#392 - 📑 Journal Club - The Complete Episode from January 10th 2026

Hello friends 👋

This week on The Incubator Podcast, Ben and Daphna review several recent studies in neonatal care. They start with a JAMA trial comparing expectant versus active PDA management in preterm infants, noting a survival signal favoring expectant care and discussing how this fits within current practice. They then review outcomes of 21-week gestation infants from the University of Iowa, focusing on resuscitation strategies and survival at the limits of viability.

The conversation continues with the ICAF trial, examining whether extending caffeine therapy through 41 weeks postmenstrual age meaningfully reduces intermittent hypoxia and for which infants this may matter. A large national cohort study on antenatal corticosteroids between 21 and 24 weeks gestation is also discussed, highlighting practice variation and implications for counseling.

The episode closes with a Neo News segment on legal liability in the NICU following a recent $32 million NEC settlement. Ben, Daphna, and Eli consider informed consent around nutritional care and how evolving legal pressures may influence communication and clinical decision making.

This compilation brings together research and policy discussions from the week in a single long-form episode.

----

The articles covered on today’s episode of the podcast can be found here 👇

Expectant Management vs Medication for Patent Ductus Arteriosus in Preterm Infants: The PDA Randomized Clinical Trial. Laughon MM, Thomas SM, Watterberg KL, Kennedy KA, Keszler M, Ambalavanan N, Davis AS, Slaughter JL, Guillet R, Colaizy TT, Cotten CM, Dhawan MA, Bose CL, Talbert J, Smucny S, Benitz WE, Rysavy MA, Ohls RK, Baserga MC, DeMauro SB, Jaleel M, Jackson WM, Carlo WA, Puopolo KM, Hibbs AM, Katheria A, Sánchez PJ, D'Angio CT, Patel RM, Johnson BA, Chock VY, Bhatt AJ, Merhar SL, Moore R, Laptook AR, Ghavam S, Fuller J, Vyas-Read S, Kicklighter SD, Steinbrekera B, Anderson K, Reynolds AM, Wyckoff MH, Montoya C, Das A, Do B, Chang S, Higgins RD, Walsh MC; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network.JAMA. 2025 Dec 9:e2523330. doi: 10.1001/jama.2025.23330. Online ahead of print.PMID: 41364689

Outcomes of Infants Born at 21 Weeks' Gestational Age. Hyland RM, Mat HD, Boly TJ, Thomas BJ, Stanford AH, Harmon HM, Bermick JR, Davila RC, Colaizy TT, Dagle JM, Klein JM, Greiner AL, Bell EF, McNamara PJ; University of Iowa Neonatology Program.JAMA Netw Open. 2025 Dec 1;8(12):e2548211. doi:10.1001/jamanetworkopen.2025.48211.PMID: 41385227 Free PMC article.

Intermittent hypoxia and caffeine in infants born preterm: the ICAF Randomized Clinical Trial. Eichenwald E, Corwin M, McEntire B, Knoblach S, Limperopoulos C, Kapse K, Kerr S, Heeren TC, Ikponmwonba C, Hunt CE; ICAF Study Group.Arch Dis Child Fetal Neonatal Ed. 2025 Nov 24:fetalneonatal-2025-329230. doi: 10.1136/archdischild-2025-329230. Online ahead of print.PMID: 41285561

The Effects of Antenatal Corticosteroids on Extremely Premature Neonates Born between 21 and 24 Weeks. Yao R, Tritch N, Vedhanayagam K, Ali N, Reimche-Vu H, Gedestad I, Karageuzian S, Contag S.Am J Perinatol. 2025 Nov 6. doi: 10.1055/a-2722-8107. Online ahead of print. PMID: 41086871

WTNH News 8. (2025, December 15). Parents, attorneys emphasizing importance of informed consent to NICUs following baby’s death at Yale New Haven Hospital. WTNH. https://www.wtnh.com/news/connecticut/new-haven/parents-attorneys-emphasizing-importance-of-informed-consent-to-nicus-following-babys-death-at-yale-new-haven-hospital/

----

Watch this week's Journal Club on YouTube 👇

----

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

[00:00.942] Ben: Hello, everybody. Welcome back to a new episode of Journal Club on The Incubator podcast. Happy New Year 2026. Daphna, Happy New Year.

[00:08.709] Daphna: Wow, same to you. It's a big time for transitions and change and looking ahead.

[00:18.188] Ben: Yeah, it's a big transition all around. We're continuing to evolve. I'm very happy actually that our podcast has never really been the same every year. We always take feedback in and we always make small adjustments. For the people who haven't had the chance to listen to our end-of-year episode, definitely go check that out; we go over a lot of the different changes. What's important for you to know if you're listening to Journal Club is that we are creating a new podcast, "The Incubator: Journal Club." It's basically just a repository of the past Journal Clubs for the people who have been asking for a faster way to access them. If scrolling is too much to ask and if the website is too far away—I'm a victim of that sometimes, you know? I just want it right there and then.

[01:09.701] Daphna: Too arduous, too arduous to look through everything. Yeah.

[01:16.226] Ben: Then go subscribe to that podcast and the Journal Clubs will be there. As we said on the end-of-year episode, the exciting thing about 2026 is that through this podcast, what we'll be able to do is potentially offer CME (Continuing Medical Education) credits for Journal Club episodes. We're working actively on making this happen. Whenever that comes through, we will let you know. Until then, stay tuned.

[01:55.417] Daphna: So many articles to review.

[02:03.109] Ben: Well, until that happens, it looks like we're going to hit the ground running with some pretty interesting articles from the end of 2025 as we're opening the new year. Yeah, I mean, you know, we were hoping to take it easy at the end of the year, and every week that went by, you saw one of these papers being released and you're like, "Damn, that's another one we must review."

Okay, so the first article this week that we're going to spend time on is an article published in JAMA (Journal of the American Medical Association) that has made the rounds. It is about the PDA (Patent Ductus Arteriosus). It's called "Expectant Management versus Medication for Patent Ductus Arteriosus in Preterm Infants," or the "PDA (Patent Ductus Arteriosus) Randomized Clinical Trial." First author is Matthew Laughlin, authored with a team from the NIH (National Institutes of Health) NRN (Neonatal Research Network).

