#471 - 📑 Journal Club - The Complete Episode from October 17th 2026
Hello friends 👋
This week on The Incubator Podcast, Ben and Daphna dig into four studies testing what simple, evidence-based care looks like at the bedside. They open with a Swedish national cohort of nearly 4,700 very preterm infants, where a three-strain probiotic was linked to a 72% lower risk of death or NEC and near elimination of surgical NEC. Daphna then presents a study of banana puree as a natural thickener for NICU dysphagia, with oral intake reaching full volume within a week. Ben reviews the AIROPLANE trial, comparing 21% and 30% initial oxygen in infants born at 32 to 35 weeks. Daphna closes with the new AAP UTI guideline, covering shorter antibiotic courses, early oral transition, and imaging before VCUG.
----
The articles covered on today’s episode of the podcast can be found here 👇
Probiotics and Risk of Death, Necrotizing Enterocolitis, and Sepsis in Very Preterm Infants. Mitha A, Söderquist Kruth S, Sinervo E, Domellöf M, Brindefalk B, Abrahamsson T, Johansson S, Rakow A.JAMA Netw Open. 2026 Sep 1;9(9):e2635560. doi: 10.1001/jamanetworkopen.2026.35560.PMID: 42776528 Free PMC article.
We're going bananas! The effectiveness and safety of banana puree as an intervention for improving oral feeding in a level IV NICU. Carr CB, Ricciardi SE, Bery S, Levin R, Wilson-Costello D, Yalcinkaya G, Ryan RM.J Perinatol. 2026 Sep;46(9):1635-1640. doi: 10.1038/s41372-026-02765-z. Epub 2026 Jul 7.PMID: 42414579 Free PMC article.
Oxygen vs Air at Birth for Moderate- to Late-Preterm Infants: The AIROPLANE Cluster Randomized Crossover Trial. Peart SR, Manley BJ, Cheong JLY, Oei JL, Grobler AC, Moore CL, Allen RL, Alda MG, Bell G, Bhola K, Carman K, de Waal K, Dombagahawaththage CAN, Fan WQ, Fang AYW, Goyadi PK, Houghton KA, Jain KF, John J, Joshi D, Kunjunju A, Marshall IR, Mills JF, Nichol J, Parkhurst K, Priyadarshi A, Roberts CT, Thomas N, Tickell DJ, Walsh T, Yates KM, Ziffer J, Davis PG, Owen LS; AIROPLANE Trial Investigators and the IMPACT Clinical Trials Network for Mothers’ and Babies’ Health.JAMA. 2026 Sep 9:e2614201. doi: 10.1001/jama.2026.14201. Online ahead of print.PMID: 42714881
Clinical Practice Guideline for the Diagnosis and Treatment of Urinary Tract Infection in Children From 8 Days to 5 Years of Age. Alverson BK, Hains DS, Downs SM, El Feghaly RE, Forster CS, Jain S, Johnston A, Mattoo TK, Nelson CP, Olakanpo O, Otero HJ, Peters CA, Poppinga NM, Saha D, Schroeder AR, Senerth E, Pilcher L, Flinn SK, Morgan RL, Mustafa RA.Pediatrics. 2026 Oct 1;158(4):e2026078565. doi: 10.1542/peds.2026-078565.PMID: 42803578
----
Watch this week's Journal Club on YouTube 👇
----
The transcript of today's episode can be found below 👇
Ben Courchia, MD (00:00)
Hello, everybody. Welcome back to The Incubator Podcast Journal Club. Good morning, Daphna. What’s up?
Daphna Yasova Barbeau, MD (00:06)
Good morning. I’m good, but I got a full night’s sleep at home, and you’ve been working in the unit all night.
Ben Courchia, MD (00:12)
I had a good call, to be very frank with you. I’m recording from the hospital. It’s now, what, 8:40 in the morning, so I’ve given sign-out.
Daphna Yasova Barbeau, MD (00:21)
It must be because the day person took such good care of the babies before you came
on.
Ben Courchia, MD (00:25)
Yes, very much so.
Daphna Yasova Barbeau, MD (00:28)
But in fairness, you came in early for me, so it’s partly because you took good care of the babies during the day.
Ben Courchia, MD (00:32)
And you’re a good person to follow, I’m not going to lie. Things are addressed. And you want to know the truth, in perfect transparency? From a clinical standpoint and from a family-centered care standpoint, you’re a very good clinician to follow. The one thing I don’t like is that we have very different documentation styles. I say “style” because you document well; we just document differently.
Daphna Yasova Barbeau, MD (00:51)
Thanks, buddy. That’s nice of you to say.
Ben Courchia, MD (01:01)
So when you’re on for a week and I come after you, I need about 48 hours to make the notes mine.
Daphna Yasova Barbeau, MD (01:06)
To totally revamp my notes.
Ben Courchia, MD (01:10)
Yeah, but whatever. That’s why I said “style.” You tend to write a lot of your thought process into your notes. I tend to be a lot more cursory and factual: this is what we did, and I’ll explain my thinking if something warrants it. There’s no good or bad, but that’s the only thing that takes me some adjustment. For everything else, when I come after a Daphna week, I think, this is good.
Daphna Yasova Barbeau, MD (01:39)
Appreciate that.
Ben Courchia, MD (01:41)
For sure. And to be transparent with the audience, that’s not the case with every clinician I work with. Everyone has their own management style, but ours are similar, so that helps.
Daphna Yasova Barbeau, MD (01:52)
That’s probably it. Well, that’s really nice. Sometimes it’s good to get some feedback.
Ben Courchia, MD (02:03)
This is not what we’re doing, people, by the way. This is not an official feedback session.
Daphna Yasova Barbeau, MD (02:07)
Not official feedback. Well, I feel the same way about you, buddy. We’ve got a good team, and I wouldn’t say it if I didn’t believe it. The people know.
Ben Courchia, MD (02:13)
You don’t have to say that. Do we have any announcements? Actually, I know you have one, and I’m going to make one more, so we have two.
Number one: last week we recorded a very off-the-cuff, emergency podcast with Patrick Myers and Brooke Vergales about the ABP (American Board of Pediatrics) basically rolling back the whole plan for fellowship to go to two years. There are still some changes coming. It was a very interesting conversation, and we’ll probably keep talking about it, because ONTPD (the Organization of Neonatal-Perinatal Training Program Directors) is going to really focus on how we modernize neonatal training. It’s an ongoing conversation, but it was breaking news, and we were very happy to cover it on the podcast.
Daphna Yasova Barbeau, MD (03:09)
Breaking news.
Ben Courchia, MD (03:12)
It’s one thing to break news and talk about it ourselves, but it’s another to get guests who will come on the very next day. We’re very thankful to Dr. Myers and Dr. Vergales, who came on right after the AAP (American Academy of Pediatrics) meeting. There was a whole ONTPD meeting at the AAP this year, so that was cool. The other thing...
Daphna Yasova Barbeau, MD (03:29)
I was going to say thank you. I couldn’t make it work, so thank you for making it work and getting something recorded in the middle of the night, or the middle of the morning.
