#462 - 📑 Journal Club - The Complete Episode from September 5th 2026

Hello friends 👋

This week's Journal Club. Monday, nasal CPAP versus NIPPV as the bridge to minimally invasive surfactant in infants born before 30 weeks, first author Dr. Hui Zhang. Tuesday, the PAEAN trial on high-dose erythropoietin for neonatal HIE, led by Dr. Helen Liley. Wednesday, extended infusion versus bolus feeding for weight gain in extremely-low-birth-weight infants, from Dr. Haribalakrishna Balasubramanian's team in India. Thursday, the DROP-ROP trial, testing dexamethasone eye drops against treatment-requiring ROP, led by Dr. Ann Hellström. Friday, Neo News: Eli and Ben on The Pitt's viral boundary-setting scene and what it says about burnout in medicine.

----

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

Nasal Continuous Positive Airway Pressure vs Nasal Intermittent Positive Pressure Ventilation in Preterm Infants With Respiratory Distress Syndrome: A Randomized Clinical Trial. Zhang H, Zhang Y, Zeng L, Tong X, Piao M, He H, Zhao C, Xie H, Zheng Z, Cui Q, Lai Y, Wang H, Wang L, Liu H, Tian X, Wu H, Kang L, Han T.JAMA Netw Open. 2026 Jun 1;9(6):e2619785. doi: 10.1001/jamanetworkopen.2026.19785.

Erythropoietin for Neonatal Hypoxic-Ischemic Encephalopathy: A Randomized Clinical Trial. Liley HG, Hunt RW, O'Connell RL, Battin MR, Novak I, Wu YW, Ballard R, Askie L, Badawi N, Ghadge A, Jacobs SE, Juul SE, Sebastian L, Wagh D, Simes RJ; PAEAN Study Investigators.JAMA Pediatr. 2026 Jul 27:e263082. doi: 10.1001/jamapediatrics.2026.3082. Online ahead of print.

Effect of Extended Infusion Versus Intermittent Bolus Feedings on Hospital Weight Gain in Extremely Low Birth Weight Infants Born at <29 Weeks of Gestation: A Randomized Noninferiority Clinical Trial. Balasubramanian H, Sahoo M, Chhadva D, Choubey R, Sakharkar S, Pillai A, Kabra NS.J Pediatr. 2026 Jul 13;298:115235. doi: 10.1016/j.jpeds.2026.115235. Online ahead of print.

Dexamethasone Eye Drops to Prevent Treatment-Requiring Retinopathy of Prematurity: The DROPROP Randomized Clinical Trial. Hellström A, Petrishka-Lozenska M, Sjöbom U, Wallander J, Nilsson AK, Löfqvist C, Ley D, Gränse L, Öhnell HM, Hård AL, Jakobsson G, Sävman K, Domellöf M, Löfgren S, Larsson E, Smith LEH, Pivodic A, Lundgren P; DROPROP Collaboration Group.JAMA Pediatr. 2026 Aug 3:e263247. doi: 10.1001/jamapediatrics.2026.3247. Online ahead of print.

Haridasani Gupta A. On "The Pitt," She Clocked Out on Time. In Real Life, It Started a Debate. The New York Times. May 11, 2026. https://www.nytimes.com/2026/05/11/style/the-pitt-joy-kwon-irene-choi-work-life-balance.html

----

Watch this week's Journal Club on YouTube 👇

----

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

Ben Courchia, MD (0:00)Hello everybody, welcome back to the Incubator Podcast Journal Club. Daphna and I are in the studio today discussing the latest evidence. What's up, Daphna? How's everything going?

Daphna Yasova Barbeau, MD (0:14)Are you with me? You look pooped. I know you're doing some traveling, I know you've got a lot going on.

Ben Courchia, MD (0:18)Exhausted, but I am here.

Daphna Yasova Barbeau, MD (0:22)Journal club waits for no one. I'm doing okay, except I

didn't turn on my focus mode, so you might hear my notification dings.

Ben Courchia, MD (0:26)That's right. I don't hear anything right now, but if you say so.

Daphna Yasova Barbeau, MD (0:40)They're a little annoying to me, but otherwise we're doing good. Started a new semester, kids back at school, getting back into the swing of things after summer.

Ben Courchia, MD (0:56)We're doing a lot of work at the podcast to try to really improve even further the quality of what we can deliver to you all. I think there's a desire for even more transparency, and one of the things we've done recently is make our statistics public — you can actually see the kind of impact the podcast has on our website.

One of the things people have been asking us a lot is, how do you guys pick articles for journal club? We've organized our process into a database, and we might release that database, updated every week or so, if there's interest — we could post it on our website. Basically, articles come in through a larger database and then we filter them: we either dismiss them, not because they're bad articles, just not something we'll have time to review; or we select the article, meaning it's on the list of potentially selected articles that we should review if we can; or it goes on a short list of articles that must be reviewed based on the level of evidence. I've done it that way for some time now and it's working quite well, but let us know what you think. We'll have some more stuff to announce down the pike, so stay tuned. We're working very hard.

Daphna Yasova Barbeau, MD (2:28)I think we're always trying to take a step further, give people more of what they want and need — a lot of that comes from our end-of-year survey. Big thanks to anybody who did that. We're excited about some of the changes, or additions I should say, on the horizon.

Ben Courchia, MD (2:48)For sure. I'll get us started today. I have a very nice article from JAMA Network Open called "Nasal Continuous Positive Airway Pressure versus Nasal Intermittent Positive Pressure Ventilation in Preterm Infants with Respiratory Distress Syndrome." The title's a bit on the nose, but it's a very nice trial.

Daphna Yasova Barbeau, MD (3:15)Did you intend that pun? No, probably not.

Ben Courchia, MD (3:17)No, but it worked out great. First author is Hui Zhang, coming out of China. It's a very interesting article about where we are today in how we treat extremely low birth weight infants, and about the paradigm shift in how we administer surfactant — from invasive approaches involving intubation to less invasive approaches. At some point as a field we're going to have to decide how we want to call this, because the article itself mentions that whether you call it LISA (less invasive surfactant administration), MIST (minimally invasive surfactant therapy), or now this even newer term, MISA (minimally invasive surfactant administration) — it's basically alphabet soup for the same thing. I don't really understand why we keep coming up with new names.

