Join us for an in-depth conversation with Dr. Ryan Sadeghian, System CMIO at University of Toledo Health, as he shares his unique journey from healthcare consultant to practicing pediatrician to AI implementation leader. In this episode, we explore how academic health systems are uniquely positioned to drive healthcare AI innovation, balancing the dual mission of education and patient care while building practical AI solutions that solve real workflow challenges. Dr. Sadeghian discusses his organization's approach to developing internal AI capabilities, managing vendor relationships, and creating sustainable change management strategies that ensure successful AI adoption across clinical and administrative teams.
Ryan Sadeghian, MD, MBA, MSc, System CMIO, University of Toledo Health
Janae Sharp, Founder, The Sharp Index
[0:01] INTRO: Welcome to Digital Health Talks. Each week we meet with healthcare leaders making an immeasurable difference in equity, access, and quality. Hear about what tech is worth investing in and what isn't as we focus on the innovations that deliver. Join Megan Antonelli, Janae Sharp, and Shahid Shah for a weekly no BS deep dive on what's really making an impact in healthcare.
[0:29] JANAE SHARP: I'm Janae Sharp from the SAP Index, and I'm thrilled to be speaking today with Doctor Ryan Sadeghian and join us for an in-depth conversation about lots of things. This is part one with our leadership, so he's the system CMIO at University of Toledo Health, and he'll share his unique journey from a healthcare consultant to practicing pediatrician to AI implementation leader. And in this episode, we will explore how academic health systems are uniquely positioned to drive healthcare AI innovation, balancing the dual mission of education and patient care while building practical AI solutions to solve real workflow challenges. Doctor Sadeghian discusses your organizational approach to developing internal AI capabilities, managing vendor relationships, and creating sustainable change management strategies to ensure successful AI adoption across clinical and administrative teams. And I was really impressed too when we met, you had so many ideas, like, so everyone listening today, you should know, he's an ideas guy. And we're gonna start talking about your ideas and how you got there. Would you like to introduce yourself to the audience and tell us a little bit about your background?
[1:48] DR. RYAN SADEGHIAN: Sure, thank you very much. Thank you for having me, Ryan Sadeghian, Chief Medical Information Officer at the University of Toledo Health, also practicing pediatrician. My background in Clinical IT and IT goes back to the 90s and early 2000s when I started as a consultant. And you know, I always had a passion for this type of technology and when I was a student, you know, a medical student, actually, Obamacare went into effect and the incentives for electronic medical records. So I told myself over the next 15-20 years, we need providers, we need physicians with IT background, which even to this day, we don't really teach them in medical training. We have a fellowship in clinical informatics, but, you know, it's almost like you have to go through the medical training and then postgraduate and if you want to do a fellowship, that's another 2 years. So, there was a need, I could see the need for CMIO type position that again didn't exist back in the day. So I went through the training with NIH and biomedical informatics where we really focus on a lot of what today people talk about AI. You know, AI is not something that is new. So we work on a lot of technical parts from machine learning, natural language processing, predictive analytics, programming, and it kind of helped me navigate those knowledge throughout my medical training, professional training, and also professional senior leadership work to bring it to what today we're talking about, you know, Gen AI and how we can incorporate that into the clinical practice. I'm very excited to be here. Happy to answer any question.
[3:40] JANAE SHARP: Yeah, well, that's awesome. It's nice when people are passionate about this, you know, and that there's that perspective that we've been doing AI for a while, and healthcare, it's really hot right now and people want to know more about it. So I'm glad we can talk about this. I also want to start just kind of talking about the background here for some of the decision making that you're doing. Like when people are trying to decide about AI capabilities within their organization or within their job, how do you decide if you want to build something or buy something? And why is that decision critical for you as a CMIO?
[4:23] DR. RYAN SADEGHIAN: Yes, that is very important. And you know, I've been asked that question so many times, not only from our organization, and the board versus from people selling you stuff too, right? Yeah, yeah, when I go to the conferences. So I would say because AI is moving from pilots to enterprise use, and CMIOs have to decide, you know, CMIOs with their CIO partners, have to decide whether to invest in capability they own or depend fully on vendors, and that is a big decision. The choice really determines not just cost, you know, that choice that we're going to make, but also controlling of our data, workflow integration, the ability to evolve quickly as regulations and reimbursement shift. And you know, at the University of Toledo, you know, I've seen that decision directly affect whether AI becomes a scalable asset or just another subscription, and I think that is very important to consider.
[5:33] JANAE SHARP: Yeah, that is important to consider. So I'd like to know, like from your experience, like, how do internal builds measure up? Are they better with security or, you know, I've heard some people say, if you talk to Microsoft, they have a giant, we love Microsoft, by the way. Like, you know, they have this giant data science team, they have all this data, like, what does that mean in healthcare, you know? Like, what do, it sounds like you like internal builds.