Ben Courchia, MD (03:43)
Afternoon, post-call.
Daphna Yasova Barbeau, MD (03:58)
We try to get it done. It’s just so great that people in our community are willing to drop everything, hop on for a bit, and update everyone. It also means they’re still really open to feedback and conversations about what’s happening, so people still have the opportunity to weigh in.
Ben Courchia, MD (04:02)
The teaser for that episode is that ONTPD is saying the rollback isn’t enough. The question at heart is how we train excellent physicians, and that question needs to be answered for the modern age. I think that’s precisely right.
Daphna Yasova Barbeau, MD (04:20)
And it wasn’t just a discussion of whether we need two years or three. There’s still some overhaul that needs to happen.
Ben Courchia, MD (04:25)
Exactly. There’s still a discussion to be had, and they’re going to put together a committee. Listen to the episode; there’s a lot going on.
The other thing I wanted to mention: you may have seen on our website that we’re welcoming a new sponsor, Pediatrix. We’ll have a whole separate episode on this partnership. Every time we bring on a partner, we want to be very transparent with the audience about what it means for us, what it brings to the platform, and what it brings to you. We’ll also have some very cool things to announce from this partnership. It’s a huge deal for us, because it lets us focus more on the podcast. And it’s good for the audience, because maybe CME (continuing medical education) is finally coming. Maybe we’ll tease that we have a date. So stay tuned. That episode is a bit meta and more behind the scenes than our usual content, but I feel like people want to know about the platform. They trust it, and transparency is a big deal for us.
Daphna Yasova Barbeau, MD (05:36)
I feel like one of the first handful of episodes we ever did was “How the heck are we doing this?” I hope people feel we’ve been really transparent since the beginning. We’ve only worked with people who’ve given us full content authority and freedom.
Ben Courchia, MD (05:53)
Yeah. The good thing about our platform, which I learned from some of the YouTubers I enjoy, is that there’s no middleman. If at any point something were fishy about the platform, people would notice right away. It doesn’t happen behind closed doors. We’re on the microphone, we talk, and you listen, so you can tell right away exactly what’s happening. That will continue, obviously.
Daphna Yasova Barbeau, MD (06:22)
And I’m not sure people realize that when you email the podcast or post on social media, we see nearly every message and respond personally.
Ben Courchia, MD (06:36)
The messages we respond to, it’s you or me, basically. We try not to make that too obvious, so we don’t end up in very long conversations, but it’s either Daphna or me. We tried having someone manage our social media community for a while, but the questions are medical, and that wasn’t going to work. Anyway, it’s already six minutes in and we haven’t even started Journal Club, but those are the announcements. You have one quick announcement. Go for it.
Daphna Yasova Barbeau, MD (07:00)
I do. We wanted to highlight our colleagues running the Neonatal Insights conference, January 29th through 31st, 2027, in Houston, Texas. You can attend in person or virtually. Their early bird registration ends October 14th. If you’re listening on release day, today should be October 12th, so you’ve got about 48 hours to look at what they’re doing and sign up. Plenty of time.
Ben Courchia, MD (07:30)
No, you have two days. All the procrastinators listening are going to say, “Plenty of time, 48 hours.”
Daphna Yasova Barbeau, MD (07:54)
Hold on. Neonatalinsights.com will take you right to the website, and you can click the big pink button. What I think is really interesting is that the conference uses a pro-con debate format on hot topics.
Ben Courchia, MD (08:09)
It’s a very innovative way to design conversations at a conference, and they have a very impressive faculty list. Definitely check them out.
Daphna Yasova Barbeau, MD (08:18)
For sure. If you’re familiar with one of the major board prep groups, they’re now also hosting Neonatal Insights. Some of the big clinical debates: the case for increasing calories or volume; should we add fat, add protein, or add the kitchen sink; should we use cord blood for platelet transfusions; how wasteful are our labs; and how do we balance cost and efficacy in the NICU (neonatal intensive care unit)? Lots of interesting topics.
Ben Courchia, MD (08:52)
All right. Thank you, Daphna. Let’s get into Journal Club.
Daphna Yasova Barbeau, MD (08:59)
You first.
Ben Courchia, MD (09:00)
Me first? No problem. I’m going to JAMA (Journal of the American Medical Association) Network Open, for an article by Ayoub Mitha titled “Probiotics and Risk of Death, Necrotizing Enterocolitis, and Sepsis in Very Preterm Infants.” I feel like I’ve talked about probiotics a lot over the past couple of months, but the data keep coming, and they’re very powerful. Regulating the use of probiotics is a priority for our field, because we believe they provide a significant benefit for our babies.
The reason I want to review this paper is that each of these papers is important on its own, and they’re not the same. I remember, as a trainee, getting a wave of articles on one topic and thinking, “Where does that leave me?” So I’m hoping that breaking them down individually helps with that. And at the end of this review, I’m hoping we can bring it all together a little.
The introduction doesn’t say anything we haven’t discussed before. Over the past decade, there’s been a growing body of evidence that specific probiotic strains reduce the incidence of NEC (necrotizing enterocolitis) and all-cause mortality in preterm infants. But results have varied across studies because of differences in strain selection, dosing, timing of initiation, duration, study population, and so on. There are also concerns about product quality and safety when implementing probiotics in clinical practice, as we’re painfully aware in the US. So probiotic use varies a lot across NICUs worldwide, including which strains are chosen and whether probiotics are used at all. At the AAP, we did an informal survey, asking people to raise their hand if their unit uses probiotics, and basically no one did. Maybe one hand went up.
This study comes out of Sweden and uses the SNQ (Swedish Neonatal Quality Register), which we’ve seen in other studies. It captures all admissions to the 38 neonatal intensive care units in Sweden and records data prospectively, either at discharge or continuously during the stay. They extracted records for preterm infants born from 28 weeks up to 31 weeks and 6 days, between 2017 and 2024. They excluded babies who died within the first three days of life, who died at any point from congenital or chromosomal abnormalities, or who had missing data.
During the study period, early feeding of very preterm infants was at the clinicians’ discretion. It was mostly mother’s own milk or donor breast milk; formula use, they say, was rare. Enteral feeding starts within a few hours of delivery and increases gradually depending on tolerance. Fortification is tailored individually, based on breast milk analysis done weekly or every two weeks with the help of computer software. Very modern, those Swedes, with their tech.
Daphna Yasova Barbeau, MD (12:35)
So sophisticated.
Ben Courchia, MD (12:41)
The exposure is interesting. Before 2020, probiotics were not used in Swedish NICUs. In 2020, probiotic supplementation became standard of care for very preterm infants under national guidelines from the Swedish Neonatal Society, which were based on recommendations from ESPGHAN (European Society for Paediatric Gastroenterology, Hepatology and Nutrition). They used the same three-strain blend we’ve seen before: Bifidobacterium infantis, Bifidobacterium lactis, and Streptococcus thermophilus, totaling 1 billion CFU (colony-forming units) in maltodextrin powder. This is the ProPrems product by NeoBiomics, diluted in 2 to 3 mL of breast milk and given once a day. Per the national guideline, it’s started as early as possible and stopped at 34 weeks.