Daphna Yasova Barbeau, MD (4:22)I think it depends what trial you were following, and that's what you wanted it to be called.

Ben Courchia, MD (4:27)The authors also note that current guidelines recommend, as much as possible, avoiding mechanical ventilation in a small preterm infant when we can, and that there might be benefits to less invasive surfactant administration. What they're really highlighting is the difference between CPAP (Continuous Positive Airway Pressure) and NIPPV (Nasal Intermittent Positive Pressure Ventilation) — knowing that NIPPV is a more robust form of support, but it's associated with the risk of abdominal distension, vomiting, and so on. They point out that NIPPV as an initial respiratory support with LISA or MISA remains largely unstudied in patients born less than 30 weeks — we don't have a lot of data for that population. A multi-center trial in 200 infants with gestational ages between 26 and 32 weeks showed that NIPPV with MIST reduced invasive mechanical ventilation within 72 hours and decreased the incidence of BPD (Bronchopulmonary Dysplasia), but no significant reduction in invasive mechanical ventilation was observed in a pre-specified subgroup of 100 infants born before 30 weeks.

So to address that gap, they conducted a multi-center non-inferiority randomized clinical trial comparing nasal CPAP with NIPPV as a primary mode of respiratory support, with both groups receiving minimally invasive surfactant administration, across 11 tertiary centers in China. The protocol was prospectively registered. They included infants born between 24 weeks and 29 weeks, 6 days, spontaneously breathing at birth, requiring non-invasive respiratory support, with a diagnosis of RDS (Respiratory Distress Syndrome) within two hours after birth. To be clear — a baby who was born needing intubation was not placed on non-invasive support for the purpose of this trial; that baby got intubated and mechanical ventilation, and isn't part of this population. They excluded babies with major congenital anomalies, those transferred to another hospital, or those needing intubation for resuscitation or surfactant administration.

Babies were randomized one-to-one to nasal CPAP or NIPPV on admission. They received standardized non-invasive respiratory support in the delivery room and during NICU (Neonatal Intensive Care Unit) admission, using a PEEP (Positive End-Expiratory Pressure) of 6 cm of water and an FiO2 (Fraction of Inspired Oxygen) of less than 40%. After admission, they were randomized to CPAP — PEEP between 6 and 8, FiO2 between 21% and 40% — or NIPPV, with PEEP of 6 to 8, PIP (Peak Inspiratory Pressure) of 15 to 20, inspiratory time of 0.3 to 0.4 seconds, respiratory rate 20 to 40 breaths per minute, FiO2 21% to 40%. If maximal parameters were insufficient to maintain a preductal saturation of 90–95%, FiO2 was gradually increased up to 40%, and surfactant was given. They used a calf pulmonary surfactant from Beijing Double-Crane Pharmaceutical, administered via MISA within 120 minutes of birth to infants with RDS on either CPAP or NIPPV.

The idea was to see whether babies could do as well on CPAP plus surfactant as on NIPPV plus surfactant, with both groups receiving minimally invasive surfactant, to isolate the effect of the ventilation mode. A 1.67 mm catheter was inserted 0.5 to 1 cm below the vocal cords via direct laryngoscopy, connected to a 5 mL surfactant syringe, and the surfactant was given as mini-boluses over about two to five minutes. The initial dose was 100 mg/kg. Repeat dosing was permitted within 72 hours if it was at least four to six hours after the prior dose.

Criteria for escalation to intubation and invasive mechanical ventilation within 72 hours or seven days included: severe respiratory acidosis, hypoxemia, recurrent apneas, a Silverman-Andersen Score (SAS) increase of more than two points or an SAS of five or more — I'll flag that this scoring acronym isn't fully defined in the source audio, but it's a measure of the severity of respiratory distress — pulmonary hemorrhage, cardiopulmonary arrest, or clinical judgment. Extubation criteria required a mean airway pressure of 8 and FiO2 under 30%. Weaning off non-invasive support required resolved respiratory distress, with nasal CPAP PEEP between 3 and 5, or NIPPV with a mean airway pressure of 6 to 7 and FiO2 of 25% or less.

The primary outcome was non-invasive ventilation failure within 72 hours after birth, defined as requiring intubation and invasive mechanical ventilation. They had a long list of secondary outcomes — pulmonary hemorrhage, pneumothorax, BPD, IVH (Intraventricular Hemorrhage), NEC (Necrotizing Enterocolitis), and more.

The trial randomized 312 infants — 153 to CPAP with surfactant, 159 to NIPPV with surfactant. Median gestational age was 28 weeks, median birth weight 940 grams, 55% boys. Baseline demographics were similar. The trial had to be stopped early for futility after these 312 infants, based on a pre-specified interim analysis, because CPAP with surfactant was clearly inferior to NIPPV with surfactant.

The primary outcome — non-invasive ventilation failure within 72 hours — occurred in 26% of the CPAP group versus 13% of the NIPPV group, a statistically significant difference. In a sensitivity analysis, the posterior probability that the true risk difference exceeded 10% was 98.7%, suggesting a high likelihood of nasal CPAP inferiority even if full recruitment had been achieved. Non-invasive ventilation failure within seven days remained significantly higher in the CPAP group — 27.5% versus 15%. Surfactant redosing rates were similar between groups. No significant differences were seen in hemodynamically significant PDA (Patent Ductus Arteriosus), persistent pulmonary hypertension, pneumonia, or severe ROP (Retinopathy of Prematurity). NEC was numerically higher in the CPAP group — 18% versus 10% — though this didn't reach statistical significance. I was honestly shocked by that; going by the authors' introduction, I expected NIPPV to distend the belly more and put those kids at higher NEC risk, but it went the other way. Other outcomes — pneumothorax, pulmonary hemorrhage, BPD, IVH, PVL (Periventricular Leukomalacia), late-onset sepsis, in-hospital mortality — showed no significant difference. Median duration of non-invasive ventilation, supplemental oxygen, and corrected gestational age at discharge were comparable, but the CPAP group had longer invasive mechanical ventilation duration and extended hospital stays.

