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

(Intro)

We're all going to have to mature to have a comfort in a gray zone, understanding your best ETA, Bill, if you're running our hyperscaler. I think we'll be up in two hours, and at three hours, I'm not saying, Bill what the heck. You're saying, I gave you my best estimate, now it's not that,

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) Alright, it is Keynote and today we are joined by Dr. Andrew Rosenberg, CIO for Michigan Medicine. Andrew, it is always a pleasure to have you on the show. Yeah, you too, Bill. Thanks. looking at your study. By the way, it is the perfect study. It should be like, if we were going to rebuild the game Clue, that's what the study would look like.

Yeah. Yeah. I in fact just two weeks ago, I gave a very big lecture to School of Public Health, Clinical Informatics. I was asked to give an AI lecture on AI and health to undergrads at the University of Michigan.

I thought that's a pretty broad topic. So I had to divide the topic and do, let's just define health. and healthcare, defined AI, and then I very quickly got into just example after example with lots of pictures, and it was great. It was an hour and a half lecture, and the questions were really great, as you would expect from undergrads, but I still teach, I I stopped practicing critical care about a year and a half ago, and I miss that personally, deeply, but I still will go up in the ICU and Do a sort of senior guest rounds and I'm also part of our critical care journal club.

So I stay very involved in that.

But at some point, they just look at it and go man, there's just so many demands for your time. It's sometimes hard to keep that up. Yeah, the best who

I Think. Two people I know perhaps better than some of the other physician CIOs is Jeff Farenti at Duke and Mike Pfeffer at Stanford.

They both are still practicing. Jeff, in particular, is a neonatologist. And what they'll say, and it makes a lot of sense, they say they have great teams, they have incredible trust, so the reality is they can delegate day to day work and it works. And For me, I needed to do five to six weeks of critical care to be competent in our advanced cardiac surgical ICU.

And I just couldn't find five or six weeks in addition to everything else I wanted to do. So I had to make that separation, but they seem to do it well. Maybe they're just, incredibly better doctors and CIOs than I am.

Give us an idea of the journey. think anytime you're putting a building in, that's top of mind. But anytime you're doing a consolidation, that's top of mind. I'm not sure what's top of mind. What do you focus on? What does it look like bringing health systems together and building a new building?

Yeah, I think what's going on at Michigan is not uncommon at the large health systems right now. We're seeing probably about a 10 year trend, and I think we're starting to see the near end of this massive wave of consolidations across U. S. healthcare. And we at Michigan Medicine are certainly part of that.

Or,

systems within systems are rationalizing and integrating or consolidating along strategic applications and platforms and infrastructure. That's been the work that is shaping up what the current CIOs have been doing. And we are deeply in that work right now.

interviewed your peer over at Corwell.

So Jason Joseph was on. I And we spent a lot of time on this topic because he was right in the center of bringing those two institutions together. And it was interesting. It was very intentional. It was to say from the get go, it's intentional to say we are a new company. We're not going to take the old names of either of the two companies.

We're going to become a new company. And that permeated their thought process. We're going to get to a single instance of the EHR. We're going to get to a single ERP. But one of the things he said to me was, man, it took a lot of time. It took longer than I think anyone anticipated to consolidate all those things.

It's a really good example, and I'm trying to think of examples that other people would want to hear about or know about so that we can learn from each other. Jason and I are actually quite close. Here's the irony, I'm speaking fairly unfiltered, so there's a little bit of vulnerability here, if you will.

is perhaps one of our most, what you would call biggest competitors in the healthcare market in the state of Michigan. I think we're doing our consolidations almost as diametrically opposed as one can find, and I talk to Jason a lot, and I don't mean that in a snarky way. I think that there's no one right way, but I tend to use Jason's teachings to me, the examples of Corwell to try to help inform what we're doing.

So it would be a wonderful business case report, 5 or 10 years down the road, to take a look at two premier health systems in the same state, look at what they did, and then look at the outcomes that were achieved, the ROI, the total cost of ownership, the work efficiency, whatever metrics one might choose, it would be a wonderful business case study that the U of M Ross Business School should do.

Yeah, it's, it really is fascinating. I remember when Epic came in to St. Joe's and had a conversation with us and essentially told us we weren't ready. Really, it was a, our culture wasn't ready to make the transition. And I appreciated them laying it out. Like you're not willing to make these decisions or move forward.