There are two primary outcomes: a composite of death and/or NEC, and a composite of death and/or culture-proven late-onset sepsis, with death treated as a competing risk. I won’t go over the outcome definitions; they’re pretty standard. Late-onset sepsis probably needs a little clarification. It’s at least one culture-proven sepsis episode after day three of life, defined per the SNQ handbook as:
clinical symptoms
one or more positive cultures deemed significant
lab results compatible with sepsis
IV (intravenous) antibiotics for at least five days
no NEC diagnosis
They used Poisson regression for the composite outcomes, adjusted for gestational age, birth weight z-score, five-minute Apgar score, surfactant treatment, year of birth, and hospital clustering.
With all that in mind, the results. The SNQ included 4,922 very preterm infants born in Sweden at 28 to 31 weeks between 2017 and 2024. After exclusions, the analytic cohort was 4,695, about 4,700 babies: 45% female, 54% male, with a mean gestational age of 30 weeks. The probiotic group included 1,571 infants and the no-probiotic group 3,124 infants, 66% of the cohort. Compared with infants who didn’t get probiotics, supplemented infants:
were more often born at a university hospital
had a lower gestational age and a lower birth weight z-score
were less likely to have an Apgar score below 7 at five minutes
spent less time on invasive ventilation and more time on noninvasive ventilation
In total, 1% of the cohort died, at a median age of 11 days, and 18 of those deaths were from NEC.
It’s important to remember that this study uses a historical cohort as the control, because practice changed in 2020. That’s the whole point of saying they changed practice after 2020. Figure 1 shows trends over time in probiotic exposure, death, NEC, and culture-proven late-onset sepsis. It’s a beautiful figure: you see the probiotic uptake and then the results. After the national guideline came out in 2020, uptake rose rapidly, from 0% before 2020 to 16.9% that year, and to an average of 65% in 2021 to 2024. A very dramatic uptake.
Comparing 2017–2019 with 2021–2024, the periods before and after the national guideline, the overall rate of death and/or NEC fell from 3.4% to 1.5%, more than cut in half. Broken down by probiotic exposure, the rate in unexposed infants was 3.4% in 2017–2019 and 2.9% in 2021–2024, while the rate in exposed infants in 2021–2024 was only 0.8%. The point is to check that outcomes didn’t improve over time for other reasons that could explain the decrease. And as you can see, there wasn’t a significant change for unexposed babies. The rate of death and/or late-onset sepsis was 4.5% in 2017–2019 and 3.5% in 2021–2024.
Now the outcome risks. Babies who received probiotics had a lower risk of death and/or NEC, with an adjusted relative risk of 0.28 compared with babies who weren’t supplemented. Looked at separately, probiotic exposure was associated with a lower risk of both NEC and death. Similar reductions held when they split the babies by gestational age:
28 to 29 weeks: adjusted relative risk of 0.25 for NEC and 0.20 for death
30 to 31 weeks: adjusted relative risk of 0.35 for NEC and 0.09 for death
For surgical NEC (Bell stage 3) alone, probiotic exposure was associated with a particularly large reduction, an adjusted relative risk of 0.04. Very impressive.
In terms of sepsis and safety, probiotic exposure was not associated with the risk of death and/or culture-proven late-onset sepsis. They also tried to look at whether any sepsis cases were caused by the probiotic strains themselves, but the database didn’t capture that, so they don’t know. They did survey all the Swedish NICUs and ask whether any had seen a sepsis episode matching the probiotic strains, and the answer was no. You have to be careful with that information, because we’re moving away from the study’s original methods toward a survey. How was it checked locally? Did a medical director just say, “I don’t recall,” or did they actually look through some kind of database? But overall, it’s positive. I would think you’d remember an episode like that.
Daphna Yasova Barbeau, MD (20:19)
Agreed.
Ben Courchia, MD (20:20)
It would be etched in your brain. But it’s important to clarify.
Finally, they did a mediation analysis to explore how probiotic exposure affected mortality. They extended their joint Bayesian model to include NEC as a mediator of death, so the probiotic effect on mortality could work through NEC or through other pathways. The question is: does mortality improve because NEC is reduced, or because of something else? Probiotic exposure was clearly associated with lower mortality, and the exploratory mediation analysis suggested that much of that reduction runs through lower NEC rates. In other words, the mortality benefit is tied to the drop in NEC. Other papers we’ve reviewed couldn’t really make that connection. Maybe the sample size here is larger, I don’t know, but it’s very interesting.
The authors conclude that giving a multistrain probiotic to very preterm infants is associated with a reduced risk of death and/or NEC. These population-based findings may help inform shared decision-making between clinicians and families. And importantly, the results shouldn’t be generalized to other patient populations.
A few things to discuss. First, our good friend Stefan Johansson is the second-to-last author on this paper, and it’s disclosed that he’s the chief medical officer of NeoBiomics, which makes the probiotic. I think that’s an important disclosure. It’s in the paper, and you can interpret it as you wish, but we’re going to report it too.
Second, we’ve covered a lot of probiotic studies, so it’s worth asking where the evidence really lands. A week or two ago, we reviewed a study from the Netherlands that used essentially the same three-strain product in babies under 30 weeks or under 1,000 grams. NEC went down, overall mortality was unchanged, and non-NEC mortality was actually higher. The Swedish paper I’m reviewing today looks at babies 28 to 31 weeks and found that both NEC and death went down. And the Connection study we reviewed, which used a single strain of L. reuteri, found the mirror image: lower mortality and no change in NEC. So the NEC signal for these multistrain products above 28 weeks seems very consistent and quite large. What happens to mortality depends on the population, and I guess on the product. And sepsis has gone in a different direction in every study we’ve looked at.
Ravi Patel has a very nice editorial in this issue of JAMA Network Open, and he makes the point I’d like us to end on. Severe NEC was nearly eliminated in this Swedish cohort: one case among almost 1,600 exposed infants, versus 40 among the unexposed. He also raises immortal time bias: infants had to survive long enough to be supplemented, so the mortality estimates in particular may look somewhat better than they really are. But the direction of the effect agrees with the broader literature. And he argues that the benefit far outweighs the risk of invasive infection, which he puts at roughly 1 in 2,500.
So it’s important to understand where these populations differ. The Swedish cohort was larger, and the Dutch study we did a couple of weeks back included smaller babies. You can see the size of the effect shrinks as we move between populations. Also pay attention to the products being used. The Swedish and Dutch cohorts used the same product. The Connection study in the US used a very different, single-strain product, so it may not even be comparable. And it may well be that in the US we’ll get that single-strain product first, because of how probiotics are now regulated. I hope this helps.
Daphna Yasova Barbeau, MD (24:58)
It’s an interesting conversation, and that’s always been part of the “problem”: which combination do we use? Is one safer than another? But at this point, it seems like we should just pick one, study it first, and get some really good information.
Ben Courchia, MD (25:14)
In the US, it looks like every single strain will need its own study to be approved.
