Finding the right innovation partner is not easy. In fact, the reality is that when you vet as many companies as UCHealth has as part of its ongoing goal to disrupt and reinvent healthcare, you’re going to “kiss a lot of frogs.”
But if the goals and values are aligned, and the right intent and expectations are established upfront, it can result in a great relationship, said Richard Zane, MD, who holds the dual role of Chief Innovation Officer at University of Colorado Health (UCHealth) and academic chair of the Department of Emergency Medicine at the University of Colorado School of Medicine. In a recent interview, he spoke about the ultimate objectives of digital health and intelligence, the guiding principles with any technology solution, and the one component that matters most when working with clinicians.
Dr. Zane also discussed what it was like to be UCHealth’s first Chief Innovation Officer, why leaders shouldn’t always target the “sexy” projects, and the symbiotic relationship he has established with Chief Information Officer Steve Hess.
- Dual roles: chief innovation officer & academic chair
- Focus on digital health & intelligence
- “It’s making better decisions around healthcare.”
- Key questions with innovation
- Guiding principle: “It has to be easier, not harder.”
- Deploying rapidly to meet provider needs
- “Iterate and validate.”
- Investing in startups & participating in the process
LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED
If one of the other 4,000 hospitals or 400 health systems in the country has solved this problem, I want to know how they solved it. I want to learn from them, and I don’t have to reinvent the wheel or the algorithm to do that.
While I don’t employ a huge number of data scientists, I have a lot of great healthcare providers who are willing to change and adopt new ways of doing things.
It’s the concept of valuing clinician and provider input, rapidly iterating, rapidly validating, and having zero qualms about saying, ‘this did not work.’
It enabled us to see things like, do doctors use it? Do they opt out of it? Is it so user friendly that they don’t even know they’re using it? One of our guiding principles is that we will never ask a provider to leave their clinical workflow.
The reason why infusion centers were a great first step is, from a cultural change management perspective, it didn’t require providers to act differently, and it didn’t require patients to act differently.
Gamble: Hi Dr. Zane, thanks so much for taking the time to speak with us about the work your team is doing. Let’s start by talking about your roles and how you balance them.
Zane: Sure, I have two roles. One is innovation officer for University of Colorado Health (UCHealth). We’re one of the larger systems in Colorado, with 12 hospitals up and down the Front Range. We’re a relatively new healthcare system, formed in 2012. I’m also the academic chair of the Department of Emergency Medicine at the University of Colorado School of Medicine. In that role, I have oversight for a large clinical department, which includes probably 180 faculty members, a residency training program, research program, and educational program.
In my chief innovation officer role, I work almost exclusively in the area of digital health and intelligence. Interestingly, digital health and intelligence have come to mean everything and nothing to people. I like to say that we work on intelligence sort of writ large, meaning we focus in a very pragmatic and ethnographic way at the frontline of healthcare delivery and how we can help providers, executive managers, make better decisions. That’s how I describe it. Others might describe it as digital health, intelligent clinical decision support, AI, augmented intelligence, and other things. But for us, very pragmatically, it’s making better decisions around healthcare. That could be anywhere from one end of the spectrum to the other. It could mean anything from helping an oncologist pick the right chemotherapeutic agent for a patient with acute leukemia, to how does the CEO think about deploying capital in an environment where there’s both virtual care and brick-and-mortar care, to how to best schedule operating rooms and infusion centers. It’s really the spectrum around intelligence and decision-making.
Gamble: I like how you present it in a pragmatic way. I think when it comes to digital health, there’s still this perception of gadgets or tools, when really it’s more about helping to make better decisions.
Zane: Right. And sometimes it’s a digital tool; sometimes it’s a checklist. Sometimes it’s something that has yet to be invented, and sometimes it’s the application of existing technology in a different environment. We start by thinking about what we’re good at and what we need, and then we identify partners to build those types of solutions. And although virtually every university, including ours, has the ability to develop internal IP and do things like pharmaceutical therapeutic devices, that’s not what we do. We focus on decisions and intelligence. We do an ethnographic stance where we look into what are the challenges that we have, and we think about whether we actually need to be innovative — in other words, do we have to create something that’s new to the world. Because if one of the other 4,000 hospitals or 400 health systems in the country has solved this problem, I want to know how they solved it. I want to learn from them, and I don’t have to reinvent the wheel or the algorithm to do that. But if it’s a challenge for us and it’s a challenge for everybody else.
And so we’ll do an environmental scan. We’ll look at the ecosystem and identify two, three, sometimes more potential partners. Very often, partners will have come to us, and we’ve thought about who we can work with, and what we’re good at. And what we’re good at is that last mile; that pragmatic last mile of delivery. And so, while I don’t employ a huge number of data scientists, I have a lot of great healthcare providers who are willing to change and adopt new ways of doing things. I have a healthcare system with a visionary CIO, with whom I partner very closely.
