It’s hardly controversial that hospital information technology has a lot of room for improvement. But despite widespread consensus on the need for interoperability, better usability, and better patient experience, change continues to come slowly or not at all.
That’s down to a combination of inertia, misaligned incentives and more than a few faulty design assumptions, according to an assembled panel of “next generation CIOs” at the AMDIS-HIMSS Physicians’ Executive Symposium on Monday, moderated by Allegheny CMIO Dr. John Lee.
Health IT and how it got that way
“The impetus for how EHRs came to be, how they came to be adopted led to this piecemeal system where we have so many different EHRs and they don’t talk to each other,” said Dr. Veena Goel Jones, medical director, digital health at Sutter Health. “They weren’t built to. It’s something we’re still trying to solve for, and the regulations are trying to help us, but we’re not there.”
Healthcare information systems struggle to replicate the achievements of sectors like banking and retail not only because of the increased regulatory scrutiny, but also because incentives are more complicated.
“It’s not an ‘I’m trying to sell you something, you’re trying to buy something’ one-to-one relationship where you’re free to choose,” said Dr. Stephanie Lahr, CIO and CMIO at Regional Health.
“We have payers in the middle of that construct and that totally changes the dynamic of how those patients can come together and makes it difficult for us to look at airlines and banking and things like that [for examples],” said Lahr. “There’s a middle person with their own agenda and goals … That’s one of the things that makes this difficult, because it’s not a free market.”
“The answer to every question is always time, money and motivation,” said Dr. Yaa Kumah-Crystal, assistant professor of biomedical informatics and pediatric endocrinology at Vanderbilt University Medical Center. “So for the ones who are in power to make these changes, you need to understand how their time is spent, where the money goes, and how they can have a shift in motivation to make those changes. And once you understand those three things you can solve the problems.”
This difficulty in making progress is also self-sustaining, Lahr said, because it colors physician expectations of technology.
“A lot of our clinical staff, they’ve lowered their expectations of what we’re delivering with technology because we haven’t delivered anything that they really feel helped them,” she said.
“And I think if we could start delivering in small ways on things that really change the structure and make people believe we do have tools that will make their lives better, these things will start to adopt themselves because we will have developed the trust that what we’re bringing to them will actually be an improvement for everyone, not an improvement for the patients at their expense or an improvement for the billing department at their expense. We’ve been weighing them down.”
Reducing friction with technology
So what would change that physicians can believe in look like? Stephanie looks to Star Trek: The Next Generation‘s Dr. Beverly Crusher, and the verbal interface of the Enterprise computer, as an example.
“I want to be Beverly Crusher,” she said. “I want to be able to interact with the computer in a meaningful way that isn’t always about screen time. And I think it’s all about automation and intelligent tools that are going to get us there.”
Voice interactions with the EHR are one example of using technology to reduce friction. But there are many other possibilities. For instance, computer vision and machine learning could enable what is essentially passive documentation, removing a major burden for clinicians. She used the example of a system that detects when the nurse turns a patient to prevent ulcers, than confirms its assessment verbally before automatically entering it into the chart.
“A lot of our challenges with data is it’s all based on structured elements,” she said. “If we really start sensing the environment, if we start using this natural language processing and computer vision and these sorts of things, can the whole concept of “the chart’ and the structured elements go away and we start building analytics around what happened organically as opposed to a box that I checked that then goes into this model.”
Ultimately, a low-friction technology solution for prroviders can be matched by a low-friction care paradigm for patients.
“I would like for our jobs to look like an experience for the patient to get care wherever they are and they only need to come to the walled garden when things have escalated to the point where someone needs to touch them,” Jones said, noting that a combination of telehealth, remote monitoring, and retail clinics can fill the gap. “It’s less about ‘I need to see you every six months’ and more about ‘You’re well, but you’re at risk for developing this condition based on these factors so we’re going to work with you to change your care in a way that makes sense’”
Getting from there to here
Overcoming the forces responsible for resistance to change will be challenging, but Kumah-Crystal said she hopes the pandemic has already forced some self-reflection that will help grease the wheels.
“COVID-19 helped us sort of look under the rock of all those inefficiencies in the system,” she said.
One strategy to affect change is to not let the perfect be the enemy of the good, and accept that some change will be incremental, Lahr said. For instance, it may make sense to adopt machine learning in lower-risk areas like documentation before attempting it in higher-risk areas like clinical decision support.
Another strategy is to identify efficiencies that can give you breathing room to deal with reimbursement challenges, as in the case of her system’s nurse triage program.
“We never had centralized nurse triage ever before,” she said. “We rolled it out during COVID and now of course the community’s like We’re not going to undo that, we want this.’ But it doesn’t have a funding source. We don’t charge for those visits. The fee for service model does nothing in that space and where it does it’s so cumbersome its not worth the time and effort.
“But if I can take refills and a whole bunch of other mundane tasks and centralize and automate them, then I can pay for some of these other things that I want to do. And I think we have to start thinking about that because the behemoth of payment models, that’s a huge thing that’s going to take time to shift. We can’t wait for that to happen to do some of these other things.”
An inside look at the innovation, education, technology, networking and key events at the HIMSS21 Global Conference & Exhibition in Las Vegas.