Daphna Yasova Barbeau, MD (25:20)
In its own right, which it should. But at some point, we’ll need enough data on one of them to move forward. It’s wonderful that other countries keep giving us data, because we’ll have retrospective data. Like you said, it’s coming out, and hopefully all that data will get us a product we can use for our babies.
----
Daphna Yasova Barbeau, MD (26:02)
I have a little refresher today after a nice long review on probiotics yesterday, so thank you for going through that. Sometimes, as a neonatologist at the bedside, it feels like, well, what do we do about it as individuals? I just hope we’ll keep reporting these. And we have lots of colleagues doing the advocacy work, so there are lots of people to join if you’re interested in doing more on the probiotics discussion.
Ben Courchia, MD (26:40)
And it is a priority at the national level. The AAP Section on Neonatal-Perinatal Medicine is looking at it and trying to help, because at an individual level, at the bedside, it’s tough to make a dent in terms of advocacy. Just know that these organizations are actually quite active on that front.
Daphna Yasova Barbeau, MD (27:01)
Yes, and there’s power in numbers. It’s never too late to join the effort, so that’s a good reminder.
My palate cleanser today is an article we’ve been looking forward to sharing. It’s called “We’re Going Bananas! The Effectiveness and Safety of Banana Puree as an Intervention for Improving Oral Feeding in a Level IV NICU,” in the Journal of Perinatology. The lead author is Kara Carr and the senior author is Rita Ryan. It’s essentially a pilot study evaluating the safety and feasibility of banana puree as a thickener in NICU patients with dysphagia.
A little background. This is something we’ve been talking about a lot on the podcast and in our own unit, probably for the last six to twelve months. So many of our NICU resources go to the first few weeks of the most critical patients, and we should be spending a lot of time and effort there. But the majority of our unit is made up of these quote-unquote “feeder-growers.” We come in, we check in, we ask, how is feeding going? Studies have consistently shown that the majority of babies, even moderately preterm babies (they quote 29 to 33 weeks), have prolonged hospitalizations, typically related to whether they can feed well enough to go home. They cite a figure of 65% of babies between 29 and 33 weeks staying in the NICU longer than we expect because of feeding.
Ben Courchia, MD (28:50)
I was going to say, I thought you were going to say we’d spoken a lot on the podcast in the past few months about bananas, and I was thinking, I don’t remember that. But we did talk about certain populations being underrepresented in studies.
Daphna Yasova Barbeau, MD (29:04)
Yes. I’m not sure we’ve ever talked about bananas on the podcast. So what’s the deal with bananas? We know that thickened feeds are the mainstay for managing dysphagia in adults and children, but we still have very few thickeners we use in neonates, mostly because of concerns about NEC, and whether babies can digest and tolerate them.
The authors said they were on the hunt for a viable thickener, and they had a few qualifications. They wanted it to be sweet, which makes a lot of sense: if a baby doesn’t want to eat or has oral aversion, maybe we can entice them with something sweet. They also wanted it to be homogeneous, safe, and of course effective.
It turns out lots of people are looking at bananas as a thickener. Dole, the fruit company, actually has whole pages on Dole.com dedicated to using banana as a home thickener. Bananas contain a starch with a dense hexagonal structure of double helices. That structure keeps amylase, both in the banana and in the patient, from quickly breaking down the starch, which allows a slight thickening effect over time, particularly with human milk. And for generations, banana has been considered easy to digest, as part of the BRAT diet (bananas, rice, applesauce, toast). So that’s why bananas.
Ben Courchia, MD (30:44)
Wow, the BRAT diet.
Daphna Yasova Barbeau, MD (30:54)
I know. It has fallen out of favor, coincidentally, but I don’t think that means bananas aren’t still a key component.
Ben Courchia, MD (30:55)
We live in South Florida. It’s a staple of the diet here.
Daphna Yasova Barbeau, MD (30:58)
Yes, all kinds of bananas and their cousins down here. I think our families would be thrilled if we said plantains, bananas, we can have them all. And I know one of our clerks would be thrilled to start making banana purees.
Ben Courchia, MD (31:18)
We have a clerk who likes to make fruit juices. She’s a fruit juice expert.
Daphna Yasova Barbeau, MD (31:22)
We’ve got a fruit juice bar in the NICU lounge.
All right, back to the serious stuff. This was a single-center retrospective study conducted at University Hospitals Rainbow Babies & Children’s Hospital at Case Western Reserve University. It’s a small cohort, and I’ll admit it’s not homogeneous. There are lots of different kinds of babies, which I’ll get into, and I think that’s important. I’d also like to point out, for anyone already rolling their eyes, that this is a fairly mature group of babies, which is another important point.
Ben Courchia, MD (32:06)
The 22-weekers are not getting their first feed with some bananas.
Daphna Yasova Barbeau, MD (32:14)
Not yet, anyway.
The study included patients at 36 weeks’ postmenstrual age or more who received banana puree as a feeding intervention in the NICU between 2015 and 2023. Patients eligible for the banana intervention, which they call BI, were identified by the neonatal feeding therapists in their unit: the OTs (occupational therapists). These babies had to have failed to improve their oral intake with other standard feeding strategies, which we know include changing the bottle, changing the nipple, changing the pacing, and so on. Infants were not considered BI candidates if they had a known latex allergy, a family member with a banana allergy, confirmed hyperkalemia, or a history of surgical NEC.
They looked at:
the postmenstrual age at each baby’s first oral feeding and at the first banana feeding
why the babies got bananas
oral intake per day at that time
what the enteral feeds looked like (human milk or formula)
growth parameters
discharge characteristics
Additionally, as a routine safety measure in their unit, the OTs trended each patient’s 24-hour saturation histograms and charted significant events. Per the implementation protocol, they would have stopped the intervention for any baby if there were concerns it was affecting respiratory status.
Importantly, full oral intake was defined as a minimum of 120 mL per kilo per day of human milk or formula. That was the goal for babies who were really having difficulty feeding. I think we’ve all been in that scenario: can we reduce the volume we expect this baby to take by increasing calories, so they still get enough for growth? So just know that the target was 120 mL per kilo per day.
They collected data on growth, respiratory support, and need to go NPO (nothing by mouth) before and after the intervention. They were particularly interested in apnea, bradycardia, and desaturation events during feeding that required a change in respiratory support. They assessed patients at several points: seven days before the banana intervention, the day before it started, the day of the first banana, and 7 and 14 days after starting bananas. They also looked at a number of discharge characteristics.
The median gestational age of infants who received banana puree was 29.7 weeks, with a median birth weight of 1,320 grams. Notably, the youngest baby was 25.4 weeks’ gestational age and the oldest 37.6 weeks; the smallest baby weighed 810 grams and the biggest 2,890 grams. The first oral feeding in this group happened at about 38 weeks’ postmenstrual age, which I thought was interesting. So this was clearly a group of babies who, for whatever reason, couldn’t feed until an average of 38 weeks, which is a little late. A high-risk group. The earliest first oral feeding was at 35 weeks, and one baby was almost 43 weeks at the first oral feeding. So that’s their group.