It was a privilege to be able to talk to Matt at Hot Topics. If you haven't listened to this conversation, please go check it out on the podcast. It's a few weeks back and it's kind of neat to be able to get his thoughts. It's a 10-15 minute conversation where he walks us through the paper, but we didn't really do a "Journal Club" on it then—we didn't dive deep into the numbers. So that's what I'm hoping we can do today.

PDA management, obviously, is something that we know is very controversial. There's lots of variability. The paper starts off by trying to ground itself in hard evidence, mentioning the Cochrane Review looking at active versus expectant management before seven postnatal days. It shows that early treatment of the PDA probably results in little to no difference in mortality based on evidence that is of "moderate certainty."

Follow-up meta-analyses published between 2010 and 2024, limited to trials of active treatment begun in the first two postnatal weeks, reported that active treatment was associated with increased mortality, although no individual trial really demonstrated a statistically significant difference. Despite these findings, there continues to be wide variation in clinical practice. One-quarter of infants born at 22 to 28 weeks of gestation in the US receive pharmacologic treatment for PDA closure.

Taking into account this ongoing uncertainty, the investigators initiated this randomized clinical trial including preterm infants with a protocol-defined PDA to evaluate the risk and benefit of expectant management compared to active pharmacologic treatment using any medication that the clinical team chose.

In terms of the study design, this was a registry-embedded, comparative effectiveness randomized clinical trial conducted at hospitals participating in the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Neonatal Research Network. Infants were screened between 2018 and 2024. We're going to discuss the immediate short-term outcomes, but there's ongoing neurodevelopmental follow-up being collected as we speak, aiming for follow-up at about two years of age.

Regarding the babies they were looking at: first, they had to be born between 22 weeks and 28 weeks and 6 days. Second, they had to be between 48 hours and 21 days of age at the time of screening. And third, they had to have a hemodynamically significant PDA defined per trial protocol using echocardiographic and clinical criteria.

The diagnosis of a PDA is quite well explained in the supplemental material. Matt mentioned how much the supplemental materials are a treasure trove of information. There are criteria there, including a modified version of the criteria from McNamara and Sehgal. Infants were categorized as having no PDA or a symptomatic PDA—symptomatic meaning echocardiographic criteria of mild, moderate, or severe with a small or medium-sized PDA, or mild to moderate criteria with a large-sized PDA.

Then they had this definition of "cardiopulmonary compromise." I invite you to review these criteria. You have both clinical and echocardiographic criteria. When looking at clinical criteria, FiO2 (Fraction of Inspired Oxygen) is a big component; PPV (Positive Pressure Ventilation) or positive airway pressure is another. When you get into the moderate and severe categories, hypotension becomes a factor. For echocardiographic criteria, there's PDA size, ductal flow, and the LA/Ao (Left Atrium to Aorta) ratio.

In the severe form, you have diastolic reversal of flow in the abdominal aorta. Regarding cardiopulmonary compromise, these infants were excluded. That's a big deal. They decided not to look at these babies who were quite sick from their PDA. They also excluded babies who had a known congenital heart defect other than an ASD (Atrial Septal Defect) or VSD (Ventricular Septal Defect), pulmonary malformations like congenital lobar emphysema or CPAM (Congenital Pulmonary Airway Malformation), or if they had prior pharmacological or procedural treatment of their PDA.

The rationale for excluding kids with cardiopulmonary compromise was that most neonatologists involved in the trial did not have equipoise to assign those babies to expectant management. Randomization was one-to-one, stratified by gestational age.

The intervention was interesting: in the expectant management group, treatments to close the PDA were not to be administered unless the baby reached 36 weeks PMA (Post-Menstrual Age) or developed cardiopulmonary compromise. In the active group, they basically allowed the teams to do whatever they wanted—acetaminophen, ibuprofen, or indomethacin. The choice of agent, dose, and route was left to the clinical team. Initial treatment was to begin within 48 hours of a trial-defined PDA diagnosis.

Surgery, meaning PDA ligation or cardiac catheterization with device placement, was also left at the discretion of the clinical team if pharmacological treatment did not succeed.

The primary outcome was death or BPD (Bronchopulmonary Dysplasia) assessed using a physiologic definition at 36 weeks PMA. Secondary outcomes included individual components of the primary outcome, NEC (Necrotizing Enterocolitis), ROP (Retinopathy of Prematurity), growth, etc.

The investigators hypothesized that expectant management would reduce the incidence of death or BPD from maybe 50% to 40%—an absolute reduction of 10%. That's how they powered the study. However, the trial was stopped early, so they never reached full enrollment. A total of 482 infants were randomly assigned: 242 in the expectant group and 240 in the active treatment group.

In the expectant management group, 25% still received treatment before 36 weeks. You might think these kids got really sick, but only 4 out of those 60 were treated for cardiopulmonary compromise or after 36 weeks. A lot of these babies were actually just a "protocol deviation," which is the bane of these trials.

In the active treatment group, 99.2% received treatment. About 26% received acetaminophen, 35% received ibuprofen, and 29% received indomethacin. About 9% received a combination. Of all these infants, 19% eventually underwent surgical closure. Cardiopulmonary compromise occurred in 14% of the infants in the expectant group and 9% of the active group.

There's no primary outcome, basically. The trial was halted in 2024 by the NICHD for both futility in observing a difference in the primary outcome and for a safety concern related to substantially increased mortality in one of the groups. At the 50% interim analysis, the conditional power for the primary outcome was only 2%. Death or BPD occurred in 81% of the expectant group and 79.6% in the active group—an adjusted risk difference of 1.2%.

In terms of secondary outcomes, this is where the bulk of the discussion is. 4.1% of infants in the expectant group died compared with 9.6% in the active treatment group. The adjusted risk difference was negative 5.6% with a p-value of 0.01. Interestingly, the NNT (Number Needed to Treat) is 18—meaning for every 18 infants managed expectantly, one more survived. I think that's staggering.

Why was mortality so high? The occurrence of adverse events was similar between groups, however, infections resulting in death occurred in 0.8% of the expectant group compared to 3.8% in the active treatment group. This was statistically significant. There is a suspicion that the increase in mortality is driven in great part by this risk of sepsis.

