Mining Data to Aid in Rare Disease Research and Improve Quality of Care - Dr. Michael Mbagwu, Senior Medical Director at Verana Health





Michael Carrese: Hi everybody, I'm Michael Carrese, and today I'm happy to welcome Dr. Michael Mbagwu so that we can continue an ongoing exploration on Raise the Line about the ways that medicine is being changed by the ever-growing amount of data available and the new technologies which allow clinicians and others to analyze and use all of that information. Dr. Mike, as he's known, is Senior Medical Director at Verana Health, a digital health company that taps a real-world data network of more than 20,000 healthcare providers to deliver insights to clinicians and medical societies to improve quality of care and quality of life. 


Another important user group are life science companies, and we'll be learning how Verana's data is helping research into treatments for rare diseases as well. Dr. Mike's also an ophthalmologist with the Department of Veterans Affairs Palo Alto Healthcare System and an adjunct clinical instructor of ophthalmology at Stanford University School of Medicine. 


Thanks very much for joining us today. 


Dr. Mbagwu: Thanks for having me. I'm happy to be here. 


Michael Carrese: So, we always start with trying to get some personal and career background highlights from our guests, and we'd love to know how you first got interested in medicine, and then particularly, ophthalmology?


Dr. Mbagwu: Sure. Well, long story. I'll keep it brief-ish. I was a self-taught computer programmer, coder and built my own computers and that sort of thing growing up, and then when I went to college, I didn't really know what I wanted to do in terms of a major. So, actually, my brother was a pharmacist and when I went to Ohio State, I kind of followed in his footsteps, basically, in terms of picking the major. But I wasn't really sure if it was going to be pharmacy. I just knew something science. But about the end of, I would say, about second or third year of college, I decided, well, what does one do with a biomedical science major?

And med school was the next option, just because I think a lot of my friends and all the prerequisites were kind of doing that course. So, it guided me toward that and then when I went to med school, you go through the different rotations, you see what jives with you. For me, I intended to do internal medicine, but I really loved surgery when I started doing the rotation, and of the surgical specialties, it was ophthalmology that called out to me the most. So, here I am. 


Michael Carrese: Why did it call out to you, do you think? 


Dr. Mbagwu: I really like the precision of the surgeries. I love precise things. I really also liked the idea that what you do makes a huge difference. I mean, it’s vision. So direct, tangible outcomes and really benefits for people who undergo different surgeries. Frankly, I had a mentor that was just really great, really ushered me through, really opened his arms out to me -- Dr. Chambers at University of Washington. He's an oculoplastic surgeon, but he definitely was that great initial first contact. 


Michael Carrese: We hear that so often from guests that it was one mentor that led them down the path that they chose. It's a very, very powerful relationship. 


Dr. Mbagwu: Yes, and it's humbling to me just because I still remember shadowing him. I remember going to the operating room and him just saying, “Hey, come on, let's do this, let's do this.” As a med student, you're so worried about messing up and maybe getting good grades, but here’s somebody who just said, “Here. It’s your turn, surgeon. Let's go.” You know? 


Michael Carrese: That's pretty cool. So, Verana Health, let's turn to that. Help us understand more about the company and what you guys are trying to accomplish. 


Dr. Mbagwu: Of course. So, Verana Health, is a health data company and we are the stewards of a few registries. We work with American Academy of Ophthalmology and the Iris database. We also work with the American Academy of Neurology with their Axon registry and finally, we work with the American Urologic Association and its the Aqua registry. All of these clinical registries were built to improve quality of care and answer different research questions, as well as help physicians report quality metrics to payers. 


With that, comes opportunity because you have this great database with lots of clinical data already getting collected to report quality outcomes for payment purposes and you can repurpose that for research and other things. So, we're the data partner and the steward of this registry on behalf of these societies, and we also work with life science partners to sell insights. 


Michael Carrese: Can you give us a couple of concrete examples of how you folks think it has improved quality of care? Has it led to new standards of care or anything like that?


Dr. Mbagwu: Definitely. You know, these quality measures -- MIPS measures -- are ones where they look each year to see how many percent of physicians are meeting, you know, X, Y, and Z. And each year, they revisit this and they sometimes raise the bar or they might even retire a measure. They also use a lot of measures to inform the next one. So, let's just say I wanted to take blood pressure at every visit. Okay, that's just a yes or no binary. The next step on top of that would be to ask that of the patients with hypertension, how many are controlled to 120/80? So, that's like the next step. 