About four weeks after their first oral feeding, patients were introduced to banana puree. So they really gave these babies enough time to work on oral feeding first. The median postmenstrual age at first banana puree was about 41.7 weeks. Patients received bananas for a median of nine days, with a median of six therapy sessions that included feeding with the banana intervention. However, all of the feeds contained bananas; the remaining feedings were done by bedside nurses or parents.
So which babies got the banana intervention? Mostly, the OTs identified babies whose feeding behaviors were contributing to poor oral intake: oropharyngeal incoordination, reflux, and oral aversion. About 53% of the patients started bananas because of general incoordination with oral feeds that persisted despite all the standard-of-care interventions. The authors stress that bananas weren’t used to speed up the normal maturation of feeding skills. They weren’t trying to use bananas early to get babies feeding sooner. They were really targeting babies who had failed the typical feeding interventions and had clinical signs of dysphagia.
In total, they had 43 infants. Before starting bananas, about 47% were receiving some amount of human milk, and about 37% were receiving human milk as most of their feeds. But more commonly, feeds were preterm, term, or elemental formula. The percentage of patients receiving human milk as the main part of their feeds stayed stable after the banana intervention, so babies didn’t have to come off human milk. At discharge, the percentage receiving any human milk went from 47% to 42%.
At the start of the banana intervention, patients were receiving a median of 145 mL per kilo per day of enteral feeds, but most were taking only about 70 mL per kilo per day by mouth, so only about 50% of their feeds. Remember, most of these babies had already been working on feeding for about a month. By day seven of the banana intervention, the median oral intake had increased to 117 mL per kilo per day. That was statistically significant, with 79% of patients reaching the goal of 120 mL per kilo per day of oral feeds.
Ben Courchia, MD (39:04)
That’s incredible.
Daphna Yasova Barbeau, MD (39:05)
I know. By day seven, the median percentage of feeds taken by mouth went from 50% to 100%, which was also statistically significant.
Fifteen patients were already taking full PO (by mouth) feeds before starting bananas. I think that’s important to point out, but these babies were having complications during feeding that kept them from being discharged:
persistent incoordination: five babies
apnea, bradycardia, and desaturations during PO feeds: four babies
reflux: five babies
an “unknown” reason: one baby
What about weight gain? They were concerned babies might gain too much weight, since bananas could be considered a high-calorie addition. Or, if babies had been on higher-calorie feeds, adding banana might reduce the calorie density. But there was no significant difference in weight gain velocity in the week before versus the week after starting bananas.
Patients were on all types of respiratory support: room air and nasal cannula, and since 2021 the unit has allowed feeds on high-flow nasal cannula. There were no perceived adverse respiratory effects. One patient needed more respiratory support to improve feeding endurance, but that was felt to be unrelated to the bananas. Most patients on high-flow nasal cannula on day one of the banana intervention were weaned to lower support during that first week of bananas. So it didn’t seem to be costing them anything respiratory-wise.
About 53% of patients were discharged home on bananas. Most patients who weren’t on bananas at discharge had stopped during their hospital stay after achieving improved, consistent oral intake; that was about 15 babies. Other reasons for stopping were parental preference in one baby, a baby still refusing all oral feeds in one, and lack of functional gains in one. One patient stopped bananas because of no perceived benefit, but no patients stopped because of any adverse reaction.
At discharge, 81% of patients were taking 100% of their feeds by mouth. Three patients (7%) went home with a G-tube (gastrostomy tube), and five patients (12%) with an NG (nasogastric) tube. Of the eight patients discharged with feeding tubes, three had reached the goal of 120 mL per kilo per day of oral feeds, but their intake was still felt to be inconsistent or unreliable. Of the 15 patients who started bananas for reasons other than failing to reach full PO feeds, 63% were discharged home still on bananas.
In conclusion, the authors say the study demonstrates the feasibility and safety of banana puree for dysphagia in NICU patients. Infants who had difficulty swallowing thin liquids improved within seven days of banana puree being added to formula or human milk, again under the supervision of a feeding team. I’m not sure it’s ready for prime time, but I think it’s something we’re going to have to look at.
Ben Courchia, MD (42:48)
This is a great paper to review. It lets us tease that in the coming months we’ll release a five-episode series on biostatistics, about how we interpret data, with Dr. Joe Kaempf. It’s really good. We’ve recorded two episodes already, we’ll record the third soon, and they’ll come out as a five-part series. And the point of that series is exactly what this paper shows. It’s not saying banana is the be-all and end-all, that it trumps everything else and you must use banana. The idea is: is this another tool you can add to your arsenal? Is it safe, and at the very least no worse? And it seems like it is. And as you mentioned, the intervention is very low-tech.
Daphna Yasova Barbeau, MD (43:34)
Low-hanging fruit. We love a low-tech, inexpensive intervention, right?
Ben Courchia, MD (43:43)
You go to the store, you get the Gerber banana thing, and that’s it.
Daphna Yasova Barbeau, MD (43:48)
I would want to make it fresh, but still.
Ben Courchia, MD (43:52)
My point is that a lot of the products we have in the NICU are very expensive, and with this, you could Instacart that stuff.
Daphna Yasova Barbeau, MD (44:03)
You could Instacart that stuff. And parents can go home on it, right?
Ben Courchia, MD (44:06)
Yes. So from all those standpoints, it’s very neat. And this is where I take off my scientific hat, but it sounds like something you’d try at home as a parent anyway. You go home, you’re struggling, somebody says, “Just give him some bananas,” and you say, “Sure,” and then he loves the bananas. So why not in the NICU? We shouldn’t be so sophisticated.
Daphna Yasova Barbeau, MD (44:34)
I love it, I really do. I’m hopeful they’ll roll out a randomized controlled trial; we’ll see how that goes. But do we need one? I don’t think we do.
Ben Courchia, MD (44:48)
To try banana in a full-term-equivalent baby.
Daphna Yasova Barbeau, MD (44:52)
A full-term baby, and you say, let’s give it a go. I think it will be case by case, but it makes sense.
Ben Courchia, MD (45:00)
For sure. Anyway, that’s time. More resources should be allocated not to bananas but to the PDA (patent ductus arteriosus). No, I’m kidding.
Daphna Yasova Barbeau, MD (45:09)
Maybe we could just treat the PDA with bananas.
Ben Courchia, MD (45:14)
With bananas.
----
Ben Courchia, MD (45:59)
There’s a very interesting article published in JAMA. The first author is Stacey Peart, and it’s called “Oxygen vs Air at Birth for Moderate- to Late-Preterm Infants: The AIROPLANE Cluster Randomized Crossover Trial.”
Daphna Yasova Barbeau, MD (46:28)
Love it. Oxygen people are very good at coming up with... what’s the word? Not mnemonics. Good trials with good names.
Ben Courchia, MD (46:37)
And good trials as well, I guess.
Moderate- to late-preterm infants, born at 32 to 36 weeks, represent 85% of our preterm population. But in the resuscitation literature, they’re not as well represented as other preemies. The question of the optimal FiO2 (fraction of inspired oxygen) for starting respiratory support in the delivery room has been a focus of research. We know that too much FiO2 during resuscitation can cause harm in both term and preterm infants through the generation of oxygen free radicals. In term newborns, data have shown that starting respiratory support with 21% (room air), compared with 100% oxygen, is associated with reduced mortality. I invite you to listen to our Giants of Neonatology interview with Dr. Ola Saugstad, the investigator who pioneered these studies. It’s a very interesting conversation.
That evidence underpins current recommendations from ILCOR (the International Liaison Committee on Resuscitation) to start respiratory support at an FiO2 of 21% for babies born at 35 weeks or later. In contrast, for neonates born before 32 weeks, an ILCOR-commissioned meta-analysis found that almost all who started respiratory support in the delivery room at 21% ended up on more oxygen by 10 minutes of life. The 2025 guidelines from ILCOR and the European Resuscitation Council recommend starting respiratory support for neonates under 32 weeks at an FiO2 of 30%.
So where’s the gap? It’s the babies in between, 32 to 35 weeks, and there are minimal data on this population specifically. Four small studies of initial FiO2 included preterm neonates born up to 35 weeks, but most participants were under 32 weeks. So no published study has targeted the optimal initial FiO2 for neonates born at 32 to 35 weeks. ILCOR guidelines currently state there’s insufficient evidence to make recommendations for this age group, and that lack of evidence is leading to inconsistent clinical practice. I could not agree with that statement more. An international survey across 21 high- and middle-income countries found that 42% of respondents start resuscitation of moderate- to late-preterm infants at 21% FiO2, 35% use 30%, and the rest use a higher FiO2.
So the objective of AIROPLANE (Air or Oxygen for Preterm Newborns, an embedded trial) was to determine whether, in neonates born at 32 to 35 completed weeks who receive respiratory support starting in the first three minutes after birth, beginning at 21% or 30% reduces the need for ongoing support.
Now the methods. This was an open-label, cluster randomized crossover trial conducted in 26 Australian maternity hospitals. Hospitals in Victoria and New South Wales caring for neonates born before 36 weeks were eligible. They included all neonates born at participating sites between 32 weeks and 35 weeks and 6 days who:
received active care
had no known major cardiorespiratory or craniofacial anomalies
received respiratory support in the delivery room within three minutes of birth
They randomized the sites one-to-one, using simple block randomization, to start respiratory support at either 30% or 21%. Halfway through the trial, each site crossed over to the other initial FiO2 after a one-week washout period. So they randomized the sites, not the babies.
Delivery room respiratory support, including its timing and mode, followed the ILCOR guidelines. Clinicians were not blinded to the initial FiO2. Eligible newborns had a pulse oximeter secured to the right hand or wrist and started respiratory support at the randomized FiO2. If respiratory support was ongoing, the randomized FiO2 was continued until three minutes of age, or until at least one minute of support had been given, whichever was later. This reflects the guidelines’ emphasis on effective ventilation rather than oxygenation in the first three minutes after birth. If the heart rate was under 100, positive pressure ventilation was given, as all the guidelines recommend. If the heart rate was under 60 or the baby needed intubation, supplemental oxygen was increased per standard resuscitation guidelines. After three minutes, FiO2 was titrated to saturation targets.
The primary outcome was receipt of ongoing respiratory support when the baby left the delivery room. They chose this pragmatic outcome on the hypothesis that if additional oxygen improved transition, newborns would need less support leaving the delivery room. Newborns who died in the delivery room were analyzed as if they had needed more respiratory support, so that this population didn’t go unaccounted for. That’s unfortunate, obviously.
The prespecified secondary outcomes were:
Apgar scores at one and five minutes
all delivery room resuscitation interventions
maximum FiO2 used in the delivery room
type and duration of ongoing respiratory support beyond the delivery room and at any time during the hospitalization
surfactant administration
length of hospital stay
need for transfer to another hospital
death before discharge
There’s a very nice statistical section. They estimated they needed about 1,200 newborns, but I’ll let you read it, because we’re short on time.
Twenty-nine sites were randomized, and three withdrew before recruitment, so 26 sites recruited from December 2022 to September 2025. A total of 1,828 infants received the study treatment. Ten were excluded after treatment, leaving 1,818 newborns in the analysis: 964 randomized to 30% FiO2 and 854 to 21%. Newborn characteristics were similar between groups. The mean gestational age was 34 weeks, mean birth weight was 2,165 grams, and 44.7% were female. The randomized treatment was correctly given to 92% of newborns in the 30% group and 96.3% in the 21% group.
So, outcomes. Ongoing respiratory support when leaving the delivery room was given to 72.6% of newborns randomized to 30% FiO2, compared with 73.3% of those randomized to 21%. That difference was not statistically significant. Most newborns who left the delivery room on respiratory support were on some form of noninvasive support. There was no evidence that the treatment effect differed by gestational age or hospital capabilities. So on the primary outcome, 21% appears to be just as good as 30%.
Now let’s look at the secondary outcomes and see whether anything might push us in a different direction. Newborns randomized to 30% had lower odds of receiving higher levels of support in the delivery room than those randomized to 21%. That included noninvasive positive pressure ventilation, endotracheal or supraglottic airway ventilation, and chest compressions or epinephrine.
The mean maximum FiO2 during stabilization was 52% for newborns randomized to 30% and 55% for those randomized to 21%. That’s interesting, because you might think that if you start at a higher level, you’d end up at a higher maximum FiO2 once you turn it up than if you’d started lower. But clinicians were clearly able to titrate to each baby’s needs, because both groups ended up around the same maximum, roughly 50%.
A lower proportion of newborns randomized to 30% received endotracheal ventilation after leaving the delivery room: 6.4% versus 9.3%. There were two deaths among newborns randomized to 30% and seven among those randomized to 21%. There were no differences in the other secondary outcomes.
So the trial found no difference in ongoing respiratory support when leaving the delivery room for moderate- to late-preterm infants started at 30% versus 21% FiO2. The secondary finding, that babies started at 30% were less likely to need higher levels of support in the delivery room and less likely to receive endotracheal ventilation afterward, raises questions about the optimal starting oxygen concentration for this population. Because these are secondary outcomes, the authors say they should inform international practice recommendations and lay the groundwork for further trials.
I think this is quite interesting, because it touches on a subject we’ve already covered a lot on the podcast: what oxygen concentration to start at. We reviewed the TORPIDO 30/60 trial, which looked at a different population, the smallest of the small, 23 to 28 weeks. What’s interesting to me is that the question of what FiO2 to start at for this population still isn’t resolved. You’d think that in a maturing field like ours, we’d know how much FiO2 to start with for a 34-weeker. And the answer is that we really don’t. I find it quite baffling that we’re only getting to this question now.
It seems we still don’t have a good answer for what the final recommendation should be. Based on the primary outcome, starting at 30% or 21% seems equivalent. It’s unclear whether we should change our practice based on the secondary outcomes, but it’s definitely interesting. It also highlights a gap in research for this specific population on this specific question, which not everyone may be aware of.