This is a very interesting paper — it fills a real gap in how we approach these very tiny babies under 29 weeks, 6 days. They mention the 2022 European consensus recommending nasal CPAP as first-line therapy for preterm infants with RDS, and it really begs the question of whether we need more clarity for these smaller infants who don't need CPAP and should probably just be stabilized on NIPPV. This goes back to something we discussed with Roger Soll on the podcast, about having extremely low thresholds for these very tiny infants. And the great thing about this paper is that surfactant isn't a confound — both groups received the same surfactant dose, so it's not a case of one group getting surfactant and the other not.

The authors conclude that in this trial of extremely preterm infants with RDS, synchronized NIPPV with surfactant significantly reduced non-invasive ventilation failure within 72 hours and was associated with shorter invasive mechanical ventilation duration compared with nasal CPAP. These findings support reevaluating current guidelines to consider NIPPV as a primary strategy, particularly in high-risk populations. However, given the early termination and limited events, results should be interpreted cautiously, and future research is needed to confirm long-term benefits.

Daphna Yasova Barbeau, MD (16:05)I agree. We have a lot of kids on NIPPV, we use it a lot in our unit, and as a group we've found it reduces intubations for babies who need just a little more support than they can get on CPAP. But yeah, on the abdominal distension and "CPAP belly" front, I was surprised too.

Ben Courchia, MD (16:42)It satisfies my curiosity that most of these infants deserve surfactant. I don't want to be the one deciding which 24-weeker doesn't get it — in an ideal world I'd like pretty much all of them to get it, because the longer you wait, the less effective it can be. That's purely from an experiential standpoint. So it's interesting to see a study design where surfactant is no longer a variable, and you can truly assess the modes of ventilation with everything else equal. So many trials leave that lingering question of, well, if that non-invasive group had gotten surfactant, maybe they'd have done better — I think this answers that particular question. It reinforces the perception voiced by a lot of key opinion leaders on this podcast, that the smallest infants may need a bit more support than nasal CPAP alone can provide. Some people will argue that well-delivered bubble CPAP, at centers of excellence that do non-invasive ventilation extremely well, can get 23–24 weekers through on bubble CPAP with great results. That's actually something we find really interesting, and we're going to have a special series this fall on how some of these centers do non-invasive ventilation so well. But for most centers, I think this data is pretty representative of what we all encounter.

----

Daphna Yasova Barbeau, MD (21:36)I'm going to tell you about a study in JAMA Pediatrics: "Erythropoietin for Neonatal Hypoxic Ischemic Encephalopathy: A Randomized Clinical Trial." Lead author Helen Liley, senior author John Simes, for the PAEAN study investigators. Some people might ask, didn't we already look at EPO (Erythropoietin) for neonatal HIE (Hypoxic Ischemic Encephalopathy)? There's still this hope that EPO will work, because it's looked good in animal and preclinical models. PAEAN stands for Preventing Adverse Outcomes of Neonatal Hypoxic Ischemic Encephalopathy with Erythropoietin.

It's a phase 3, multi-center, double-blinded, placebo-controlled randomized clinical trial across 24 NICUs (Neonatal Intensive Care Units) in Australia, New Zealand, and Singapore. Interestingly, it was designed in collaboration with investigators from the HEAL trial (High-dose Erythropoietin for Asphyxia and Encephalopathy), using almost identical eligibility criteria and identical study drug and dosing regimens. So again, in late preterm and term newborns receiving therapeutic hypothermia for HIE, they wanted to know whether repeated high doses of EPO over the first week reduce death and disability at two years of age.

They recruited babies from 2016 to March 2021. Eligibility included inborn or outborn infants at 35 weeks, 0 days gestation or greater, with one or more indicators of depression at birth — consistent with other HIE trials, so I won't review them all. Additional criteria: therapeutic hypothermia initiated by six hours, study treatment startable within 24 hours of birth, family consent, and anticipated availability for follow-up at two years. Exclusion criteria included contraindications to EPO, a plan to give EPO or another stimulating agent within two weeks of birth, birth weight less than 1,800 grams — which in this country has tended to rule babies out for cooling — suspected major chromosomal or congenital anomalies, microcephaly with head circumference less than the third percentile, or consideration of withdrawal of care.

Like the other trial, they used 1,000 units/kg of EPO, five doses: the first within the first day of birth, then as close as possible to 24, 48, 96, and 144 hours after the first dose, or until death, withdrawal of care, or NICU discharge if that came first. They could give the last dose 24 or more hours early if IV (intravenous) access was being removed, or on discharge or transfer.

The primary outcome was a composite at two years: all-cause mortality, or survival with moderate or severe developmental deficit (motor or cognitive), based on standardized assessments including the GMFCS (Gross Motor Function Classification System) and the Bayley-III. Secondary outcomes included death from randomization through two years, and among survivors: cerebral palsy, cognitive deficit, respiratory support, nutritional support, cortical visual impairment, hearing impairment, epilepsy, and healthcare costs/utilization through two years, plus adverse events up to a month after the last treatment. Tertiary outcomes included the global Dubowitz optimality score before hospital discharge, death or moderate/severe encephalopathy at the last Sarnat assessment, and overall severity across four categories: normal; mild motor or cognitive deficit; moderate or severe motor or cognitive deficit; and death.

They enrolled 313 infants — 18 sites in Australia, 5 in New Zealand, 1 in Singapore (75% of babies in Australia, 23% in New Zealand, 2% in Singapore). At two years, mortality status was unknown for about 6% of infants, and the primary outcome was unavailable for about 10%, similar between groups. Baseline characteristics were similar, including HIE severity — 78% moderate / 22% severe in the EPO group, 77% moderate / 23% severe in placebo — and similar across sites. Mean gestational age was 39.4 weeks (EPO) versus 39.3 weeks (placebo); mean birth weight 3,457 grams (EPO) versus 3,327 grams (placebo). Inborn/outborn status, onset of labor, mode of delivery, and presence of a predefined intrapartum sentinel event were all similar. Mean 10-minute Apgar scores were 4.6 (EPO) versus 4.8 (placebo), and mean pH values were also similar. Therapeutic hypothermia was commenced at a median of 1.5 hours from birth (EPO) versus 1.8 hours (placebo), with similar time to target temperature and device types.