The conclusion is that in extremely preterm infants with a protocol-defined PDA, death or BPD did not differ between expectant management and active treatment in the first 21 days of life. However, survival was substantially higher with expectant management. That's the big takeaway. Daphna, I'm curious to get your thoughts since you weren't with me at Hot Topics.

[15:08.695] Daphna: I know, I missed out! This PDA debate has been going on for decades. I think this is a huge step—I don't know in which direction, but it's going to give us pause. There's some confusion, right? Not all PDAs are the same. Even babies who weren't supposed to get treated still got some treatment. We've all seen babies who really seem to respond to treatment. I couldn't quite get a feel for the most common cause of death for the babies in that group. Does this mean the answer is that some PDAs need treatment, but maybe it's the medication itself that's the problem? We're learning more about individual medications in this postnatal age group. Does this open up the opportunity for more device closures?

[16:29.578] Ben: You're touching on a few things. Regarding different medications, they found it didn't really matter which medication was used; the outcomes were similar. Even though the medications are different, the way they close the PDA is somewhat similar.

I think what you said is very astute. We're focusing on the treatment modality, but maybe we should flip it on its head—it's the patient. Can we understand the patients better? Understanding every baby's physiology and the kind of shunt they have is probably the best way to find a modality that makes sense.

[17:41.349] Daphna: I was just going to say, for trainees—and this took me a long time to figure out—they may ask: "Why do we keep trying to close the PDA if it doesn't make things better and sometimes makes them worse?" But there's all this underlying data that exposure to the PDA is probably not good. We're trying to figure out how to get rid of it without making things worse, and we haven't gotten there yet.

[18:08.462] Ben: And remember, we don't know what happened to these PDAs. It's a pragmatic study. They treated them, but didn't tease out whether they actually closed, got smaller, or didn't close at all. That's a whole different group.

[18:28.168] Daphna: Yeah. So I'm very sorry to report: the discussion continues.

[18:35.887] Ben: That's what we say after every PDA paper. Sorry, guys! We'll see what's coming next.

[18:55.867] Daphna: There's another article at the end of 2025 that really got people talking. It's from JAMA Network Open (Journal of the American Medical Association Network Open), the pediatrics section. It's entitled "Outcomes of Infants Born at 21 Weeks Gestational Age." Lead author Rachel Hyland and the senior author Patrick McNamara.

Basically, the University of Iowa wanted to report on their experience with 21-weekers. And the question they were posing is: what are the outcomes of infants born alive at 21 weeks in a center that attempts resuscitation in selected cases?

So I guess for our colleagues outside of the US, even across the US—I mean, even since we trained—that line of where we offer resuscitation has certainly moved to kind of 22 weeks. The upper limit of where we move to full resuscitation for all babies, that's still a little vague and that is not the topic of today's conversation.

[20:10.626] Ben: Now I invite people to check out the discussion we've had on the podcast with Dr. Edward Bell from Iowa, because I mean, the highlight of my conversation with him was when I was asking him whether he was going to hit at one point a physiological wall for viability. His opinion was no.

[20:27.195] Daphna: Well, I think for every decade of neonatology, they thought that they had hit the physiologic wall. History is proving him right. But it seems impossible. And yet, here we are.

Well, I think—yeah, on the Board Review podcast, we've reviewed the development of the fetal lung, right? And at some point you look at it and you're like, how can we sustain respiration with a lung that has literally no alveoli yet, right? No even saccules. But we're doing it. We're pushing the limit.

And so I think all of us have been faced—I mean, if you're a unit that's routinely resuscitating at 22 weeks, then you've been faced with this discussion about why not 21 and 5/7? And then when you do a 21 and 5/7, people say, why not 21 and 3/7?

And so Iowa obviously has been resuscitating small babies for some time. And they wanted to give us a report on what they were doing for the 21-weekers. Part of why they wrote this—and we've already kind of discussed it a little bit—is there's significant variability in survival at this gestational age.

Even in the kind of 22-week, 23-week group, however, reported survival as low as 0% and as high as 83% across international cohorts shows that there's variability in resuscitation and there's variability in outcomes.

And so, University of Iowa reported to the VON (Vermont Oxford Network) from 2014 to 2024 at 22 weeks gestational age. They have a 66% survival to discharge among those 22-weekers who are resuscitated. So that was kind of their benchmark moving into this discussion about 21-weekers.

And they are, of course, interested like the rest of us are: what does that mean for not just the short term—survival to discharge—but what are the long-term physical and intellectual morbidities?

They highlight in this paper that US cohorts born at 22 weeks and followed up in the 18 to 22 months period—which again is not that long—report a wide range: 39% to 55% with no or mild impairment and 18% to 33% with severe impairment. Again, those are babies at 22 weeks.

And so they wanted to look at all the infants born alive at a gestational age of 21 0/7 to 21 6/7 days between January 1st, 2010 and February 28th, 2025 at the University of Iowa. Those were the babies that were included. And just to be clear, live birth was defined as any evidence of activity or heart rate present.

Resuscitation was broadly defined as any attempt to revive or stimulate. This did not require the presence of a neonatologist or placement of an ETT (Endotracheal Tube). And NICU (Neonatal Intensive Care Unit) admission was defined as admission to the NICU for any time period.

They talk about their care briefly of the extremely preterm infants. They have some papers on that that you can also review. And basically what they're saying is that all of the care strategies that they give to the 22 and 23-weekers were extrapolated to those infants born at 21 weeks.

They talk a little bit about the antenatal counseling, which included shared decision-making between the obstetrics professionals, the neonatal staff, and the parents. They discuss the outcomes of this gestational age are basically unknown, that ability to resuscitate may be limited by the size of the patient and the equipment, and that there's basically unknown short- and long-term maternal/fetal risks.

Resuscitation was not offered at 21 weeks if there were significant congenital anomalies, and it wasn't recommended in multiplicity greater than twin gestation, though they do highlight that this has occurred historically. Now, if resuscitation was not desired by the family, if comfort care was elected by the family, this was fully supported and comfort-directed care without resuscitation was provided.

If there was no heart rate response with effective PPV (Positive Pressure Ventilation) through an endotracheal tube, further resuscitation like epinephrine and chest compressions was not recommended, but they say was individualized. And the decision to provide antenatal corticosteroids—another area of interest—was made by the obstetric team based on maternal and fetal considerations.

And so there was still decision making, one, in resuscitation and two, in the administration of antenatal corticosteroids.

[25:22.35] Ben: No, no, I mean, I was just trying to say like, it's really uncharted territory, basically.

[25:29.146] Daphna: For sure, yeah. And they didn't describe this, but I think they could write a whole other article on what did the counseling look like? I mean, in full disclosure, we were having these discussions in our unit with families admitted at this gestational age, but it's not been written about.

Saying to families like, "We really have no idea what's going to happen," I think is a different way to provide informed consent. So during their study period, they had 22 infants born alive at 21 weeks gestational age. During that time period, they had an additional 230 fetuses at 21 weeks classified as stillbirths. So they had 252 deliveries at the 21st week, but 22 infants of those were born alive.

17 of the 22 liveborn infants, that's 77%, were offered resuscitation in this group. 47% were female and 53% were male. So we're down to 17 resuscitated infants in the 21st week. The median age—I think this is important—is 21 weeks and 5 days, though the range was from 21 0/7 to 21 6/7.

Now of these 17 infants, 6 or 35% were discharged home from the NICU. An additional one baby, so 6%, still remained hospitalized at the end of the study. And 10 or nearly 60% died. So what did that look like? The 10 deaths: 3 in the delivery room, 7 in the NICU.

Rates of resuscitation and NICU admission increased throughout the study period. That's not a surprise to anybody who's been following along. During the first 10 years, 2010 to 2019, 6 infants were liveborn; 3 or 50% underwent resuscitation, but none of the infants in that decade survived.

Now in the next five years, 2020 to 2025, 16 infants were liveborn, and 88% underwent resuscitation. They include demographics for the 17 resuscitated infants. The birth weights ranged from 250 to 450 grams. 41% of the 17 patients were SGA (Small for Gestational Age). Seven of the patients were part of a multiple gestation. Again, most of them were twins, but there was one set of DC/TA (Dichorionic Triamniotic) triplets.

Survivors were less likely to be part of a multiple gestation. Survivors were more likely to have received a complete course of antenatal steroids (3 of the survivors versus 0 non-survivors). However, similar numbers of survivors and non-survivors received no antenatal corticosteroids (2 out of survivors and 2 out of non-survivors).

One- and five-minute Apgar scores were low for all patients, with a median score of 1 at one minute and 3 at five minutes. Maternal characteristics and morbidity were similar between survivors and non-survivors, including frequent PPROM (Preterm Premature Rupture of Membranes) in 47% of the total cohort. All patients underwent inborn vaginal delivery, even though 41% were breech. 29% received limited DCC (Delayed Cord Clamping), meaning 15 to 30 seconds.

I thought this was also interesting—intubation attempts were performed by the most senior neonatology fellow or staff. Success at first attempt occurred only in 6 infants (35%). Multiple attempts were required in 65% and successful intubation took a median of two attempts, but up to four attempts, and was unsuccessful in one patient because the mouth did not accommodate the laryngoscope.

Time from birth to NICU admission was 20 minutes or less for all patients except one who did require extensive resuscitation in the delivery room. They did attempt umbilical access in all patients. UAC (Umbilical Arterial Catheter) access was achieved in 57% of patients, and UVC (Umbilical Venous Catheter) placement was achieved in 93%.

Surfactant was given to all patients after ETT position was confirmed on chest X-ray at a median of 1.4 postnatal hours. Their goal is within one hour; this only occurred in one patient. Again, I think that speaks to how complicated these babies can be.

On admission to the NICU, all the infants were connected to HFJV (High-Frequency Jet Ventilator) support. Their RSS (Respiratory Severity Score) varied widely during the first 72 hours from 1.3 to 11.4, with the scores being lower in infants who survived.

An initial high oxygen requirement was common. The mean FiO2 (Fraction of Inspired Oxygen) on admission was 92%. However, the mean FiO2 during the first 72 hours was 47%. Not surprisingly, early hemodynamic instability was common. They had frequent vasopressor use in 36%, inotrope use in 64%, and iNO (Inhaled Nitric Oxide) use in 64% during the first 72 hours.

That is interesting given the data that we have on the use of nitric oxide in ELBWs (Extremely Low Birth Weight infants). 12 of the infants in the NICU (86%) received one or more therapies as previously discussed. The highest VIS (Vasoactive-Inotropic Score) was 8 in survivors and 25 in non-survivors.

Non-survivors tended to have a lot more medication need than survivors. Interestingly, they were not using routine doses of nitric. The maximum iNO dose in survivors was 5 ppm (parts per million) compared with 20 ppm in non-survivors. And most of those non-surviving babies, it seems, died on maximal nitric support.

All surviving patients had bedside echo (echocardiogram) performed within the first 12 hours (86% in the first four hours), while only 43% of the non-survivor group had cardiac echo performed. Among these 10 patients who underwent echo, cardiovascular phenotypes included acute pulmonary hypertension in 90%, cardiac dysfunction in 40%, and a hsPDA (Hemodynamically Significant Patent Ductus Arteriosus) in 50%.

Many of the infants (90%) experienced multiple or transitioning phenotypes over this period. I think that's why this whole discussion and why people like you are going to get additional TnECHO (Targeted Neonatal Echocardiography) training is crucial—even if you document one phenotype, the next few hours the baby might have a totally different cardiopulmonary phenotype.

They looked at early mean blood pressure measurements across the first 72 hours. There was wide individual patient ranges. Survivor SBP (Systolic Blood Pressure) variance was about 7.7 mmHg, and the diastolic BP variance was about 9.2, compared to non-survivors who had 24.5 variance in systolic and 16.6 in diastolic. Hypotension occurred at least transiently in 86% of infants.

All patients were treated with hydrocortisone for presumed adrenal insufficiency and/or relative hypotension. I'm not sure that is standard of care in all units.

The seven non-surviving patients admitted to the NICU died on postnatal days 0 to 6. In all cases, the cause of death was listed as extreme prematurity with complicating factors including respiratory failure in all infants, some pulmonary hemorrhage in two, and IVH (Intraventricular Hemorrhage) in four.