How are you? How are you viewing the of the CIO today, how are you seeing that integration of AI, digital, and whatnot, or , are these different disciplines now?

When you've been sitting in a seat long enough, you start to have different opinions about the exact same question.

I remember vividly 10 years ago, scoffing at the idea of digital. I thought it was a platitude or as a marketing, or it was meant to be sound really cool, but it didn't land well when we were all just struggling to get.

Now, I think the word digital is exactly the right word. I think it encompasses a series of both. pragmatic, practical, concrete, and nuanced issues that we deal with. And you mentioned a few. I think the modern leader of technology for a large health system should be a chief digital Because technology, and especially IT within technology, is only a component of where the business needs to change the way it works, and how to shift the way it works.

They have to understand the business. They have to understand prior auth and where. Payers are going with the speed they're using ML and AI, and where the work has to change, it's not about removing jobs, it's about shifting work. AI fits into that, but AI, I think many of us would agree, is a layer that cuts across all of it.

It's not this individual, unique thing, and if anyone tries to unitize it and make it its own division, it will fail in the same way that I would argue today. If anyone tries to hold on to IT, how many times have we all come across someone who just bought a SaaS with their credit card and are working with it?

Some are in organizations that accept it right away. Some of us are trying to articulate it, what it should be. But I do think the Chief Digital Officer is the right role going forward where technology is enabling things. And one might argue, ironically, that we go back to having a CTO run the IT portfolio of IT services and products and infrastructure, especially now with more and more cloud work that I hope we can talk about.

That's what the CDO should be doing.

We had this conversation a little while ago when Ascension hired. their CDO and the CIO reported into the CDO. And it's interesting because I think that the CDO was more of a chief strategy officer, more of a chief business strategy officer is how it was described to me.

And that's almost what you're describing here. It's somebody who really has a deep understanding of the technology, but that's not their focus. Their focus is on the operation, on enhancing. The use of the technology and utilizing technology to enhance the outcome of the organization, either quality metrics, financial metrics, efficiency metrics, whatever they happen to be.

Or what was it, healthcare, IT news the paper. Okay. And you would see these titles sometimes even on the cover, how to be a strategic CIO. Right there. You hear this conversation. I remember years ago, I hope Janet Guptil sees this. We were at Scottsdale Institute. I was A new CIO, I think, yeah.

And there was a really good panel of chief innovation officers, and they were really good, they were impressive. And the instant whoever was leading the panel said, okay, any questions? I just bolted up to the microphone and Andrew Rosenberg, Michigan medicine. And I was being a little snarky, but I said, I think the CIO.

related, but not necessarily perfectly aligned work. So right now, whether it's the chief AI officer, the chief innovation officer, the chief disrupt the chief transformation officer, the chief strategy officer, the chief analytics officer, the chief data officer, All of those roles are overlapping.

And one could say, yeah, roll it all up to a chief digital officer. I don't know of many places where that would happen. Because it's more of a political reflection of an organization and its culture, not any perfectly aligned one. But from my point of view, because the formal technology in many ways is becoming more simplified it's similar to the point of whoever would build their own iPad today.

really is looking at contemporary and modern methods to achieve infrastructure, services, and products. But the Chief Digital Officer should not be a strategy officer. I think it's very deliberate. The Chief Digital Officer should have a really good understanding of how the place works. It's almost a combination of a Chief Operating Officer and a traditional CIO together.

But, with the lens of where digital and technology enablement will help new work. That's where I think this role is.

And we're adopting these new models. Are you guys thinking about the cloud? The last we spent a lot of time on the cloud, the last interview I did with you, because you guys were right on the cusp of that journey.

Yeah, so we've done some really cool things. We're by no means the first, or even in the absolute top five percent, but we're very quickly be in the next five percent, and at the scale we're doing some massive stuff.

And within two years, maybe even faster, ideally faster, we will be out of our secondary data center entirely. We'll be moving sector to the cloud. And we'll be moving full Epic onto Azure, full Epic. At the scale and at the size that we're doing, all of that prep work I've talked about before, we have all of the.

