#466 - 📑 Journal Club - The Complete Episode from September 19th 2026

Hello friends 👋

This week on The Incubator Podcast, Ben and Daphna dig into four studies reshaping bedside decisions. They open questioning whether routine DAT screening for ABO and RhD incompatibility still earns its place, finding transcutaneous bilirubin alone predicts phototherapy need just as well. Next, a Swedish national registry of over two million births links missed intramuscular vitamin K, and rising oral substitution, to a near-doubling of bleeding and intracranial hemorrhage risk. Ben then presents the PHILODENDRON pilot trial out of Milan, testing initial delivery-room CPAP of 5 versus 8 cmH2O in infants under 30 weeks, feasible and safe, but not yet enough to settle the physiology question. Daphna closes Journal Club with a SART-linked study tying IVF conception to significantly higher odds of VACTERL association. The week wraps with Ben and Eli on Neo News, unpacking a study showing birth certificates miss nearly 40% of true NICU admissions, with troubling racial disparities.

----

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

Utility of DAT in Low-Risk Neonates With ABO or RhD Incompatibility. Tsai TL, Ma T, Nester T.Pediatrics. 2026 Sep 1;158(3):e2026076989. doi: 10.1542/peds.2026-076989.PMID: 42642037

Vitamin K Prophylaxis in Newborns and Bleeding in Infancy. Simatou E, Tsamantioti E, Hallström A, Stephansson O, Razaz N, Persson M, Bolk J.JAMA Pediatr. 2026 Sep 1;180(9):995-1004. doi: 10.1001/jamapediatrics.2026.2606.PMID: 42440325

Initial CPAP of 5 versus 8 cmH2O during delivery room stabilisation of very preterm infants: the PHILODENDROOM pilot randomised trial. Cavigioli F, Bresesti I, Castoldi F, Gatto S, Lupo E, Rossi S, La Verde A, Bianchi S, Meneghin F, Viaroli F, Pivetti V, Manfredini V, Cannata G, Bastrenta P, Fontana P, Stucchi I, Daniele I, Chiera M, Lista G.Eur J Pediatr. 2026 Aug 17;185(9):673. doi: 10.1007/s00431-026-07322-6.PMID: 42608615 Free PMC article. Clinical Trial.

In Vitro Fertilization and VACTERL Birth Defects. Tark JY, Richard MA, Schraw JM, Fisher SC, Betancourt D, Stone SL, Forestieri NE, Baker VL, Cameron K, Eisenberg ML, Belva F, Lu Y, Williams CL, Sutcliffe AG, Lupo PJ, Luke B.JAMA Netw Open. 2026 Aug 3;9(8):e2629355. doi: 10.1001/jamanetworkopen.2026.29355.PMID: 42616500 Free PMC article.

Birth certificate data substantially misrepresent actual NICU admissions, including among most vulnerable. Hughes CS, Lorch SA, Schmitt S, Passarella M, Phibbs CS.J Perinatol. 2026 Jul;46(7):1307-1312. doi: 10.1038/s41372-026-02726-6. Epub 2026 May 27.PMID: 42204354

----

Watch this week's Journal Club on YouTube 👇

----

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

Ben Courchia, MD (00:00.91) Hello everybody, welcome back to the Incubator Podcast Journal Club. We're back for this week of new evidence. Daphna, good afternoon — good night, really. It's very late tonight for this recording.

Daphna Yasova Barbeau, MD (00:12.256) It's very late. I was just thinking, we haven't needed to do a really off-hours recording in some time. We used to get stuck sometimes recording at six in the morning, or ten at night.

Ben Courchia, MD (00:21.75) Ha, that's right. Six in the morning is worse, I believe. It's now nine PM, almost — and it's just the only time there is between shifts and everything else. So we do journal club then.

Daphna Yasova Barbeau, MD (00:33.059) We couldn't miss it.

Ben Courchia, MD (00:44.3) No, journal club waits for no one. We had to do it. And it's not bad.

Daphna Yasova Barbeau, MD (00:47.042) That's right. Well, all things considered, I'm so glad that we are doing this. I'm glad that we keep doing this. And I'm glad that people listen, so that we feel totally obligated to get on and do this — because it's part of why we made this. I mean, it's part of why you wanted to do the Incubator, right? So that we'd keep up with the literature. So thanks to all of you for being our — keeping us accountable, that's right. All right.

Ben Courchia, MD (01:23.724) Yeah, exactly. Our peer pressure. So, without further ado, we're going to dive into our first paper. It's a paper I found in Pediatrics called "Utility of DAT in Low-Risk Neonates With ABO or RhD Incompatibility."

