This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

The 229 Podcast: CMIO 3.0 - The Role Nobody Trained You For with Veena Lingam

[00:00:00] There's a discussion you're gonna wanna see at himss. It's about one of your biggest security vulnerabilities, not the network, but the service desk at hims. Security experts are pulling back the curtain on how major healthcare breaches often start, and why the service desk has become the new front line.

Join Nordic and Clear on Tuesday, March 10th at 1:00 PM booth, 6 31 for a candid round table. You won't wanna miss.

Bill Russell: Today on the 2 29 podcast.

Veena Lingam MD: there is some, what we call informatics deserts. If you look at the map of all the academic centers, um, and that's maybe where we need more innovation and telehealth and remote monitoring and all these things, that This group has the skillset to make a difference.

Bill Russell: My name is Bill Russell. I'm a former health system, CIO, and creator of this Week Health, where our mission is to transform healthcare one connection at a time. Welcome to the 2 [00:01:00] 29 Podcast where we continue the conversations happening at our events with the leaders who are shaping healthcare.

Let's jump into today's conversation.

all right. Today we're having another 2 29 podcast. Today I'm joined by, uh, Veena Lingam, Dr. Veena Lingam, I'm sorry. Uh, A-C-M-I-O at Moffitt Cancer Center. Uh, Veena. Welcome to the show.

Veena Lingam MD: Thank you. Excited to be here.

Bill Russell: we actually sat next to each other at the Tampa City Tour dinner, and a lot of the conversation that night was around ai.

Would you agree with that? That was a fair amount of conversation around ai.

Veena Lingam MD: Yeah, it could not be avoided.

Bill Russell: Um, yeah, it's, it seems like it, it cannot be avoided, uh, anymore. I, I think last year at this time, wondering like, Hey, you know, why are we talking about AI so much? It's not really there yet. It really couldn't do math problems without making mistakes.

It couldn't, whatever. And we fast forward a year from, uh, even one year from, uh. At that moment, it can [00:02:00] now do math pretty well. It has chain of thought reasoning. It's doing a lot of things. Um, and it's really having an, an impact on, uh, clinical informatics and, and how we, how we, uh, think about the role moving forward, which. and you and I started talking about, uh, your role at, um, as chair for, uh, for the, uh, training and curriculum at at amia. Um, give us an idea of, of what you do over there and, and what, uh, what the chair role is about at amia.

Veena Lingam MD: So I think it starts closer to home. I am the program director of the Clinical Informatics Fellowship Program at USF, more Sunny College of Medicine and Moffitt Cancer Center. And prior to that I was the program director for the fellowship program at Stony Brook University for five years. Um, so I came into the field of informatics kind of by, oh my goodness, this is so cool, but there's so many things wrong with it and I wanna be the one I wanna be part of fixing it.

Um, and then kind of [00:03:00] fell into the education role because I was educating myself. And then we had a program now am a, or American Medical Informatics Association is. The professional home, I would say for academic clinical informaticians. Um, and, uh, it also kind of houses most of the educational programs, whether that's graduate master's, undergrad, or fellowships.

Um, and I'm the chair for the Clinical Informatics Fellowship Program Directors. Um, this group represents about 60, maybe 61 academic programs that have clinical informatics fellowships. The fellowship is a two year training program for physicians who've completed any primary residency, radiology, surgery, medicine, family medicine, peds.

There's a lot of pediatricians in this field. Um, and so we take them for two years and kind of train them through all the things a future CMIO or, or a health data science researcher needs to know. Um, so so this. [00:04:00] Group of program directors really is thinking about how can we prepare the current trainees and current physicians to lead in the future where AI is unavoidable.

Bill Russell: clinical informatics is. if I thought about it, clinical informatics used to be about translating the needs of the clinicians into the EHR. I'm probably oversimplifying it, but that was a lot of the job. I mean, we spent a lot of time there. Um, and that will, that will continue to be part of the role moving forward. But more and more we're asking questions like, Hey, how are we gonna monitor these AI models and how are we going to. measure AI drift, and how are we gonna make sure that there's, a, uh, you know, a kill switch on it in case it goes rogue and we need, you know, all, all the, I mean all those, uh, governance and, and, and measurement and, and quality.