In four cases, care was redirected to comfort after finding IVH in the setting of significant clinical instability. Two additional patients were redirected after prolonged desaturations at three and four hours of age, even before head imaging. One patient died after cardiorespiratory failure with refractory hypoglycemia despite code resuscitation on postnatal day 3.

So what is the data for the survivors? All infants were extubated directly from high-frequency jet support to non-invasive NAVA (Neurally Adjusted Ventilatory Assist). Successful extubation occurred at a median age of 79.5 days (approximately 32 weeks PMA [Post-Menstrual Age]). Grade 3 BPD (Bronchopulmonary Dysplasia) occurred only in one patient; the remainder had Grade 2 BPD. No infants required tracheostomy.

There were no cases of NEC (Necrotizing Enterocolitis). All surviving patients were treated medically for hsPDA and 71% received interventional PDA closure—transcatheter closure in three patients and bedside surgical ligation in two.

57% (four patients) were diagnosed with chronic pulmonary hypertension, two of whom were discharged home on enteral sildenafil. IVH occurred in 71% of survivors, but severe IVH (Grade 3) only occurred in two infants (30%). No neurosurgical interventions were needed. One patient had seizures managed with anti-seizure medications.

All discharged patients were followed up with serial developmental assessments, but they just weren't able to be followed that long yet. Only four were older than 6 months corrected age. One patient was older than two years and their assessment was "normal." The remaining three were delayed in their milestones but progressing; each scored borderline to high risk in the Bayley-4 (Bayley Scales of Infant and Toddler Development, Fourth Edition) screening test.

Two patients were diagnosed with spastic cerebral palsy, though neither used orthotics. Three of the six patients required rehospitalization for things like viral illness, inadequate weight gain, or placement of a G-tube (Gastrostomy tube).

The authors highlighted that resuscitation rates at Iowa for 21-weekers increased to 87% in the last five years, and in 2024, it was 100%. ACOG (American College of Obstetricians and Gynecologists) guidelines suggest consideration of steroids at 22 weeks, but nobody has made recommendations for 21 weeks because of lack of data.

They noted that even in a highly skilled unit, they struggled to intubate and place umbilical lines because the babies are so small. They used nitric oxide judiciously; in survivors, they only needed about 5 ppm and were able to wean off it in the first 24 hours, likely titrated using echoes to avoid opening the PDA further and causing pulmonary overflow.

[41:17.920] Ben: Yeah, I think it's very interesting. It shows that for units that have strong protocols, you can do quite well even with babies as low as 21 weeks. We have to be careful because we're talking about a very low number—17 resuscitated infants total. It's a seed for potential future work.

[42:17.102] Daphna: It will be interesting to see how families look at this. We're here at the behest of their wishes. Since the line has been pushed to 22 weeks, we've resuscitated gestational ages we didn't in the past. It's happening; the question is, will it happen consistently?

There's an invited commentary by Dr. McElroy in JAMA Network Open entitled, "What Do We Do When Limits of Viability Shift?" He says we are between the obligation to push limits and the obligation to protect infants from harm. We should cautiously applaud Iowa, but we need considerably more data before adopting this as a new standard.

[43:52.462] Ben: Very interesting. Daphna, we're going to talk a little bit about caffeine.

[44:12.124] Daphna: Ooh, one of your favorite topics—both personally and professionally.

[44:15.566] Ben: I think you're going to be excited about this paper because I know you like to give caffeine.

[44:21.244] Daphna: I know, I've been looking forward to seeing this paper.

[44:25.824] Ben: Today I'm taking a look at an article published in the Archives of Disease in Childhood: Fetal and Neonatal Edition. It's called "Intermittent Hypoxia and Caffeine in Infants Born Preterm: the ICAF Randomized Clinical Trial." The first author is Eric Eichenwald and the ICAF Study Group.

Intermittent hypoxia occurs very frequently in the babies we care for who are born preterm. It is secondary to the persistence of immature respiratory control, which really doesn't reach a level similar to term infants until approximately 42 weeks post-menstrual age (PMA). There are animal and human studies showing that intermittent hypoxia causes oxidative stress, free radical production, and pro-inflammatory cytokines, eventually leading to central nervous system injury.

Furthermore, secondary analysis of the Canadian Oxygen Trial (COT) demonstrated a link between severe intermittent hypoxia episodes over the first 10 postnatal weeks and motor and neurocognitive impairment at 18 months. Knowing that caffeine reduces the incidence of apnea of prematurity, the ICAF study group previously reported that clinically inapparent intermittent hypoxia is common after discontinuing caffeine and can be ameliorated by extending therapy through 38 weeks PMA.

This specific study was a multi-center, masked, randomized placebo-controlled trial. The primary objective was to assess whether caffeine affects intermittent hypoxia through 43 weeks PMA. Secondary objectives included assessing caffeine's effects on inflammation-related plasma biomarkers and head imaging.

They screened infants born at less than 30 weeks gestation between 2019 and 2023 across 16 US sites. Eligibility included being between 32 and 36 weeks PMA, the ability to tolerate enteral medication, and being on room air for at least 12 hours without ventilatory support. Exclusion criteria included severe intraventricular hemorrhage (IVH), congenital disorders, or conditions like cardiac arrhythmias and renal dysfunction.

Infants received once-daily caffeine at 5 mg/kg or a placebo. The placebo used was the same suspending vehicle, Ora-Plus. At 36 weeks PMA, the dose was escalated to 5 mg/kg twice daily and continued until 42 weeks and 6 days. This escalation was intended to account for the increased clearance of caffeine as the baby matures. Salivary caffeine levels were monitored periodically to ensure they were within a therapeutic range.

The primary outcome was the number of seconds per hour where oxygen saturation (SpO2) was below 90% during recorded periods. Secondary outcomes included seconds below 85% and 80%, as well as biomarkers like TNF-alpha (Tumor Necrosis Factor-alpha) and MRI findings. They also tracked safety metrics like the need to resume supplemental oxygen or caffeine therapy.

Regarding the results: the study enrolled 170 subjects. It didn't reach the planned sample size of 240 due to funding issues and the pandemic, but the results remain very interesting. For the primary outcome, infants on extended caffeine had significantly fewer seconds per hour below 90% SpO2 from 34 through 41 weeks PMA.