Daphna Yasova Barbeau, MD (01:46.023) I mean, already when you read that you're like, "Wait, I didn't know we had to question the utility of the DAT." Gosh, here we go.

Ben Courchia, MD (01:53.807) Here we go. So — the first author is Tsung Lin Tsai, and this is about hemolytic disease of the fetus and newborn, or HDFN (Hemolytic Disease of the Fetus and Newborn) as they refer to it in the paper. We know HDFN can lead to severe hemolytic anemia, and it can lead to subsequent neurotoxicity.

Historically, anti-D alloimmunization secondary to RhD (Rhesus D) incompatibility has been the leading cause of hemolytic disease of the newborn. Now, thankfully, because of good prenatal care — because mothers receive Rh immunoglobulin (RhoGAM) during pregnancy — there's been a shift from RhD incompatibility to ABO incompatibility becoming the most frequent etiology of neonatal hyperbilirubinemia and neurotoxicity.

To assess the risk of hemolysis and guide early clinical intervention, many institutions routinely employ the DAT (Direct Antiglobulin Test) on cord blood for newborns born to mothers with specific blood group types. Some centers do it on mothers who are just group O, or mothers who are Rh-negative, basically to detect antibodies on the neonatal red blood cells.

Now, according to the AAP (American Academy of Pediatrics) guidelines, mature newborns above 35 weeks of gestation with isoimmune hemolytic disease are at higher risk of neurotoxicity. So for the trainees in the room — when we talk about DAT, it's essentially the same thing as Coombs, right? These are interchangeable terms. And the reminder is that it detects antibodies, or complement, already bound to the surface of the patient's red blood cells. The way this is done is that the red blood cells are washed, and then anti-human globulin (the Coombs reagent) is added — the idea being that this reagent binds to the antibodies on the cells, causing them to agglutinate. So if you see agglutination, you have antibodies — that's how it works.

Now, why is routine DAT being questioned? The authors mention that at their institution, they do DAT screening on mothers who are group O or who are Rh-negative. At our institution, we do a DAT on everyone. So I think there are three tranches to this: there's our center, and centers like us, where we do blood typing and DAT on every single baby; there are centers that do it only on a specific category; and there are centers that maybe don't do it at all.

Despite the historical precedent, the clinical utility of routine DAT screening is increasingly being scrutinized because of advances in prophylaxis and non-invasive monitoring. Although Rh immunoglobulin has successfully prevented true alloimmunization, it can actually cause a positive neonatal DAT solely via placental transfer of passive maternal anti-D. So mothers who've received RhoGAM can end up with babies who are DAT-positive, even though that's just passive transfer of anti-D.

The 2022 AAP guidelines state that routine DAT is unnecessary for infants of Rh-negative mothers who have a negative antibody screen. Furthermore, newborns testing positive on the DAT can be managed as DAT-negative babies if the mother tests positive only for passive anti-D. This update is supported by the extreme rarity of clinically significant hemolysis caused by prophylactic RhoGAM. Both of these statements rest on the same logic: a positive DAT only matters when it signals actual immune hemolysis, and in these two scenarios, it usually doesn't.

Despite these clear recommendations against routine testing, a lot of hospitals persist in doing routine cord blood DAT screening. And the predictive value of routine neonatal DAT for ABO incompatibility — which occurs, for the students in the room, when a group O mother delivers a group A or group B infant — is actually debatable in the modern era. Although ABO incompatibility can lead to isoimmune hemolytic disease, the AAP guidelines do not recommend routine DAT for these infants. Moreover, the widespread availability and reliability of transcutaneous bilirubin monitoring gives clinicians a direct, non-invasive screening tool for the infant's clinical status. Given these monitoring capabilities, it remains unclear whether a positive DAT result meaningfully enhances the ability to predict the need for phototherapy beyond what transcutaneous bilirubin monitoring alone can do.

So in this study, the authors investigate the necessity and clinical utility of routine DAT screening in two distinct neonatal cohorts: mature newborns with ABO incompatibility, and babies who are ABO-compatible but RhD-incompatible, born to mothers without active alloimmunization. What they're trying to determine is whether routine DAT testing effectively guides clinical management, or represents an unnecessary test. Dum dum dum. Okay.

