Welcome to Ed Scoop's podcast series on IT Modernization in Higher Education, under it this week by Microsoft. I'm your host Wyatt Cash, and in this week's episode, we'll take a look behind the scenes at how a leading university medical research center made the move to cloud computing, and we'll look at how that's helping researchers accomplish their work faster while also reducing compliance and security risks. Here to talk about those lessons today is Casey Frankenberger, Chief Research Informatics Officer at Rush University Medical Center in Chicago. Casey plays a leading role in facilitating institutional research at Rush University for the benefit of the university, as well as commercial, government, and nonprofit partners.
He played an instrumental role in moving the institution's research platforms to the cloud. Casey, thanks for joining us, and welcome to the program. Thanks Wyatt, great to be here. Casey, let's start by having you just briefly describe the nature and scope of research activities at Rush University's medical division.
Yeah, thanks. You know, essentially at Rush, in terms of research informatics, we touch a wide variety of technologies and platforms and individual investigators across the university. Everything from physician scientists who is practicing, but also finds ways to improve care and wants to look at, let's say, medical data on a cohort of patients, things like here in the time of COVID, we manage annotating all the data in our biorepository for COVID, and also making data available for feasibility of research. So just understanding, you know, anybody that wants to do a project, do we have the information of the data available for you to really complete that project in the scale that they're interested in as well?
We also manage data governance. So we've just established a data governance committee to really make sure that we're patient, first mentality, or actually making sure that everybody's got an informed consent, things along those lines. So really it's access to data processing, data analytics, as well as data governance for the institution and two of the vectors that that fits under as research as well as innovation. So we think about commercial partners as innovation.
We work with the University of Chicago and around 13 other organizations across the Chicago land area. We have dozens of projects where we're doing all kinds of analytics on data and working on sharing the identified data for those projects. So Casey, prior to moving to a cloud computing environment, what were the biggest challenges that researchers and medical staffs faced in conducting their work? And then how did your IT department help support them in that transition?
Really prior to the cloud and prior to COVID last year, we were running almost all of our research activities off of a single research server that we had to constantly perform administrative tasks in terms of pulling data off and archiving the data to do new projects. And this was a constant struggle. So, you know, moving from there into the cloud, everything in the cloud is scalable and on demand. So we can now get access to data through our clinical data warehouse.
We can de-identify that data and then we can make it available on the cloud for either creating a data mart, just creating access to data so that people can analyze it themselves. Or in addition, we can share that data with other institutions for collaborative projects. Take us back a little bit in time. What was the tipping point that led you to say it's time to move to the cloud?
Yeah. So I actually came from a background. I was at the University of Chicago before I was at Rush and they definitely had a lot of virtualization and cloud computing there. I worked in industry also prior to coming to Rush in this role.
And we did basically everything in the public cloud at that organization, you know, coming to Rush and just seeing what the pain points were, really getting access to data and then getting access to manipulate the data with compute in an orderly manner. And you know, with the cloud, we can provide that sort of diverse array of resources and data and package it up to each individual investigator's needs. Talk us through what were the biggest fears or concerns that you encountered with your management about moving to the cloud or maybe even your researchers. And then what helped win the confidence that your organization to make the move?
Yeah. Well, so the biggest concerns are largely around security and perception of data governance. And so, you know, one of the big steps that we did was set up a data governance committee with all of the essential data leadership across the organization. So people from research administration, people from legal, people from compliance, people from cybersecurity.
So they all help essentially ensure that the data has been through the proper review processes, things like IRB. So a big hurdle was just simply having them see that we have a process in place. The process is audible and well documented. Then transitioning over into the cloud, we've done it basically on some of the leading investigators, people that are already doing some of these activities, you know, whether they're with collaborators at other institutions or, you know, doing sort of really progressive things inside of that research server that I mentioned.
So now they're moving into the Azure natural language processing pipelines, they're moving all that data over into data marks in the cloud. And frankly, what we're just seeing is these economies of scale or, you know, reproducible frameworks that if one person has used it, another person feels confident that they've seen that they can use it. I think the third thing, it's that confidence that we're working within the framework of cybersecurity and compliance, it's seeing leaders perform these tasks at Rush. And then I think additionally, it's just coming back to really understanding how we can deliver on democratizing access to all that data, making it so that other people are using those reusable data sets.
And we're really seeing like a big uptick in the usage across the university, you know, with researchers. So it's very exciting. Sounds like it. So we're going to talk about how we can make the data more accessible to the cloud providers out there.
How did you go about selecting a cloud provider? And what were the biggest issues you faced in actually transitioning to the cloud? Yeah. So we're a multi cloud environment.
We use all of the different providers. I think what we found with Microsoft is just that there's some organizational experience in regulated spaces. So we've had a relationship with Microsoft and Azure and essentially they, as part of our agreement, have given access to some cloud architects. And then in the interim, we also hired some key staff.
So we hired a chief cloud officer. We hired a cloud architect. And we hired a database specialist. And we hired a team that was able to do all the projects, deterministic in their operationalization.