Specifically, the caffeine group spent 57.5 seconds per hour below 90%, compared to 142 seconds in the placebo group. That is a median reduction of about 60%. This effect was consistent across the entire study period, with the greatest differences seen between 35 and 39 weeks PMA.

Statistically significant decreases were also observed at the 85% and 80% thresholds. At 80%, the caffeine group had a median of 2.1 seconds per hour versus 7.2 seconds in the placebo group. This suggests caffeine prevents not just mild "dips," but also more severe hypoxic events.

Looking at biomarkers, TNF-alpha levels at 38 weeks were 23% lower in the caffeine group compared to placebo (p < 0.02). Other measured biomarkers like IL-6 (Interleukin-6) and MRI findings at 43 weeks did not show significant differences between the two groups.

For safety and clinical outcomes, fewer infants on caffeine required restarting supplemental oxygen (4 infants vs. 14 in the placebo group). Interestingly, the caffeine group had a significantly shorter time to discharge, with a median of 17 days after enrollment compared to 27 days for the placebo group—an 8-day difference.

The authors noted that caffeine was well-tolerated. There were no differences in weight gain, heart rate, or sleep patterns between the groups. This addresses a common concern that extended caffeine might interfere with growth or cause excessive tachycardia.

They also pointed out that while we often stop caffeine at 34 or 35 weeks, this study shows that infants continue to have significant intermittent hypoxia well past that point. By continuing the caffeine, we may be protecting the developing brain from the cumulative effects of these hypoxic episodes and the associated inflammation.

[55:20.000] Daphna: That discharge data is actually pretty impressive. An 8-day reduction is huge for families and for the unit.

[55:30.000] Ben: It is. And while the primary intent wasn't to look at discharge, it makes sense. If you have fewer desaturations and you don't have to restart oxygen, the baby is ready to go home sooner.

I think this is very interesting. To be honest with you, we're already quite reluctant, you and I both, to discontinue caffeine at 34 weeks. So the question has to be asked: should we just extend it and really reduce these events? Because I'm very afraid of these intermittent hypoxic events. I think they are very—

[56:14.573] Daphna: That's what I was going to say. I think it's a reminder that intermittent hypoxia is not benign. Not benign at all. Especially when we're talking about cumulative inflammation across a NICU admission and beyond. We often say to families, "Well, if the desaturations are just brief and self-limiting, they don't matter," but that's not necessarily true.

[56:24.962] Ben: It's not.

Which is what we often tell ourselves, and many times, by the time you actually get to the bedside, the event has resolved. So it's difficult to address. Obviously, we're not talking about a baby who is persistently in the 80s and you're not doing anything about that. Anyway, I think that, yeah, it looks like caffeine is the gift that keeps on giving.

[56:48.561] Daphna: Right. Everybody gets more coffee, I guess.

I have an article that relates to our discussion from Tuesday. We talked about 21-weekers and touched on antenatal corticosteroids (ACS) at that time. This study, published at the end of last year, is titled "Effects of Antenatal Corticosteroids on Extremely Premature Neonates Born Between 21 and 24 Weeks." It's from the American Journal of Perinatology. The lead author is Ruafan Yao and the senior author is Stephen Contag, coming out of Loma Linda in California.

[57:51.886] Ben: Yeah, an extension of the discussion on the Iowa paper, looking specifically at the administration of steroids. Very interesting.

[57:54.396] Daphna: Exactly. They wanted to see the impact of antenatal corticosteroids prior to delivery on neonatal outcomes in extremely preterm neonates delivered between 21 and 23 weeks of gestation. As we mentioned, ACOG (American College of Obstetricians and Gynecologists) now suggests we should consider giving steroids at 22 weeks, whereas previously it wasn't recommended before 23 weeks, mostly due to a lack of data.

This was a retrospective cohort study of all live-born, non-anomalous, singleton neonates delivered between 21 0/7 and 23 6/7 weeks at a single Level IV NICU (Neonatal Intensive Care Unit) between 2010 and 2021. They excluded any infants who didn't receive active resuscitation.

They categorized infants into three groups: those who received no steroids, those who received a partial course (one dose), and those who received a complete course (two doses). Their primary outcome was neonatal death, defined as death before NICU discharge. Secondary outcomes included common NICU morbidities like Grade 3 or 4 IVH (Intraventricular Hemorrhage), NEC (Necrotizing Enterocolitis), and BPD (Bronchopulmonary Dysplasia).

During the study period, they had 252 infants meeting the criteria. Interestingly, even at 21 weeks, 18.5% of mothers received some steroids. At 22 weeks, that went up to 48%, and by 23 weeks, it was 90%. This shows that even before formal guidelines changed, clinicians were already "pushing the envelope" based on clinical judgment.

[01:01:30.000] Ben: It's that "gray zone" where clinicians feel they have to do something even if the evidence is still catching up. What did the mortality data look like?

[01:02:43.278] Daphna: The results were striking. For the entire cohort, any exposure to antenatal corticosteroids was associated with a significant reduction in neonatal death. Specifically, the risk of death was 51% for those with no steroids, but dropped to 39.1% with any steroid exposure.

When they broke it down by gestational age, the 22-weekers saw a massive benefit. In the 22-week group, the mortality rate was 53.6% without steroids versus only 33.3% with steroids. Even at 21 weeks, although the sample size was small, there was a trend toward improved survival, though it didn't reach statistical significance for that specific week alone.

Regarding secondary outcomes, they found a significant reduction in the incidence of severe IVH (Grade 3 or 4). In the no-steroid group, the rate was about 34%, compared to only 16% in the group that received a full course of steroids. This is a huge finding because IVH is one of the biggest drivers of long-term neurodevelopmental impairment.

[01:06:30.000] Ben: Did they see any differences between a partial course and a complete course?

[01:07:00.000] Daphna: Yes. A complete course was superior to a partial course, but even a partial course (a single dose) provided more protection than no steroids at all. This reinforces the idea that if you think delivery is imminent, you should still give that first dose of betamethasone or dexamethasone.

