This Week Health

Don't forget to subscribe!

July 24, 2024: Tom Stafford (Healthcare CTO at CDW), Chris Akeroyd (Executive VP and CIO of Children’s Health), and Theresa Meadows (SVP and CIO of Cook Children’s) join Bill for a Solution Showcase. They discuss how initial cloud enthusiasm meets real-world challenges, such as data retrieval issues and vendor lock-in, and how these experiences shaped their current strategies. What lessons can be drawn from early cloud adoption missteps, and how do healthcare organizations ensure security and flexibility in their cloud architecture today? As they navigate the shift from traditional data centers to cloud-based solutions, how do they balance innovation with the practical needs of healthcare delivery? 

Key Points:

  • 00:52 Early Cloud Adoption Challenges
  • 02:26 Strategic Cloud Migration
  • 04:47 Cloud Security and Governance
  • 16:25 Staff Transition and Training
  • 22:15 Future of Data Centers and Cloud

Subscribe: This Week Health

Twitter: This Week Health

LinkedIn: 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.

This episode is brought to you by CDW Healthcare. CDW understands how to connect the latest technology solutions and services to advance your healthcare mission while keeping patient and staff at the forefront. Whether you're looking to upgrade your EHR, redesign your infrastructure, improve care delivery at the bedside, or elevate the home health patient experience, Our dedicated team of industry tested experts leverage software, hardware, and comprehensive services to orchestrate cost effective solutions that leads to positive patient experiences, higher quality care, and more efficient clinical workflows.

For more information, check them out at ThisWeekHealth. com slash CDW.

Today on Keynote

(Intro)  the cloud and mobility and the way you can present today is actually keeping our caregivers closer to the patient.

Correct.

Because that's where they need to be,

My name is Bill Russell. I'm a former CIO for a 16 hospital system and creator of This Week Health, where we are dedicated to transforming healthcare one connection at a time. Our keynote show is designed to share conference level value with you every week. Now, let's jump right into the episode.

  (Main) are from a 229 project meeting. We're hanging out with Tom Stafford with CDW, Chris Aykroyd children's Dallas. I'm sorry. You guys kill me with these names.

And you should be in Illinois somewhere. I should be. We're south of Illinois. Teresa Meadows with Children's. Do you get that a lot? Yes.

Every place I go.

really want to talk about the cloud journey. Tom, you and I were talking earlier. There was a time where the forward thinking CIOs were saying cloud first. I remember thinking back then, it's that doesn't sound like it has a lot of logic to it other than there's a new thing out there we're going to go.

But the world's changed in 10 years. Yeah, because that

was like back in 2018 at HIMSS, right? The banners, cloud first. We're all like pumping our fists and we go home and we didn't do anything. That's not what really happened.

But the cloud now has become a strategic asset for most health systems and we're moving there.

I'd love to Teresa, we'll start with you. love to hear what is, what did your cloud journey look like over the last 10 years? I don't know, 5 to 10 years.

Yeah, it's super interesting. Were early adopters of cloud. In 2010, we took, we were one of the first pediatric customers to use Athena Health as our pediatric EHR in the cloud.

And, we learned a lot of hard lessons through that process. And when we decided to leave Athena to go to Epic, We made a decision that we were not, we did not like cloud. Because we had some issues where we weren't able to get all of our data back.

Oh, at the end of the contract. Yes,

and the data that we did get back, we got a million PDFs that then we had to then load

our new instance of the EHR.

And so there was a very bad Taste everyone's mouth about cloud. And so we were probably on the later end of the curve of going to cloud. And really what

So you were on the early end and the later end. And the late.

Yeah, we were both. And what really got us started is Cloud fickle. Yeah, for sure.

Some of it was just naturally being driven there. We knew we were going to Workday, which was a cloud based product. We knew we were going to go to 365. We were early, late adopters of 365 even. We still have on prem Microsoft. Because we were so scared about going to cloud. We're in that, we're in that transformation.

