized on Cisco UCS? Am I interconnect? And can I go buy a Cisco, you know, four socket solution that's gonna suffice for the operational database that I can manage the same exact way, besides minor specifics around locking down the system for access for my operational database that I can for the rest of my epic and clinical application infrastructure.

I, I think the biggest part of all that is, is coming up with a cost model. So most organizations, if you were to ask me to come back to you and say, Hey, you know, bill, my CIO asked me to price this out for three years, the first thing I'm gonna do is. Not just look at the hardware pricing, but figure out what the soft costs are.

Find out what the licensing prices are, depending on what platform I'm running. Figure out VMware licensing. Right? Figure out Red Hat licensing. Figure out support for three years, and then extend that out to five years and figure out, based off what you as ACIO is telling me, if we're gonna onboard any additional users,

hospitals.

Another part of it's skillset. This is a very good opportunity for me to come to you and say, Hey, look, we're short staffed, right? We're doing a lot with very little people, and we need to sharpen our skillset a little bit. Let's send some of the Epic support staff to training, not just for VMware, but go to Epic training, go to VMware training, you know.

Changing platforms

is a should expand out beyond operational database. Consolidation wise, platform consolidation, right? When am I refreshing my hardware for my operational database? When am I refreshing my hardware for hyperspace and Citrix and VMware Horizon? When am I refreshing my hardware for everything else application server wise?

Can I get into a predictable refresh model where I can come to you and say, here's the cost for updating 33% of my clinical application infrastructure hardware, including Epic every year. To make that or are we gonna get stuck in this where I knock on your door every three years and say, Hey, we do hardware refresh.

We have two IX of dollars, need three, Josh.

Uh, an answer to that as well. Next, we're gonna go into the actual migration, by the way. So this is where we're really gonna nerd out, and I'm gonna ask you what it's gonna take for me to do this migration to, to a new platform. In fact, I'll hit some of these questions we got in the webinar because they're, they're very specific and I think it's, it's, it's really interesting.

But Josh, what am I going to, am I gonna the CEO and saying, look, we just have to do this. Epic told us to do it. Or is, is there, you know, we heard a couple things there. We heard consolidation skills where consolidation of, of equipment. We heard opportunity for, you know, improved training and those kind of things.

You know, am I gonna be able to, one of the i's things is it's, it's cloud deployable, so aws, Google Cloud platform, Azure. I mean, am I gonna be able to do some, some interesting things there from an availability and other, other standpoint? Well, well I'll be the, uh, I guess I get be the negative Nancy on these, but.

For a couple things. The, the cloud piece is really all dictated on what the public clouds are able to give us for performance today. Right. So I, I can't, I can't move that transactional data store to, you know, to a cloud, you know, too far away. 'cause I have too much latency. Well, even just the number of storage iops that they can pull off for the operational database of Epic.

We're, we're looking at somewhere probably eight to 10,000 concurrent users as your max size client that can fit in a public cloud today. Today. That's, you know, that's increasing every day. But so there's limitations there that we have to be careful of. The consolidation of platforms within the organization that you're supporting, the vendors that you're working with, the support locations that you know, that you're contacting, those are all things that are good benefits of moving to that consolidated compute platform and VMware and those pieces all make a lot of sense to do that.

I think. I don't know that it gets us away from, you know, the way we manage some of these is special systems and they're still highly important. The central, you know, system for all of Epic, for example, runs on that core database, right? So you can still pretty much point to the server that, you know, is epic production in the data center type of thing.

So it's not, it's not abstracting it so far away that it's, you know, not a special server that we're, we take a lot of extra care in. So there's, and there's a lot of things that we have to do that are kind of particular to those environments versus just standard Windows compute or Linux normal compute servers that are sitting in, in the data center.

But I really think the consolidation of Red Hat skills and getting down into VMware and just kind of reducing the number of different things that people have to keep track of and understand, and.

Savings and streamlines the operations. Well, well, thanks Josh. Thanks for for pulling me back down to the Earth. I was really hopeful here that, you know, we were gonna see a public cloud version of Epic sometime in, in the very near future. Actually, we'll get to futures in, in a little bit. I'm gonna come back to that.

Let's talk about the migration. So. You know, there, there's, there's a lot of things to, to take into account, but a lot of it has to do, uh, with the data. And we, we got a bunch of questions on, you know, you know, will an IRIS database mirror with a cache database allowing us, you know, shortened downtime. That was one of the questions we had questions on.

Uh, migrating, you know, what is the data conversion from Power AIX to Linux Red Hat look like? Give us an idea of, of what the, the process is for moving one to the other and some of the considerations. I can kick that one off. So the way I look at it, for the most part is it's almost like a standup of a new Epic environment.

So we basically stand up as much as we can, all the Red Hat environments ready to go. Just as if we were going into a net new epic go live. And that allows us to make sure that we have all the user access and all those things defined as necessary. And then the real piece that comes is, is building a, your first mirror and IRIS and cachet can mirror together.

So that's a, that's a thing, but that'll allow us to mirror across to a Linux based of our database.

Snapshots or clones, whatever you have for storage technology on the backend. And we can use those to create our sub tap. Other environments that are needed that are full size in the environment and there is a conversion. So there is a process of us copying that data over, initially running through a conversion of from big Indian, the little Indian, and that gets us to the point where we can start nearing.

And then from that point it really becomes a bit of a, a strategy game with your. Environment teams to kind of schedule when you can move environments to the new servers and make those final cutovers. And then finally, you know, essentially your prod cutover is almost kind of try to tell people to treat it like a Dr.

Failover. But that's basically you're going through the same steps. You're promoting your mirror to a production and demoting your current production mirror down to a Dr. Mirror member. And if things were to go completely south too, the good news is, is you're able to . Uh, revert back in a similar fashion just by switching out which one's the, uh, primary mirror member.

So it really, for me, it's, it's not really any different in a lot of cases, the how we've used strategies to migrate from physical server to physical server in the, um, power world for the last number of years as well for the operational database. And timing is really driven. I, I feel like we can, you can set up that core Linux environment in a pretty short amount of time.

It's really driven by the environment team and getting those downtime scheduled, the cut environments, not a lot of other technical effort in that mix. So again, questioning from the CIO perspective give you an idea on, so how much, how much testing are we gonna do? How much.

You know, I mean, what, what does that look, I mean, I mean, in.

Are we gonna have to test every aspect of beaker and every aspect of, of, of every module out there? The big things to really hit on are things like printing, making sure that those things are able to, to work, but a full beaker, probably not. 'cause as long as, again, some, some work goes in ahead of time to make sure that just like in a Dr.

Switch, or if we were switching from one server to the other, like most organizations have had to do in the last five years. We've had to kind of abstract away, right? Epic production is known by ADNS alias. Typically, it's one of the things that Epic likes to do too. So we can repoint users and other endpoints to that, that alias, which allows them to communicate without having to go directly to like a known IP address.

And that allows a lot of. Flexibility to, to make that migration with a lot of effort. Most of the time I don't see organizations probably having to do a lot more testing than what we would do during a, a switch out from server to server, just because Epic has done so much work with. Linux and then, you know, past few years, performance testing and everything.

I, we've seen a lot of switches happen without significant issue performance wise or new things crop up because we're on Linux versus another oss. That's fantastic. The, the biggest key part is for the end users, the clinicians, and even to an extent, the application folks that are supporting the applications like Speaker right or ADT, they're gonna have the same exact experience, uh, where there's gonna be major differences.

Gonna be for the ECS team,

operational database. Their experience is gonna change and their, the way that they manage, you know, downtime, standard maintenance is gonna be before be. Take and do a good enough job to run through the migration for your PLC test dev environments. Document everything you know, sometimes blow things up, fail back, make sure you really understand how it works.

It should be just like a regular downtime for your end users when you do the actual full production cutover. I is this, just let it it. Is this any harder than the normal updates we're doing on Epic on a quarterly or now? It seems to be more common to be semi-annual. Is this any more complicated than that?

I mean, I see from a, from ACIO perspective, I see the opportunity. I mean, I can re-platform, you know, we were using their, their HIE product and whatnot. It was all built. It's now all built on top of IRIS. The InterSystems platform was all built on top of IRIS, and that was our, that was our integration engine.

Platform you're using for ensemble? Yeah, we were using it for APIs and all sorts of other stuff. So there's, there's a lot of benefits and opportunities there, I would assume. And so I see the opportunity, but is, is the migration, is the upgrade any more than, you know, just the quarterly, normal quarterly that we we're used to doing?

I, I mean, I think that there's risk associated with anything you do. Right. But I don't see that this is, I. An application upgrade or you know, in parallel to doing their normal quarterly upgrades. I don't see anybody that I've been working with yet, like pausing epic upgrades or anything to be able to take on this migration.

It's just been happening in parallel for the most part. Some of the biggest things from an end user perspective that we see is likely is changes, is a lot of, uh, people on other technologies had like local users and passwords. And maybe synchronized between all the servers, but you know, as they go to Linux, it's all, you know, single username and password backed by active directory type of thing.

Maybe ROS passing directly in, they're logging in. So there's just some changes like that. And depending on how they've structured their old environments, they might break out into more servers. But at the end of the day, that's really just a mapping for their analysts. And those people, not like end clinicians, they don't, they're not gonna notice any difference.

Maybe faster, just 'cause they're a newer tech. But that would be hopefully the only thing they experience. . Well, that's a number more question I hear. Is this really gonna be faster or is there gonna be any performance improvement, cost reduction, that kinda stuff. But let, let's talk about what's next. I guess the question from, from my perspective, again, I'll just go back to this is what, this is essentially what IRIS is saying about their own platform.

And this was back in.