In both groups, more than 96% received the first and second doses, more than 93% the third dose, 85% the fourth dose, and 56% the fifth dose. No doses of study drug or placebo were withheld for suspected treatment-related adverse events. Non-adherence was mostly due to decisions to remove intravascular cannulas, clinical recovery, death, or withdrawal of support, with no difference in the distribution of causes between groups.

Primary outcome: death or survival with moderate or severe developmental deficit was evaluable in about 281 participants (nearly 90%), occurring in 34.1% of the EPO group versus 28.7% of placebo — not statistically significant. Secondary outcomes: 15% of the EPO group died versus 12% of placebo, not statistically different, with no difference in timing of mortality by group. Among survivors, there were no differences between EPO and placebo in cerebral palsy, motor deficit, or cognitive deficit. No infants required supplementary respiratory support; no differences in nutritional support, cortical visual impairment, hearing impairment, or epilepsy. Mean Bayley scores were 99.1 (EPO) versus 97.8 (placebo) — no major differences in composite, language, or motor scores.

For tertiary outcomes, among children who could be assessed, there was a small shift in the two-year outcome distribution — actually an increased odds of a worse outcome category in the EPO group. No major differences were observed in the combined outcome of death within seven days, moderate/severe encephalopathy at the last Sarnat assessment, or the global Dubowitz optimality score. For safety, there was no difference between groups in predefined safety outcomes within 30 days of the last treatment, or in adverse events resulting in treatment interruption or discontinuation. They were especially interested in thromboses — 2.6% in the EPO group versus 0.6% in placebo, not statistically significant. In a post-hoc meta-analysis combining this trial, the HEAL trial, and the NEATO study for major thromboses, they found 2.3% of EPO participants versus 1.2% of placebo, a relative risk of 1.96 but a p-value of 0.21.

No clinically important differences were seen between groups in highest FiO2, respiratory support needs, seizure medications, or inotropic support during treatment. Peak lab findings — hemoglobin, creatinine, prothrombin time, PTT (Partial Thromboplastin Time), and lowest platelets — also didn't differ. The distribution of encephalopathy severity by treatment day was similar between arms.

They conclude that, consistent with the HEAL trial, the PAEAN trial found that high-dose EPO given early in the evolution of neonatal HIE, in infants receiving therapeutic hypothermia, did not improve primary or secondary outcomes compared to placebo. Disappointing — we're still looking for something to add to our arsenal alongside therapeutic hypothermia.

Ben Courchia, MD (34:41)I want to go back to something you mentioned early on, because I think a lot of listeners might ask, isn't that data we already had? You addressed that — but as you said, it's by design, these studies were done in parallel. I think it's good that the question pops into our heads, "isn't this basically the HEAL trial again," but it's also a sad state of affairs in neonatology, because replicating a study to confirm its results is something that's foreign to us, and it shouldn't be.

Daphna Yasova Barbeau, MD (35:06)Yeah, I agree, it shouldn't be. Being able to replicate data is valuable — we've kind of moved away from that.

Ben Courchia, MD (35:15)There are some subtle differences between the trials, specifically in how they define neurodevelopmental impairment, but the intervention and population are pretty much identical. I think the idea is, and the authors mention this somewhere, that running two trials gets you to the answer faster than one large trial, and allows for further analysis down the line — an individual patient data meta-analysis type of thing. And never underestimate the differences in conducting trials in different regions of the world — we're not all the same, genetically, environmentally, and so on. The HEAL trial was mostly US sites; this one is Australia, New Zealand, and Singapore. So I found that interesting on its own. Like you, I'm disappointed we're not seeing an effect and don't have an additional tool in our arsenal — but it is also a testament to the HEAL trial having been a good study.

Daphna Yasova Barbeau, MD (36:19)Yeah. To your point, I think everybody wants to be first, but we need all the information — people shouldn't stop working on something just because somebody else already published it.

Ben Courchia, MD (36:32)No — and by the way, if I'm not mistaken, a number of the authors on this trial are also on the HEAL trial. Sandra Juul, Wu, Ballard, Juul — several names are on both. These are examples of groups collaborating, not competing.

Daphna Yasova Barbeau, MD (37:02)Mm-hmm.

Ben Courchia, MD (37:02)Interesting stuff — go check out this paper.

----

Ben Courchia, MD (39:01)This was a very interesting paper I found in the Journal of Pediatrics: "Effect of Extended Infusion versus Intermittent Bolus Feeding on Hospital Weight Gain in Extremely Low Birth Weight Infants Born at Less Than 29 Weeks of Gestation: A Randomized Non-Inferiority Clinical Trial." Long gone are the days titles were snappy — I'm getting tired of these long, purely descriptive titles.

Daphna Yasova Barbeau, MD (39:36)What are you implying, people are ChatGPT-ing their titles?

Ben Courchia, MD (40:01)This really stems from a discussion all of us have in our NICUs daily. Intermittent bolus feeding and continuous feeding are both standard methods of enteral feeding in preterm, very-low-birth-weight infants. We all agree bolus feeding mimics physiological feeding patterns and promotes cyclical release of gastrointestinal hormones that can stimulate gut motility and maturation. Continuous feeding, as a slow infusion via syringe pump, has been used in extremely premature infants with concerns about feeding intolerance or respiratory compromise exacerbated by bolus feeding, and in those with pre-feeding hypoglycemia. But currently available randomized controlled trial evidence is insufficient to recommend one strategy over the other, or to conclude they're equally effective. There's also the whole issue of milk fat residue in the tubing — a lot of innovation has gone into pump angle, syringe type, tubing type. Bottom line: in every unit you have people who believe in bolus feeding and people who believe in continuous feeding, and everybody's usually pretty entrenched, so it leads to fiery conversations.

Daphna Yasova Barbeau, MD (41:35)Well, at least a lot of back-and-forth.

Ben Courchia, MD (41:38)Right. But if the babies could talk, they'd probably say, can you please get together on this and stop switching me back and forth every day between bolus, intermittent, and continuous feeds? The authors say it remains important to determine whether the rate of hospital weight gain with extended infusion feeding is not unacceptably worse than intermittent feeding in extremely-low-birth-weight infants —

Daphna Yasova Barbeau, MD (42:05)Yeah, just pick something.

