#453 - [Journal Club] - 📌 What Happened to NEC When Centers Stopped Using Probiotics?

When the FDA warning landed in late 2023, probiotic use in high-use NICUs collapsed from 86% to under 7% almost overnight. What happened to NEC? This week Daphna brings Tolia and Patel's new Journal of Perinatology analysis of the Pediatrix Clinical Data Warehouse, a natural experiment across 347 NICUs and more than 10,000 extremely preterm infants. Centers that stopped saw NEC climb from 2.7% to 4.4%. Centers that never used probiotics saw nothing change. Ben and Daphna work through the difference-in-differences model, the demographic imbalances, and the uncomfortable question underneath it all. When does data stop and advocacy begin?

----

Probiotics and necrotizing enterocolitis in preterm infants after the food and drug administration warning actions. Tolia VN, Bennett MM, Handler D, Canvasser J, Greenberg RG, Ursprung R, Ahmad KA, Patel RM.J Perinatol. 2026 Jun 2. doi: 10.1038/s41372-026-02712-y. Online ahead of print.PMID: 42225922

Transcript

Ben Courchia, MD (00:00.632)
Hello everybody, welcome back to the Incubator Journal Club. We're back today for another episode. Daphna, you have something in store for us.

Daphna Yasova Barbeau (00:09.996)
I do. I think we've had an action-packed week, haven't we, of information? But this is a really neat article. It's entitled "Probiotics and Necrotizing Enterocolitis in Preterm Infants After the Food and Drug Administration Warning Actions," in the Journal of Perinatology. Lead author Veeral Tolia and senior author Ravi Patel.

Basically, using the Pediatrix Clinical Data Warehouse, they wanted to look at NEC.

In two ways. They wanted to look at NEC between high-use centers of probiotics and low-use centers of probiotics. And then look at how the incidence of NEC changed before and after the FDA basically prohibited the use of probiotics. We've done a couple episodes on that, if anybody wants to take a listen to that discussion.

Including some things with the NEC Society as well, and people who are international, not in the States, who have continued to use probiotics. Okay, so I think this is exciting because people are asking, how are we going to study probiotics? But this is certainly one way that we are able to do this.

Ben Courchia, MD (00:12.288)
It's for sure a clever use of the Pediatrix Data Warehouse. Granted, for disclosure purposes, we work for Pediatrix. But it's true that when this FDA memo came out and it suddenly dropped, the opportunity to have a pre and post dataset that's accumulating in real time was quite nice. I spoke to Veeral about it, and he was telling me that he was looking into this, and I was actually eager to see what the data would show.

Daphna Yasova Barbeau (00:15.759)
Yeah. For sure.

Daphna Yasova Barbeau (00:22.16)
That's right.

Ben Courchia, MD (00:42.072)
Granted, our center was not part of that at that time. Yeah.

Daphna Yasova Barbeau (00:42.32)
Yeah.

Daphna Yasova Barbeau (00:48.266)
Right, right. When our center became part, when our team became part of Pediatrix, it was after this data collection.

Ben Courchia, MD (00:55.232)
Mm-hmm. Though we were using probiotics before the FDA memo came out, in our units. But I do feel like we have a baseline rate of NEC that is quite low, thank God, so I don't know how much significance our data could have contributed.

Daphna Yasova Barbeau (00:59.812)
In our unit.

Daphna Yasova Barbeau (01:13.158)
So it was a multicenter retrospective observational cohort study of preterm infants discharged from their group of NICUs from October 1st, 2022 through March 31st, 2025. And if you're not familiar with the Pediatrix Clinical Data Warehouse, it is a multicenter clinical database which uses prospective collection of clinical and outcome data. Basically, all of our notes are extrapolated.

Into the clinical database. Okay. In this group, they included all infants 23 to 29 weeks gestation at birth, and they excluded any infants with major congenital anomalies. All right. So it's a short paper, which is lovely. They looked at 17,279 infants.

