The U.S. healthcare system is ripe with opportunity to reimagine clinical and operational processes that more closely resemble remnants of the 1980s, not 2023.
"COVID-19 really opened our eyes that we built the healthcare system for 1980 — we didn't build it for 2080," Aaron Miri, senior vice president and chief digital and information officer of Jacksonville, Fla.-based Baptist Health, told Becker's.
To meet patients and clinicians' growing expectations for convenience and efficiency, health systems must transform these decades-old clinical and operational processes.
Baptist Health is embracing artificial intelligence across all areas of care to streamline workflows and sustain lasting improvements, including in the operating room. The health system was using a decentralized and fragmented OR block scheduling process, which hindered visibility into operating room availability and left surgeons struggling to reserve OR time, even during periods of low utilization.
Baptist partnered with LeanTaaS to implement a predictive and prescriptive analytics platform over its existing OR scheduling system, which allowed for more efficient block scheduling. As a result, Baptist saw a 16 percent increase in prime-time OR utilization systemwide and a 9 percent increase in the number of case minutes completed during prime time. The system also cut abandoned blocks by 22 percent.
Becker's recently spoke with Mr. Miri about these efforts, why AI is so crucial for health systems' transformation work and how Baptist is working to put more actionable data into clinicians' hands.
Editor's note: Responses have been lightly edited for length and clarity.
Question: Can you tell us about your views on EHR capabilities and why they are not enough for fast transformation and guaranteed outcomes?
Aaron Miri: Electronic health records are designed for a specific purpose, and they are doing a great job of being that. But they're not designed to support comprehensive care, which is where consumerization is leading us to. We now have a much more digitally native consumer audience — our patients, our physicians, etc. EHRs are great partners on the journey, but they're not going to lead the journey for us. We have to lean into change. Hence the advent of AI tools, which simplify and automate the workflow.
COVID-19 really opened our eyes that we built the healthcare system for 1980 — we didn't build it for 2080. And so our focus has been preparing our hospital operations and entire healthcare continuum for what our patients are asking us for by leveraging new tools in a new way.
Change is hard in healthcare. Clinicians learn through repetition, so changing our workflows cannot be done overnight. We have to look at care delivery and make sure that it is safe and effective care. But the most successful organizations right now are really leaning hard into operations like we did, looking at our OR efficiency. The reality is that's where healthcare's at.
Q: How are you empowering clinicians in transformation efforts?
AM: Our board of directors and executive team not only empower us with additional tools, but empower our caregivers at the point of care. Right now, we're laser focused on delivering those data points at the point of care so that clinicians aren't waiting on IT. They get those action alerts at their fingertips, and they're able to do something with it.
We're the first health system in Florida to go live with automated robot assistance. We have leveraged and deployed a low-code platform throughout our entire health system where our clinicians can write their own code and their own apps. We have done numerous things on digitizing the entire care continuum.
It's my philosophy as CDIO that I don't want to be in your way. I want to give you every tool and every ability and every data — appropriately, responsibly — to do your job as a clinician. Clinicians have been doing this for years in Excel or access databases. Why not give them a secure way of delivering their own app? They can write their own apps and deploy them to their own mobile devices. They should be able to work where they want it, how they want it. That's empowerment.
Q: Where do you see AI most affecting healthcare today and going forward?
AM: We're already seeing hard ROI from AI. What's amazing is once you have one success, it permeates through the organization. You have many more successes. I'll give you an example. We deployed a conversational language agent throughout our entire health system. We started with automating things like password resets for our patients and providers. Within six months, we achieved about 150 percent ROI of time saved because they don't have to wait in a queue to reset a password. We've expanded that now to numerous disciplines and subspecialties, leveraging high-end artificial intelligence to be able to do these things.
That's coupled with looking at things like LeanTaaS and our OR efficiency, being able to really optimize our care operations. AI for us is not just how do you generate more money or how do you become more efficient caregivers, it's also internal operations. It's looking at things like your workflows for material management or revenue cycle management. Why does prior authorization have to be so complicated? Why can't we automate it? We automated all radiology prior authorizations and immediately recognized almost a 300 percent ROI in terms of time saved. So, AI can influence all levels of care — operationally and clinically. When you put that all together, it's a recipe for magic.
Q: Where are things headed in the future with ChatGPT and Bard? What excites you? What makes you nervous?
AM: I think conversational linguistic AI agents are the future. But ChatGPT will not replace people. It will replace very basic functions that you would spend hours having to do. It's not gonna be able to deliver care at the point of care — not yet at least.
We're looking at ChatGPT and other tools as potential ways to get more effective patient information at people's fingertips. How do we answer very simple questions for our patients when they say, "Hey, how do I get around Baptist Health?" Because we're a very large health system. Or "Can you tell me how many different appointments I have today?" Think about it as a personal assistant helping guide you through all the levels of care. ChatGPT is immensely powerful, and it's only gonna make our caregivers much more efficient and effective.
The only thing that really makes me nervous is it could come out with a very erroneous result if the wrong data is fed to it. So how do we proof that output so that it is a democratized and a trusted data source? Because what will kill this faster than anything else is if people use this for nefarious reasons. That will kill public trust, and we don't want to do that because we need these conversational high-end tools in our toolbox.
Q: Any final thoughts you'd like to share?
AM: The time is now for digital officers and information officers to step up and start to answer these basic questions across the health system. We need to make sure that we don't sit on our laurels and be comfortable — we have to drive the ball forward.