Now steps moving forward. We're in final contracting with Azure and the system integrators in order to do that work. We're working with AHEAD and Optimum. Excellent partners to do this work. And that's not me shilling for them. That's me giving specific details of what we talked about.

I have the same mindset now, and that's still in many ways, while we're going to do some cool, innovative things, I still think we're still just talking about run And systems of record, to be frank, so that we can use Fabric to do really cool precision health and connect with many more groups so that we can bring in our affiliates and partners into this complex system of consolidated work so that we can actually do it cost efficiently.

I was really skeptical about the cost of cloud. Now, I'm much more reassured that it's probably going to be neutral at worst, and I think we have opportunities to distinctly show where we can add savings, whereas many of my peers, and I talk to them a lot, have said, I don't necessarily see the ROI yet.

We have. Upskilling of our talent we have so many pieces to do, but we're moving at pace now. It's not cloud first, cloud smart, it's to me contemporary strategic enterprise platforms in order to do all the new work that we all know we need to be doing and to do it as efficiently as we can.

give me a learning from this, something that as you were moving into it, you recognized that maybe you didn't know as you were going into it.

It's such a, it's so obvious now, I can't believe I missed it, but it is.

This is that, really holding back on trying to not throw out Old, tired cliches and platitudes, but one that immediately came to mind is this is a great question that shows benefiting from people like you and others where we learn and we stand on the shoulders of giants is the cliche, but it came to mind.

We started three years ago telling our people what we were going to do repeatedly. I've been in front of the entire IT group. Half a dozen times easily saying, this is what we're planning to do. This is where the future is going. Here are all sorts of cloud education certifications that you can do.

But, some of our staff said we want to wait to see which one before we get into the training. And I said, do the training and do the training. And some did not avail themselves of it. It's not a big problem, because now they are. They were the people who said, no, I want to wait. Is it this one or that one?

No, it's this one. Great. I'll start working on that. That's actually not going to be our problem. People are actually excited and , I don't want to dismiss or under emphasize the importance of what you brought out. We just did that one reasonably to very well. Not that we don't have people with angst and worry.

Our challenge is how do we do that and all the stuff we need to do. It would be a very, tactical issue, not a strategic or existential at all. The learnings, I would say, would be two. And partly, we're not there yet enough for me to really feel confident yet, but I think I've heard this from our vendor partners enough that I think it's true.

We did an unusually detailed service design document with an enormous amount of work that we did with Deloitte and NetApps. To do app server mapping for exactly what are our workloads that we're looking to move into the hyperscale or not, we have Epic and we have.

Our system integrators have a really good idea of exactly what we're looking to not the general. The amount of effort we put in, and we will publish, on what it took to get those levels of detail slowed down our process by probably a year from what I thought it would have been. I thought we would have been where we are now a year ago, but I suspect that level of pre work will pay off.

The GRAFs, things like that, the latency issues with our networking. The challenge will be our 150 other associated applications. We don't have those white papers. We don't necessarily know where all of the plugins work in that large ecosystem, that's an example talking to Jason. Corwell did that super well.

We'll learn from them. Those would be two immediate, and then the third would be what is the local business driver to be doing this work? I've struggled with that. Because I did not lead with, oh, we want to save money, so we're going to move to the cloud. I think that's potentially absolutely wrong.

Do we want to spend X number of millions of dollars to upgrade HVAC, electrical, and even then, not have the electric, the power and size to run, that others would say, we, we don't have that issue. Our primary data center is not that case. So each organization is going to have to look at the variety of business use cases, business speed, we don't really need business speed that a startup software company would need to use cloud.

You have to find those that really matter for your organization, where the CEO, the CFO, the regents, whomever can see the value. Ours saw the value of getting out of a old data center from a business continuity point of view and to some degree a modernity point, contemporary IT work point of view.

Yeah, you haven't lived until you're sitting in front of the board having a conversation about needing a new data center and all the equipment in the data center. And essentially they say, what are we going to get when we're done? And you're describing it and the board member looks at you and goes, So you're telling me we're getting a new roof.

Yeah, when you're done, it's going to run your systems and all that, it's, so that's challenging. I do want to hit the business continuity. It's been interesting. We recognize this right out of the get go. We were in Southern California, so we had a lot of conversations about if we move things to the cloud, if we move these workloads to the cloud.