We have the same in ophthalmology. Of the patients with glaucoma, how many have had their eye pressure controlled at this certain level or have had annual testing? So, one builds on top of the other, and that's definitely a process that's done today. 


Michael Carrese: Yeah. So, let's talk a little bit about the rare disease applications. We were discussing some of this before we started recording, but how does the use of real-world evidence have the potential to overcome obstacles to research into drugs that end up being treatments for rare disease? 


Dr. Mbagwu: I think the data that Verana has is really suited for rare diseases in part, at least, and the reason is because you're really looking for a needle in a haystack. You're looking for certain keywords or certain genetic tests. As we know, some of these rare diseases have a prevalence of only 100 people in the entire United States of America or even less, sometimes. The way we categorize a lot of our data in medicine -- ICD codes and such -- might not capture the exact granularity of a condition, but clinical notes do. So, if you wanted to find somebody with a specific genetic defect -- a specific condition for which an ICD code doesn't even exist -- clinic notes represent really the only place you could even find that information. We work with a lot of life science companies for just this reason. 


Michael Carrese: Ah, so the database includes the clinician's notes? 


Dr. Mbagwu: That's right. The entire note, everything from your subjective symptoms to what the clinician then wanted to do, to what she is thinking about in her differential diagnosis, what procedures she thinks are suitable...kind of going behind the veil and really getting that complete 360 view of clinical care. 


Michael Carrese: Oh yeah. That's a really different dimension and a really rich vein of information. I can totally see that. So, how is real world evidence helping to accelerate and streamline rare disease clinical trials? 


Dr. Mbagwu: We have the Verana Research Network and some other initiatives whereby we can help people when they are looking for something specifically.  What we do is we will meet with a client to talk about what disease or what treatment they want to look into. I myself am an ophthalmologist and we have other consultants that we work with in each of the subspecialties. When we take the request in, we then do our own internal consultation saying, “Okay, as a practicing ophthalmologist, how would I document this? How would we expect this to appear? What are the different permutations?” And then we'll go through our notes and we'll print out a list and say, “This is how many people are showing up.” Then, if we choose to engage, we'll say, “Okay, let's maybe find out where these people are, what practices are they going to, what are their treatment patterns? What have they done before in the last year to five years? What is the standard of care,” and really just give insights to where these people are, but more importantly, what's actually being done for them? It's one thing to say what you should do, and what we say we do, but what are we actually doing? The notes really help us figure that stuff out. 


Michael Carrese: Yeah. And this is a big obstacle in rare disease research because, as you were saying before, some of these are ultra-rare conditions and for a clinical trial, you obviously need enough people to make it meaningful as research. 


Dr. Mbagwu: That's right. I think that's probably one of the most unfortunate things is that we have a lot of people and we have a lot of good ideas but you need to go that last mile of how you connect somebody with a rare disease to a trial that maybe even their doctor isn't familiar with or knows of. That's really what we're trying to solve here at Verana. 


Michael Carrese: So what are some rare diseases that might be impacted most through this? 


Dr. Mbagwu: I can speak for ophthalmology just because that's kind of my home court, if you will, but there's dozens and dozens. When I think of retina diseases, we can think of things like Stargardt, we can think of things like macular telangiectasia... all of these other very, very rare things. When you think of rare disease, there's rare genetic and then there's rare that maybe isn't genetic necessarily...diseases whereby some people can have a specific phenotype.


There's something called neurotrophic keratitis where, because of a virus or because of damage, the front of your eye loses that sensation. Really that is rare, because there's a lot of steps between with us thinking, “Is this really just a severe dry eye or have they formally lost this sensation in the eye,” and that can be from a variety of causes. We were talking earlier about retinitis pigmentosa and the first FDA approved genetic treatment with Luxturna. So, that's one camp of genetics. There's the other camp of people with more common diseases, but who just have this end terminal manifestation of it. That flexibility to have the notes define your phenotype, define who you're looking for, is really the only way you're going to even find these people. 