Daphna Yasova Barbeau, MD (58:30)
So we keep doing what we’re doing for now.
Ben Courchia, MD (58:38)
I think most people will wait until ILCOR, NRP (the Neonatal Resuscitation Program), or whatever guideline you follow incorporates some of this data. And I agree with you: the data aren’t so striking that you’d say, “I’m switching practice right now, before the recommendations come out, because this is very concerning.” So for now, you can keep doing what you’re doing, whether that’s 21% or 30%. Consistency of care is probably your best bet. Stay consistent with your practice, and only change once there’s enough data to move the needle.
----
Ben Courchia, MD (59:25)
We’re finally getting to that paper.
Daphna Yasova Barbeau, MD (59:40)
Well, it’s not that we’re behind. It’s very new.
Ben Courchia, MD (59:44)
Very new. No, I mean this week. We could have opened the week with this article.
Daphna Yasova Barbeau, MD (59:48)
That’s right, we could have opened the week with it. So why are we excited about it? This October, the AAP released a new clinical practice guideline for the diagnosis and treatment of urinary tract infection in children from 8 days to 5 years of age. The lead author is Brian Alverson and the senior author is Reem Mustafa. Before I forget, I’ll also mention that the guideline comes with three accompanying technical reports, so I encourage everybody to take a look.
What I’ll try to do is highlight what’s new and different, and then go over the key action statements. They recognized a gap for babies younger than two months, who weren’t covered by the 2011 guidelines. So we’re very thankful they’ve been thinking about us neonatal professionals.
The other thing I’ll say is that this document is chock-full. It’s 75 or 78 pages long, packed with great background information and data on risk factors for urinary tract infection and on anomalies seen with urinary tract infection. If you’ve got a rainy afternoon, I think everybody should read it cover to cover to really understand how we got to these guidelines.
Again, this is an update from 2011, which is a long time to be using the same data. The 2011 guidelines focused on febrile infants aged 2 to 24 months without anatomic problems. The current guidelines cover 8 days to 5 years, so they also reach up to toddlers just entering school age. They focus on medically non-complex children, but I think they give us a lot more data on children who are medically complex and may have renal anomalies.
In 2011, it was really about febrile infants or young children with no apparent source. In the 2026 guideline, they still don’t specify much for the 8-day to 2-month group, so it’s probably still febrile infants between 8 days and 2 months. For 2 to 24 months, it’s really fever without a source. And it now includes children 2 to 5 years who may not have a fever but have symptoms consistent with a UTI (urinary tract infection).
On timing of evaluation, they’re more explicit that we should consider UTI even in the first 72 hours of fever, because the sooner we start treatment, the better, with fewer long-term consequences for the kidneys.
How are we diagnosing it? A clean-catch urine still isn’t felt to be the best method; we want a catheterized urine if at all possible. Interestingly, they’re changing the culture threshold for catheterized urine. We previously used 50,000 CFU or more. Now they’re telling us to look for 10,000 CFU or more, plus some evidence of inflammation. I think many of us were already considering that in practice, so it’s nice to see.
Ben Courchia, MD (1:03:33)
And the other parameters, I’m assuming you’ll talk about them, like the UA (urinalysis)?
Daphna Yasova Barbeau, MD (1:03:37)
Yes, the UA. Before, they used a vague “significant pyuria and/or bacteriuria.” Now they’re giving us age-specific thresholds for the urinalysis, which is exciting. They’re also giving us age-specific guidance on antibiotic duration and route. And they give more guidance on imaging. The real change is that we get a renal ultrasound, and VCUG (voiding cystourethrogram) really isn’t recommended unless the renal ultrasound shows abnormalities. They give us a much better definition of what counts as abnormal enough to proceed to VCUG. So that’s the gist. I’ll also touch on antibiotic prophylaxis.
Ben Courchia, MD (1:04:37)
We reviewed a paper on antibiotic prophylaxis showing it didn’t do much for these kids and created more resistance. So it’ll be interesting to see what they say.
Daphna Yasova Barbeau, MD (1:04:45)
That’s exactly right. So I’ll start with the key action statements.
Key action statement one applies to medically non-complex children: those 2 to 24 months with fever without an identified source, and older children aged 2 to 5 years with symptoms of a urinary tract infection. We should evaluate them for UTI within 72 hours of symptom onset. Previously, they recommended a longer period of observation. And like I said, we know the sooner we treat, the better.
Key action statement two: in children with UTI, the AAP guideline panel suggests a shorter course of antibiotics rather than a longer one. Previously, more than seven days was typically recommended. Now they hope we can treat most UTIs with less than seven days of antibiotics, though I’ll be very specific about age. They have a really nice summary table for this key action statement:
8 days to 2 months: still considered high risk, particularly with a febrile UTI. They still recommend a prolonged course of 7 to 14 days, so that hasn’t changed.
Older than 2 months with a febrile UTI plus known GU (genitourinary) abnormalities, bacteremia, or poor clinical improvement on antibiotics: also high risk, and still treated with a prolonged course of 7 to 14 days.
2 to 24 months with a febrile UTI, no known GU abnormalities, no bacteremia, and a good response to antibiotics: low risk, so treatment comes down to 5 to 7 days. Even with pyelonephritis, the course comes down to 7 days.
2 to 5 years with a non-febrile UTI: low risk, with an antibiotic duration of 3 to 5 days.
Ben Courchia, MD (1:06:59)
So you’re looking at the box on page 14. It’s very nice; maybe we’ll post it on social media. Interestingly, it’s not age alone that makes them high or low risk. Age is a component, together with whether they have a febrile or non-febrile UTI. And the presence of a fever is obviously a key component.
Daphna Yasova Barbeau, MD (1:07:12)
It’s the risk category. Age is a component of the risk category.
Ben Courchia, MD (1:07:29)
So you could have a low-risk child between 2 and 5 years old, but you could also have a high-risk patient in that same range, and they tailor treatment to that. For low risk, you said three to five days, and for the highest risk, seven to fourteen days.
Daphna Yasova Barbeau, MD (1:07:49)
Which is potentially longer than we treat.
Ben Courchia, MD (1:07:53)
Yes, because the theme in UTI research has been fewer antibiotics. At least that’s how it feels to me: a lot of papers come out saying you don’t need as long, you can get away with three days. It’s interesting to see that for some populations three days may be enough, but some babies with a febrile UTI will need 14 days.
Daphna Yasova Barbeau, MD (1:08:14)
And to be clear, if you really get into the supporting statements, especially given what we’ve reviewed this past year on ELBW (extremely low birth weight) infants and neonates in the NICU with UTI, it’s not that they need 14 days. It’s that we’re not sure yet whether they can have less than 7 to 14 days. I think we may get there, but the AAP isn’t ready to recommend it yet, if that makes sense.
Ben Courchia, MD (1:08:42)
Fair enough.