[6:02] DR. RYAN SADEGHIAN: So it really depends, right? When we build in-house and you're right, you know, we control the model logic, the data we feed into the system and how the outcomes are going to measure, they're all in our control, right? This means that we can make quick adjustment without waiting on the vendor roadmaps, right? We're not waiting for the vendor to say over the next year, we're going to release this update so you can use it. Financially, you know, by doing that, we might be able to cut costs, and I'm being very cautious about that, but we might be able to, you know, cut some cost by targeting very specific pain points. You know, for instance, at the University of Toledo, you know, I built some denial reduction tools. I built some coding accuracy app to help with the denial reduction. So without having to go through any, you know, vendors or getting a license and it works really well and trying to expand it to other areas. But on the other side, there's some tools that, for instance, it would not be worth the organization to invest, for instance, if you want to do ambient AI documentation. There are really robust companies out there that they're doing a great job and for the cost that you have to pay for the subscription, it is still worth pursuing those vendors than, you know, investing and building something that may take years and cost tremendous amount of dollars and with resources that the organization may not have. So you have to really kind of weigh and see which area is the area that I can build in-house and I control those, or there are some areas that I can rely on a vendor because they're doing a better job.
[7:50] JANAE SHARP: Right, it's about your resources, and I've heard that too. It's about resource management, who's at your organization, and claims denials is huge, you know, prior authorization, that's a huge issue for everyone in healthcare. Can you give us some other examples like what are examples where building internally has allowed you to have better alignment with your clinical workflows?
[8:14] DR. RYAN SADEGHIAN: Very good question, you know. We do a lot of work with our College of Medicine Life Science, you know, we have a school of medicine, I think it's about 175, 174 students per year. There are a lot of tasks for our physicians who also take role in the College of Medicine as associates to go over medical students' SOAP notes, you know, they have to submit their objective finding from seeing patients, and these can be very time consuming for them to sit down and review. So being able to build something for them to review those on the fly and provide feedback, constructive feedback based on a very robust rubric and logic, you know, is quite important. I'll put an example of, you know, the CPT coding and prior authorization. Those are very high demands that, you know, anyone potentially could build for the organization. We're also looking at basic science and again college of medicine to see what other teaching companion and study guide we can create. We've worked on 3 different subjects now, that the student can use as a companion to augment their learning. And finally, we're looking at some Q-bank, you know, how can you have question bank on the fly as a medical student that you can generate questions based on different topics on the fly, on your phone and you can practice. So I think those are very remarkable things that as part of day to day operational, it's very helpful to the students and residents and fellows, and we're trying to achieve that.
[9:56] JANAE SHARP: I like that, because it sounds like things have really, do you do that too, so people could log in when they log in every day, they have to answer something from the Q bank?
[10:04] DR. RYAN SADEGHIAN: Well, we're piloting it to see the quality of questions by faculty. We have not rolled out to the student yet, but I can tell you that everybody has been blown away by the quality of these tools.
[10:16] JANAE SHARP: Does it help? Yeah, we'll have to see if it helps them get better scores too. Everybody at the school. Cause that's the name of the game. Having enough time to study and making it more available is huge. I want to talk a little bit more, maybe you could tell us more about that relationship. Like, it sounds like you work together with the academic side and healthcare delivery side. Not everyone has that. Like, do you feel like that's an advantage? Like, what does that look like?
[10:45] DR. RYAN SADEGHIAN: And that is very important to have, especially if you have that relationship, for instance, you know, the University of Toledo Medical Center, we are under the umbrella of University of Toledo itself and the University of Toledo just like any other university, we have different departments, colleges, deans, and chairs, right? So we have College of Engineering and, you know, we have a very good relationship with the dean and the chair. And working on building an organization-wide governance structure that focuses on AI as an initiative for the entire university with this in mind that, you know, College of Medicine and the hospital may have different needs because we're more clinically driven. And so being able to work with those engineering, you know, gives us access to technical expertise beyond what we already have. It gives us that talent pipeline without hiring an entire AI department from scratch. We bring the clinical problems, they bring the modeling and technical depth beyond what the EMR company offered to us, or again some of the tools that we built within the hospital and clinical from my department. By doing so, we build tools that are faster and more cost-effective. It also creates the academic feedback loop. Students, faculty get real world projects and we get fresh innovation aligned with our needs. I had some MD PhD student the other day that, you know, although they're under college of medicine and they're working with, you know, some of the engineers, and because of my role as a CMIO, they want to be able to have access to electronic medical record and see how we can integrate. This is just a beautiful example of how 3 colleges are coming together, including the division within the hospital to be able to have someone, you know, as an MD PhD doing research to be able to get something to the finish line.
[12:48] JANAE SHARP: Yeah, I think it's also great for students because a lot of times in school, you know, you hear the criticism that has nothing to do with your real world job. And just imagine if you were in school learning about things you're actually gonna do. That are actually useful like. I don't think everybody could imagine that. Like, so you start them early and I think that's great. Let's talk a little bit about more about this vendor collaboration and internal development. Like you've built an environment of collaboration with academic medicine with students. Do you do the same thing with vendors like, or in building internally? How do you strike that balance?
