Collaboration, Clinical Input Drive Analytics Innovation at UCHealth
As the healthcare industry comes to grips with the scale of disruption it will be facing over the next few years, a handful of forward-thinking organizations are aiming to lead the change instead of being swept along in its wake.
UCHealth, a non-profit network of ten acute care hospitals serving patients throughout Colorado, southern Wyoming and western Nebraska, is among those taking a proactive stance towards adopting data-driven analytics strategies to improve the quality of clinical care and the patient experience.
At the CARE Innovation Center, founder Dr. Richard Zane is developing innovative partnerships with clinicians, developers, and early-stage companies to solve for some of healthcare’s thorniest problems.
We’re fairly certain that if we identify a pain point in healthcare, we’re not the only ones who are experiencing it,” said Zane, who is the Chief Innovation Officer at UCHealth and a practicing emergency medicine physician.
“And we also know that we’re not the only ones who haven’t come up with a set of solutions yet. So we believe it’s very important to partner with industry and co-develop solutions that can be applicable across organizations to help us – and our peers – provide better care across the board.”
Large healthcare systems, especially academic ones, provide the optimal environment for testing and refining solutions to real-world problems, Zane told HealthITAnalytics.com.
“We have the ability to use our large healthcare system as a vertically and horizontally integrated learning laboratory,” he explained.
“We have a team of clinicians, subject matter experts, engineers, implementation scientists, and healthcare economists that we convene to think about ways we can partner with outside companies, usually very early stage startups, to develop and deploy novel technologies.”
CARE prioritizes ideas, not size, when choosing its industry partners, he continued.
“We’ve got partners that range from three people in a garage in Palo Alto to some of the large multi-national players. We try to connect with companies that are on the right track with an early-stage product, but can benefit from some help navigating the peculiarities of healthcare,” Zane explained.
“We’ve been focusing particularly hard on the quality of the patient experience and the efficiency of the system, and how to apply insights from data science and prescriptive intelligence to the issues we’re facing.”
Unlike other innovation ventures that primarily serve to commercialize internally developed products, CARE is focused on creating solutions that can have immediate impacts on the way UCHealth operates while simultaneously assisting the developer partner in creating a viable product for marketing.
“We don’t want to specifically start companies or invent products for the sake of staring companies and inventing products. We want to create things that we will use,” Zane said. “We have a great opportunity to be a co-development partner and an equity partner, through a fund set up by our board.”
“What we do is work together to develop what we think is a minimally viable, deployable product, which we put into a test environment. Once we’ve worked through any limitations and designed the workflows, we’ll bring it into the actual clinical environment to partner with engineers and providers to deploy the product.”
Most of the entity’s projects have focused on clinical decision support, artificial intelligence, and prescriptive analytics, Zane said.
While these technologies have significant promise for supporting more efficient and effective care, they are also among the most difficult to get clinicians to use.
Physicians, nurses, and other front-line staff are generally averse to changes in their workflows, and are often highly skeptical of new technologies that purport to make their jobs easier.
Part of the challenge for innovation experts is identifying the human factors that contribute to the success or failure of a new tool.
“One of our guiding principles is that nothing works if you say ‘thou shalt,’” Zane stressed. “You can’t say to a group of doctors that you must use this tool or this device, because typically you’ll get the opposite reaction.”
“No one is going to use a tool that makes their life harder. It has to be fewer clicks instead of more clicks; it has to be elegantly embedded into their workflow; and it has to open up a path of even less resistance, otherwise it’s going to gather dust.”
The quality of the tool being presented to them is key for gaining providers’ trust, Zane observed.
Clinical decision support tools must be “absolutely bulletproof,” he said. “If it is intended to assist a provider in making a decision, it has to be right every single time.”
Providers must be involved in the development process from the very beginning – something CARE takes very seriously, he added.
“The providers are the ones telling us what pain points we need to focus on,” he said. “That’s where we get our ideas to start with. Then we ask them to use whatever solution we’re working on. We don’t tell them to use it – we ask, and then we watch what happens.”