Take their capabilities to the next level. So it's cloud provision, it's an interoperable platform, API management, analytics and backend. Business Inte Intelligence and NLP AI and machine learning. I mean, and these, these were not, hey, we're, you're gonna, you, you do these things in this platform. It's, we're gonna connect to the most common systems that are out there.

What, what is the potential? Where's Epic going with this? Potentially again, and all this stuff's gonna be forward-leaning because it's just your opinion based.

IRIS now gives Epic what is possible where, where could this potentially go over the next 3, 5, 7 years.

So right now, what Epic is essentially doing is taking what they have for Cache for the data tructure and dataset. They're essentially migrating that over the IRIS without any major updates, right? They're treating it as the core database for.

It will be interesting now that ensemble has transitioned away from being ensemble and now it's part of IRIS and it's part of IRIS for Health, where from the research that I've done and looking into what InterSystems itself is doing, if we take Epic outta the picture, if they have built a database platform and data.

It. They're trying to push for a truly integrated data system and data platform for healthcare where portions of your I and data exchange are built in. You know, you have advanced analytics that are running on the system, and if Epic takes that and develops against it. I think it's gonna create some great innovations and it's also gonna take some time because it might involve significant changes to the databases themselves.

And Epic does have their clarity code or reporting environment anyway, right? It's what portions of that might move over to IRIS platform over time. Is it gonna be more efficient to run that on IRIS to actually extract additional data's there? It'll also be interesting to see what happens to ensemble still, it's still fairly widely adopted and it's out there.

You know, not everybody moved over to Corepoint or another system. What are they gonna do when they go through this upgrade? What's gonna be different for them and how is that gonna integrate into Epic? And then what's Epic gonna do with the platform itself in the next couple of years down the road to the next 10 years?

Because Cache was around for a very long time. They had significant improvements on how they were using Cache, and now they're gonna do the same for IRIS. I think it's gonna be up to the healthcare systems to get their hands on some of the development capabilities to see what they can do outta the box.

Besides using as a base database, you know, engine for Epic itself, and if there's gonna be other vendors, other healthcare, innovative vendors that take IRS and start developing it out. Yeah. Josh, where, where do you think Epics going? I think the most near things that we'll probably see as benefits. Epic's been doing a lot of work to try to make things easier on the administrators, um, particularly trying to use Kuper to do more and more, uh, effort even on the operational database side.

So I think some of those API driven benefits might be, uh, first exploited in Kuper to, to help manage the, the database through that tool more. And I. Realtime access to the data because of the fact that we're ETL out to the other CODO environments, some more realtime action and activity in the database to get insights, you know, F point of care or in a quicker, less retro collective way.

I think it's probably where I would see the next couple of things happening, more so than anything else in my point of view. Any, any, any changes to how Dr. Do you think there's gonna be d dr, uh, new options available? I, I don't know. I, I would like to believe that at some point, from an HA perspective, instead of having single operating system and single, um, storage subsystems and all those sorts of things only being available.

In caching mirroring first brought it in to where you could do, uh, synchronous mirroring with AV between two servers and twos separate storage subsystems underneath those things. We do have clients that are using that for their ensemble environments. However, epic has never really been able to make that jump yet.

Maybe IRIS will have enough performance gains to allow us to run something like that in a synchronous where they could always take over for each other and, uh, have more. Resiliency built into the environment. I think that would be one of the first things I would see other than I think Dr. Today will stay fairly similar.

Some, some of the clients that are really drilling it and doing good can fail that environment in 15 minutes or less. So I think that's fairly. Fairly good target to shoot for. Well, fantastic. I, I appreciate Josh, Marty, I appreciate you guys coming on. If people are, uh, interested in, in, in, in, you know, talking to you guys more about this, they can, uh, reach out to their serious rep or, uh, hit serious healthcares.

Healthcare website as well. You can, uh, track these guys down and they can help you with this migration. It's not for the faint of heart. It's, uh, I know if I were ACEO I'd be making phone calls to find people who have done this before and have the experience. It's too, too critical of a move and, and quite frankly, I think the opportunity is, uh, significant to, to replatform.

The fact that we had 900 applications and we had to support multiple operating systems, all that stuff was a cost to me. And it was, it was hard to manage those things. You had to keep special skillsets around and this is an opportunity to really reduce those and, and get much more efficient around that.

So. Marty, Josh, thanks for coming on the show as always, I learned a ton. Thank you. Thanks, bill. What a great discussion. If you know of someone that might benefit from our channel from these kinds of discussions, please forward them a note. They can subscribe on our website this week, health.com, or you can go to wherever you listen to podcasts.

Apple, Google, overcast, that's what I use, Spotify, Stitcher. . We're out there. You can find us. Go ahead and subscribe today or send a note to someone and have them subscribe. We wanna thank our channel sponsors who are investing in our mission to develop the next generation of health IT leaders. Those are VMware, Hillrom, and Starbridge advisors.

Thanks for listening. That's all for now.