The authors did a retrospective cohort study of newborns who underwent DAT screening between 2022 and 2024. During the study period, the nursery used a hybrid screening protocol for neonatal hyperbilirubinemia. Transcutaneous bilirubin measurements were routinely obtained for all newborns between 18 and 24 hours after birth. The protocol for umbilical cord blood testing and DAT depended on maternal blood group characteristics. For babies born to Rh-negative mothers, blood typing and DAT were universally performed. For those born to Rh-positive mothers, DAT was not automatically performed — instead, testing was ordered at the clinician's discretion based on specific triggers: primarily the clinical development of early jaundice, a maternal group O in the setting of an elevated transcutaneous or serum bilirubin, or when the cutaneous or serum bilirubin approached the threshold for phototherapy.

All infants who underwent DAT screening during the study period were screened for eligibility. They excluded babies born before 35 weeks, babies transferred to a higher level of care right after birth before bilirubin could be assessed, and infants born to mothers with prenatal alloantibodies other than passive anti-D. Infants born to non-group-O, Rh-positive mothers were also excluded from further analysis.

The definition of the two cohorts was clever — the population was stratified based on maternal-infant blood group compatibility. Cohort one, the ABO-incompatible cohort, was defined as a group O mother with a group A or group B infant. Non-group-O maternal-infant mismatches — like a group A mother with a group B infant — were not classified as ABO-incompatible in this study, because the antibodies made by group A and group B individuals are IgM and don't cross the placenta, unlike group O individuals, who make IgG antibodies that do cross the placenta. The second cohort — ABO-compatible but RhD-incompatible — was defined as an Rh-negative mother with an Rh-positive baby who was ABO-compatible.

There's a full flowchart of patients in Figure 1. One interesting detail: if a mother was group O and the baby was also group O, but they were RhD-incompatible, that baby was categorized into the RhD-incompatible cohort.

The DAT was performed manually using a gel microcolumn assay, and data were extracted from the medical record — maternal variables including age, gravidity, ABO and Rh status, and so on. They were also able, through additional testing, to determine whether the antibodies detected in the Rh-negative mothers were passive or due to active alloimmunization.

The primary clinical outcome was the requirement for phototherapy. This outcome included phototherapy administered during the initial birth hospitalization, and any subsequent hospital readmission for hyperbilirubinemia. Transcutaneous levels measured between 18 and 24 hours of life were also recorded, and everything was based on the most recent recommendations from the 2022 AAP guidelines.

During the study, the authors identified 1,354 infants who underwent testing. After exclusions, the final analysis included 618 infants — 310 in the ABO-incompatible cohort, and 308 in the RhD-incompatible cohort.

Looking at the bivariable analysis of transcutaneous bilirubin in the ABO-incompatible cohort: the authors assessed the association between maternal and fetal variables and neonatal transcutaneous bilirubin measured between 18 and 24 hours after delivery. The analysis revealed a significant positive correlation between transcutaneous bilirubin and maternal Rh-positive status, gestational age at birth, infant Rh-positive status, and a positive DAT result. In addition, infant blood group B was significantly associated with higher transcutaneous bilirubin levels compared to group A — meaning a group O mother with a group B infant tended to have worse bilirubin levels than a group O mother with a group A infant.

The authors also found that the neonatal DAT tended to be positive if the infant was female, had a higher birth weight, a more advanced gestational age, or was Rh-positive.

To evaluate predictors of phototherapy requirement, the authors conducted a multiple logistic regression analysis assessing all study variables, and found that DAT and transcutaneous bilirubin were the only significant independent predictors in the ABO-incompatible group. Other maternal and fetal variables — including birth weight, gestational age, and fetal ABO/Rh status — did not reach statistical significance in the multivariate model.

Transcutaneous bilirubin alone yielded an area under the curve (AUC) of 0.887. When DAT results were incorporated alongside transcutaneous bilirubin, the AUC barely moved — rising only to 0.896. These results suggest that transcutaneous bilirubin alone is a robust predictor of the need for phototherapy in ABO-incompatible infants.

In the second study group — the RhD-incompatible cohort — all mothers were Rh-negative, and by study design, mothers with active alloimmunization were excluded, so this cohort represented an obstetric population receiving standard antepartum RhoGAM prophylaxis. Within this group, a positive prenatal antibody screen was observed in 63.6% of mothers. This positivity was exclusively attributed to passive anti-D from RhoGAM administration — which is quite something, that they were able to tease that apart. Among neonates born to these mothers, 14% had a positive DAT (Coombs test) at birth, likely due to passive anti-D transferred from the mother — meaning it was the RhoGAM itself, passed from mother to baby, that produced the positive Coombs test.