Ben Courchia, MD (42:03)— who are vulnerable to feeding intolerance and postnatal growth faltering, yet this subgroup makes up less than a third of the study sample in trials evaluating extended infusion feeding strategies. So they ran a non-inferiority trial comparing the two. First author is Haribalakrishna Balasubramanian, out of India — an exceptionally well-designed trial, very interesting methodology. Single-center, open-label, parallel randomized controlled trial with 1:1 allocation, conducted at a level 3 NICU in western India. Participants were infants born before 29 weeks' gestation, weighing less than 1,000 grams, at the time trophic feeds were initiated within the first seven postnatal days.

Infants with major congenital anomalies or surgical conditions, and those with severe hemodynamic instability, were excluded, along with infants who couldn't start feeding within the first postnatal week and outborn infants admitted after 24 hours of age. Infants were fed expressed breast milk as a slow infusion through the orogastric (OG) tube via syringe pump — 20–50 mL syringes connected through a 10 cm tubing extension primed with milk. Over 24 hours, they used a four-cycle schedule: five hours of continuous infusion followed by a one-hour break — so five on, one off, giving 20 hours of infusion and 4 hours of breaks. Feeding breaks also allowed intermittent venting of the gastric tube in infants on NIMV — I'll flag that acronym, since it's not fully spelled out in the source, and it may be a variant term for non-invasive ventilatory support similar to NIPPV. Infants who reached 1,200 grams at 30 weeks' gestation were transitioned to bolus feeding over 12 to 24 hours. I've never done that particular schedule before — I've done 10 on, 2 off, 24 around the clock — but never five on, one off.

Daphna Yasova Barbeau, MD (44:20)Yeah, it never occurred to me to do it over six hours.

Ben Courchia, MD (44:33)For the intermittent bolus arm, feeds were given every two hours through the OG tube over 10 to 20 minutes.

Daphna Yasova Barbeau, MD (44:44)But that means much smaller feeds than every three hours.

Ben Courchia, MD (44:48)Right, which is interesting — another thing I've never done that doesn't sound so strange once you think about it.

Daphna Yasova Barbeau, MD (44:56)Isn't that what we tell parents of term kids — feed them less, more often?

Ben Courchia, MD (45:03)Exactly. They chose two-hourly feeding rather than three-hourly as the standard for this study, citing evidence supporting better feeding tolerance with two-hourly feeds in infants under 1,000 grams — and that's their unit's practice. Reading the protocol, I kept getting turned off at first because it's single-center, thinking "this could be biased" — but the more I read, the more exceptional it seemed.

On TPN (Total Parenteral Nutrition), briefly — fairly standard: starting at 80–100 mL/kg/day, amino acids at 2 g/kg/day, lipids 1–2 g/kg/day to start, increasing by 20 mL/kg/day up to 150–170. Minimal enteral nutrition started at 10–15 mL/kg and was advanced 20–25 mL/kg/day if less than 26 weeks, or 25–30 mL/kg/day if more than 26 weeks — different advancement rates based on maturity. They also allowed slower advancement for growth restriction, absent or reversed end-diastolic flow in the umbilical artery, mechanical ventilation requiring 60% oxygen, or hemodynamically significant PDAs on medical management. Enteral feeds were advanced up to 180–190 mL/kg/day. Feeding used mother's own milk or pasteurized donor milk, fortified with Lactodex HMF (Human Milk Fortifier), an Indian product — started once infants reached 100–150 mL/kg/day of enteral feeding, continued until discharge. The feeding assignment continued until 30 weeks and 1,200 grams, at which point everyone transitioned to bolus feeding.

Feeding intolerance was measured by days of feeding interruption in the first 28 postnatal days. Feeding was interrupted for one major criterion — hemodynamic instability requiring vasopressors or inotropes, hemorrhagic gastric aspirates, or blood in the stool — or two minor criteria: emesis, abdominal distension, abdominal wall discoloration, bilious aspirate, or visible intestinal loops. With just a single minor criterion present, the clinical team could choose not to hold feeds, at their discretion. They also used probiotics — three strains of Lactobacillus, started on minimal enteral nutrition and continued until discharge or 37 weeks postmenstrual age.

Sometimes you see a study coming from India, an LMIC (Low- and Middle-Income Country), and think, "here we go" — and then they're doing things better than a lot of us. No preconceived notions, please.

On outcomes: the primary outcome was rate of weight gain, or growth velocity, at discharge, evaluated for non-inferiority using both intention-to-treat and per-protocol analyses, with pre-specified secondary outcomes of mortality and various morbidities. That's enough on methodology — onto results.

They randomized 132 infants — 68 to the infusion group, 64 to the bolus group. Mean gestational age at birth was 26.4 weeks, mean birth weight 750 grams; 22% of the cohort was small for gestational age. Of the 132 infants, 62% were born under 27 weeks, 38% between 27 and 28 weeks, 6 days. Median time to feed initiation was three days. Extended milk infusion was provided for an average of 47 days. Kaplan-Meier analysis showed time to full enteral feeds was similar in both groups, at 10 days — on par with some of the best centers around the world. Full feeds beyond 10 days were achieved in 36% of infants, and 12% achieved full feeds later than 14 days.

Primary outcome, analyzed for 124 infants: postnatal weight gain at discharge was not significantly different — 15.3 g/kg/day in the infusion group versus 15.7 g/kg/day in the bolus group. Results were concordant with the intention-to-treat analysis, which showed a mean difference of 0.35 g/kg. Feeding interruption days were significantly fewer in the infusion group — 6.7 days versus 8.7 days in the bolus group, over the first 28 days. Seven infants had focal intestinal perforation, four had NEC stage 2 or higher; two of the perforation cases and one NEC case required surgical exploration. Overall NEC rates were very low, under 3%.

They conclude that extended infusion feeding was non-inferior to intermittent bolus feeding for hospital weight gain in extremely-low-birth-weight infants with potential for early enteral feeding, and that results should be replicated in multi-center studies and settings with higher maternal milk usage. Adaptive trial designs randomizing infants with failed bolus feeding, or post-GI-surgical states, to infusion-based feeding could more pragmatically evaluate the effects of extended feeding, and extended follow-up for nutritional and developmental outcomes will be needed to confirm the findings.

Interesting, huh?

Daphna Yasova Barbeau, MD (52:01)What do you think?

Ben Courchia, MD (52:03)I think it's very interesting. I'll spoil a bit of upcoming content — we're doing a special series on the golden hour, and I'm excited to bring back Dr. Tarah Colaizy from the University of Iowa to talk about nutrition during golden week, if we can call it that. In the episode we already recorded, she said — I'm paraphrasing — find a system where you interrupt feeding the least.

Daphna Yasova Barbeau, MD (52:32)Meaning it's better to keep feeding than to have to advance and then stop.

Ben Courchia, MD (52:38)Exactly — better to stay on partial enteral feeds for longer than to hit a pre-specified goal and then go NPO for seven days dealing with a NEC workup. We talk about it in that episode, I highly recommend people check it out when it's out. To me, this is data confirming a discussion we've had on the show, that you'll feed these infants for much longer if you're not resorting to bolus. I had no data to support that, I just believed it — happy to see some well-deserved confirmation bias.

Daphna Yasova Barbeau, MD (53:20)Isn't that the best when that happens? In full disclosure, we're not always looking for papers that confirm our bias.

Ben Courchia, MD (53:29)No, clearly — otherwise we'd have told you EPO was a great medication. But sometimes you're like, I knew it, I always believed this worked better.

Daphna Yasova Barbeau, MD (53:42)I think the fear was that these babies would end up at a real nutritional deficit — and maybe not.

Ben Courchia, MD (53:53)Yeah.

----

Daphna Yasova Barbeau, MD (54:30)

This is coming out of JAMA Pediatrics: "Dexamethasone Eye Drops to Prevent Treatment-Requiring Retinopathy of Prematurity" — the DROP-ROP randomized clinical trial. Lead author Ann Hellström, senior author Pia Lundgren, for the DROP-ROP Collaboration Group.

The background people should know is that there was hope that topical dexamethasone, a glucocorticoid commonly used for ocular inflammation, might mitigate some of the inflammatory response believed to underlie ROP development. They're specifically looking at treatment-requiring ROP, which is an important feature of this paper. Let me tell you about the methods: it's a multi-site, double-masked, randomized controlled trial conducted at six university hospitals and eight county hospitals across Sweden. Infants were enrolled between September 2022 and October 2025, born before 30 weeks' gestational age. They were eligible if they had pre-threshold ROP — stage 1 or 2 without plus disease in zone 1, or stage 2 or 3 in posterior zone 2 without plus disease. That matters because it's really what we're looking for in babies developing treatment-level ROP. Infants were excluded if they had an ongoing external ocular infection, or if the neonatologist felt eye drops were unsuitable for that infant.

Daphna Yasova Barbeau, MD (56:18)Intervention infants received topical dexamethasone, 1 mg/mL, or a preservative-free saline placebo, in identical single-dose containers, administered by the bedside nurse — one drop per eye every other day for stage 1 or 2 ROP, one drop per eye daily for stage 3 ROP. The intervention continued for up to 12 weeks, adjusted in masked fashion based on ROP progression or regression: if a baby's ROP worsened, drops were increased; if it improved, drops were decreased. The number of drops was tapered to one drop per eye every other day over one week at discontinuation, and the intervention was stopped if any invasive ROP treatment was initiated. Some babies were at home, and caregivers recorded eye drop doses in an electronic diary.

The primary outcome was the proportion of infants requiring invasive ROP treatment for type 1 ROP — meaning intervention is needed: in zone 1, any stage ROP with plus disease, or stage 3 ROP without plus disease; or in zone 2, stage 2 or 3 ROP with plus disease. Eyes meeting this criteria should generally be treated within 48 to 72 hours, with laser and/or anti-VEGF (anti-Vascular Endothelial Growth Factor) therapy. They also used RetCam imaging as an adjunct — really cool — reviewed by the entire steering committee; patients were masked, but three of the four steering committee ophthalmologists had to agree on the diagnosis before invasive treatment.

Secondary outcomes included time to type 1 ROP, time to type 1 ROP among infants without regression, and recurrence after laser or anti-VEGF therapy. They also looked at pre-specified adverse events, most notably elevated intraocular pressure, since all steroids are known to raise it — plus systemic growth (did steroid use reduce growth over the treatment course?), hyperglycemia, hypoglycemia, and pulmonary, cardiovascular, GI, infectious, and neurologic complications.

They enrolled 100 infants, all included in the intention-to-treat population — 50 per group — with 97 in the per-protocol population: 48 in the intervention group and 49 in placebo. One infant was excluded for receiving fewer than three doses, one for a protocol violation in eye drop administration, and one was withdrawn by the parent. The safety population, for adverse events, included 51 infants in the intervention group and 49 in placebo — I'll flag that this doesn't quite reconcile with the 50/50 intention-to-treat split and is worth checking against the source paper.

Mean gestational age at birth was 25.1 weeks, mean birth weight 712 grams — 25 weeks and 716.4 grams in the dexamethasone group, 24.9 weeks and 709 grams in placebo. Neonatal background and maternal history were similar between groups, though there was a non-significant overrepresentation of IVF and preterm premature rupture of membranes in the placebo group. The mean number of investigational eye drops received was 71, and 90% received one drop per eye per day at some point during the study.

On the primary outcome, in the intention-to-treat population: 29 of 100 infants were treated with laser or anti-VEGF for type 1 ROP — 10 of 50 (20%) in the dexamethasone group, 19 of 50 (38%) in placebo. The unadjusted odds ratio was 0.41; adjusted for gestational age and site, 0.44, with a p-value of 0.08 — so nearly half as many cases in the dexamethasone group, but not statistically significant. In the per-protocol population, type 1 ROP occurred in 9 of 48 (18%) in dexamethasone versus 19 of 49 (39%) in placebo — unadjusted odds ratio 0.36, adjusted 0.4. There were a handful of babies with pre-threshold ROP who underwent invasive treatment based on a local ophthalmologist's call and who wouldn't have met criteria per the steering committee, though the paper doesn't give a lot of detail on those cases.