[13:31] DR. RYAN SADEGHIAN: Yeah, I would say, if I were to give an advice, because that's what what I do, right? I would say don't think of it as all or nothing. You need to use vendors when they can provide a scalable infrastructure or commodity features like EMR integration. You can reserve your internal builds for high value workflow unique to your organization, unique to your institution. One of the lines, one of the quotes that I have that most of the conferences I went to, they take it and put it on LinkedIn and social media is, I tell them all the time, don't outsource your thinking. You know your organization better than any other vendor, you know your workflow. So if you have your unique workflow to your organization where customization, control, ROI matters the most, you may want to consider building something for yourself if you have the resources, right? And to me, balance is about knowing where differentiation is critical and where it is not. There's so many vendors out there. I have a great relationship with them. I would love to bring to our organization, you know, some of them we're working on integrating into our system, and there are times that, again, you may be better off doing a good job using your internal resources if you have without going through the vendors. So having that balance, I think and differentiating where the critical items are is crucial.
[15:15] JANAE SHARP: I love that. I think you can tell why people would use that as a tagline. So your advice to CMIOs is don't outsource your thinking, which really I feel like is applicable to everyone. Like I can, I can give that to anybody, you know. Next time I have a fight with my teenager, I'm gonna tell him that. When they ask me what chore they should be doing, you know. But it's funny how those things are, when you really think about it, when it boils down to it, it's simple advice, but then the practical application isn't always obvious. What advice do you give to vendors? This is not, this wasn't on our question list guys, so I'm popping this on you right now.
[15:57] DR. RYAN SADEGHIAN: What advice do I give the vendor? The same thing, and then I tell vendors that they need to look at the system as their partner, and I always want to build a relationship with the vendor for a long term, right? This is a very sensitive time in healthcare. It is sensitive on many different levels. Number one, the generative AI is fairly new. So if you have something that is using this technology, there's not much return on investment statistics on it, right? So this is something we're working together, we're building that journey together, right? So, transparency is key to me. Number 2, they really need to understand organization needs. We hear a lot on the news about things that have gone wrong with the large language model from hallucination to, you know, recent Change Healthcare data breach, Salesforce data breach, all these companies that, you know, they're massive and they put their information on the same type of cloud that we're using large language model and other information that are being processed on the cloud. So building that culture of trust is very important between the vendor and the organization. And along the same line of, you know, don't outsource your thinking, the vendor needs to really get the organization to focus on their workflow and understand that. And if I were on the vendor side, I would not want to come in and tell the organization what they need to do. I would like for them to understand their own operation and advise them about the gaps because most of the time when people come from outside, they give you this interview brochure and, you know, they leave and you're left out with all this information. Most of the time people don't read it. And they continue the same old trend and workflow. So I think it's very important for the vendor to understand how they can partner and guide the organization, especially at this time that requires even more change management than what we needed for the EMR implementation.
[18:13] JANAE SHARP: Yeah, that's right. We don't just need technology, we need change management and they go hand in hand. So to close, I have one more thing. We're looking forward. So 5 years from now, what do you think the ideal mix of build versus buy will look like in most health systems? And if you don't want to say it for everybody, blanket, you can just say it for you.
[18:39] DR. RYAN SADEGHIAN: I would say I expect the hybrid model, especially organizations that may have some resources to do this because a lot of tasks have been so much easier with using, you know, when using the new technology and generative AI. So core infrastructure and compliance framework will come from the vendor, but the real innovation, the AI that touches coding, education, patient engagement or especially workflows will be still built internally or in partnership. This is what, you know, we think at the University of Toledo because again, we have the resources. Now, any of those items that I mentioned, if you're a small health system with no affiliation to the university or College of Engineering type department, then that would change the scope. And I think health systems that master both sides will be the ones delivering the most sustainable value.
[19:46] JANAE SHARP: Yeah, people who understand partnerships. Well, thank you so much for your time and for meeting with us today. And thank you also for your work. So, we will meet again and discuss more about the future of AI and the future of healthcare.
[20:03] DR. RYAN SADEGHIAN: Thank you very much. Thanks for having me.
[20:06] OUTRO: Thank you for joining us on Digital Health Talks, where we explore the intersection of healthcare and technology with leaders who are transforming patient care. This episode was brought to you by our valued program partners Automation Anywhere, revolutionizing healthcare workflows through intelligent automation. Nara, advancing contactless vital signs monitoring. Elite groups delivering strategic healthcare IT solutions. Cello, securing healthcare identity management and access governance. Your engagement helps drive the future of healthcare innovation. Subscribe to Digital Health Talks on your preferred podcast platform. Share these insights with your network and follow us on LinkedIn for exclusive content and updates. Ready to connect with healthcare technology leaders in person? Join us at the next health impact event. Visit Healthimpactforum.com for date and registration. Until next time, this is Digital Health Talks, where change makers come together to fix healthcare.