So we know what the right steps are. We have the people that can do these various tasks. And I think that that's just made it a lot more predictable as to how long things will take, you know, what are the specific resources required and that everything's well documented that just provides a lot more confidence for the organization. And as you think about it, how long did it take you to make the transition to where you are today?
We were really fortuitous in that we began the transition to cloud literally last January. So right before COVID hit and it just forced, it was a forcing function. We found that the team was just as productive working from home once we got a few things in place. Everybody had their equipment for their home offices.
But at that point, it really became the only solution. And everybody was actually as productive or more productive working from home. And our capacities just continue to expand. So there's really no sacrifice.
And for the foreseeable future, we continue to be remote even though the entire team has been vaccinated. And many people are already back to work at Rush. That's a great story and nothing like timing. Tell us a little or give us some examples of the kind of research and analytics work that your researchers are able to accomplish now using Microsoft Azure platform and perhaps how that's also benefiting the larger community that Rush University Medical Center serves.
Yeah. Well, first, there's a few different data marks that were set up. There's a few different data marks that were setting up. There's some partnerships with life science companies.
There's some organizational collaborations across various institutions. We're collaborating with public health. We've built an analytics hub for all the data in Chicago. And then we've also built a lab component to that project called the Regional Innovation Public Health Labs.
So both of those are the city of Chicago and focused around COVID. But one project that we're working on right now that I'm really excited about is a data commons. So if you look at the NIH, they've really charted a course to have these things called Gen3 data commons that house data and then they have a searching function inside of it. They allow self service analytics in an environment that's all very secure and it's a framework that's called fair.
It's findable, accessible, interoperable and reusable. And so with the data commons, we've been working with the Microsoft team to construct that inside of our Azure space. And it's eventually, you know, over the next three to six months, going to house data on a, I think, around a million patients. It'll all be de-identified for internal use.
It'll have things like medical history. It'll have genomics data. And it'll also have imaging data. And what that enables is sort of an AI ready framework.
So if people want to run machine learning algorithms or deep learning algorithms across sort of multi-modal data at Rush, we will be well positioned to do that. And we just see consistent outreach from both commercial and academic collaborators asking for those kind of participation, especially in the area of COVID, things like image analysis relative to the flu or COVID analysis where, you know, looking at which social determinants of health most impact patients. And so, you know, we even have some algorithms that on the point of patient presentation at Rush, we have a pretty good chance of understanding whether they'll be on a ventilator if they've tested positive for COVID. So just many ways in which to enhance diagnostics improve clinical action ability and even just do some basic translational research from biospecimens and our biorepositors.
Very impressive. Finally, Casey, what recommendations would you give to your peers about moving to a cloud platform? And given the budget issues that many higher education institutions are facing, how might they get started in light of that? Yeah.
So one of the things when I joined Rush was that we had a fairly significant budget. It was over a million dollars for the team that I managed, which is essentially a team of statisticians, bioinformaticians, and clinical analysts. And we were able to establish a chargeback mechanism and reliable services. So what that's allowed us to do is just to continue to expand that team.
And part of what enabled that was really putting together a strategic plan. So I'd say one of the most important pieces of all of this is putting together a clear strategic plan and then running it by the essential stakeholders in our ecosystem. So research affairs, cyber security, legal compliance, and then scientific leaders and physician leaders at the organization. So making sure that we have that buy-in and that everybody has sort of had a say in the discussion.
And then from there, honestly, this is the best solution because what we define at research and innovation at Rush is largely things that are not being paid for by operations. So if we have commercial partners that want to do projects or if we want to have academic partners, having a way in which to have very line item billing for individual projects or data marks that are associated with projects and then slowly bringing on the staff to manage these things. So like I mentioned, hiring a cloud architect, hiring a chief cloud officer, and then hiring these more specific roles like a database specialist. And what I found, it was interesting is that as soon as we hired those roles, they were completely booked.
So those people's time is booked for the next year and a half. I just thought that was pretty amazing because it was a struggle to get the budget allocated for them and as soon as the budget was allocated, they actually all of their time was bought from grant funding. A lot of it is due to the cloud. Like we don't have to invest a million dollars to buy a high performance cluster that people will use 25% of the time.
Like we have the advantage at Rush of building to the future instead of supporting our legacy infrastructure. So it worked out really well. Some very interesting and insightful points. I'm afraid that we're going to have to leave it there for today.
Casey Frankenberger, thanks so much for joining us to talk about how Rush University Medical Center moved to the cloud and how that's benefiting your researchers, the university, and also the communities that your institution serves. Thank you very much, Wyatt, for the time to talk today. I really appreciate the work that we're doing with Microsoft Azure Team. We want to thank Microsoft as well for underwriting today's episode.
Look for more of our coverage of IT Modernization in Higher Education on EdScoot.com and subscribe to our EdScoot radio channel on Apple Podcasts, Spotify, Google Play Stitcher, and TuneIn. This is why I cash thanks for tuning in.