They also looked at the "steroid-to-delivery interval." The maximum benefit for survival and reduction of IVH was seen when the steroids were administered between 2 and 7 days before delivery. If given less than 24 hours before birth, the benefit was less pronounced but still present.

[01:09:33.806] Ben: I was discussing this with a family today. It's striking how powerful this intervention can be, specifically for these 22 and 23-weekers. A 51% reduction in the risk of neonatal death compared to 39%—that's a huge difference to a parent.

I agree with you. The guidance for the 21st week is still vague, but using shared decision-making with the family and Maternal-Fetal Medicine (MFM), it's perfectly reasonable to offer steroids based on this emerging data, especially if you have a sense the patient won't make it to 23 weeks.

[01:10:26.107] Daphna: Absolutely. I'm not a researcher by trade, but it seems like it would be impossible—and perhaps unethical—to do a randomized controlled trial at 21 weeks now, randomizing babies to "no steroids." We have to rely on this high-quality retrospective data.

[01:10:43.832] Ben: Agreed. Very interesting.

[01:10:52.262] Eli: We are talking about another really interesting, really complicated article. I didn't know what to feel about this article. But I know it's important, which is why we're going to talk about it. And I feel like many of our colleagues may also have conflicting feelings about this. But that's the point—to engage with those feelings and to know that patients are thinking about this. So this article is about a family that was awarded a $32 million settlement after Aries Rain Peterson, a 27-weeker, got necrotizing enterocolitis and unfortunately passed away after being initiated on a cow's milk-based formula and fortifier without parental consent. This court hearing and the reporting around it gets really deep about this broader conversation around consent, autonomy, and shared decision-making. Including related to decisions that often we're not formally consenting parents for in the same way we might an LP or a thoracostomy tube.

I think it's also important because including families in even the smaller details of decisions, such as what sort of nutritional supplementation we are using when we do not have enough maternal breast milk or in other cases where you just need to bridge a form of nutrition—those micro-decisions are the foundation for building trust with families. The failure to include families in those decisions is one thing, but to the extent that impacts trust, we find ourselves in really thorny situations like with this lawsuit and ultimately the settlements. It's not just about this one family and Yale; it's about the entire ecosystem of neonatology when one family at one hospital gets a $30 million penalty awarded to them. Guys, how did you think about this? The question of legal liability around using cow's milk-based formula. And then as you read about the settlements, how did you process that in your brain, both as a doctor and thinking about how your patients would interpret this decision?

[01:13:04.796] Daphna: Yeah, I mean, this has been a reality that we've been facing in our unit. And to underscore what you said, Eli, you almost don't know what parents would want to be consented for because we've had parents who really don't want donor milk, right? They do not want their baby to get a donated product for sure. But they don't want the baby to get cow's milk product. Different families want different things. And to your point, Eli, there are lots of things that we don't consent parents for, but they want to know about. As soon as babies are admitted now, like extremely low birth weight infants, I say, "By the way, a lot of parents find out in week two, three, or four that their baby is on caffeine or multivitamins or that we use Epogen." Things like that are part of our routine policies and don't feel like a big deal to us, but they do feel like a big deal to parents.

And so now I'm trying to—mostly because we keep getting caught in that. Like, "You started sodium supplements on my baby and I didn't know about it." So I think the best way to protect ourselves from that is to inform parents when those things are happening. I don't think that means a consent form for absolutely everything, obviously, or we'd never get anything done, but at least informing. We wouldn't start things on an adult patient without telling them what we were going to start them on. We don't do it on the Peds floor without telling patients what we're going to give them. We shouldn't do it in the NICU. But especially around these dietary concerns, I think there potentially is a consent requirement here. It's something we're talking about in our unit for all of the options because there are pros and cons to each thing and different families have different priorities and different concerns.

It feels like you could be sued for giving almost any product right now, except potentially mom's own milk theoretically, and we don't have paid parental leave. So how can we expect moms to give enough milk for a whole admission? And then it still leaves the concern about fortifying that milk and picking a fortifier that is appropriate for the baby and is in line with the family's needs. So I think—and this is something we've talked about in our group—that potentially we're going to move to consent-based feeding policies. I guess I'll let Ben, you're the medical director boss these days.

[01:15:56.153] Ben: This article made me so mad. So mad. The pragmatic overview of this issue, from a medical director standpoint, is that if you are in a position of leadership in your unit, you should probably consider creating consent forms for alternative methods of nutrition, whether it is donor milk, whether it is human milk fortifiers versus bovine fortifiers versus formula. I think that all these should be—we're going to get there. It does raise the point that you guys have both mentioned, which is: at what point am I allowed to practice medicine in the best interest of my patient? Do I need to tell patients, "Hey, we're going to put your baby under humidity"? And the risk of humidity is that maybe it could be an environment for infection. So do you want the humidity? Do you want the isolette? Do you want the breathing tube? Are you okay with oxygen or just pressure? At what point are we going to say, let us practice medicine? Everything we do has risks.

Every single thing we do has risk. And what infuriates me the most is when attorneys step out of their role and start talking about medicine, because I don't talk about the law. I am not an attorney, but in this article, when you read about an attorney mentioning how the risk of NEC goes up by 80% when you add cow's milk product, it is bullshit. It is just wrong. This data is in quotes, so it's his words, but it is wrong. Is it true that the risk of NEC is a bit higher with cow's milk-based product compared to human milk product? Yes, but the rates of NEC are 7%. They're not 80% in a population of 27-weekers. And many 27-weekers get cow's milk-based product. So the information is not factual. That's a concern.

I feel very ambivalent. On the one hand, I believe that we should now, to protect ourselves, have these consent forms. And on the other, I am reluctant to give in to that because then what will be the next consent form for? There is nothing that we do to these infants except probably even skin-to-skin. Everything we do has risks and there is not enough time in the day to consent parents for every intervention that we do. But it is a $30 million lawsuit and it was won by the defense. And I feel it's a very tragic story and we should not forget that a baby passed away. The broader conversation is distressing for our field and for how attorneys and legal conversations are getting in the way of our care. Kind of the way it's already done with HIE, for example, where we do a lot of things that are no longer evidence-based because of the fear of lawsuits. So it's very frustrating.