And as we started looking, all of our major platforms were going to cloud. And our strategy was really about, okay, let's be cloud smart about the things that we're selecting. But, if we thought cloud was a choice, it really wasn't. When we were looking at a new ERP, our choices were between two cloud ERPs.

It wasn't something where we could say, oh, we want to do that on prem. There really wasn't an option. And that really got us thinking about, okay, what would be the smart way to do cloud with the things that we have that are remaining. that are not cloud based. And we started the journey early of looking at how do we put at least some EPIC instance in the cloud, primarily for disaster recovery.

So that's really where we've started migrating things that were on prem to off prem or cloud based services. But, we're still going through that. Does it make sense? But almost everything we're purchasing is in the cloud. So It's less of a choice and more of, okay, let's figure out the best strategy and how we do this.

Plus, our data centers aren't the newest from a construction building standpoint. And they're too large. If you look at the amount, like, when we went to, We had 70 servers that we decommissioned out of our data center. So if you can imagine the big gaping hole of space that created, and then you move other things out.

The data center is too big. And it's prime real estate. It's in the basement of our hospital.

The data center is an interesting journey because started with virtualization. Like all of a sudden we were just running along and we were proud of our data centers.

We were putting stuff in. And we'd give tours and we'd be like, look at this is amazing. And then virtualization happened and all of a sudden, for a while there, we weren't investing. Then it started to go back up, and then cloud hit. And, have you built a new data center?

We used to be thinking about when are we going to build our new data center, and I probably haven't done that in a while.

We haven't. We actually got a data center business years ago. We co locate both our primary and secondary data centers.

Yeah, so tell us about your cloud journey.

It began, so we years ago we adopted a SaaS first model, so if the vendor could host it and maintain it, we'd lean that way.

Driving reasons behind that for us was pretty rapid, or hopefully rapid, software releases staying more current where we struggled for a long time moving forward. It gives you some added governance, right? It does. It forces us to adopt change. And so it was a nice message on that side of things.

Say we have to make the upgrade. So it was an organization that got us, into that change mindset. But our move to native cloud Pick your, we're in the two main, we're not in GCP, but the first thought of it was really around scalability. in a huge growth mode. What is it, if we're going to acquire these things rapidly or flex volumes, how can my static infrastructure flex without capital funding, etc.?

So it was a good idea, but then I think it's, I think you said it really was it smart cloud or something? Yeah, it's, it sounds good, but what's the, we don't understand the cost model. We don't understand, and we've learned this more recently and have to really try to embed it. We could technically spin it up, great.

But have we really gotten through the security model? Do we really have the up skilling we need? We're getting better at it, but where we are now is the same place of let's get Epic up there a DR fashion. Let's understand how does it work? Where is the data gravity? Where do you need your presentation layer?

So start answering some of these architectural questions where staff is working through it. The only financial upside I can foresee right now, like the ethics side, is the DR. Can I start getting rid of my secondary colo facility completely? Can I make, can that colo just become AWS or Azure?

Because that plus one is expensive,

right?

Think about all your server, your compute, you've got to power and cooling all that, and by moving at least your DR workloads to the cloud, now you're paying consumption, but You don't always have to run it. And if you're not running it, the cost is minimal. And so at CDW, we're seeing most of our customers, their first move is to do non prod workloads and DR to the cloud.

And you actually can achieve savings, but you can still fail over and, still get an A on your report card for Epic. But you don't always have to run it. And it does give you that flexibility. And I really saw that happen during COVID. Because everybody had to scale so fast, so a lot of more folks moved their workloads and started using things in the cloud because you didn't have to go purchase servers and wait for them to be delivered and rack and stack them and then turn everything on.

Now you literally just

a lot of people will think that's what CDW does. We place our orders, you send us the stuff or whatever. How does CDW, your business model in the cloud?

Obviously, we filled a lot of data centers over the year with the right gear, so you keep running on prem, but, we started seeing the erosion in the data center from SAS based and now cloud native, moving cloud native workloads up there.