Subgroup analysis showed the effect of dexamethasone eye drops was numerically greater in infants with birth weight over 660 grams (odds ratio 0.23), and in infants who received parenteral nutrition for less than 14 days (odds ratio reported as 0.04 — I'll flag that as an unusually extreme value worth double-checking against the source, in case it was meant to read 0.4).

Secondary outcomes in the intention-to-treat population: no difference in time from randomization to type 1 ROP or to end of eye drop treatment between groups. Nine infants had ROP recurrence requiring a second invasive treatment — two after laser, seven after anti-VEGF — four of those nine in the dexamethasone group, five in placebo.

Risk ratios for adverse events overall, and adverse events of special interest including intraocular pressure, were either close to 1 or had very wide 95% confidence intervals — so really no major differences even in the babies who got the steroid, which I found a little surprising. Total adverse events were similar between groups: 61 events among 36 infants (70.6%) in the dexamethasone group, 55 events among 36 infants (73.5%) in placebo. Serious adverse events occurred in 64% of the dexamethasone group and 73.5% of placebo. There were three sepsis cases in the dexamethasone group versus two in placebo. Salivary cortisol was lower during treatment in the dexamethasone group but normalized after treatment stopped.

So in conclusion: topical dexamethasone numerically reduced the risk of progression from pre-threshold ROP to treatment-requiring type 1 ROP, but this observed 47% relative risk reduction didn't reach statistical significance. The authors are hopeful a larger study might reach significance.

Ben Courchia, MD (1:06:51)Steroids to the rescue.

Daphna Yasova Barbeau, MD (1:06:53)We hope. More steroids — but topical, not systemic.

Ben Courchia, MD (1:06:54)The next study we'll review is dexamethasone suppositories. I'm just joking around — I feel like if we could lather babies in steroids —

Daphna Yasova Barbeau, MD (1:07:13)You're saying that because I convinced you to start steroids on a baby you didn't want to start steroids on.

Ben Courchia, MD (1:07:18)No, no — that's not why. I'm just making fun of our field, that our one lever is dexamethasone for everything and anything.

Daphna Yasova Barbeau, MD (1:07:23)Adults use dexamethasone for lots of things too, so.

Ben Courchia, MD (1:07:31)I agree. But the results here are interesting, right? I think this idea about ROP progression is something to consider.

Daphna Yasova Barbeau, MD (1:07:42)Yeah — it feels like if we're observing them every one to two weeks, isn't there something we could do before it becomes treatment-worthy? So it's exciting.

Ben Courchia, MD (1:07:51)Absolutely. Especially now that there's more data coming out about Avastin and about recurrence. This could move quickly — I feel like this is always the case with anything ROP: one study comes out and we jump on it, because it's such an important topic for parents and families, and then we find out we should have studied it a bit more. Again, this isn't a small study — 100 infants — but it's not the biggest either, and the results are encouraging. So you might say let's do it — but it's also something we already give systemically, so.

Daphna Yasova Barbeau, MD (1:08:38)Right. And so often these are kids who are doing well, on low support, getting ready to go home — and then they need an invasive therapy, or have to come back after discharge. That's tough for families, tough for us, really tough for families.

Ben Courchia, MD (1:08:52)Yeah — and that graph, figure two, we'll post it — it's a nice graph. The incidence of type 1 ROP being whatever, sixty-something percent in placebo and thirty-something percent in the dexamethasone group. All right, story told.

Daphna Yasova Barbeau, MD (1:09:07)I know, I know. Hopefully

----

Eli (1:09:20)Ben, good to see you — I think we're both a little bleary-eyed. I'm fresh off two months of service, we're in that fellow transition period. I did just go to Greece, but my flight got delayed a bunch, so I'm still kind of bleary-eyed despite all the salty air and the sea bass.

Ben Courchia, MD (1:09:44)Good to see you, Eli. Take two seconds and tell people why it's important to commit to traveling even when you don't have a lot of time — does it just refresh and reinvigorate you for the grind ahead?

Eli (1:10:03)It's interesting — there's literature on how changing your environment is one of the most productive things you can do, both transiently, in terms of going away and getting perspective, and for creativity specifically — changing your environment is incredibly productive, as opposed to staying in the same environment trying to crank things out. It's also been shown — and this maybe only slightly has to do with my mother's urging — that changing your environment can be productive from a romantic standpoint too, if you're stuck, having trouble finding a partner, or your personal relationships aren't where they need to be. Sometimes getting away, temporarily or permanently, can be useful. However, when you're from New York City, the idea of changing your environment is sort of anathema. So there's the rub — but here we are.

Ben Courchia, MD (1:11:10)Here we are. That was a much longer answer than I expected, but thank you for that.

Eli (1:11:15)And that's our first article — literature-based. Anyway, why don't we dive in? Our first article does actually have to do with this. The first story I want to discuss takes us back to something everybody in medicine is having fun talking about, and certainly we're having fun with here on Neo News — The Pitt.

One of many conversations that arose out of this season, season two of The Pitt, is born out of a very brief interaction on the show, between a third-year medical student named Joy Kwan — an exemplary, exceptional student who's very strong and also has very clear boundaries. As people who watch the show know, every episode is one hour; twelve hours after Joy shows up on the show, her shift is over and she tries to go home. One of her supervisors — a guy who's actually sort of a mensch on the show, one of the good guys — tries to pressure her to stay. She says she's finished her shift, she's not being paid, and says to Dr. Langdon, this younger, mensch-y guy, "maybe all you lunatics need to learn how to set some boundaries." And she leaves, and the show goes on, quite literally.

This interaction sparked a lot of conversation on social media around boundary-setting in medicine, work-life balance, and all the things a lot of people in medicine think about in theory but have trouble imagining what it would actually look like to stand up for themselves in a situation like that, or set boundaries. Ben, how did you process this segment on the show, and what's your read on the broader conversation it's evoked around work-life balance — both from the standpoint of preventing exhaustion and burnout, and the conversation it's evoked among patients, who don't want doctors who are overworked or burned out, who want people who have a life so they can be the best version of themselves when they show up to work?