Daphna Yasova Barbeau, MD (1:08:44)
Now, this is exciting. Key action statement three: in children aged 8 to 28 days diagnosed with a UTI, the panel suggests parenteral antibiotics with early transition to oral antibiotics after clinical improvement, rather than parenteral antibiotics alone. Yes, I know. And they have a nice flow sheet for it. There are a bunch of flow sheets at the end of the paper, which I’ll review if we have time. And in children older than 28 days and younger than 5 years, the panel recommends oral antibiotics rather than parenteral or sequential antibiotics requiring hospitalization. This is one of the strong recommendations, and there aren’t many of those, so that’s exciting.
Ben Courchia, MD (1:09:31)
What a pain it was. We had babies on IV antibiotics, and if we weren’t a NICU but an inpatient pediatric unit, we would have transitioned them to oral after a while. Instead, I’d have to stick this baby for a new IV because I didn’t have the evidence to support transitioning to PO. So it’s exciting.
Daphna Yasova Barbeau, MD (1:09:34)
Just a few more days, and they keep losing their IVs. Love that.
Key action statement four: in children younger than 5 years diagnosed with a first febrile UTI, the panel recommends renal and bladder ultrasonography rather than skipping it and going straight to VCUG. This is really well outlined, especially for our population, in one of the flow sheets: figure 4 in the paper. Basically, you get your imaging, your renal and bladder ultrasound. Are there abnormalities? If yes, go to VCUG.
So what are those abnormalities? Hydronephrosis, which includes all the terms radiologists use: pyelectasis, pelviectasis, caliectasis, calyceal dilation, and pelvicaliectasis. All of those count as hydronephrosis and potentially should be worked up with a VCUG. Any dilation, really.
Ben Courchia, MD (1:10:54)
Say that again? In figure 4, it sends you to box 3 whenever you have renal or bladder ultrasound anomalies. These include hydronephrosis, ureteral dilation, renal parenchymal anomalies, bladder or urethral structural anomalies, renal ectopia, hypoplasia, dysplasia, or an abnormal size discrepancy. That’s what they list.
Daphna Yasova Barbeau, MD (1:11:27)
Yes. I think we all felt that ureteral dilation, parenchymal abnormalities, structural anomalies of the urethra, and renal ectopia were clearly abnormal. But they give a much better definition of what to consider hydronephrosis in that first line: even just dilation of the renal pelvis. I think that’s interesting.
I also want to highlight this. If you have one of these abnormalities, you go to VCUG. If you don’t, and it’s the first UTI, some babies still qualify for VCUG despite a normal renal and bladder ultrasound: babies less than one month old, which is a lot of our babies; babies whose infection isn’t E. coli; and uncircumcised males. If you can answer yes to any of those, it’s an option box.
Ben Courchia, MD (1:12:22)
Yes, it’s an option box at the bottom. And interestingly, there’s the timing of the renal ultrasound: after the acute symptoms resolve.
Daphna Yasova Barbeau, MD (1:12:42)
That was always a point of discussion: when exactly is that? But give it some time, I think. I don’t want this to take too long, but these are very useful flow charts, so I want to make sure we hit some of them. Let’s get back to the key action statements.
Ben Courchia, MD (1:13:07)
We did number four, and now we’re getting to five.
Daphna Yasova Barbeau, MD (1:13:09)
Key action statement five ties back to number four: it recommends against routine VCUG after a first febrile UTI in children younger than 5 years who have a normal renal and bladder ultrasound and none of the additional conditions we discussed.
Six: in children with low-risk VUR (vesicoureteral reflux), the guideline panel suggests against routine CAP (continuous antibiotic prophylaxis). Again, a conditional recommendation. They define low risk as lower-grade VUR, grades 1 to 2, and no bowel and bladder dysfunction. I think that’s an important piece. Maybe not for our NICU babies, but we’ve all experienced a child like this, some of us in our own families. Functional constipation is excluded, but any child with bowel and bladder dysfunction would not be considered low risk. So they must have low-grade VUR and no bowel or bladder dysfunction. For babies with high-risk VUR, CAP is an option. Grade 3, they say, is uncertain; they don’t have a good recommendation.
Ben Courchia, MD (1:14:28)
So CAP being continuous antibiotic prophylaxis. They’re basically saying grade 3 is a gray zone. If you feel this baby is at high risk, you might want to give continuous antibiotic prophylaxis, but if you say no, that’s okay too.
Daphna Yasova Barbeau, MD (1:14:52)
Yes. And even with high risk, they’re saying CAP is an option. I think they’re leaning toward yes, but it’s not strong.
Ben Courchia, MD (1:14:57)
So you feel like they want to say no, but they can’t?
Daphna Yasova Barbeau, MD (1:15:00)
That’s exactly right. Because high-risk VUR includes grade 4 and 5 reflux, intrarenal reflux, renal scarring or nephropathy, renal or ureteral structural or cortical abnormalities, recurrent pyelonephritis or bacteremia, and again bowel and bladder dysfunction. These are complicated kidney features.
So those are the key action statements. I think we’ve covered the antibiotic courses and the VCUG situation pretty well. But I’d like to highlight figure 1 as well: the 8- to 60-day-old febrile, well-appearing infant with no evident source of infection, who should be evaluated immediately. You know that what we do for the febrile infant is changing too, but right now we’re just talking about UTI.
If the baby is less than 22 days old, they should get a urine culture and UA by catheter, along with the other labs in the febrile infant guidelines, and we should start antimicrobial therapy. Any culture in these babies that grows 10,000 CFU or more of a pathogenic organism should be treated for more than seven days, according to the organism’s antibiotic sensitivities.
If your baby is 22 to 60 days old, you can obtain urine for urinalysis by catheter, bag, or stimulated void, and follow the guideline. Their UA parameters are five or more white blood cells, or trace or more leukocyte esterase. If yes, you culture the urine and proceed as above. If no, meaning fewer than five white blood cells and less than trace leukocyte esterase, UTI is unlikely, even in this group.
Ben Courchia, MD (1:17:20)
This goes back to understanding what you’re treating. You have a fever: is it a fever of unknown origin? They’re basically saying that if you don’t have five or more white blood cells per high-power field or trace leukocyte esterase, it’s still of unknown origin. If you do, then maybe it’s a UTI, and it’s not so unknown after all.
Daphna Yasova Barbeau, MD (1:17:33)
Make sure you’re looking for something else. So I think this provides a lot more guidance than before. One place I’ve always felt there’s a gap is the afebrile infant. We’ve basically used the guidelines for febrile infants with UTI for afebrile infants with UTI, because we don’t have anything else. So, AAP: do you want us doing the same thing for children under two who have a UTI but no fever?
Ben Courchia, MD (1:18:14)
They did a very good job, I must say, with these tables and flow sheets. They’re printable, and you can post them around your NICU right now. They’re ready to use. I really enjoy that. It’s very comprehensive, and they’ve done a good job lining up with the data that was already out there, so none of it comes as a shock. I think this is very important. Everybody should read it, and for trainees, it’s a must.
All right, buddy, this was great. Good job reviewing that in a short amount of time.
Daphna Yasova Barbeau, MD (1:18:52)
That’s a record for me: reviewing a guideline in 20 minutes.
Ben Courchia, MD (1:18:53)
Twenty minutes. All right. See you
Comments