Michael Carrese: You were saying when we were chatting before the recording started that rare disease is actually kind of prevalent in ophthalmology. How come? 


Dr. Mbagwu: You know, in medicine, we all have our rare conditions. In ophthalmology, not only does it happen enough, but it's frankly something really drilled in our heads as residents. We have these books that we study for. It's an in-service exam. Basically every year you take something called the OCAPs and what they test us on is almost always the exception. It's never the rule or what you see every day in clinic. So, this stuff is really very much top of mind. I took my boards last year and I passed and I’m all done with that, but a lot of my study for boards was drilling rare diseases. 


The other reason is a lot of ophthalmic conditions have systemic manifestation. It's one thing to have an eye disease. It's another thing to say, “Hey, by the way, those patients have a higher risk of cardiac issues or having a brain tumor or something like that.” So, at least in our field it’s really drilled in our heads to keep that on our mind. 


Michael Carrese: That's great. I wish more disciplines could say the same. So, let's dig a little bit more into how artificial intelligence can improve the identification of patients and trial sites for rare disease clinical study, which is related as I understand it to how it can pull information out of the clinician notes we were talking about before. 


Dr. Mbagwu: Yes. So, in the end, all of these notes are written in prose the way I would actually write it. It's not going to be “patient has this and got this.” A lot of it might be me discussing it or thinking about it and ruling things out...saying less likely this disease, more likely that disease. That’s the value of clinical notes that we extract. It's the thoughtful curation of that and doing that at the disease level. 


For instance, if I treat a lot of patients with glaucoma, there are certain things that we always look for. We look at the drainage angle of the eye and assess eye pressure and how thick your cornea is. In something like a diabetic retinopathy, we're looking at different parameters -- what's their hemoglobin A1c, et cetera. The idea is that each one of these diseases have different notes and different texts that are important. So, using natural language processing and machine learning to thoughtfully harvest those notes and meaningful insights are really what we have to do. 


Michael Carrese: And AI is involved from the very beginning, as you're saying. 


Dr. Mbagwu: From the very, very beginning. I am consistently humbled talking with my AI colleagues here at Verana where they have all these fancy methods and I'm just always wowed with what they have. I see why it's important because some of the things that we just assumed were never possible, or just were kind of hopeless endeavors with AI, we're now for the first time able to actually see that. Again, it's one thing to say this person has retinitis pigmentosa. It's another thing to say they're no longer driving, they're having issues with night vision, et cetera. If you want to recruit people for a clinical trial, you need to know what level of severity they’re at and we can then kind of inform that and make our own levels of severity based on what we expect the language to look like at each stage. 


Michael Carrese:  Can you characterize for our audience the impact of all of this? How would you sum that up? 


Dr. Mbagwu: You know, it’s huge. In all seriousness, it is. I think for me, one of the most unfortunate things is you have these books that you study from and you read that this condition has a one-in-a million prevalence or one-in-two million or something like that. And you think about not knowing where they are. I don't even know what could even begin to characterize these people in the first place. 


But with what we have at Verana, I think this is how we will ultimately solve for rare diseases. In the end, a lot of these rare diseases have heterogeneity that hasn't yet been researched or tapped into. It’s important to be able to identify if something is a different manifestation of a disease and what the implications are of that. Are they inherently two different patient groups? Is this even a new disease that we never thought of? That's really what AI is going to make possible.


Michael Carrese: Yeah. The complexity of all of this is a little bit mind blowing. 


Dr. Mbagwu: Oh, yeah. 


Michael Carrese: And that's why it is so impressive that progress is being made. It's just a big mountain to climb. So, as you may know, Osmosis is a teaching company, and we love to get direction from our guests about a subject to focus on... something that you care a lot about. It could be related to what we've been discussing or not. It could be generally in the field of ophthalmology or something entirely different where you would say, “Osmosis, you guys should make a video about that.” What would that be? 


Dr. Mbagwu: You know, one of the passion projects that I did when I was in medical school was a fellowship year between my third and fourth year where I was looking at disparities in care. I was focused on Chicago. One thing that I would like Osmosis to do is work on this recognition that every population has disparities. We're not homogenous. Disparities can be from rare diseases. It's not just ethnic or racial. Disparities is just disparities and can be geographic, et cetera. 