And I mean, do you see that the role, is the role fundamentally shifting or is it, is [00:05:00] just the tool sets shifting that we're focused on?

Veena Lingam MD: I think the tools available are shifting, so we're having to learn how to work with those tools. At the end of the day, we're still doing the same thing, which is implementing and kind of translating at the, um, kind of, confluence of medicine and technology and trying to innovate.

That's still the heart of everything, but what we do it with is very different. I like to say. CMIO 1.0 was like really after the HITECH Act and everyone needed to adopt the EMR. And the biggest ask there was change management, getting people on board, changing workflows. CMIO 2.0 I'd say would be like maybe five years ago, like pre-COVID or maybe around that time is when now they have a lot of data.

How do we use that data? A decade of healthcare data to run analytics, dashboards, create visualizations that make sense and [00:06:00] can tell the story and how we can improve the learning health system. Uh, and now it's really 3.0 a lot of us are having to tackle how do we govern, uh, ai, existing AI that we already have and we maybe never paid attention to.

Um, and. All the vendors out there that are trying to sell products, how do we evaluate and how do we govern? Um, especially if there's no onus on the vendor or the product owner to govern or, you know, manage that.

Bill Russell: I love that 1.0, 2.0, 3.0 1.0 use the technology. So you helped organizations gain adoption, uh, training, uh, you know, matching the workflows. I mean, we had to take those paper-based workflows and turn 'em into digital workflows. That in and of itself was a major. Uh, lift, uh, 2.0. We, we, uh, were inundated with the data.

I'm just trying to make sure I have this right. So inundated with the data. All right. Now let's, let's use this data in meaningful ways to impact care, to impact quality, uh, across the board. Uh, and then [00:07:00] the third is this, this age of ai and I, the distinction here is that. AI is actually summarizing data.

It's actually making meaning from the data. but that's where we're at today. Where we're at today is it's, it's sort of looking at all the data and saying, Hey, be a physician, could be somebody in billing and saying, Hey, this is the right code, or this is the right, potential path, or, Hey, here's, I type up a note for you. you approve of this note? Um. But in the future, we're looking at this age agentic thing. Uh, age agentic ai. I don't know why I say age agentic thing. It's, but it, it is right that it's right around the corner. And what we're seeing is these, uh, AI tools that are being connected directly into uh, the operating systems. I mean, we're seeing this over and over. I mean, uh, epic now has their, uh, their agent factory which sits in there and you can actually have it start to. [00:08:00] Move down the move down the pike. know, what were the difference in skills that a clinical informaticist would need to have in 1.0, 2.0, 3.0 1.0 is really an organizational change project, isn't it?

Veena Lingam MD: Yeah. 1.0 is organizational change. Uh, willing to be that person that can help configure the EHR and be the early translator with no. Um, no support system, really. Everyone's kind of doing it on their own. There's no, uh, real framework or, uh, precedence in a lot of cases. Um, but they were the innovators of that time where they took on the challenge and it was really getting people on board and rethinking processes.

Um, and some things really worked and some things didn't really pan out 'cause we stuck too much to replicating the P workflow. Uh, 2.0. I think again, we had to kind of start crunching numbers and knowing some data analytics. And I think 2.0 [00:09:00] also kind of got caught up with, uh, it's the, you said you used the word meaningful and that made me think of meaningful use.

And all the things we had to do and continue to have to do, but I feel like that became a big part of what we needed to know. So really requiring the, you know, understanding the regulations, the metrics, how would you capture it, how do you capture it without causing more pain to the clinicians. Um, that was more of the skillset.

I think 2.0 really needed 3.0. I think some of that cognitive burden of processing the data and getting the. Kind of, you know, summarized wisdom of it, or at least knowledge of it, maybe not wisdom, um, is getting easier. Um, things that we worked with data scientists in the biomedical informatics department back at Stony Brook, like problem by problem, uh, now is more accessible without having to be a data scientist.

So if you know the problem and you have the, uh, data or the workflow, you can, you know, you can [00:10:00] do it without having. To have that data science talent entirely. Um, you can get farther, I would say. Um, but we also need to think of, it's still, we can't forget the core principles of that change management and how to fit into the workflow.