To evaluate the clinical factors associated with hyperbilirubinemia in this cohort, the authors again ran a bivariable analysis. The results showed that a positive neonatal DAT was significantly associated with a positive maternal antibody screen. Transcutaneous bilirubin measured between 18 and 24 hours of life showed no significant correlation with the neonatal DAT result, nor with the maternal antibody screen. Instead, transcutaneous bilirubin only showed a weak but significant negative correlation with fetal sex and birth weight. These results suggest that the DAT result is not a determining factor for the development of hyperbilirubinemia, or the subsequent need for phototherapy, in these infants.

Furthermore, the incidence of clinically significant hyperbilirubinemia in this cohort was extremely low — only three of the 308 infants, about 1%, ultimately required phototherapy. Because of this exceptionally low event rate, a multivariate logistic regression for predicting phototherapy wasn't performed — the base rate was simply too low.

I think these are very interesting results, and the discussion section was interesting too — I want to read you some of it. They mention that their institution uses this hybrid approach to DAT screening: infants born to Rh-negative mothers are universally tested, but those born to Rh-positive mothers are not — DAT orders for that group are driven by the clinician based on the clinical presentation. They say their cohort likely represents a population with a higher baseline risk for hyperbilirubinemia than a true universal screening program would capture, since they only screen based on risk stratification — compare that to a center like ours, where every baby gets a DAT. And the fact that DAT provided minimal additive predictive value over transcutaneous monitoring, even in this clinically targeted, higher-suspicion cohort, strongly reinforces, in their view, the argument against universal DAT screening. This aligns with the 2022 AAP guideline, which does not recommend routine DAT screening for ABO incompatibility, provided adequate bilirubin surveillance is in place.

Because DAT positivity remains a recognized risk factor for neurotoxicity, they suggest it should be reserved specifically for infants who develop significant hyperbilirubinemia — in those targeted cases, the test serves as an essential tool for risk stratification and setting treatment thresholds. For the Rh-negative cohort, they found no evidence that passive anti-D caused clinical hemolysis, consistent with findings from other papers. Consequently, they conclude that routine DAT screening in newborns born to Rh-negative mothers without active alloimmunization is unnecessary — though safely implementing that practice requires being able to clinically differentiate passive anti-D from active sensitization, which they say was a key strength of their study. I don't know that we'd be able to do that as easily at our institution.

Bottom line: they conclude that transcutaneous bilirubin monitoring alone is sufficient for predicting the need for phototherapy in ABO-incompatible infants, and that passive maternal anti-D does not appear to pose a significant hemolytic risk to RhD-incompatible neonates. Both conclusions align with the AAP guidelines. So routine DAT in this population may be unnecessary, and shifting to a targeted testing approach could optimize neonatal care and lab resources. That said, it's essential to emphasize that the safety of this targeted approach relies heavily on complete maternal antibody screening — a DAT should still be performed whenever maternal screening is incomplete or unavailable.

How interesting, huh?

Daphna Yasova Barbeau, MD (20:26.882) So what do you think?

Ben Courchia, MD (20:29.342) I think that if any hospital administrator gets wind of this, they're going to say, "Hold on, we could do fewer tests."

Daphna Yasova Barbeau, MD (20:33.57) For sure. Yeah, and I mean, it really would be a lot fewer tests.

Ben Courchia, MD (20:42.383) Yeah. I think it's interesting because the paper bundles together ABO and Rh-negative. I think the case for ABO incompatibility is quite strong and makes sense. The Rh side seems to require a little more planning and strategizing with local resources. But again, that could just mean you're only testing for ABO incompatibility when the patient develops hyperbilirubinemia — and you could still DAT all the Rh-negative mothers' babies if you want, since that's not a huge population anyway. In comparison, we have a lot more Rh-positive patients. Anne Hansen, one of the medical directors at Boston Children's Hospital, says: when you want to implement something, don't be the first, but don't be the last. So I'm going to wait for some—

Daphna Yasova Barbeau, MD (21:40.223) Well, there are some institutions that screen like this already.

Ben Courchia, MD (21:45.083) Yeah, clearly the institution in the paper is one of them, so why not. This is coming out of the US — the state of Washington, I believe.

Daphna Yasova Barbeau, MD (22:01.868) I'm realizing that at the very least, it would be nice to know if your Coombs-positive result is specifically anti-D — that would at least be useful in making decisions about bilirubin screening.