Ben Courchia, MD (1:13:39)Absolutely. I want to reference — I feel bad I don't have it in front of me — this year at the Delphi Conference, we had a TEDx speaker talking about sleep and exhaustion. One of her first slides was: if you knew your son or daughter's school bus driver hadn't slept in 20 hours, would you put your kids on the bus? The obvious answer is no — and it should be the same for medicine. I think that's very interesting, because there are a lot of things implicated in this conversation. Number one, there are the patients — I think patients should come first, and they deserve clinicians and healthcare providers who aren't working exhausted, who are in full grasp of their capabilities to make the best decisions for their patients. I don't know if people will agree, but unfortunately we've all worked tired, and for me, when I work tired I'm a lot more conservative in my decisions. How long can you watch a baby on non-invasive ventilation before intubating, when you're full of energy versus exhausted? Are you going to have a lower threshold to intubate because you yourself are tired, versus the baby truly requiring it? It's a fine line — we're talking about the edges here — but yeah, that's something I've witnessed myself. So the decision to create boundaries is very important.

Then there are two other aspects: work-life balance, and the moral distress that can come from working long hours. Creating boundaries for ourselves is important to make sure we're not getting overworked — but I do think there should be a little more flexibility, meaning that as committed individuals, if the ask isn't too great, could you extend yourself a bit? I think the answer should generally be yes, because we work as teams, and I think that's usually a sign of good team dynamics.

The third aspect is education, which is the other key piece here, because the character in the show is a trainee. Should we be more flexible with our trainees about the hours they're working? I think the answer is yes — but it goes both ways. The frustrating thing about medical training is that patients don't operate on a curriculum. When you're learning on the job, they come when they come. Sometimes it's 5 PM, time to clock out, and something valuable for your education shows up — I think it's a miss not to expose a trainee to that. To be honest, my opinion is, forget duty hours, make sure you learn from this — but in exchange, we need to be kinder with our trainees when it's not their time to clock out. For example, I've seen it so many times where it's 2 PM, nothing's happening, and the trainee just waits for something to happen. Send them to go lie down for an hour, because if something shows up at five, I'd like you to be slightly rested so you can stay. If we keep squeezing trainees for every minute of every day, we can't expect them to extend themselves either — but if we extend kindness and say, find opportunities to rest — I had attendings when I was a resident in New York who would do that, say "give me the pager, go take ten minutes." That's real protection, because I knew that by giving away my pager, no one was going to bother me, and my attending would say, "I've got it, go get a coffee, get some fresh air, come back in 15, 30 minutes, an hour." And nothing bad happened — it was just, "we're not that busy right now, this is an opportunity." I think these things aren't hard to do, especially for attendings — it's not a big ask to take your resident's pager and say, go lie down for an hour, because at five I might want to show you something. I think the conversation would be very different if that paradigm shift happens. Long-winded answer for a brief question, but I hope I answered it.

Eli (1:20:16)Yeah, you answered it in so many different ways. I think framing work-life balance not merely as something nice to do for people, but as a patient safety concern, a training pipeline concern, these bigger systemic priorities everybody's behind — I think that's a great framing. As you said, we all have our vulnerabilities when we're tired — mine are myriad, and I've done a lot of reflection, coaching, and therapy to understand when I'm potentially not at my best, and what my tendencies are in those moments. For me, when I'm tired, one thing that happens is: if you get a call to a bedside, do you actually go, or do you try to triage it? I know that when I get a text in the middle of the night, my willingness to assess it at baseline is lower. So I have reminders on my phone, and a mantra I say to myself: everything you'd go to the bedside for during the day, go to the bedside for overnight. I tell myself that so that on night shifts, if there's any inclination not to do a bedside assessment of work of breathing, or the baby looks a little punky, or I'm not sure the abdomen is a little distended, I go and do that assessment — because the literature shows there are clinical exam findings that are earlier than what's reflected in all the numbers we're watching. There's ample literature that nurses' intuition about how a baby, or a patient more broadly, is doing, is one of the biggest predictors of negative outcomes, and plenty of evidence that validating our multidisciplinary peers — respecting their opinion, valuing their expertise, going to the bedside when they say "I'm worried" — promotes longer-term working relationships and better communication, so that the next time they're worried and I'm on overnight, they still escalate the concern. It wasn't NEC this time, but maybe it's NEC next time. So I think everything you're saying, even taking a rigidly pragmatic view of work-life balance, the ripples of promoting it are far-reaching — more than a nice-to-have, more than a romantic notion that Gen Z doctors are coming in with different priorities than the workaholics of the millennial generation and the boomers. The question, I think, as you said, is how do we start bending the culture of academic medicine away from a culture of stoicism and heroism — knowing that with that stoicism and heroism, there are going to be shifts where you're awake all 24 hours because that's what you need to do, there's no other option — and that treating two hours of downtime as if it must be educational, or that it's important for someone to stay awake through it, may be its own form of hubris, in terms of the longer-term outcomes for that person in training, and for the babies and families we care for.

Ben Courchia, MD (1:23:41)Yeah, I could not agree with you more. The article also has a lot of interesting data about the current state of burnout, which is obviously very alarming. It reminded me of what Jimmy Turner, one of our early podcast interviewees, said — that the hospital is not going to love you back. So it's very much the right thing for our trainees to say, I need to protect myself. But at the end of the day, with the hospital not loving us back — whether that's the administrators or the attendings — and trainees delineating boundaries, I think we're all sort of neglecting the patients in the middle. I don't think it's going to take a revolution — it's going to be small steps that end up being revolutionary. But yeah, I agree it's a form of respecting other people, and our ability to generate assessments and opinions really depends on that.

Eli (1:24:44)Absolutely — we could talk about this all day. I love that you French are always talking about revolutions.

Ben Courchia, MD (1:24:52)Famous words — I forget which advisor — when Louis XVI asked, "is it another revolt?" and they said, "no, sir, it's the revolution."

Eli (1:25:02)That's it — heard it here first, Neo News, breaking news. Anyway, we'll be back in your headphones next week. In the meantime, thanks for listening — let us know if you have any feedback, any thoughts, we'd love to hear from you.

Ben Courchia, MD (1:25:20)Thanks, bye.