So that knowledge of healthcare delivery, physician patient, uh, clinician practices is still the most valuable part. Uh, and a lot of implementing, like a lot of pilots currently being implemented with these AI tools. Tend to not get that far because they weren't implemented in a way that's meaningful and useful, or they're not solving the problem.

Nobody asked the why at the beginning. so you know, that's still core of all. Even though there's all these fancy ways of doing things, the core is still understanding how best to implement something.

Bill Russell: Uh, give me more of your journey into clinical informatics. I mean, what, what was the thing that, you looked at it and you said, man, that's, you know, you talked about wanting to be a part of the solution, being part of the, uh, of, of the change.

Is [00:11:00] that what was your journey and do you think the journey changes as the tools change? Or is it still the same kind of. Kind of drive and the same kind of people that would gr gravitate towards clinical informatics.

Veena Lingam MD: I'll answer the second part first. I think it's still the same kind of people. Um. But how they get in is changing. So CMI, oh 1.0 2.0 was more like right place, right time, right. Interest and opportunity to get involved. Three point ohs can be very, uh, meaningful about building the path to this career. Um, for my own career, you know, I did medical school and I grew up for most of my life in India, where there was, at that time at least, there was no electronic medical records.

You just had scribbles. There's. Patients traveled with their own records. So it's really on them if they don't have the right information, and you kind of piece the story together and hope you're making sense of it and treating the right problem. When I came here for [00:12:00] residency, uh, at Stony Brook, we had Cerner, uh, power chart, but it was kind of semi implemented.

You had orders and results, but not really notes. Um, and I was like, oh my goodness, this is amazing. I can actually see what happened the last time they were here. I'm not guessing based on what they're telling me, but then there was also so many holes and like half built workflows that, um, were aggravating.

And I could see why people were complaining, but I didn't. I was of the opinion like, we can't throw away the baby with the bath water. There's a lot of potential. We just need to make it better. Um, and I met the person doing the work, one of the associate CMOs there, and then got involved in projects and I was like, oh my, this is more fulfilling.

Um, or just as fulfilling, but more impactful on a wider scale. When I fix some things in the EHR, it helps patient safety and it helps clinician kind of efficiency and burden. Um, so. Just as I graduated, so it's right time and right place. Right. Um, is when A-C-G-M-E, [00:13:00] uh, approved the subspecialty of clinical informatics as a, you know, A-C-G-M-E accredited subspecialty.

So it's kind of, you could, I could think of it on par with, well, would I do GI or cardio or any of those things. And I started kind of investigating how do I get into this? And my learning pathway started with attending a. Conference, I think a hospitalist conference in the city and they had an informatics track.

So obviously I went to that and, you know, talked all day about how we could make things better. And then that, um, the person running that I believe was the Mount Sinai CMIO, um, said, oh, go do the Amy 10 by 10 course with Bill Hirsch at OHSU. So I did that 'cause that was an, an a smaller entry point for me to really understand the scope of what it is I'm actually thinking I wanna do.

that was great. It gave a great overview and at the end of that course you get to go to an AMIA conference and [00:14:00] have like a get together of the group. And every, when I went to that first AMIA conference, every single workshop and all the titles 'cause it was like, oh, all our like thoughts and. Key points are stuck in free text.

How would we ever get that out without building more check boxes or dropdowns? But then there was a whole work group working on NLP back then. I was like, I love everything people are working on here. I found my people and this is the path I wanna take. Um, so then I went and did a graduate certificate course with Columbia.

Um. And that really gave me more of the technical knowledge, the interoperability, the regulatory requirements, kind of the lingo I needed to be part of it. Um, then got a role as a deputy CMIO with 20% protected time. And within six months of that, the biomedical informatics chair at Stony Brook got accredited, got the accreditation for a fellowship at our program.

Um, so, and they needed someone to kind of teach and run it. That opened up, so I, I've just kind of again, right time, right place [00:15:00] and kind of being prepared with maybe the right knowledge and interest.

Bill Russell: So you didn't start with a, a, a strong technical background. You developed it over time.

Veena Lingam MD: Yeah. And I would still say I'm not a super strong technically, I just think I can translate and understand.

Bill Russell: a translator.

