#408 - 🔵 [NEO CONFERENCE] - How does structured bedside data power multicenter comparative effectiveness research? (Dr. Veeral Tolia)

Hello friends 👋

How can a database tracking 20% of all US NICU admissions change the way we practice neonatology? Live from the NEO Conference, Ben and Daphna sit down with Dr. Veeral Tolia to discuss his groundbreaking work with the Pediatrix Clinical Data Warehouse. Dr. Tolia dives into the power of leveraging decades of observational data to supplement randomized trials—from analyzing the 50-fold increase in Precedex usage to studying natural experiments like the vitamin A shortage. The group also looks ahead to the Newborn Express dataset, exploring how socioeconomic metrics like the Child Opportunity Index might help us understand the alarming rise in neonatal vitamin K refusals.

Link to episode on youtube: https://youtu.be/_HvujP3ZTpE

----

Short Bio: Veeral Tolia, MD is a clinical neonatologist, Director of Neonatal Clinical Data Warehouse Research for Pediatrix National Medical Group, and Director of Neonatal Research at Baylor University Medical Center in Dallas, Texas. He also serves as co-director of the DFW Collaborative, a regional quality improvement workgroup spanning more than 17 NICUs across 6 health systems. He holds an assistant professor appointment at Texas A&M Health Science Center College of Medicine, where he directs the NICU MS4 sub-internship, and serves as core faculty at Baylor University Medical Center. Dr. Tolia completed medical school at the University of Texas Southwestern Medical School, pediatric residency at Ronald McDonald Children's Hospital of Loyola University, and a neonatal-perinatal medicine fellowship at Northwestern University's Feinberg School of Medicine and Lurie Children's Hospital. His research interests include neonatal abstinence syndrome, bronchopulmonary dysplasia, quality improvement, and health services research, with publications in the New England Journal of Medicine, JAMA Pediatrics, and other leading journals.

----

The transcript of today's episode can be found below 👇

Ben Courchia MD (00:00.398)

Hello, everybody. Welcome back to the Incubator Podcast live at the NEO conference in Nevada. Daphna, good morning. Day two, first recording of the day, and we have the pleasure of having on with us, Dr. Veeral Tolia. Veeral, welcome to the podcast.

Daphna Yasova Barbeau MD (00:06.35)

Good morning, we're in day two.

Veeral Tolia MD (00:15.288)

Thank you. It's really exciting to be here. It's nice to meet you both.

Ben Courchia MD (00:18.016)

Yeah, same. I feel like we've met through your various papers. We've reviewed a lot of your articles on the show in the past. You have a very interesting role within the Pediatrix medical group where you are at the core of leveraging the database.

Daphna Yasova Barbeau MD (00:23.824)

That's right.

Ben Courchia MD (00:45.302)

That Pediatrix has built and translating some of that data into meaningful research, quality improvement projects, and so on. Is that an accurate representation?

Veeral Tolia MD (00:56.556)

Yeah, absolutely. The Pediatrix Clinical Data Warehouse is an incredibly valuable resource. It goes back more than 20 years. It covers about 20% of babies born in the United States that receive care in NICUs. It's quite extensive. We were talking about Reese Clark earlier. Reese Clark obviously founded the data warehouse and developed it. He really felt—and I've carried this through—that we have a moral obligation to learn from the data we collect. It is observational data, so it's not the same as randomized trial data, but randomized trials are hard to do. There are logistical issues. While we develop those trials, they should be informed, and even some of our practice can be informed by what we can learn from large observational studies.

Ben Courchia MD (01:56.782)

And it doesn't mean that you have to learn from a single source of information as well. Randomized control trials have a tremendous role to play in how the field is evolving, but just because that is true doesn't mean other things cannot be true at the same time.

Veeral Tolia MD (02:11.778)

That's right. As all of us practice neonatology, I used to do a thought experiment. If you go through a day of rounding, how many of the decisions you make can you back up with a randomized trial that is specifically applicable to your baby and your question? There are too many questions to answer. I don't think we provide definitive evidence, but I think it's useful and helpful to know what people are doing, how that may be affecting our patients, and what we can learn from that.

One of the strengths of the Pediatrix Clinical Data Warehouse is that we cover more than 300 NICUs. One of the ways we really try to leverage that is we say, "If some NICUs are doing this and other NICUs are doing that, what can we learn from that?" Because as you know, there's a wide variation in care and a wide variation in outcomes. So the centers that are doing really well, why are they doing really well?

Daphna Yasova Barbeau MD (03:21.198)

I love that concept of this moral obligation. As an individual professional, I learn from every patient that I see throughout my history. But this is a way for me to learn from your patients, Ben's patients, and patients that I don't get to see. This concept is almost like you're not randomizing the group, but you have multiple groups getting different interventions just by nature of where they are and who's practicing. It's really neat.

Ben Courchia MD (03:50.734)

From a data standpoint, obviously the scale of what you're talking about is almost a little bit terrifying. But I feel like you have a unique expertise. For people listening who are working—whether for Pediatrix or anybody else—we all gather data in one way or another. I believe the difficult part is how we go from data collection to leveraging that data for clinical information. I'm just curious, if somebody is listening and they say, "We're not part of Pediatrix, but we have a small unit and I have this data, but I don't really know how to go from there." What is your advice on how to translate the information we enter into the computer into actual data that we can query and question?

Veeral Tolia MD (04:47.778)

So you've actually touched on, Ben, my favorite part of data warehouse and big data research, which is that because the data are already collected, you don't have to spend a lot of time collecting data. Instead, you get to be really thoughtful of how you're going to use it. The fun part for me is figuring out how to take data we already have and use it to answer a question. It's fun; it's almost like a puzzle trying to figure out, "Okay, we have this, we want to answer that, how do we ask the question?"

The most fundamental thing, which I learned very early from one of my research mentors, Karna Murthy, who is part of the CHNC, is that the first thing you have to do is really define your question. We see this now in PICO with frameworks for defining your question. It is even more important in data research. You have to be very precise because every decision you make after that should be predicated on exactly how you define your question. What exactly are you trying to answer? What's the patient population? How does choosing your patient population and your cohort potentially affect what you're doing? How is the outcome that you're looking at collected? Is that a good way of collecting that outcome? Is it a bad way of collecting that outcome?

BPD is a perfect example because there are so many different ways to define it. You have the old NRN definition of oxygen support at 36 weeks corrected. You have the new Jensen criteria. You have the diagnosis—like if you use an ICD-9 coding of BPD, which is actually not a very good indicator of actual disease. If you're really trying to ask yourself, "What is the outcome that's meaningful to me? What is the outcome that's meaningful to parents?" You have to know precisely what that is. The most important thing is you have to define the question that you're answering as precisely as you can.

Ben Courchia MD (07:08.386)

Let me follow up then with something that I think I know what you're going to say. What if this is not always possible? Meaning you say your clinical question is too good, and the data doesn't follow. With the data that we collect, do you think there is any room to do this a bit backwards where you say, "Here's the data that we've gathered quite well."

Daphna Yasova Barbeau MD (07:18.03)

That was going to be my question.

Ben Courchia MD (07:36.436)

"And what question does this data lend itself to potentially answering?" in order to make sure that we can make use of the data. You could also find yourself in a situation where you say, "This is the question that I want to answer, and my data is actually not equipped."

Daphna Yasova Barbeau MD (07:52.29)

Or is it even possible that there are problems we haven't thought about, but because you have the data, they reveal themselves?

Veeral Tolia MD (07:58.798)

So yeah, one of those methods is the exploratory method. Asking the right question is important, and sometimes that question can be really broad. I'll give you an example. About every eight to ten years, we look at medication use in the data warehouse, and we just look at trends in medication use. That's entirely exploratory. It's predicated on the idea that we collect medication data probably as well as or better than any other large dataset because we have both the medication and the day of use over a large number of years. So we can say, "Okay, what are the medications that are increasing? What are the medications that are decreasing?"

That's a purely exploratory research question that is not super specific, but it lends itself very well to the data we collect. Once you say, "Okay, we have this 20 years of data and we're seeing this medication increase or this medication decrease," what does that mean from a patient perspective? Who's getting this medicine and why are they getting it? And are there things we can learn from how it's being used?

Ben Courchia MD (09:14.574)

I just always love those papers. It's Amp, Gent, surfactant, and Lasix somehow.

Veeral Tolia MD (09:21.154)

Caffeine. Caffeine is usually one of the top five.

Ben Courchia MD (09:24.206)

But it's always interesting to see. So Amp, Gent, caffeine, you're like, "Okay." And those that come in position four, five, six, you're like, "My God, I was not..."

Veeral Tolia MD (09:34.38)

Well, we've also done some fun stuff, primarily led by Rachel Greenberg, one of our really excellent collaborators. I don't know if you guys have met Rachel at Duke, who is incredible. One of the really interesting things is not necessarily the most common medications, but which medications have changed the most. So the big one right now is Precedex. Precedex use has really increased.

Ben Courchia MD (10:00.236)

Yeah, it's not used so widely, but the change in its use is astronomical.

Veeral Tolia MD (10:06.84)

Absolutely. So then you can ask the question. You take the good data we have on medication use. The question is how has medication use changed? Then you look at the changes and say, "Okay, Precedex is being used a lot more. Who's it being used in? How is it being used?" I think that's what you're talking about. You're designing the question based on the data you have available. We had a paper looking at the use of Precedex versus other opiates and who's getting it. It's really interesting because if you look at the clinical trial data on Precedex, it's pretty sparse.

Daphna Yasova Barbeau MD (10:47.63)

I was going to say, I think I can find it. Let people know what episode that is.

Ben Courchia MD (10:51.662)

I believe we reviewed it, and I believe it showed that it was interesting to see the relationship with opioid use.

Veeral Tolia MD (10:58.452)

That's right. It was the opioid paper. Yeah, it was a JAMA Network Open paper.

Ben Courchia MD (11:01.934)

In reduction in opioid use. It was very interesting. All right, do you have another question?

Daphna Yasova Barbeau MD (11:06.73)

No, you can finish your question. It was this paper, just so people know: "Use of Dexmedetomidine and Opioids in Hospitalized Preterm Infants." It was Journal Club number 164. Precedex use has increased 50-fold.

Ben Courchia MD (11:25.24)

Yeah.

Daphna Yasova Barbeau MD (11:29.966)

We do like Precedex in our unit. Yeah, for sure. Well, I have a question. Okay. I'm wondering, you've been doing this for some time. You've answered a lot of questions, a lot of papers. I didn't realize how many publications have come out of the warehouse until we spoke about it yesterday. What has been your favorite project?

Veeral Tolia MD (11:58.68)

That's different.

Veeral Tolia MD (12:03.242)

It's a little bit—it's not like choosing a favorite child because there are definitely some that I did not enjoy.

Daphna Yasova Barbeau MD (12:11.778)

Parents lie. They have favorites.

Ben Courchia MD (12:15.266)

No, he's saying that there's something he actually dis-

Veeral Tolia MD (12:17.09)

Yeah, dislike. We have a publication right now under review that I think is really important, looking at the use of probiotics after the FDA warning statement. Right now, that's my favorite because it mirrors my first paper in the Data Warehouse, which is probably also one of my favorites just because it was my first.

When I first started at Pediatrix, I joined a unit and the BPD rate was kind of high. I was like, "We've got to bring our BPD rate down." The only evidence-based therapy I could find to use was vitamin A. So I went to all my older partners and I said, "Look, I think we really need to try this." I spent three months trying to convince people to make this change. I pushed through a protocol. I was the young new guy telling people what to do, which was probably pretty obnoxious. So I got everyone on board, we implemented the protocol, and then there was a shortage; there was no vitamin A.

I was really frustrated. But the little germ of an idea came out: "Okay, there's a shortage now. This is a natural experiment. Let's see what happens when we have a drug that is indicated to reduce BPD, and it goes away. Does BPD increase?" I still like that study because I think it's a great use of our data. It was in a context where an excellent clinical trial had been done, but the data were 20 years old. We leveraged our data and showed that there were centers using vitamin A, then the shortage happened, they didn't have any more vitamin A, and there wasn't really any change in BPD rates. I really like those kinds of studies because they're clever and they leverage what we collect to answer questions.

Ben Courchia MD (14:24.382)

Absolutely. And in this case, it took your frustration with not having this drug available and turned it into a provocative study, an opportunity that presented itself.

Veeral Tolia MD (14:38.976)

As practicing clinicians, my favorite papers are the ones that help me make decisions on babies I take care of. And I think that one really did. I went into it thinking we were definitely going to see an effect. We tried as hard as we could to find an effect, and we couldn't. As a consequence, I think there's been less use of that medication. That medication is like 12 sticks. So maybe there are 12 fewer needle sticks for some of these babies. I like that paper. Similarly, the probiotics paper we're working on right now, I'm very excited about.