Ben Courchia, MD (22:13.468) Yeah. And you know, sometimes we don't have all the testing on the mother, and we don't really know. So you have this Rh-negative mother who received RhoGAM, and the baby comes back Coombs-positive, and you think, "My god, we still have to watch this baby closely" — but in truth, it may just be the RhoGAM showing up on the DAT. These infants often make us nervous, and then we find their bilirubin never really rises above low levels.

I find that frustrating, which is part of why I was excited when we discussed that tool with the team at Capnia — because it feels like— anyway, I'll come back to it, it'll come to me. The point is, there are so many babies who, according to the textbooks, should have a high bilirubin and don't. And then some babies may not be hemolyzing much at all — they could even be Coombs-negative and still have some degree of hemolysis. So there's a lot of unpredictability in hyperbilirubinemia, and I think a more tailored approach makes sense.

Daphna Yasova Barbeau, MD (23:38.796) Yeah, I agree — I think you were referencing CoSense, from Capnia. Okay, all right. I also heard we should be doing a lot more transcutaneous bilirubin measurements.

Ben Courchia, MD (23:53.626) I was thinking that was going to be the first thing you'd say — that we could lean on transcutaneous alone, which I think is true. But here's what's even more interesting, the point you should have made and didn't: very often, if we have a DAT-positive baby, we go straight to a serum bilirubin, because we think, "Oh no, Coombs-positive," and we get a reticulocyte count, a CBC, or a hematocrit — when in truth, you could hold off. You don't even need—

Daphna Yasova Barbeau, MD (24:24.886) Yeah, but why, you know? And that's probably the baby who really deserves transcutaneous monitoring, right? Because if you're checking serum levels every six or eight hours, that's a lot.

Ben Courchia, MD (24:43.14) Right — at least until you actually start phototherapy, in which case it's a different ballgame. But that's what I thought you were going to say — except—

Daphna Yasova Barbeau, MD (24:51.574) Except, remember, we learned about those skin stickers.

Ben Courchia, MD (24:55.494) Yeah, I don't believe in that. Do you think so? I don't know — I have yet to see it. Maybe you're right. Like these little shields.

Daphna Yasova Barbeau, MD (25:09.71) They're coming. So you could keep using transcutaneous. I hope we'll hear from people who say they're almost exclusively using transcutaneous monitoring — let us know.

Ben Courchia, MD (25:21.402) Yeah, but I think the new AAP guidelines give you such a good threshold to check a serum level — why not.

— — — —

Daphna Yasova Barbeau, MD (25:41.230) I'm giving you this JAMA Pediatrics paper: "Vitamin K Prophylaxis in Newborns and Bleeding in Infancy." Lead author Eleni Simatou and senior author Jenny Bolk. This is coming out of Sweden, because of course they have a national birth database. So they wanted to look at the temporal trends of prophylactic newborn vitamin K administration and its association with bleeding diagnoses during infancy. I think this is something we're all dealing with, in parallel, across the ocean. So I wanted to take a look and have more contemporary information on bleeding risk — I think that's very valuable as we counsel families.

Ben Courchia, MD (29:18.259) I feel like I'm watching the Titanic. I kind of know how it's gonna end.

Daphna Yasova Barbeau, MD (29:22.458) You know how it's gonna end. We know, we all know. But — I told you they have a national registry — this was a nationwide cohort study, including all births in Sweden between January 2003 and December 31, 2021. They excluded births before 35 weeks and 0 days' gestation, births with no recorded gestational age, stillbirths, and births with missing identification numbers. They also excluded infants born in one area of Sweden in 2021, because that region had a documentation error affecting vitamin K administration records. Good for them that they have most of the country documented — I think that's pretty good.

The data source — I told you every baby in Sweden has a Swedish personal identity number. Sorry, the mothers and the infants have Swedish personal identity numbers, and they were cross-linked from the Swedish Medical Birth Register and data from the National Patient Register. They also drew on a number of other registers: the Swedish Neonatal Quality Register, the Cause of Death Register, the Total Population Register, and the Education Register. The Swedish Medical Birth Register, for example, has prospectively collected information on prenatal, obstetric, and antenatal care for all births in Sweden since 1973, with more than 98% coverage of births and high validity. That's quite a database for them.

The exposure they were looking for was non-receipt of intramuscular vitamin K administration at birth, based on a checkbox for newborn intramuscular vitamin K administration in the mother's delivery chart. Since January 2012, the delivery chart includes two checkboxes — one for intramuscular and one for oral administration — so they did look at IM (intramuscular) versus oral, we'll discuss that. But the main exposure was whether you got IM or anything else — so oral counts too. Oral is still not IM, you understand what I'm saying? Okay, good.