Veena Lingam MD: Exactly, and I, I give, I guess credit or I blame my brother who never taught me how to use the computer. He had one earliest when I came to visit. Um, so I had to figure it out myself. So I kind of more self-taught.

I kind of dabble, but I love, I, I wanna know what the capabilities are and how it could be used more than the actual building of it. That's kind of more of my aspect of this. Um, and then I think that's often a fear. A lot of, uh, physicians, clinicians think they need to be an engineer or data scientist or a programmer before getting into the field, but you really don't.

You need to be able to work with people who do and kind of it's [00:16:00] interdisciplinary. And I was saying, you know, the journey of how people are getting to this point is changing. So for I think CMIO 1.0 2.0, it was all like my story where it wasn't really a planned pathway, but now with a decade of the fellowships having been around, starting with four, and now there are 60, this could actually be a planned pathway for someone who's doing medical school residency and go into fellowship.

And be trained because learning on the job is really hard to do these days because there's so much to learn. Uh, so I think a focus to your fellowship where you get to work on real world projects with CMIOs and other, uh, you know, in other team members, you come out with a portfolio of things you have actually done and accomplished.

Uh, and preparing you for, you know, kind of leading in another program or the same.

Bill Russell: you talked about cognitive burn [00:17:00] burden. wanna talk about the how the curriculum, how the, the training sort of changes as you move forward. It, it used to be if you were gonna do a cancer, uh. or, or those kind of things. You had to able to, you know, navigate all these systems, pull all this information together, organize it and that kinda stuff.

And I, I'm, I'm, the more I'm out there interviewing, the more I'm hearing these stories where it's like, no, now we have, you know, these systems are able to reach out and they're pull it all in. And the nice thing is they're in a lot of cases, they're trusted systems. I mean, they're built into Epic, they're built into these other tools that we can monitor to them.

And we have all those, all those kinds of things. how do you, how do you see AI changing, the, the practice of medicine going forward and what does the clin, clinical informaticist need to, to know? I mean, you talked earlier about the standing at the intersection that, uh, I think you said confluence of, uh, uh, technology and, and,

Veena Lingam MD: Healthcare.

Bill Russell: and, and delivery of [00:18:00] care. Um. Is it, are we still, is that still the place It's standing except just the technology is changing.

Veena Lingam MD: I think so. That is still the case 'cause we are not, um. We're not leaning too far into technology. Yes, we do, but that's not the answer. We still need to be at that confluence and be the translators to make sure this really cool and, you know, innovative, uh, software or applications actually, you know, make a difference at the bedside.

Uh, and we don't want all that to go to waste. So I think we still stand at the Confluence. It's just we're having to Healthcare is always not as fast to adopt change. Uh, we tend to practice in historically this is how things were done and it's, and although we're getting better, um, they're slower.

So I think we clinical informaticians there for clinical informaticists. Um, we have to keep up with the pace of the technology. In terms of our knowledge, and then recognize the problems on the health system side and be the funneling of or [00:19:00] connector of those two pieces. So try to get the healthcare system to catch up, but in a meaningful way.

Bill Russell: Uh, talk to me about AI literacy. The role of the, of the clinical informaticists to promoting or helping an organization with, with literacy around in some cases, broad topic, and, and what it can do.

Veena Lingam MD: if you can see the curriculum, it in the, in the original iteration of the curriculum, there was not even a data analytics domain. It was all changed my. Management, leadership, project management, and you know, healthcare systems knowledge. And then we got the data analytics piece, kind of that domain added, and now AI kind of lives under there.

Um, and we're, we're thinking maybe it does need to be separated out. Um, but I think really for most of the world, AI is synonymous with when open AI came out and there's chat, GPT, there's a before and after.

Bill Russell: I ask it a question, it gives me an answer.

Veena Lingam MD: that, that's just [00:20:00] one kind of ai. But clinical informaticians, data scientists like, and, and in the healthcare world, we've been working on several levels of this already.

Uh, you know, all deep learning, machine learning in these are the things behind it. And they all fall under kind of the term of ai and it's just more accessible now. So I think we have. A role, um, you know, AI literacy. For us, we need to recognize what is actually going on inside the software, um, and what data I was trained on and what are the risks.