Ben Courchia MD (15:23.084)

Which journal can we expect the paper from?

Veeral Tolia MD (15:24.782)

We don't know yet. We're still in that phase. We presented some of the data already, and it's provocative. It shows that in units using probiotics in most of their babies, once the FDA actions happened, there was a massive drop in probiotic use—from 80% to essentially less than 5% of their babies. And in those units, the NEC rate increased by a relative amount of 50 to 60 percent.

Daphna Yasova Barbeau MD (15:56.884)

I think it just highlights the utility of the data you have, because that is the only way now that we could answer that question in this country. In Canada and Europe, they're looking at it, but we can't use it. So the only way to study it is this way.

Veeral Tolia MD (16:13.518)

And I think that action—I don't know if you guys were at centers that were using probiotics or not—how else are we going to inform future trials on probiotics? That's a medication where I think a lot of people have already lost equipoise on whether it was helpful.

Ben Courchia MD (16:35.83)

Yeah, it's one of these things where the legal context basically halted everything. We had a lot of episodes on probiotics, and it's one of the rare instances on the podcast—giving people a little behind the scenes here—where we even had episodes scheduled that we had to pull pre-release because people were like, "With this FDA memo that came out, I would rather we put a hold on everything."

Daphna Yasova Barbeau MD (16:59.096)

Wait... wow.

Ben Courchia MD (17:00.146)

Yeah, because the FDA tells you, "Hey, this is a risk of mortality." And now it's a dangerous game to play. But we also had people like Ravi Patel and others come on where we discussed the fact that the evidence is there. Taking it away is absolutely not based in evidence.

Veeral Tolia MD (17:28.017)

So Ravi has been one of our collaborators on that paper. He has another similar paper that's also a favorite because it's what we're working on right now, looking at how common probiotic-related sepsis is, or rather probiotic organism sepsis. What was fascinating—and this is where it's really great because we have literally two million babies we can look at—we ended up with a cohort of about 10,000 babies, and there were two babies that had lactobacillus sepsis that were not exposed to probiotics. It just happens naturally, but unless you have a big enough group of babies, you won't necessarily see that.

Ben Courchia MD (18:12.694)

And it is truly the intersection of clinical research, neonatology, and ethics. It's such an interesting idea. Even if it increases the risk of sepsis by a certain amount, when you put that in comparison to the number of lives you could save from NEC, what is the right decision?

Daphna Yasova Barbeau MD (18:34.038)

I mean, that's the conversation about any medication, right? All medications have risk. What's the risk-benefit ratio? I have one more question.

Ben Courchia MD (18:44.726)

No, no, I know it is very interesting. I'm looking forward to the paper coming out. These are the types of papers that I print, sit on my couch, and dive into. It's so much fun. Go ahead.

Daphna Yasova Barbeau MD (18:54.094)

Well, we would be remiss if we didn't mention that you are moderating a session today... tomorrow? Today.

Veeral Tolia MD (19:06.568)

Is it today?

Daphna Yasova Barbeau MD (19:08.792)

Today at one o'clock.

Veeral Tolia MD (19:11.214)

About ethics and parent refusal. The updated schedule does not have any moderating yet, but I'm glad I verified, and I was getting my bio ready. But that's another area where I think it's fascinating. I don't know what your question is, but...

Ben Courchia MD (19:21.528)

He's good, you know?

Daphna Yasova Barbeau MD (19:33.698)

Well, my question is, I think that's a very important topic. Dr. Mercurio was giving an ethics talk on parental disagreement on medical treatment. I think it's so pertinent to our current status quo.

Veeral Tolia MD (19:46.382)

It is incredibly pertinent. One of the things that we're doing with our data that I think is really important is looking at hospitals where Pediatrix is covering newborn services. The clinical data warehouse for neonates is based on a program called Baby Steps, but there is also a program for newborns called Newborn Express.

Ben Courchia MD (20:00.226)

I mean, we're doing that.

Daphna Yasova Barbeau MD (20:12.856)

We're familiar with it.

Veeral Tolia MD (20:16.002)

We're developing a dataset of newborn hospitalist medicine. And one of the first questions we're going to ask...

