← Becker’s Healthcare Podcast
Becker’s Healthcare Podcast
Turning Healthcare’s Toughest Operational Challenges into Breakthrough Solutions: Part 2
00:14:22
Loading summary
Transcript21 lines
- [00:00]
A
Hi, everyone. This is Lucas Voss with Becker's Healthcare. Thanks so much for tuning in to the Becker's Healthcare podcast series. It's great to have you. Today I'm joined again by Ben Fichna, America's president at UiPath, and Jason Bui, who leads the healthcare industry practice at UiPath as well. I want to make sure you know about our first podcast episode together where we discuss some of the ways healthcare organizations are already applying AI to solve those operational challenges. Make sure you don't miss that episode. It's fantastic. This, however, is part two of that conversation where we focus on what leaders are learning from those deployments and what it actually takes to scale these capabilities across the organization. Ben and Jason, thanks again for being here today. It's so great to have you.
- [00:45]
B
Happy to be here and love your pronunciation of my last name and Bowie's last name, for that matter.
- [00:50]
A
There you go. It's that ie that ie that gets
- [00:53]
B
people I before E except after C, as I tell too many people when I'm checking into hotels, but there you go.
- [01:00]
A
There you go. Well, I want to hop into the conversation here. We touched on some of the operational pieces in our first episode, too. And Jason, I'll start off with you when we talk about this aspect of real operational change. Right. Again, I'd love to have you touch on what that actually looks like inside of healthcare organizations from your perspective right now.
- [01:25]
C
Yeah, thanks for the question and thanks for having us back here, Lucas. You know, when we say real operational change, the way that we talk about this with our customers is that it needs to reduce tasks that weren't meant for humans in the first place. Right. There's a lot that we've let grow and permeate throughout healthcare over the years, and we're finding that we were living with it because we have to. And that change can be hard and is hard. And we now, with agentic AI combined with generative AI, we're seeing that we can now eliminate those tasks in a clean way and then elevate the human in the loop to be more of a decision maker, more of an oversight role, and then practice what they were originally meant to do, especially in the clinical space. And we're seeing that when you combine and coordinate agents, robots, and humans together, you're achieving a new level of efficiency that healthcare never had. And so these are aspects that I consider when we see real operational change happening within the payer and provider organizations we partner closely with now.
- [02:37]
A
Ben, one of the things that we didn't touch on in our first part of our conversation of our podcast here is we didn't talk financials, we didn't talk, we didn't talk ROI when it comes to AI. And Ben, I'd love to know how you think about ROI differently when moving from these isolated AI solutions that we're seeing to then really the scale deployment piece?
- [03:04]
B
Yeah, I think, you know, a lot of organizations look at ROI in kind of simplistic measures. I'd say like where they introduced it, is it time saving, reviewing a chart or pulling to the right documentation? But that's. Those are like shavings that certainly make up a pile. But that's probably not where organizations are losing a ton of money. And if you look more at claims or denial management, I should say you see some of these cases sitting for days or bouncing between teams or getting reworked because something upstream didn't line up. And that's where the real ROI is, right. Where you're missing filing windows, delayed payments, write offs. So it's back to Jason's point, like how do you not just look at the one step in the process or the simple task, but looking at the end, end process or the end to end workflow and to just make sure cases are moving faster, things are not getting stuck, there's less rework. So it's not did I save time pulling that chart right? It's did the workflow move faster throughout end to end or the process end to end, seeing significant improvements and better outcomes. And I think once we've seen that from some of our top clients and that that has really transformed their organization to say, hey, this is something that really works, let's push it to scale. Right.
- [04:14]
A
And scalability or scale has almost become this buzzword. Right.
- [04:18]
B
Scalability is so crucial.
- [04:20]
A
Jason, are there specific operating models or governance approaches that you are seeing that actually enable scale?
- [04:28]
C
Yeah, definitely. I think a lot of the healthcare organizations we work with are still exploring which model will support scale. And again, there's not a one size fits all. But the common thread between the ones that are successful I are the ones that are combining their AI teams and practices along with their automation centers of excellences and combining their strategies. And so those two teams working together really accelerate the adoption of these new capabilities that are available and then also enable scale at a big level. And then I think what's critical is to have intelligent orchestration along with real observability and the controls that we need in health care. Without that, it's extremely tough to scale any of these new AI capabilities in Healthcare in general, we're very risk averse, rightfully so, and we need to vet that thoroughly so that we can springboard from that capability to attack all the largest and most complex problems in the ecosystem. And then also I find that decision committees that are willing to explore and experiment in a controlled manner help the scalable story. If they're willing to get after it, then the organization can rally behind them rather quickly and then we start to see the successes and it just compounds, the momentum builds. But if decision committees are taking a long time and really slowing down the process, then the momentum declines along with it. So those are just some of the things that I'm seeing on the marketplace today.
- [06:15]
A
Jason, in our first episode, which by the way, I highly recommend everybody listen to you, provided so many great examples of where AI today exists across the healthcare ecosystem. Can you walk us through an example where a workflow crosses those clinical and administrative boundaries and where AI either succeeded or failed because of that crossing?
- [06:42]
C
Yeah, absolutely. I talk a lot about prior auth in the marketplace because it's the place that both payers and providers are trying to optimize over the last, call it three to five years and now with agenta capabilities, they're truly able to span end to end and improve the quality of submission and the quality of decisioning and turning it around quicker so that the patient gets the care more timely. The successes we've seen start small. So for an example, let's say the prior auth comes through and there's a stack of medical records and you have to analyze, make sure that you're not missing anything to go through the review process. Agents can now look at that and ensure that everything is in place before it moves to the next step so that it's not found later in the process and have to revert back to square one. And then again, elongating the timelines and turnaround times and AI has been most effective that I've seen so far in summarizing medical records. Right. Summarizing it to the Persona that cares. So a prior auth intake coordinator has a different view that they look at as opposed to the clinician who's rendering a decision on whether or not to approve this care. And so having it tailored to that Persona has really impacted the ability to make a decision faster. And we're seeing customers leverage this capability and experience, you know, 50 plus percent reduction in their overall turnaround times and 75% plus in just reducing the review time of medical records. Another great example of where AI is impacting the Clinical space is the ability to create care plans, generate them and have them be tailored to the actual patient's experience and continuum of care. Historically, we kind of grouped folks into tiers and categories and sent out boilerplate messages or care plans. Now agents can really assess their historical claims, information, other metadata that they found in medical records records to create again that very pointed care plan that's going to be most effective for them. And then once that care plan is created, coordinating the care afterwards is now possible to do at scale. Right. You can refer them to the right programs, you can follow up with them through all the various channels that patients engage with their healthcare administrators and providers with and really impact their ability to adhere to the care plan. So a lot of innovative stuff is happening in the space and we expect it to grow over the next few years here significantly.
- [09:33]
A
Ben, Jason touched on something that he called orchestration, right. As an enabler, as a key enabler for all of these processes that we've touched on. Right. I'd love to know what orchestration means in this context, in a healthcare AI context. And what's the risk if organizations continue to deploy AI in silos without set orchestration?
- [10:00]
B
It's a great question. And I think if you look at healthcare in general, right. And healthcare workflow specifically, Jason, explain out that the challenge is in the steps. It's how things connect and how things are interoperable. Right. And move across end to end process. And we hear from a lot of clients, right. Like between tokens, models, constant change, it's really hard for them to make AI feel real and also have predictable costs. And I think our orchestration story, and being more vendor agnostic and sitting on top of a lot of these hard coded tools across healthcare and our ability to offer a lot of the prepackaged, out of the box solutions, if you will, that are designed for payers and providers has proven wildly successful. Right. And I think when you think about orchestration, this is more, as we used to say, like the Switzerland of this. Like we're not stuck into one in any specific technology. Right. We help sit on top of all of this to help connect the workflow end to end, which allows you to make sure that decisions are happening in the right place, things are staying consistent, governable, safe. Right. And really driving the consistency that's so incredibly important in an industry like healthcare that's so highly regulated and so important to get right on the first time to drive better patient outcomes. Absolutely.
- [11:11]
A
I think again, we come back to it. That's really what matters, is Making sure
- [11:14]
B
that the patient gets the care that
- [11:15]
A
they need at the right time, at the right place. And that's an enabler for set process. And Jason, I want to close our conversation today with something actionable for the folks that are listening. Right? What metrics or signals truly indicate that an organization is truly ready to scale today? What are those signals that leaders need to watch out for? Need to know about being on the
- [11:42]
C
ground level with a number of healthcare organizations. I'm seeing one common thread which is the willingness, not only at the leadership level, but all the way down to the operators and the clinicians that are operating on the front lines of healthcare. That requires a lot of education, a lot of explanation, and a lot of back and forth feedback. And so if there's the willingness, that's there, the scalability will happen. We also find that a number of healthcare organizations who have a strong foundation of automation already in place tend to be the ones to scale first. Right? They understand what automation can do, and now that agentic AI is available and scalable, they're simply evolving their automation programs to take on larger, more complex problems and finalizing steps in the process that only humans could do prior. And then being intentional about governance and observability over these autonomous capabilities is extremely crucial. Right. So if that is mature and the governance programs are in place, the scalability will come along with it. And lastly, I think early signals would be having those quick wins, tackling some of the smaller use cases and then gradually progressing from there so that we can build trust within our workforce and then have our teams realize that this isn't just an operational expense play. This is more about cost of avoidance as well as improving the quality of data that flows throughout healthcare.
- [13:26]
A
Ben and Jason, thank you so much for sharing your insights with us. So great to have you. And thank you to uipath for bringing us together for this conversation. We've covered so much ground over our two episodes here. And if you missed episode one on turning healthcare's toughest operational challenges into breakthrough solutions, make sure to listen in by visiting our podcast page at Becker's Hospital Review.com and as Ben and Jason shared, the organizations making progress aren't just introducing AI into parts of the process. They're rethinking how work moves across the system so those workflows actually run end to end. And if you're interested in having a deeper conversation about those processes and how AI and automation can help automate and orchestrate some of these workflows worth touched on. UiPath will be attending the Becker's annual meeting, April 13th through the 16th in Chicago. Make sure to stop by and say hello. We're excited to see you there.