And I think educating the decision makers on that is also a big part of the job because we are not always the decision makers. And I think that topic came up during our last 2 29. But yeah, we wanna kind of. Enable that AI literacy in a more digestible format and hopefully be at the table when decisions are being made.

Bill Russell: needs to be a basic understanding of what AI can and cannot do. And it's, it's managing the expectations of leadership who are hearing, oh my [00:21:00] gosh, look, this is what has done. This is what some other health system has done. You know, why can't we do something like this? It's like, well, not necessarily a technology problem, it's an operational problem. Like we, yeah, but we just don't have the staff maybe to respond to it. We don't have the processes, we don't have those kinds of things. So it's, it's not, um, it's not magic. I.

Veena Lingam MD: and I think there's tolerance for that. It's also how we label and market. I think the vendors always want to. Feel like this is a shiny packaged project that's ready for prime time and it's just plug and play. But the reality is it's never plug and play. There's always some fitting to do. And, uh, I think a lot of places where the, so the innovation doesn't really meet its goal is because our workflows haven't adopt, you know.

Adapted and innovated to meet in the middle. If we're still trying to plug that into old redundant workflows, [00:22:00] we're really killing the capabilities of the product. So I think the adoption has to happen on both sides, and there should be appetite for innovating and changing our existing workflows.

Bill Russell: Do we have a shortage of clinical informaticists relative to the demand for informaticists? So do you, I mean, where, where's the, where's that at?

Veena Lingam MD: There's definitely papers written a long time ago also about this expected need for clinical informaticians, uh, and uh, the workforce. Especially since the HITECH Act came out. Um, but in real world, just looking now, um, when I first got board certified and I was looking at, okay, how can I actually do this for a job and not just as volunteer free work and looked for jobs, there was very few things and LinkedIn that had anything to do with it.

Uh, and in fact, there's not even a job code and there's a pina, like a physician's group working on that. Uh, with, with the labor department. Um, but now if you look now on [00:23:00] LinkedIn, while it's still not a category of its own, I think I can find at least a dozen prominent institutions looking for CMIOs and associate CMIOs.

Even just here in Florida, we in Tampa, we need two um, so if you know any oncologists or you know, informaticists ready to lead, we have a need here. And there's also these kind of silent areas. It's become, historically it's really been, uh, the academic turf where all these, you know, programs grow and the initial CMIOs grow.

So that's why you hear. Stanford, UCSF, Columbia, and all these programs. But really there's a need at every community level. And whether the leadership recognizes or not is a different question, but there's definitely a need and there is some, what we call informatics deserts. If you look at the map of all the academic centers, um, and that's maybe where we need more innovation and [00:24:00] telehealth and remote monitoring and all these things, that This group has the skillset to make a difference.

Bill Russell: I appreciate your time. I know you guys are in the middle of a implementation right now, which is, I, I think a busy time for, uh, somebody with your title. Um.

Veena Lingam MD: Um, yeah. We're just revving up right now. We have a long path about, I think a year and a half still before we go live.

Bill Russell: you're, you're in a comfortable stage right now. Not, not real busy.

Veena Lingam MD: Getting there, it's getting busy, but I think from April is when things really kick off. So right now it's kind of like training up our workforce in the clinical informatics department and ourselves. Um, so it's, it's the preparatory work right now.

Bill Russell: Well, Veena, I wanna thank you for coming on the show. I wanna thank you for participating in the, uh, city tour dinner. It was great to, uh, sit there with you and, uh, hear your thoughts on this, and I'm glad we got the opportunity to capture some of them.

Veena Lingam MD: Yeah, and I love everything about 2 29. I wish there was, you know, I had not known about it before. Um, especially as someone who moved new to [00:25:00] Tampa with this job being able to connect with other local informatics leaders and, you know. Folks was really, it's been really valuable to me, so thank you for doing that.

Bill Russell: Awesome. Thank you.

Thanks for listening to the 2 29 podcast. The best conversations don't end when the event does. They continue here with our community of healthcare leaders. Join us by subscribing at this week health.com/subscribe.

If you have a conversation, that's too good not to share. Reach out. Also, check out our events on the 2 29 project.com website. Share this episode with a peer. It's how we grow our network, increase our collective knowledge and transform healthcare together. Thanks for listening. That's all for now.