About four or five years ago, we acquired a cloud managed service company. And the reason why we did that is we wanted to be able to quickly help our customers move, and if we were going to organically grow that, it would have taken time. So today we still help folks. There's always going to be some level of data center for healthcare, because we do have monolithic software that just doesn't make sense to put in the cloud, and you also have edge systems that you want to keep on prem.

So we'll always have that level, but Now what we're doing is focusing and helping our customers move to the cloud. And as like Chris said, it's really the cloud native stuff. Azure, GCP, and obviously AWS are big areas of focus for us today.

Two new areas that have come up for us is our cloud strategy now to incorporate security.

So our concerns of recovery, but with a lot of the acquisitions of whether it be Citrix or VMware, We need, we got vendor lock in. And the pricing models are becoming unaffordable. So that's part of what we're looking at Cloud, too. How do we offset hypervisor costs? And move to something, be less vendor locked in.

In some cases like a core

infrastructure. And that's all happening. It's, and I think the current issues that we're, that you guys are all struggling with is forcing folks to look at new ways for presentation layers. But, that's all possible in the cloud. Healthcare always seemed to take a little bit longer, right?

Than other verticals. That was very kind of you, actually. Yeah. But we're getting there.

Talk about security you talk about data lock in, you talked about really application lock in, or virtualization, the hypervisor lock in and whatnot that we're looking at. Talk about architecture.

I I think I'll start with you in terms of talking about architecture. It's really interesting because in healthcare, in many industries, we make these decisions that seem pretty wise at the time, And then we end up in a dark alley by ourselves looking at each other going, What just happened? We have to back up and do something a little different.

How can we be more forward thinking in terms of the architecture and keeping in mind that we may not always want to be with that specific cloud vendor or we might need that that ability to move between partners and that's, when I think of the hypervisor space specifically As we sit in these meetings and hear CIOs be like, Oh my gosh, my pricing just went up 40 percent or 50 percent and that kind of stuff.

They feel trapped. How do we make sure we don't get trapped?

One way is obviously to manage your consumption. And what I haven't seen, there is a potential future where you could actually have Azure production and AWS or GCP Azure backup, right? And to be able to flip back and forth and understand your consumption costs and to see Because you get to re skew in the cloud like every six months and it can drive down your pricing.

So if you do have two cloud vendors supporting architecture, you can actually see that and then make the appropriate decision as an IT leader. Because obviously, cost is huge today in every health system. Because we're the only business, when the cost of goods go up, we don't get to change our sell price, right?

That is, that

is truth.

And band aids are more expensive now than they were five years ago. Correct. And all that stuff. And then, like Chris talked about security a second ago. And one of the benefits of moving to native cloud is it's infrastructure as code. So you literally write a software program.

And the great thing about that is you can put policies for security and overall governance directly into that code so you know it's right every time. Because we're all human. And if you have someone, lighting up a server and they forget to do something, you're going to be vulnerable.

You don't even know it. So there is another benefit of that too, is that you can write it once, test it, make sure it works, and that will always be in there. You'll get those sort of security benefits.

So I want to talk to you about your current partners and how they're really drawing you to the cloud with what they do.

Cloud Native is interesting to me. At one point And it will be a long time before healthcare is cloud native. I think we can all agree with that. challenges, or the beautiful thing about cloud native, I remember sitting down with an architect and somebody who wrote code, and he said, look, it's all software.

If they start attacking this, he goes, we just dissolve it. That IP address goes away, we dissolve it. Now they're like, oh, crap, I gotta go over here and attack this one. They attack this one, it dissolves. It's all software. The application is still running because it's just spinning up new things all the time and keeps serving the workloads.

I'm sitting there going, wow, that's great. Why can't we do that over here? He goes, architecture. Like you can't, and I think about it. We have a lot of our partners are now pulling us in. It's almost like AI. Some of it, people are saying to me, it's I'm like, what's your strategy for bringing AI in?