Ben Courchia MD (20:22.476)

For the people who don't understand, Baby Steps is an electronic medical record system, but for nursery patients, obviously the degree of complexity of a well baby in the newborn nursery is so much lower. There is a pathway within Baby Steps to actually do the documentation without having to go through all the prompts that you would usually go through when you have NICU patients on TPN, ventilators, and so on. It wouldn't make sense. So for newborn nursery patients, the process is a bit more streamlined and involves a different set of links and inputs.

Veeral Tolia MD (20:53.806)

Right, it's a different interface. We had traditionally not collected that data in a de-identified dataset, but we are planning on it. The first question we're going to ask is: what are the factors related to vitamin K refusal? Because I think that that's a real problem in the newborn nursery. I mean, if you guys do newborn care, I'm sure that that's...

Daphna Yasova Barbeau MD (21:13.227)

Every day.

Ben Courchia MD (21:13.816)

We're doing newborn care in Florida.

Daphna Yasova Barbeau MD (21:18.676)

We handle vitamin K refusal multiple times a month.

Ben Courchia MD (21:23.342)

It's interesting to me because this is my way of doing clinical studies: we are getting refusal forms from parents that we have to verify, and they're being left in our inbox. Recently, I found that all the clinicians have a stack of refusal forms. It's been painful. It's very subjective clinical research that I've done, but it does look like we are seeing an uptick in the refusal of not only vitamin K, but everything else. Parents just saying, "Don't give anything to my baby and just let us be."

Daphna Yasova Barbeau MD (22:00.103)

We are having refusals of glucose monitoring.

Ben Courchia MD (22:06.35)

"Don't stick my baby in the water." No, I think that it's a very important issue.

Veeral Tolia MD (22:11.694)

I think the benefit for us is that we have a significant presence in these newborn hospitals, and there is no equivalent dataset. There are a lot of neonatal datasets—there's us, there's the Canadian Network, there's VON, there's the NRN, there are registries in Europe—but there's hardly anything for newborn care. So the question we're going to ask is really: what are the factors related to vitamin K refusal?

Daphna Yasova Barbeau MD (22:34.165)

Right.

Veeral Tolia MD (22:41.358)

One of the additions we're making to the data in the next six months is incorporating the Child Opportunity Index, which is a measure of socioeconomic status. We're going to apply that also to the newborn nursery. The question is: what are the clinical and sociodemographic factors that lead to vitamin K refusal?

Ben Courchia MD (23:09.874)

I will be very interested to read that. I have a specific theory myself because I've been saying for some time on the podcast that I feel the current political climate is leading to this voluntary discrimination, where basically people discriminate against themselves by saying, "I have decided based on a TikTok video that I don't want this for my patient." And it's everything we're fighting for. People who have some form of education and make a decision will most likely see the benefit, but it's the people that we're trying to support that might voluntarily withhold the medication. That's very concerning because it's such a bizarre way of the world.

Veeral Tolia MD (23:50.414)

It is definitely a different environment for treating patients than it was even five or ten years ago.

Ben Courchia MD (23:59.202)

Absolutely. Well, it was great talking to you, Veeral. Congratulations on all the work you're doing, and we will be looking out for whenever you're moderating, whether it's tomorrow or today.

Daphna Yasova Barbeau MD (24:09.326)

Thanks so much.

Veeral Tolia MD (24:10.166)

Tomorrow I think we're doing a session on AI, which is a whole other topic and very exciting. But thank you guys so much. I'm a huge fan of the work that you do. At PAS last year, one of the workshops that I did was on dissemination and implementation. We can publish a paper, but if no one reads it, if no one knows about it, if no one changes practice, it's just another...

Daphna Yasova Barbeau MD (24:22.766)

I'm very grateful for the work.

Daphna Yasova Barbeau MD (24:38.146)

That's the call to action then for people to share the podcast with their fellows.

Veeral Tolia MD (24:42.592)

I think that that is as important as the paper itself.

Ben Courchia MD (24:47.586)

Yeah, and I think that this is what we're hoping to do.

Daphna Yasova Barbeau MD (24:50.318)

Reduce the time from paper to bedside implementation.

Ben Courchia MD (24:54.094)

If you subscribe to the podcast and you don't listen, it's there. Then one day you're on the drive and you just decide to catch up and listen to one. So, thank you.

Veeral Tolia MD (25:01.93)

Great. Thank you guys so much.