We have a single robust instance of an electronic medical record. And while that might not seem like a big deal to some, it actually is. Across the entire healthcare system, a condition of participation is being on our single instance of Epic; we have guiding principles around how we do that. Those guiding principles are: whatever we do has to be easier, not harder. When we’re talking about providers, it has to be the path of least resistance; fewer clicks, not more clicks. We think about our healthcare system as being linked through this electronic medical record as a living, horizontally and vertically integrated electronic validation laboratory.
And so we partner with companies. We develop and build solutions. I have a team of implementation engineers, physician informaticists, programmers, and healthcare economists, and we think about how to deploy the technology very rapidly. When I say ‘very rapidly,’ it’s not an understatement. It’s hour to hour, day to day. Iterate and validate. So that when I go to a group of physicians and say, ‘We believe we have a potential solution to the problem you identified with us,’ they have that fidelity that even if it doesn’t work the first time, it will work eventually, and if it doesn’t, we’ll pull it. It’s the concept of valuing clinician and provider input, rapidly iterating, rapidly validating, and having zero qualms about saying, ‘this did not work.’
Gamble: Right. Can you provide some examples?
Zane: Sure. One of the examples we looked at was the concept of how challenging it is to order pharmacotherapy. It sounds like basic, bread-and-butter healthcare delivery, but the reality is that almost one-third of all prescriptions written are never picked up at the pharmacy. And although there are several reasons for that, the most common is that the patient can’t afford the medication, or it’s not covered by their insurance, or it’s not on their formulary. On top of that, it has to be the right medication.
And so, about three years ago, we partnered with RxRevu, a company based in Colorado, to see if we could answer that question. We built a solution that is now actually embedded in the native Epic environment. It started as an add-on, and it became so useful to clinicians that we scaled it across our entire healthcare system. Now it’s scaled across 50 other systems; even in Cerner.
This is how it works. When a clinician writes a prescription, it does a real-time check to make sure that it’s the right drug, and that it’s covered by the patient’s insurance company. After that was put into place, we went from a no-fill rate of somewhere in the 20 to 30 percent, to almost none. If I’m writing you a prescription for heart failure, and I want you to be on a medication that’s marginally better than another medication but is going to cost you $50,000 out of pocket, you and I can have a conversation about it. I might say, ‘Hey, this medicine is probably a little better, but your insurance only covers X amount, or I can give you this medication, which your insurance fully covers.’ It also surfaces those alternatives.
Gamble: And is this something where you were able to see results pretty quickly?
Zane: Very quickly. It enabled us to see things like, do doctors use it? Do they opt out of it? Is it so user friendly that they don’t even know they’re using it? One of our guiding principles is that we will never ask a provider to leave their clinical workflow, and so it appears to them as though it’s in Epic. It is in Epic now, but when we started, it wasn’t. It was actually on a bunch of Amazon servers in New Jersey, but the clinicians couldn’t tell.
We can measure right away if people are using it or not, and when people start to use it or stop using it. We can go to the bedside with clinicians and say, ‘how did it work? How did it not work? How can we make it better?’ And now that’s scaled.
Another key thing is that when we build these systems, we don’t assume everyone is like UCHealth. We think, if it’s a problem for us and no one has solved it, it’s a problem for everybody. But we have to build this thing for customers two through 2,000, not just for us. That’s how we function, and I think that’s why we’ve been successful.
Gamble: Right. The fact that so many other systems are using it shows that it’s a big need that has to be addressed.
Zane: When we work with these companies, we have the ability to make investments so that we have a lot of skin in the game and we can participate in the value creation. Another company we work with is LeanTaaS, which leverages intelligence to improve operations. It’s about bringing in process improvement and AI to help make better decisions. At UCHealth, we started with infusion centers. It may seem simplistic, but when you have a patient that needs an infusion, the doctors’ main concern is making sure the patient gets the right infusion. But patients care most about being comfortable, and having it done at a place and time that’s convenient for them. Meaning that if they show up at 9 a.m., it starts at 9 a.m., and not after a two-hour delay.
Using their technology, we were able improve scheduling at infusion centers across a large base, improve efficiency by well over 30 percent, and avoid having to build a new building. We did the same thing with the operating rooms. The reason why infusion centers were a great first step is, from a cultural change management perspective, it didn’t require providers to act differently, and it didn’t require patients to act differently. After that we went to operating rooms, which are one of the most expensive resources in healthcare. If we can make that smoother and more efficient, that’s a lot of bang for your buck. That’s what we’re looking at now. The preliminary results are pretty impressive, although they are still preliminary.
This is now in place across the country — not just here — and we were one of their first customers.
Part 2 Coming Soon…
Interview Sponsored by: