Episode 23 – How AI is Reshaping Healthcare
Jan 20 9:00:00 am
Today, Jon Knisley (the host of hello, Human and a long-time technologist helping companies adopt and utilize emerging digital solutions) talks with Matt Gustitus, the founder and lead advisor at Digital Workforce Solution.
In today’s healthcare ecosystem, every health system, hospital, and physician is being asked to do more with less. Healthcare providers face increasing pressure to manage revenue, optimize utilization, and reduce costs. At the same time, they are being asked to prevent illness, optimize care, and improve patient outcomes. Fortunately, new technologies are providing a lifeline. It is estimated that AI applications can cut annual US healthcare costs by $150 billion in 2026. A large part of these cost reductions stems from changing the healthcare model from a reactive to a proactive approach, focusing on health management rather than disease treatment.
A big thanks to FortressIQ for sponsoring the program and be sure to hit the subscribe button whenever you listen to podcasts.
- What is Digital Workforce Solution and what is their mission?
- How we can leverage technology in health care
- The current state of AI adoption in the health care sector
- Is AI adoption helping simplify the “easy things” as well as the “big shiny” objects?
- The top three opportunities for AI to impact healthcare in the next few years
- The big challenges that may slow the adoption of AI in the health sector
- The model that works best for adopting AI
- The impact of ethics and explainable AI in healthcare
Resources/Links:Digital Workforce Solution
If you enjoyed this episode, subscribe and check out our series at fortressiq.com/podcast. Thanks for joining us today on hello, Human.
Full Episode Transcript:
Jon: Matt Gustitus, the founder and lead advisor at Digital Workforce Solution joins us today on the hello, Human podcast where we discuss the latest topics in artificial intelligence and how it’s being applied in the real world. I’m Jon Knisley, the host of hello, Human, and a longtime technologist helping companies adopt and utilize emerging digital solutions. A big thanks to FortressIQ for sponsoring the program. Be sure to hit the subscribe button wherever you listen to podcast.
In today’s healthcare ecosystem, every health system, hospital, and physician is being asked to do more with less. Fortunately, new technologies are providing a lifeline to the industry. It is estimated that AI applications can cut annual US healthcare costs by $150 billion in 2026 alone. A large part of these cost reductions stemmed from changing the health care model from a reactive to a proactive approach, focusing on health management rather than disease management.
We are fortunate to have an industry leader give us his perspective and insight on the exciting uses and future of AI in the health sector. Welcome to the program, Matt. Thanks for joining us on the hello, Human podcast, and bringing your knowledge and expertise to the program. To get us started, it would be great if you could just give us a little background on Digital Workforce Solution and your journey to start the firm.
Matt: Thanks, Jon. Glad to be here. This is Matt Gustitis. I started Digital Workforce Solution in early 2020. We are building a small team now. We’re heavily focusing on helping clients really better understand how they can leverage process mining and technology in order to improve what their jobs are on a day-by-day basis. There’s a team of us, and a lot of the work that we spend on right now, especially in the healthcare sector.
We’re spending a lot of time with IU Health here in Indianapolis, helping them better understand how they can leverage technology to really improve not only their processes, but also just the ability for employees to be involved in the work that is going on to really transform how business is being done.
Jon: That’s great. You also have a bit of background in the pharmaceutical space as well, if I remember correctly. I spent five or six years earlier in my career helping with a very early online medical education program. I think having that balance between pharma and healthcare gives you a bit of a different perspective sometimes. Is that your sense as well?
Matt: Absolutely. I spent almost 20 years at Eli Lilly, and actually spent a lot of time in finance in a number of different roles, but toward the end, really helping them spearhead, getting them started on their automation journey using process automation tools, looking at how we can leverage technology to really improve business processes, especially within the finance department.
It is interesting, when you look at pharma and how they’re approaching the use of process mining tools, AI, and then now working more on the hospital side, and really seeing how there’s so much opportunity, especially on the hospital side of things within healthcare, and just the opportunities that are there. It really gives me a very different perspective, looking at it from those different lenses.
Jon: We can tackle the pharma space another day. We’ll get back to the healthcare sector now. To give us a baseline for our discussion today, how would you describe the current state of AI adoption in the health sector? As we’ve seen in many industries, the pandemic was seen as this real accelerator for adopting AI and especially automation technologies. We’ve done two years of transformation in the past two months. That line came through in a couple of different areas. Have we seen that same similar accelerator in the healthcare area?
Matt: It’s interesting, Jon. Automation has been a hot topic for I would say, a number of years all the way back when I was at Eli Lilly. It was very much a term that was used quite broadly. Healthcare companies and hospitals have been, I would say from my perspective, a little behind other industries and its adoption of automation. However, from what I’m seeing, it has the greatest potential. I would say now, there’s a big interest in hospitals and healthcare. Because of the greatest potential that they have, this can directly impact the patient even.
I think that we’re seeing hospitals and healthcare organizations realizing the potential for the use of AI and really seeing real results, especially from hospital to hospital, you’re starting to see the growth there. It definitely has the biggest upside that I’m seeing of any industry across the board. Beyond just the bottom line benefit is also (like I said) the benefits to patients, to those that can really benefit beyond just the value of it.
That really excites me seeing that big upside and really the interest from the top down, from leadership down to those that are involved in the day-to-day activities, really seeing a significant growth in interest and more people wanting to step into those types of roles today.
Jon: The healthcare industry has its own unique piece for people who’ve spent time in it. Fortunately, I’m relatively healthy and haven’t been to the hospital too many times. Even if you just go to your general checkup, it is usually pretty easy. But any visit to the ER, or anytime you need some diagnostic or procedure done, you end up getting five or six different bills from eight different organizations and how it all comes together. It seems to be that ideal area for AI where you’ve got all these different sources of data, they’re not necessarily in the same place, and how you can bring them all together and ultimately get a better outcome.
This idea of leveraging technology to shift from more reactive to a proactive approach to delivering care makes a lot of sense and sounds really cool. The reality of the situation is you are in this day to day. Healthcare systems are ungodly complex. There are lots of areas that we can fix before we get to the cool stuff. Are we starting to make some progress on the easy stuff and not trying to just chase the next shiny object out there?
Matt: That is an interesting question because this is what comes up pretty frequently. This topic brings a lot of contention because leadership always wants to find the “shiny object” and with good reason. Obviously, you would want to get the most value from your investment. However, the reality is, with anything new to the organization, it takes time, dedication, and hard work. As an example, it’s a lot easier to score runs in a game by hitting a number of base hits, but it’s more fun to hit home runs.
What we are trying to do in our organization in helping these health care units is we’re trying to help them understand that you can aim big and try to find those home runs, at the same time keep with the small opportunities in mind. We really focus on helping them understand the balance of continuing to look for those opportunities. Let’s keep things moving forward. Let’s find even the small opportunities and do our best to deliver on as much as we can because time is of the essence especially in healthcare.
Health systems are very complex, absolutely like you said, so it takes time to really understand processes. In some cases, you have to fix the process before you can do any type of automation. That’s where we really find where we spend most of our time is actually understanding the process and really helping figure out what is the best way to set up this process before we just go and try to automate it. This is why we recommend carving out a continuous improvement team, which is usually a group that’s up front, trying to help figure out these things that we can improve, maybe we have to re-engineer the process prior to starting RPA development.
It’s a lot of work that I think sometimes organizations don’t realize, but at the same time, there are huge benefits. Absolutely, to your point, we’re always looking for that big opportunity but at the same time keeping our eye on continuing to deliver on what we need to. Once you find that shiny object, you’re going to be ready to deliver it. You don’t want to find that shiny object and then say okay, now what do we need to do? You want to be ready to go ahead and deliver on it.
I think setting up shop and being ready to go gives you the opportunity to find those big opportunities and really drive them to completion. That’s what we really try to do on a regular basis.
Jon: You hit a lot of great points there. I love the sports analogy and the hits versus home runs. Like any good team, you ultimately need a balance of it, and what the right formula is to be determined. Having that combination of skills and outcomes usually drives the best deliverables.
I think the other piece that really caught my attention there in your comments was this idea of you got to re-engineer the process before you automate. We see in the companies that we deal with so often, there’s this rush and push to automation. There’s nothing worse than automating a bad process.
The second rule of automation is you’re just accelerating the dysfunction if you automate a bad process. That comes up again, and again, again, so we try to hammer on. Look, you got to discover the process. You’ve got to re-engineer the process, and then you can automate the process. You’ve got to do it in that order to ultimately be effective. That’s a great insight.
What would you say are the top three opportunities for AI to impact healthcare in the next few years?
Matt: Wow. I would say there are a lot of opportunities, there are so many different areas. From what we’re seeing, there are many excellent opportunities, especially within hospital networks. Having worked in (like you said) pharmaceuticals, there are great opportunities there. You think about the complexity. The more complexity you have, obviously it makes it harder to really figure out where you have opportunities and go after, but once you automate it, you can make significant headway in what you’re trying to do.
We are seeing great progress, especially within hospitals in the revenue cycle operations. Examples would be registration, patient indexing—there’s a lot of information that has to be captured—provider to payer, payer tracking, and billing. All of that is really key. With so many different players, so many intricacies, and the amount of time that’s involved in it, there’s a lot of data entry within hospitals and within the healthcare space. This will need to be improved and simplified.
To me, those are really a top priority for a lot of healthcare systems. Definitely as far as AI goes, those top three opportunities would probably—obviously anything facing the patient—where you can really speed up. Just getting the information from the patient when they come in, to actually getting that all the way into your EMR and other systems that you need to, is a really big opportunity that I think AI can continue to improve on.
You’re seeing it across the board. We’re seeing so many different new technologies coming out in the AI space, really trying to help that because they know the importance to hospitals and healthcare networks, as it relates to both on the patient side, then also in dealing with the provider and ultimately to the payer, as well. I think those are really some key areas that we’re seeing.
Jon: The other one I would add to that is that whole diagnostic area and the computer vision technology. I used to also do some work on the security side, and we talked about having the security guard having 50–60 screens to try to look at. How can you stay focused and figure out, okay, let me find that anomaly so I can address it. You go blind to it after a while and anytime you have those image diagnostic situations, it seems to be a similar type thing. We can use the technology just to make the improvement better to help augment the human’s ability to say, okay, that might look a little funny. Let me dig into that piece a little more closely.
Looking at the other side of the coin, what would you say are the big challenges that may slow the adoption of AI in the health sector over the next few years?
Matt: I can tell you from what we’re seeing, the biggest threat at this point would be there are so many options in different automation technologies. If people have to figure out what is the right tool for the right situation, there are also a lot of competing priorities. I think that’s always a challenge, especially within healthcare. You just have so many competing priorities.
When COVID hit, that was a big challenge. You’ve got to switch gears and you’ve got to make do with what you can. That’s always a challenge when you have those competing priorities because there’s a lot of work that obviously has to go into really setting things up for long-term success.
I would say the other threat would be a defined strategy across the enterprise. We’re seeing where there are a lot of people with vested interest, but you definitely want to have a strategy that can work across the enterprise and not just in silos. Like I mentioned, too many options on automation technologies, this can cause confusion. Sometimes we’re seeing where it slows down progress.
I remember going to and speaking at different conferences and talking with people. You’d see the same people a year later and say, how’s it going? Well, we still haven’t figured out what tool to use. You really need to get in there and you can’t overanalyze the technologies. You obviously want to make a good decision, but you’ve got to make sure that you get going. You want to get started on your journey. I think that’s really important.
We’ve seen a lot of competing priorities with clients over the last year, like I mentioned, with COVID. I think setting goals that are challenging but achievable, is very important. We see a lot of organizations struggling with just the challenge of is this achievable or is it not? You want to have it to be challenging, but at the same time, you want to be able to meet some of your goals as you go.
We have to remind organizations that the initial investment is well-worth the return on investment for years to follow. Unlike a typical IT project, sometimes where it’s maybe a one-year project and you get the initial benefit, in this case it’s really for the long haul. You’re making an investment that’s really going to transform the organization.
I think those are the top three. It would be the various technologies, competing priorities, and then a defined strategy that we’re really seeing and trying to work through a lot of that to make sure that organizations are successful.
Jon: That all makes a lot of sense. I’ve been saying recently, we hear about this challenge of transformation and only 30% of transformation projects are successful. All the analyst communities tout that number. I think some of that is around too much focus on technology being the answer, and I think you addressed it there, too. There’s this paralysis that happens because everyone’s waiting for the perfect ideal technology. But as we’ve learned in medical research, there are a lot of failures out there before you find the wonder drug that comes in. You’ve got to just test and start somewhere with this technology sometimes.
Again, people have been too focused on technology as the answer and forgetting about or not giving enough emphasis to the people in the process dimensions of any complex projects. Looking at that side a little bit, obviously there are a lot of people tackling AI in the healthcare space to really improve outcomes. Is there a particular model that seems to work better? Do hospitals need to borrow a page from the tech world and establish COEs to accelerate outcomes and values potentially?
Matt: It’s a great question, Jon, because although technology is very key to hospitals, it is not usually the focus of the organization. And rightly so. Not to say it’s not important, but it’s just when you’re dealing with patients and all the work that has to go into making sure that people are getting what they need.
From our perspective, we always recommend establishing some sort of COE, a center of excellence. Where that is housed may vary by organization. There needs to be some group that’s really becoming the experts and able to proliferate that out to the rest of the organization. You need some central area that you can really control where this is coming from. Beyond that is also just being able to train and enable the rest of the organization over time.
What is very important is that the business (from my perspective) in a lot of ways needs to lead the identification of the opportunities, and thus really has to drive the prioritization and delivery. I think this is something that will change how business operations work together in the future. There’s always a lot of conversation around where the COE should reside. More importantly than that is how IT is in the business working together.
At the end of the day, we have to match up the right technology with the right opportunity from a business perspective and make sure that we’re delivering on what the business needs are. To me, that’s really an important thing, and to your point, having a center of excellence is really a huge benefit and can really help, especially as you get out in two years. After year one, and you get into year two, and year three forward, really having a core group becomes more critical.
Jon: That’s right. I hadn’t really thought about the idea that the core mission of a hospital is so far withdrawn from a technology viewpoint that they don’t really think about it that way. I saw somewhere, a case study of a bank around automation and COEs. They had like 20–25 people who were process experts alone in the COE. They had another 25–30 who were automation engineers. And they had a full 20-person or so data science team, just for a large national bank in the Asia Pacific region. Just seeing those numbers and what the bank was focusing on, I’d be hard-pressed to find a big healthcare system with an automation COE of nearly 75 people or so, whatever I mentioned there.
I think again having that insight and drive of hey, if we don’t really adopt technology, we’re going to be out of business in a couple of years. Healthcare doesn’t think in that way, yet. Maybe in the future it will a bit more. I think that’s why you have that discrepancy around technology versus patient outcomes. Obviously, patient outcomes are the most important, but having that vision that technology can drive it (I think) is critical.
One more area to touch on, then we’ll wrap things up here. Ethical AI comes up in conversations more and more in the commercial sector. What’s the impact of ethics and explainable AI in healthcare that you’re seeing? When the technology starts deciding who does and who doesn’t get treatment, I assume the scrutiny of AI models starts to jump significantly.
Matt: Just to go back with what you had mentioned as it relates to technology and hospitals. One more comment is that, the really good news is that we’re starting to see where technology really is getting more attention. At least from my experience—I only have it for a number of years here—if you go back, I think definitely seeing the partnership now, especially with some of the clients in the healthcare space. It’s really exciting to see that IT is really wanting now to partner even more than they ever have before because they’re seeing the potential opportunities.
When you share in the benefits, when you share in the excitement of the delivery of what you’re doing and meeting your goals, I think it drives more interest. It is interesting, back to your point, that it probably wasn’t as much in the past, but I think we’re going to see a lot more as we move forward within the healthcare space. I just wanted to add that.
Related to the ethical AI, I am definitely not the expert on these, specifically within these conversations. I will say that we are seeing a number of years out from true AI in this space becoming a reality. I think that’s probably a good thing from the fact that there are probably some things that have to be worked out. Most of the automation we’re seeing is just probably at the machine learning stage at this point. There’s definitely a continuum of automation.
That’s also one reason why we say organizations should really get interested in getting on the path of automation because it’s kind of a long journey. There are a lot of new things coming in, but it will take a large amount of data, and a lot of coordinated effort and testing. Before, AI is able to really make decisions like the example that you brought up, where an AI model is able to just make a decision on its own.
The healthcare industry (from what I’m seeing) will obviously need to require a significant amount of validation and testing before this would ever become an option. In that case, they’ll probably really flush out a number of concerns or things that need to be looked at, but I think that will just improve the AI models over time.
I definitely think AI in healthcare is going to play a major role. I just think that hospitals and health care are really going to honestly help drive the capabilities within AI because of the scrutiny, and a lot of the testing and validation that will have to happen before you would ever want some AI model making a decision on your healthcare. I think that that’s definitely something that I would imagine we’ll see in the near future.
Jon: I appreciate you keeping me honest there, Matt. I think you’re right. It’s probably less of an AI replacing the human; it’s more of an augmentation. How do we really help make better decisions by combining the best of technology with the best of our human intelligence around these decisions on medical care? I think overall, it’s great insight and a great point to end on.
To recap today’s conversation with Matt Gustitus, the lead advisor and founder at Digital Workforce Solutions, healthcare providers face increasing pressure to manage revenue, optimize utilization and reduce cost. At the same time, they are being asked to prevent illness optimize care, and improve patient outcomes. Fortunately, AI-enabled technology solutions are delivering productivity, efficacy, and accuracy to traditionally human-centered processes, especially in the health sector.
That’s a wrap on today’s show. Thank you, Matt, for joining us, and FortressIQ for sponsoring. If you enjoyed it, be sure to give us that like or five-star review on whatever platform you’re listening to. I’m Jon Knisley, and this has been hello, Human.