And it's what do you mean, what strategy? It's it's starting to just show up in everything that we have. It's in Office 365. It's just popping in. So talk a little bit about how you're being drawn to the cloud. There's a lot of those partners now who are just saying, it's cloud.

You've talked about this a little bit.

Yeah, the education I've been doing in our organization is, cloud's not a choice. We just have to pick the right partners. Because when you're architecting the solution, so we've been having a lot of discussions about multiple providers and having those experts in those areas help us.

Because we don't have the internal expertise to do it. And even if we did, we couldn't manage the change, because the change is so rapid. And so having multiple providers, especially, so now we're thinking about things like, researchers love Google, so we'll put All our research payload may be

over there and then you know

and put others in other places based on the workload that people need and then we get to see how both things work and determine what the best solution is exactly what Tom said and so that's been a real kind of eye opening experience to really think about it and what's driving it for us is the data and the use of the data and being able to do real time analytics and having it.

Volumes and volumes of data, and being able to move it from one place to another quickly. If you think about MPEEDS, almost all of our patient care is research based. So sometimes I'm doing a protocol, and it's just daily patient care, and sometimes I'm doing it and it's a research. Protocol I want to separate those data payloads in two different areas.

This type of technology will allow us to do that more efficiently, because today we can't. Today we are stuck with co mingling data, having research data and clinical data co mingled. These new things help us move faster. And we're thinking about the architecture, like how would we architect this so that the right data for the right purpose and the right use is in the right place.

But we're going to do it faster than we've ever done it. I mean it's, usually a decision like that would take us ten years to figure out what to do. Now we're in ten months.

It makes it very agile and flexible. you brought up another thing when it comes to data. And we've seen this data driven organizations or research organizations.

So if you move your data to any of the cloud native vendors, you have access to their AI engine, their NLP engine, machine learning engines. To build that on prem, that's an expensive venture. And, you gotta replace that every three years, or four years, right? You can rent the compute and power you need and only use it when you need it, and then even like what Teresa's saying, she can actually decide which cloud vendor she wants to send the data to based on the data.

Yeah.

Especially with genomics, like that's

Yeah, one terabyte per Yeah, there's

no way that we could ever build the infrastructure to do genomics in a way that we would want to be able to do it, and deliver it in real time, quickly, to a physician who's trying to make a decision.

  📍 📍 📍 📍 Hi everyone, I'm Sarah Richardson, president of the 229 Executive Development Community at This Week Health. I'm thrilled to share some exciting news with you. I'm launching a new show on our conference channel called Flourish. In Flourish, we dive into captivating career origin stories, offering insights and inspiration to help you thrive in your own career journey.

Whether you're a health system employee in IT or a partner looking to understand the healthcare landscape better, Flourish has something valuable for you. It's all about gaining perspectives and finding motivation to flourish in your career. .

You can tune in on ThisWeekHealth. com or wherever you listen to podcasts. Stay curious, stay inspired, and keep flourishing. I can't wait for you to join us on this journey.  

How does this look to the end user?

Do they even know it's not in your data center at all? They just

Unless the internet's down. That's the only time. the beauty of

it. If you do it right, your end user is never going to know. At all. They're just still going to their workstation, doing their normal workflows and everything. The

one thing they do

notice, though, is the mobility aspect.

Because now they can have an app on their phone. Oh, that's true. Now they have an app on their iPad. They have a lot more flexibility about where they are and how they do the things that they do. Where before they were locked into a device in the patient room or a device, at the nurse's station or in a dictation or wherever.

Now. They can do a lot of these things almost anywhere, and that's what cloud has helped us be able to do, because in our, trying to connect to our data center that way would not have been successful.

Yeah, and to achieve it would have been extremely expensive. And then, so you actually brought up a good aspect, is that the cloud and mobility and the way you can present today is actually keeping our caregivers closer to the patient.

