Embracing Cloud Solutions in Healthcare
The discussion between Narinder Singh and Eric Yablonka highlights a persistent challenge in healthcare technology: the desire of some organizations to host software on-premises instead of embracing cloud solutions. While traditional data centers offer a sense of control and security, this mindset can impede flexibility and innovation. Yablonka points out that many organizations are beginning to recognize the benefits of cloud technology, including improved security and scalability. However, concerns about costs and the risk of stifling innovation remain. The future of healthcare technology hinges on adopting cloud capabilities while addressing these concerns to create a more efficient and interconnected ecosystem.
Listen to Narinder Singh and Eric Yablonka discuss how the hesitation to embrace cloud solutions in healthcare often stems from concerns about data control and costs, despite the significant advantages in flexibility, security, and innovation that cloud technology offers.
Video Transcript
Narinder Singh:
And I want you to give me, we’re going to try to understand the other side of the story and the good reasons or the bad reasons why this story may have existed. We go to a hospital system and they say, Hey, we really like this. It seems very compelling. We love the flexibility. Do you have the option of running it in my location? And I look at that and say, this feels like 2007 where people were asking cloud computing companies to run their software inside of the system. And I want to answer that question without going through. The reason that flexibility exists is because it’s a cloud solution. There’s no major enterprise vendor today that doesn’t do cloud solutions that would offer you that Salesforce doesn’t allow you to install the system and that’s good for you. It’s good for things like security which are paramount in the industry today. Why do you think it is good or bad reasons that we still get those kinds of questions? And I think it’s why sometimes folks from outside of the industry look at it as like, Hey, this does feel like I’m moving back a decade or a decade and a half in time in terms of the thinking,
Eric Yablonka:
That’s a great question and I’m not sure there’s an easy answer. I would think that there are some who have nice data centers with lots of racks who traditionally have filled those racks as we discussed before, perhaps they like a CapEx model better. So they can spend a lot of money every three, four or five years, whatever their cycle is, their tech cycle is, and they can manage their costs that way. They lose a lot of flexibility in doing that. And obviously the whole storage question is a great one. Do I buy three years of storage and use only a year at a time and what am I doing with that technology that’s now three years old at the end of it? And all that data that I’ve stored? And the answer really today as we’ve learned is don’t host all that data unless you have to put it in the cloud. So I can’t imagine, I mean if I was your company and they said, well, can you host it? The answer would be no. No,
Narinder Singh:
Of course not.
Eric Yablonka:
But no and why? But I think that that train has left the station. I think a lot of organizations have just given up on trying to host everything, excuse me. But it is still a weird thing where some believe that it’s whether they believe in physical or technology security InfoSec, they believe it’s safer to host it or not, or they really want to manage the performance inside. Some people may be freaked out by the recent downtime, global downtime due to a software update that probably gave people pause, but that actually had nothing to do with cloud, it had nothing to do with hosting. It had to do with a bad software update. At least that’s what was reported. So I think there’s a balance. The last person I talked to said, Hey, look, I’m just hosting some of my high performance imaging systems where I need to retrieve stuff quick, quick. Other than that, I don’t want a single thing in my data center. And I think that’s more and more where people are going. I mean, look,
Narinder Singh:
I think your answer is great because it really, I’ll pull out some of the good and bad reasons. Your last point. We also have edge compute for high performance reasons in place. It doesn’t prevent us from having a cloud-based system. I think the security pieces are examples of things that are a little bit antiquated in that I think we’ve now seen that CrowdStrike is a great example. We were up in fine. We had no downtime as a result of CrowdStrike. I think that the centralized cloud model, it’s like the entire infrastructure of the companies around optimizing that performance and aspects. And so those are maybe some of the bad reasons. Some of the good reasons though that I think you highlighted is the implied questions. Am I going to overpay you because you’re not going to innovate, you’re not going to evolve capabilities. Do I need this thing to be innovating and so am I getting ripped off through this subscription based model?
And so that’s the implication is translation of that question is into can I host it versus asking that question more directly or these kinds of notions of there absolutely has to be a thing of we should be able to show the financial ROI this model, not just have a religious statement on cloud is better. But really translating that into why. Speaking of kind of old school, new school clashes recently, a number of Oracle leaders, including the former head of CMS, have said something along the lines of Judy Faulkner, the CEO of Epic is the single biggest obstacle to EHR interoperability. Fair, unfair. How do you think about that coming from the owners of Cerner now?
Eric Yablonka:
Well, obviously it’s their competitor who is seemingly beating them. So one can understand why they might say that. I’ve worked with Epic for 20 plus years. I found Judy to be reasonable and very smart, very much understands the problems in hospitals. So in that respect, I think Epic does a fine job. I don’t have a lot of experience with Cerner, but Epic does a fine job. But getting back to our discussion whether we want a monolithic computing infrastructure or an architecture that enables interoperability, I think it’s been a challenge. Epic is fine. You can interoperate with ’em. But going back to my story, when we bought our enterprise service bus and was talking to Epic about how we connect it all up, they wondered why we even needed one. And that was a fairly standard commercial grade technology just not used in healthcare a lot. So I think Epic has an older architecture, works really well, but one must wonder in the next five or 10 years can they evolve into some newer product and what will that new product do for healthcare organizations? You can say the same thing for Cerner. It’s no different. They have dragged along a lot of legacy technology as well. Yeah,
Narinder Singh:
Oracle has not exactly been known as the bastion of openness for sure. It’s an interesting point you brought up. So I think Epic and Judy, you can proxy that has have a little bit of why don’t you just do that in Epic? And that I think is like one piece, but the second piece is it’s really hard. It’s a 20, 30, 40-year-old technology. It’s hard to make it open and connect. And so those two different perspectives can both be true and manifest themselves as not being open. But the motivation is important because how you address the challenge is a little bit different depending upon which one of those do you believe, is it that Epic doesn’t believe things should be done outside of Epic ever? Or is it that hey, they’re open to that. They could believe that, but it’s just hard because they’re on mumps or whatever and trying to make that interoperable is difficult perspective on one versus the other or both of those? Yeah.
Eric Yablonka:
Well obviously Epic operates with lots of systems, so it’s not that it can’t do it, usually it’s a challenge of getting their attention. And if a vendor walked in to Epic and it’s well understood, it had been well understood. If a vendor walked in and talked to Epic and said, Hey, I want to interoperate with you, they might not get a callback, but if two health systems went with that vendor and said, Hey, we want to interoperate, you might get some traction and you can understand that. I mean, epic can’t just respond to every organization, every vendor who wants to partner with them. On the other hand, when the healthcare organization really needs that kind of capability, really believes it to be important, then whether it’s Epic or Oracle, they really need to step up and step up quick. The other thing they can do is just really enable those services. Buyer is great, but there’s a lot more data than just what we’re seeing coming than we’re seeing today. AI is going to change that. As we talked earlier, whether it’s streaming data or very large data sets, large models, that is really challenging work. And whether Oracle and Epic can transition to those kinds of capabilities or add them, they’re usually behind a little bit from other vendors. Oracle might buy a vendor to get the capability and as we know Epic first to grow their own
And that’s fine, but things are moving too fast right now and we’ll see what happens. I think in the next 24 to 36 months, we’re going to see a sea change both on the healthcare side and the healthcare organization side and the vendor side, and how are we going to work together and advance our capabilities quickly and cost effectively and not do the dance we’ve been doing the last 20 years. I think that’s got to change.