7,000 infants were pre the FDA warning, and 6,711 were post the FDA warning. They had a washout period of about 3,500 babies. These infants were collected from 347 NICUs. They excluded 24% for major congenital anomalies. This is a pretty high rate of

congenital anomalies. But anyway, the remaining cohort was 10,083 infants, 5,268 pre-warning, 4,815 post-warning. The demographic characteristics were similar. The only significant differences were seen in maternal race and ethnicity compositions and breast milk exposure in the low-use centers. So just to

clarify that. Yeah, I'll tell you. So there was a difference in race and ethnicity. In the high-use centers of probiotics, they had, let's see, for example, fewer Black infants, both pre and post, compared to the low-use centers. They had

Ben Courchia, MD (03:10.67)
Yeah, what does that mean?

Daphna Yasova Barbeau (03:36.878)
more Hispanic infants in the high-use centers.

Pre-implementation, they had actually the same number of Hispanic families in the post-implementation period. There were slightly more white patients. This is probably not even significant. And then other/unknown, so they had more other/unknown in the high-use centers compared to the low-use centers. And then,

looking at breast milk at discharge, the high-use centers, so high use of probiotics, they had more breast milk at discharge and they had more any breast milk during hospitalization than the low-use centers. But of note, the breast milk at discharge and any breast milk during hospitalization is not different in the pre and post

for high-use centers. Does that make sense? So when we're comparing high-use centers to themselves, there wasn't a difference. It's not like they started getting more breast milk along the way. I think that's important to note.

Ben Courchia, MD (04:50.124)
Yeah. High-use centers, you said, are people who use probiotics in more than 80% of the eligible cases.

Daphna Yasova Barbeau (04:58.054)
Yes, sorry. Thank you for clarifying that. Okay, so the results. Probiotic use decreased from 86%, so more than 80%, to 6.9% in high-use centers, and from 13% to 0.1% in low-use centers. This was statistically significant for both groups.

That makes sense. If you weren't using a lot of it, you were like, okay, we'll just get rid of it altogether. But a big change, obviously, in those high-use centers. Yeah, it's like a roller coaster. Sure.

Ben Courchia, MD (05:31.0)
That graph that shows that precipitous drop, the power of the FDA, right? It's interesting that the publication of a single document. Look at that. This is absolutely insane.

Daphna Yasova Barbeau (05:45.72)
Yeah. Wow. Especially when you should acknowledge that many of the therapeutics used in the NICU are not FDA approved in the age group in which we use them. Yeah.

Ben Courchia, MD (05:57.343)
Yeah. But what I'm saying is that it's an administrative event backed by a single case. Fair enough, it's not just an administrative event, but wow. Eighty-six to seven percent. Dang it. Okay.

Daphna Yasova Barbeau (06:02.755)
A lot of power, yeah.

Daphna Yasova Barbeau (06:10.79)
Yeah. So then they looked at the incidence of NEC. The incidence of NEC among infants in high-use centers increased from 2.7% to 4.4% post the warning. And it was pretty constant, 3.6% both pre and post, in the low-use centers. Yeah. So what did that look

Ben Courchia, MD (06:34.654)
Use, yeah.

Daphna Yasova Barbeau (06:40.646)
like? It really included all NEC. So it went from 2.7% to 4.4%. The medical NEC went from 1.7% to 2.7%, which was not statistically significant by itself. Surgical NEC went from 1% to 1.7%, which again was not statistically significant by itself. They also looked at SIP, so spontaneous intestinal perforation.

Ben Courchia, MD (06:45.559)
Yeah.

Daphna Yasova Barbeau (07:08.422)
Two percent to 1.6% in the high-use centers, which was not significant. They also did not have a statistically significant change in bloodstream infection or the composite outcome of NEC or death. But for all NEC, there was a statistically significant increase in those high-use centers who decreased their use after the warning.

In the low-use centers, there were no differences for any of the outcomes. There was a, people hate when we say this, there was a trend, a p-value of 0.06, for bloodstream infection, but actually it was 9.7% pre the warning, when they were using more probiotics, and 11% post the warning. So a slight increase there. Okay.

Ben Courchia, MD (07:39.49)
Mm-hmm.

Daphna Yasova Barbeau (08:03.494)
They then used a difference-in-differences approach to look at NEC. So the pre and post difference in NEC was 1.64% in the high-use centers and negative 0.02% in the low-use centers, for an unadjusted difference of 1.66% with a p-value of 0.044.

The model-estimated difference in differences, so after adjusting for infant characteristics and any breast milk exposure, was 1.18% with a p-value of 0.045. So still statistically significant. And I told you about the differences in medical versus surgical NEC. So they conclude this was a natural experiment involving over 10,000 extremely preterm infants,

and centers that discontinued routine probiotic use following the FDA warning experienced an increase in NEC incidence, while centers with minimal probiotic use showed no change. The clinically important findings suggest that rapid discontinuation of probiotics following regulatory warnings and actions may have had unintended consequences in this population vulnerable to NEC.

Thoughts?

Ben Courchia, MD (09:21.057)
Yeah, I think that it's a bittersweet result, right? I feel like the signal is there. There's definitely an increase, specifically for centers that used it more. We were one of those centers that used it more. We were not included in that study, but we are a center that used a lot of probiotics.

Daphna Yasova Barbeau (09:31.694)
Agreed. Yeah. You're like, aha.

Ben Courchia, MD (09:48.792)
And I worry that people are going to say, 4.4% is still pretty much okay. And they're going to say that's fine. And I don't think that's fine. I'm very worried that this is going to remove an opportunity for excellence, where we were at 2% NEC rates for a network of units, right? That included how many, did you say, in total for this particular

Daphna Yasova Barbeau (10:17.284)
Babies, 10,000 babies. Yeah. That's right.

Ben Courchia, MD (10:19.097)
10,000 babies, 347 NICUs. That's really good. That's really good. Why can't we do that? And like we said in the episode where we actually discussed the memorandum from the FDA, it's frustrating that there's not a great path for probiotics to return to the NICU. And it's frustrating to see that people around the world continue to use them and we're in a little bit of a holding pattern when it comes to that. So.

Daphna Yasova Barbeau (10:47.918)
Yeah, there are studies in other, let's say our neighbors in Canada, where the use of probiotics did decrease their rates of NEC, and so they'll continue to use it.

Ben Courchia, MD (11:01.837)
Absolutely. Yeah. So it is what it is. And I don't know whether this is going to move the needle. I hope it does.

Daphna Yasova Barbeau (11:13.2)
Well, this is not the only, what's the word, major institution looking at this, right? We have other databases, so hopefully they will look into that as well, and then collectively we can maybe take that information to make change.

Ben Courchia, MD (11:34.435)
Yeah. But to answer the question, I think it was both also worrying. It's like, oh my God, it really should show a difference, right? When I was talking to Veeral about this, the question is, what if it doesn't show anything? Then how are we going to interpret that data? So I'm happy that it does show a difference. And I'm happy that it shows a difference in the direction that we physiologically postulated.

Daphna Yasova Barbeau (11:48.347)
Hmm.

Daphna Yasova Barbeau (11:51.92)
Yeah.

Ben Courchia, MD (12:04.091)
So I think that this is where data leaves us and where advocacy picks up, where now it's about us advocating for our babies. So like Ravi had mentioned, and I forget who we interviewed on this, but the data's there. The next step is no longer an RCT. The next step is advocacy.

Daphna Yasova Barbeau (12:13.028)
That's right. Well, for the science, weren't,

Daphna Yasova Barbeau (12:31.704)
Yeah, and I think it just goes to show that we need everybody, right? All the people doing different types of scholarly work. We can't move forward without all the types of scholarly work, which is interesting. But alas, here we are today.

Ben Courchia, MD (12:37.23)
Yeah.

Ben Courchia, MD (12:49.391)
All right. We will leave you off now. Tomorrow we'll be back with Neo News. Eli and I are reviewing data on the administration of vitamin K, so a very polarizing topic. Stay tuned for that.

Daphna Yasova Barbeau (12:59.106)
Oof. As I present our new vitamin K refusal form this afternoon. That's on me. That's on me.

Ben Courchia, MD (13:05.497)
Finally.

Listen, better late than never. So it is much better. Just so you know, now people are going to email you saying, could we see your vitamin K refusal form? So be ready to email that around.

Daphna Yasova Barbeau (13:14.298)
I was preparing for journal club, okay? So.

Daphna Yasova Barbeau (13:22.478)
When it is official, I will be happy to share it.

Ben Courchia, MD (13:25.187)
Sounds good. Sounds good. All right, everybody. See you tomorrow.