A few thoughts come to mind. One is I don't know where the also the old adage of you are never going to get fired hiring IBM fits into this question. I think this is new enough work that I really don't have. Certainly not personal experience, but knowing a lot of our colleagues around the country, I don't know that many where I would point to say they have a lot of experience at the scale we're going to move to know what it means to have outages in a hyperscaler predominant architecture versus an on prem.

I said, pay attention to this one. Our workloads were not so disrupted by it, but I said pay attention to how fast AWS corrects this. Because it's not going to be that we're going to have the outages. What it really is going to be is how fast they correct it versus how fast we would. And I think it's going to be very important because As more and more we get to these dependencies on not us, and you see it with SAS right now, you see it with others, I think the key is how deep do we investigate those firms to see their continuity and their response rate.

Because One could just look at several of our wonderful colleagues who went through hell when they got truly locked down with ransomware, and publicly, they're down for six to eight weeks. Privately, they're really not functional for a half a year. The costs are enormous. So I think the first thing when it comes to business continuity is to say, here's our due diligence around not just our strategic vendors, but here are Redundant architectures WAN architecture that I'm talking about will be much better than the one we currently have even with on prem.

And what we've had some outages that have been due to the fact that we still have some single threading in our region, in our

The physical network.

And then the second one, to the cloud, the challenge that I think we're going to have is, when we had the most recent CrowdStrike outage, Within a few minutes, I was able to talk to people who were in our data center. We know what it is, and we are physically fixing it. Our ETA to this recovery and then that level of recovery will be about an hour and a half to three hours, and I had just perfect confidence.

We don't have that level of communication with our hyperscalers when we have an outage, like we do locally, and that's going to be a change that we're all going to mature around. How do we get notified about those? Microsoft Variable ETAs versus, we're not telling you anything, client, until we have absolute assuredness because we don't want to be perceived as giving you bad information.

Yeah I do remember, I don't know if it was the same outage, but Amazon had an S3 outage on their storage, their block storage had issues. And I remember following that, it was really interesting. They posted this stuff on the website almost immediately hey, we know what it is, here's the issue.

Here's how we're re architecting it. Here's when it will come back up. And I remember at the time thinking, that's the model, like that's the communication model for, they have to let the world know. They have to let Netflix users know. It's

Right. And that's my point.

An example, which is unrelated, but I still think a good one. When Epic was starting to really get a lot of clients, and we'd start to say, what about this, what about that? I remember having a conversation with Carl and Sumit years ago. We were looking at UCC, we were looking at another vendor, great vendor, we were all going to be happy.

We were within days of finalizing that contract. And then UGM comes out and Epic says, oh, we're going to UCC. And I said, why aren't you telling us? Where your roadmaps are. And the answer was understandable at the time, but I think they've matured enormously since then. If we don't meet our deadlines, if we don't meet their expectations, we have clients who are really unhappy with us.

I think that's an example of a co maturation. I suspect as we go more and more hosted. Whether it's hosted, oh by the way, Epic does a great job hosting too. As we mature, I think, to your point about business continuity, the hyperscalers are going to have to mature to a level of current discomfort to say, here's our best estimate, but we're not going to be held to it, but we're giving our best effort. And thank you. That's enough because that's what we did locally.

Yeah. It's managing expectations. Yeah. Andrew, the last time we talked we ended up doing a two part series cause you and I can talk. both talkers and we We love this topic. And there's so much more to talk about. I didn't even get to broach the future with you. And I love our conversations about the future. We will have to do this again Early next year, I'd love to circle back we're just at the start of our normal conversations

It was really good. I was impressed.

Yeah, and I understand why they had to do it in two parts, my daughter's a music theater and my whole family's a big music theater.

Here's the teaser. The future is all about three things that come together. And I'm really spending more and more of my time, and when I go back to being an academic professor, I will be doing this even more. New data types. Think multimodal data. Sensor, genomic, pharmacogenomic data. True big data. The new analytics and data management needed to manage that type of data But interestingly, the infrastructure and technological enablement to make those two work shine that lens on problem areas of healthcare.

And that's where we'll see new advances in healthcare where that digital mindset will be working together to help that happen. That's the topic I would love to talk about.

Yeah. Thanks Bill. Appreciate it.

Thanks for listening to this week's keynote. 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..