I hope in the future, and we’re getting there, we can really double click on each of these populations so that whatever solutions we come up with are going to address the underlying problem for them in the context of where they are and not just a tone deaf or a one-size-fits-all approach. I think that's really where data can help. 


Can we look at the patient journey? Did they have disparities because they got in at a later stage of the disease? Is there somehow a systemic difference in the way patients are treated medically in different demographics, et cetera. You know, those sorts of things. I hope that we can double click on that and really get granularity because that's the only way we'll solve for it. 


Michael Carrese: Actually, one of the goals of the Year of the Zebra campaign is equity. You know, this class of patients has been marginalized historically -- not to cast aspersions on the medical community...there are reasons why it's evolved that way -- but if we're going to have an accessible, equitable health care system, we've got to up our game as far as paying the kind of attention to this group of patients that they deserve. 


Dr. Mbagwu: Absolutely. And to that point, it looks unique, right? A lot of people with zebras... what's one typical story? They've been to many doctors over many years and told nothing's wrong with them or “we kind of think it's this” and nothing quite helps. Then finally -- at the end of one, two, five, ten years or even more -- we then find, “a-ha, it was this all along.” That is definitely disparities in care in and of itself, and one that I hope can be solved sooner than later. 


Michael Carrese: Well, we totally agree with that. So, we have a lot of students and early career professionals in the audience and like to get our guests to offer their go-to advice for people who are about to enter the health care profession, particularly at a time when there's been so much upheaval. What is it you tell students?


Dr. Mbagwu: You know, I think that each of us has something truly -- and I'm not just saying this -- I think each of us has something unique to offer. I think to the extent that you can find what you actually want to do, what your passion is, and merge that with your career it is going to be phenomenal in terms of your personal motivation to get up every day to take care of the patients ad go the extra mile. 


I think that you should never try to make the left shoe fit on the right, as I think many of us do in medicine. We might be trying to go to a certain school or do a certain specialty because we think that's kind of what we have to do, as opposed to just being honest with ourselves and saying, “What do I like? What do I think I'm good at?” and not being afraid to take the plunge. It's how I ended up in ophthalmology. Again, my plan was to be like Dr. House, be an internist. But, you know, when I did surgery, I felt like I had another calling. At age twenty-four, twenty-five I said, “All right, well, let's just go for it.” And I did and I'm so happy I did. I hope everyone else can listen to their inner calling. 


Michael Carrese: Yeah. And you're way nicer than Dr. House. So, yeah, I don't think that comparison works. 


Dr. Mbagwu: (laughs) No, no. Yeah. No one to aspire to.


Michael Carrese: That's true. Well, it seems like you have found your passion, and one other thing I'm curious about is you have these different hats that you wear. You are teaching at Stanford, as I mentioned. You're seeing patients. You're doing this work with Verana. So, another piece of advice maybe that would be useful to the audience is how you juggle

all of that. 


Dr. Mbagwu: Sure. Carefully. You know, it's hard. When I went through residency, I saw myself as a practicing ophthalmologist working sixty, seventy hours a week. That's it. But kind of midway through my residency, I really had this nagging feeling of what I was doing before that programming that fellow year. How do I just make this? How do I own my career?  


I was blessed when I did my fellowship at Stanford. I was connected with folks who were working at Verana and they saw, “Well, you have this data background. You're an ophthalmologist. Let's see what we can do together.” For me, it's natural because this is exactly what I want to do. This is my way to find these people with rare diseases four out of five days of the week and then go to the clinic the fifth day. For me, it's been motivating to continue to practice medicine, makes me still informed of the data. It's how I can give the insights I want to, and also it's very gratifying and it's something I don't want to lose. So, you asked how I balance it. Yes, carefully, but also that I'm doing what I want to do and so I guess the motivation hopefully speaks for itself, at least for now. 


Michael Carrese: Yeah, that totally makes sense. We're going to have to leave it there, but I want to thank you very much for joining us today and wish you the best of luck with all the

different roles you're playing. 


Dr. Mbagwu: Absolutely. And thank you very much for having me on Raise the Line.


Michael Carrese: I appreciate it. I'm Michael Carrese. We want to thank the audience for checking out today's show, and remember to do your part to raise the line and strengthen the health care system. We're all in this together.