[01:19:06.939] Daphna: Yeah, and it's just the more and more distressed a family says, "How can I trust you guys if you're offering cow's milk formula?" And I actually think the NEC Society does a really good job about this. They have some statements specifically on formula lawsuits that even though they are committed to all things preventing NEC and fighting for families who have been affected by NEC, they say lawsuits are not a strategy used by the NEC Society towards our vision of a world without NEC. They're not involved in any of the litigation because they recognize that in some situations, things like formula are potentially necessary and not every state has donor milk. So what are they supposed to feed babies when mom's own milk is not available?

[01:19:58.333] Ben: We're getting into these situations where parents will be reluctant to agree to donor milk, will be reluctant to agree to formula, and are not producing enough breast milk to feed an infant.

[01:20:04.711] Daphna: That's right. This has happened a few times this month in our unit.

[01:20:10.575] Ben: Yeah. And then where's the discussion about keeping a baby with a central line and TPN? It just makes no sense.

[01:20:19.318] Eli: Yeah. And I think, despite all the frustration that any of us could feel about thinking that decisions that feel as small as ensuring a baby continues to receive nutrition when there are no other enteric options available—families are thinking about this and they're worried. In an era where there is only more information—some good, some bad, lots coming from generative AI that cannot pull from the evidence because the evidence is paywalled—how do we act as sort of translators of what we know about the evidence to families? How do we make sure they're included in those conversations and how do we lay out the risks and benefits of, say, necrotizing enterocolitis versus a CLABSI from a PICC line that needs to stay in forever? Because we need to do TPN or the distress that an infant may get from getting poked five additional times to try to get a new PICC line in or whatever the case may be.

[01:21:47.057] Daphna: Yeah, or how does suboptimal nutrition affect the immune response? There's so many things that we are weighing.

[01:21:54.141] Ben: Yeah. My issue is that the way these problems tend to unwind is through a form. And I don't really care what we say about how we get consent and how we counsel families. When there is a form in between a physician and its patient, it becomes a very transactional form of thing. It is hard. You go to the dentist and they're like, "Well, you make the decision." People are not equipped to make decisions. We barely know what we're doing with all the information that we have, with all the evidence. And now we're going to ask parents, "Well, you decide what you want to do: if you want potentially reduced developmental scores at two years versus cow's milk formula, which might raise the risk a bit of NEC. And then you let me know what you decide for your child." That's just terrible care. It's not the right kind of care.

[01:22:46.636] Eli: Yeah. I know we've talked about this before, but when I had to go to the mechanic for my car, I feel deeply ill-equipped to know what should go in various places. Maybe I'm a particularly naive audience. I grew up in New York. We don't own cars in New York. This is not a thing. So I didn't do car stuff. I like cars, but I'm not a car guy. But still, at a certain point, expertise exists, not for the accrual of power from certain parties, but expertise exists because people share that expertise with one another. And sometimes it makes more sense to have someone be really good at one specific thing and help you understand that rather than have everybody try to understand every dimension of the universe.

Anyway, this is obviously part of an evolving trend around shared decision-making—what shared decision-making looks like in this day and age with lots of different sources of information from lots of different places. We've covered it on vaccines. We have covered it here with breast milk. We will continue to cover this. We certainly welcome any listeners' thoughts on this lawsuit and how they might think about it and how their parents are assimilating this information. But an ongoing discussion.

[01:24:22.365] Ben: And it's an ongoing discussion, but I think it's an important piece for our field collectively to just be proactive about this, because I think that we might wake up one day with an all-purpose waiver, kind of like when you go to the trampoline park and they just make you sign your life away. And then they say, "You signed the waiver." And that might provide an ironclad protection against lawsuits because we made people sign at the front desk when they arrived. But is that what we want to do? Is that the environment in which we want to practice? I think that it's something where we need to be active participants. For people who are working in a large hospital setting, be involved in these conversations with your division of quality and with the administration. Your voice is welcomed and it shouldn't be something that is just turfed to risk management. Your input could be valuable and you could shape a little bit how care is delivered in your hospital. So I think it's an opportunity there for us as well. And so I hope that we can leave on an optimistic note rather than just a "don't."

[01:25:29.794] Eli: I think people have agency to have these conversations in lots of different forms, in lots of different ways at different points in time. Right. When I do my prenatal consults, I talk about donor milk in the prenatal consults. You talk about all these different things. I think there are tons of opportunities in here, acknowledging the state of the world to say we have agency, we have power. We have to over-communicate more than we've done historically, but there are lots of micro opportunities to do that in the course of the day. And there are lots of micro opportunities to build trust. That if unfortunately, as many of these 27-weekers may do—and certainly more than any of us want—develop necrotizing enterocolitis, we can have a conversation with the family that is not about fault or blame, but it's about what happens in children with immature guts. And what we're going to do about it to try to ensure, despite the fact that this has happened, that the child can survive and thrive and that we're on the same team.

[01:26:36.573] Ben: And as we're talking about trust and communication between parents and families, I think it's an opportunity as well to maybe work on ancillary aspects of care, like a parent council. I think that if parents could get information from other parents, maybe these types of frustrating lawsuits don't even happen because there's information on the front end and parents know about the culture of the unit, about the practices. Again, there's lots of opportunities to elevate our game instead of just falling for the form. So just think about that. And maybe you come up with something very creative and innovative that inspires others, and we invite you to do that.

[01:27:16.093] Daphna: Share it with us so we can all use it for sure.

[01:27:17.541] Ben: Share it with us exactly.

[01:27:17.962] Eli: And research how people are getting information. Research how people are getting information, research how people are making decisions. It is, you know, again, it is such an interesting evolving world out there in terms of how people relate to medical systems. And part of the issue is I think all of us have increasing anecdotal experience, but we don't know systematically how and where people get information. We don't know systematically best practices for managing expectations for consenting people. Informally or formally on things like cow's milk-based formula. Don't know best practices necessarily for regaining trust once something like necrotizing enterocolitis can happen. There is so much research, so much work to be done and we need to keep having these conversations. So I'm so glad that we talked about this on today's episode of Neo News. We will continue to talk about that. Thank you so much to our listeners for tuning in. And as always, let us know your thoughts. We are very eager to hear them.

[01:28:20.669] Ben: Thanks Eli. Thanks Daphna.