“If they stop using it, we will ask them why. That’s often the point where we get the most useful information.”
Once a problem has been identified, CARE’s teams of engineers, data scientists, and developers spring into action almost instantly.
“We don’t iterate quarter to quarter or year to year – sometimes it’s hour to hour, but never less frequently than every day,” said Zane.
“Because if someone tells you something’s wrong, and you don’t fix it right there and then, you’re going to lose the user’s trust and interest in renewing their use of the tool. We need to be able to say to our users, ‘you told us something was wrong, we fixed it, and now we’d like you to try again.’”
A speedy development cycle allows UCHealth to iterate and validate digital tools in a fraction of the time it usually takes to develop a new solution while producing lasting engagement and generating enthusiasm among the target end-users.
“We believe that commitment from the end-user is often lacking in other settings,” Zane said. “Healthcare projects often break down when you don’t involve clinical folks along with the engineers and designers, so we work hard to avoid that.”
“If we don't come up with the solution ourselves, someone else will, and then they're going to try to force us to use it whether or not it’s really the ideal option. So if we’re the ones embracing change and disruption instead of being disrupted, we’re in a good position.”
Choosing projects that have a direct impact on clinical care, financial outcomes, and the patient experience can help to cement the idea that innovation is something to embrace, Zane noted.
One of UCHealth’s partners is a local startup called RxRevu, which offers clinical decision support for prescribing at the point of care.
“Prescribing is a huge challenge for the healthcare system, especially as drug costs continue to spiral out of control,” said Zane. “We’re struggling as an industry to make sure the right people get the right drugs at an affordable cost, because we don’t have the data to support better habits and decisions.”
“To be honest, when a provider writes a prescription, they’re most likely to recommend a medication that they can spell, and that they’ve recommended before for someone who looks like you and had the same problem as you. That’s actually a pretty low bar,” he said.
Some drugs are simply ineffective for the person taking them, due to the nature of their condition or underlying genetic factors. Others can cause measurable harm.
And even when a potentially effective drug is prescribed, patients often fail to take the medication as directed, often because they cannot afford to purchase the therapy.
About a third of all prescriptions written are never filled at the pharmacy, Zane added. Non-adherence costs billions of dollars every year, and contributes to tens of thousands of potentially avoidable deaths.
“The ideal state for prescribing is ensuring that the right patient gets the right medication for his or her clinical needs and underlying predispositions to its effectiveness,” he stated. “The patient also has to be able to afford the medication, and it has to be available at the pharmacy when they go to pick it up. Otherwise, there’s no point in prescribing it.”
UCHealth decided to work with RxRevu to solve this widespread problem. The tool supports more effective prescribing by analyzing electronic health record data that would indicate whether one therapy is more appropriate than another.
“It looks at allergy information, of course, but it can also look at data such as kidney function, if it’s available, and coexisting heart problems or respiratory problems. Eventually, we hope to get more data on pharmacogenomics and whether a person is fundamentally aligned with the medication or not,” said Zane.
“It then goes to the patient’s insurance company and does a real-time benefits check and sends in a prior authorization if one is required. Then we ping the pharmacy electronically, and the pharmacy can confirm whether the medication is in stock or easily available.”
By closing the feedback loop, UCHealth providers are more likely to connect patients with affordable, effective prescriptions that can have long-term positive impacts on both their clinical outcomes and their experiences at the health system.
Zane credits strong commitment from executive leaders, as well as a willingness for front-line staff to engage in co-development of new tools, for the success CARE has had thus far.
“You can’t propagate engagement across the organization if the leadership isn’t absolutely committed to embracing innovation and change,” he asserted. “It just can’t work without the folks at the top. At UCHealth, innovation is one of the strategic pillars of the organization, and it’s not just lip service.”
“It’s an understanding that the next generation of healthcare providers is going to have to work differently than the generations that came before, and not all health systems are going to survive that change if they don’t adapt.”
“I think of it in terms of Amara’s law: we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. The world isn’t going to change overnight, but it is going to change. And we get to participate in that change, which is a really positive, exciting thing in my opinion.”