Correct.

Because that's where they need to be, not behind some desktop typing away all day. Or even like now with all the other products that are out there, they can pretty much do their jobs with a handheld.

In my experience, the people who push back hardest on moving to the cloud is the IT staff.

That's true. Because it's a new skill set, a bunch of unknown And potentially some things that they haven't done before. I think a lot of times they're pleasantly surprised. It's hey, if you've SQL server here, not all that different. You're moving up the stack a little bit, but not all that different than what you're doing over here.

And you get out of this, like you said earlier, purchase a server, wait for the server, install, configure the server, all that stuff. You just spin it up and away you go. Talk about transitioning your staff and what that looks like, Chris.

You're right. Years ago, when we started doing this big SaaS push, our ERP, when we moved it to hosted platform, there was panic.

My job's going to be gone. What am I going to do with it? It never happened, right? We never materialized. That helped, but a little bit of the same resistance as we moved towards native cloud. The opportunity now, and I think where people are really getting excited, is the opportunity to learn something new, something different.

So you're innovative people, you're ones that want to learn on your staff, have really embraced it. They're all in, they're up skilling, they're figuring out. We're providing as much training as we can. I can't Over emphasize the need for training anymore, but it's untrained person doing this is going to cause you headache and heartache.

Leave something sitting wide open or. actually had somebody on a non prod I think it was Azure Desktop, left something exposed to the Internet within five minutes.

Oh yeah.

It was hit. Oh really? That quick? Yeah. wasn't our production tenant, but it's okay. That just emphasizes you have to really have the right governance, the right oversight.

But it's the things they can do to spin up. It's really created some positive energy. And we're learning.

For those who are really into it, they just, oh my gosh, you're giving me all these new tools. And look at this automation we can do. a software stack.

Now all of a sudden, one of the attendees in the other room was talking about, has 45, 000 automations and essentially they've automated as much as they can think about. They're just constantly automating it. And He said it's, literally it's, keyboard and away you go and you're ready to go.

I know I'm making it sound easier, but for the staff who wants to, it's a really cool sandbox.

It really is. And that's a little bit of the challenge we're having now. We're having to pull the reins back a little bit because they've got into the software or the infrastructure's code mindset that cloud brings.

Not everything in our environment is going to be cloud based. We're going to still have on prem. What we're struggling now is I've got a group of people that want to do infrastructure as code in our private data center, our private cloud. Hold on. That's a lot of skill set we don't have. You can put

yourself at risk.

Yeah, because you're going to get, what, five, six people that can really do it well and the rest don't. And so now how are you going to manage that? So at CDW,

one of our goals is when we're helping our customers out is to help them with upscaling. still have all the knowledge of the infrastructure and architecture, but you have to know how to write code.

That's the new skill set, right? And so if you have a DevOps team in your organization, they're actually best suited to be your cloud engineers because it is literally writing code. But we do take our time and help our customers with the upscaling and training so eventually they can take over managing their own infrastructure up in cloud native.

it's been an interesting journey moving to the cloud. I'm excited about some of the things we can do in health care but I'm also a little concerned and, this is one of the things I about having the conversation with CDW here is we do need good architecture in order to make sure that we don't end up in these, narrow alleys, essentially.

And we can move around. But there are organizations, the two of them I've talked about, they haven't moved their EHR to the cloud yet. But others have, obviously.

Yeah, then like the pathway we're seeing, I think we're helping about 50 customers right now in healthcare. But everybody starts with DR.

Because that's actually real savings. As we talked about before. And then you can use those savings to start moving prod. So you can net your cost out. But what you get is what we talked about, right? Security profiles that are solid are going to work. You get the agility and flexibility.

You get access to these. AI and ML and NLP engines up in the cloud to help you do your job. And healthcare's, we're becoming more data driven every day. We're always financially driven, but again, I think COVID made the other executives and healthcare organizations realize wow, this clinical data can actually help us make the right decisions, right?

As data keeps flowing, that's going to keep driving you more to the cloud.

back in the day, we had people telling us we were gonna save money by moving to the cloud. . Remember those? Oh yeah. You remember those. We have wore gray hair than you guys

I know Miss Claral

But but back in the day, people would come in and say, look, you could save money moving into the cloud. for me, that didn't materialized at in my time as a CIO. I don't know if that has started to materialize

yet. No, I think for us it's been. Mostly budget neutral.

So it's mostly about agility and access to new kinds of tool sets and AI.

And scaling. Like you

said before, you can scale up quickly.

Where it might have saved us money, and I haven't really studied it, but having resources without adding resources to our team.

Right.

there might be cost savings there. Because if I had to build a whole infrastructure team to do what we can do in the cloud locally, I would need three times the number of people.

If I really sat down and kinda did the math, I probably could figure out that it is saving money because I'm not asking you for ten more infrastructure people or

Because you're not managing the bare metal. Correct. Correct. That exists somewhere in the cloud, right? Or the power in the you're managing a software program that's building your infrastructure for you.

Let me ask you the exit question. So the exit question for this interview is, if I'm walking in your data center five years from now, I walk in your data center today, I walk into it five years from now, does it look different? And how does it look different?

Yes.

Yes, it looks different.

It does. We're actually designing one now, a data center.

And it's being primarily designed for two reasons. One, just the age of our current data center and the locale because we've built around it and now we have location challenges. It's

a hospital data center, so it's in a bad place. Yes, it's in a

terrible place. So we're building a new one, but it's probably a fourth of the size of our current one.

And the primary purpose is running medical devices. Some telephony, things that have to have compute close to it. So

just the basics.

Just the basics. All the other things are being designed out to the cloud.

So an already old data center. The new data center you're building is a quarter of the size.

That's telling.

And it's still too big.

And you're already in two Colo data centers.

We're in two Colo. I think in five years, if I get my way, you'll see me with one Colo. How will you do continuity and

recovery?

be able to flip DR, etc. over cloud provider.

Oh, to, oh, okay. Yeah, so the inter second

call is going to be Second call will be AWS, for example. That makes sense. And many years ago, we, because we were being pushed by our board when it was a big headline of cloud saves money, go cloud. We really took a hard look at it and said, okay, where if any savings is there?

And it's Architecture, I think, I felt was their approach. So we, got rid of our three tier architecture back then. We went heavily hyper converged five years ago, maybe. Huge cost savings for us. And so I think if you look at us today, we'll look very much the same in five years. And the density, we run the most of our organization on four racks.

Isn't that incredible?

Isn't that incredible? That is incredible. Fifteen years ago, yeah, it just, you guys are saying the data center is just going to keep eroding and shrinking. Yes. That's what SAS and Cloud Native.

My footprint right now is 8 total racks in my data center running over almost 4, 000 servers, supporting 15, 000 employees.

And half of those racks are the virtual desktops we run.

Right.

So imagine, 4, 001 or 2U servers. That's a lot of space. That's a lot of heat in one place too. How does

CDW's business change over, the same kind of thing, if I walked into CDW today and you're primarily working with them on five years ago it would have been you working on their existing data centers, but now Five years from now, it's probably a significant cloud conversation and convergence conversation in architecture.

Oh yeah, and we're ahead of the

game, yeah. And really shifting away from shipping out that bare metal infrastructure into CW providing services to help our customers move and maintain in the cloud, essentially.

Awesome. I want to thank you guys. Thank you. Appreciate it. It's been great.  

Thanks for listening . If you found value, share it with a peer. It's a great chance to discuss and in some cases start a mentoring relationship. One way you can support the show is to subscribe and leave us a rating. it if you could do that. Thanks for listening. That's all for now..

Thank You to Our Show Partners

Our Shows

Related Content

1 2 3 289
Healthcare Transformation Powered by Community

© Copyright 2024 Health Lyrics All rights reserved