The primary outcome was bleeding, defined as any ICD-10 (International Classification of Diseases, 10th Revision) code for bleeding in the medical birth register or the national patient register within six months after birth. The secondary outcome was intracranial bleeding specifically. They also looked at bleeding diagnoses within the first 28 days after birth.

I looked at a number of potential confounders based on previous literature, and they had a whole bunch of perinatal and infant characteristics — you name it, it was probably included in their register. They also wanted to look at non-receipt over time and by location.

In addition to all that, they did several sensitivity analyses to check for potential risk of underreporting of vitamin K administration. They looked at excluding infants who were admitted to the neonatal ward, in case they'd gotten vitamin K from the neonatal staff instead of the obstetric staff. They excluded infants who died during the first two days after birth, in case those deaths reflected unrelated perinatal conditions that had nothing to do with vitamin K administration. They excluded the counties in Sweden with the highest percentages of non-documented intramuscular vitamin K, since those might skew the results as outliers. And they excluded infants with low Apgar scores — less than seven — thinking that those babies, who needed resuscitation, might have had suboptimal documentation of vitamin K administration and a higher risk of bleeding unrelated to whether or not they actually got vitamin K.

So, what were the results? There was a total of 2,115,352 births in Sweden between January 2003 and December 2021. They excluded a number of births — premature births, stillbirths, missing identification numbers, and the infants born in the region with the highest non-receipt. After exclusions, they used a total of just over two million live-born infants. Of these, 24,000, or 1.19%, had no documented record of receiving intramuscular vitamin K. The cohort's gestational age was about 40 weeks on average — so, a term population — and about 48.2% were female.

Information on oral vitamin K administration was available only for infants born between 2012 and 2021 — just over 1.1 million live-born infants during that time frame. Looking at what characteristics were associated with not receiving IM vitamin K: mothers of infants without IM vitamin K were more likely to be of Nordic origin, more likely to have delivered by emergency C-section, and less likely to be cohabiting with a partner. Infants without IM vitamin K had a higher prevalence of a five-minute Apgar score below seven, and were also less likely to be AGA (appropriate for gestational age) in weight.

Otherwise, the groups were pretty comparable overall — though there was a slightly higher proportion of missing data on maternal education, cohabitation, mode of delivery, and Apgar scores among births with no record of intramuscular vitamin K, so a lack of vitamin K documentation also tended to come with gaps in other recorded data.

The rate of infants with no record of intramuscular vitamin K administration at birth was 1.32% in 2003. This decreased to 0.66% in 2006, then increased to 1.5% in 2021 — a statistically significant increase. There were also regional variations in non-receipt.

What about the odds of bleeding? During the study period, the frequency of bleeding episodes up to six months of age doubled, from 0.22% in 2003 to 0.53% in 2021. Infants without IM vitamin K had higher odds of bleeding compared with those who did receive intramuscular vitamin K — the adjusted odds ratio for any bleeding was 1.52, and the adjusted odds ratio for intracranial bleeding specifically, within the first six months of life, was 2.91. Looking just at the neonatal period — the first 28 days of life — infants without intramuscular vitamin K had an adjusted odds ratio of 2.62 for any bleeding, and nearly 2.9 for intracranial bleeding, compared to those with intramuscular vitamin K. A composite outcome of bleeding or death within six months showed similar associations.

So what about oral vitamin K? Between 2012 and 2021, information on oral vitamin K administration was available for 1.1 million live-born infants. Of these, about 14,000 — nearly 15,000 — had no record of intramuscular vitamin K, and among infants without intramuscular vitamin K, about 10% received oral vitamin K instead. There was a substantial increase in oral vitamin K administration over the study period — from fewer than five infants in 2012, to 87 in 2013, to 193 infants in 2021.

Ben Courchia, MD (38:22.003) But it would have been higher.

Daphna Yasova Barbeau, MD (38:46.361) Baseline characteristics of births with versus without any record of vitamin K administration were still pretty similar. Now — infants who received oral vitamin K at birth had significantly higher odds of bleeding compared with those who received intramuscular vitamin K. The adjusted odds ratio for any bleeding within the first six months of life was 2.26, and for bleeding within the first 28 days of life, it was 3.74. So — for families who say, "well, I want oral" — this is the difference between IM and oral. They also wanted to look specifically at what was happening with home births. Between 2018 and 2021, a total of 441,000 live-born infants were recorded in Sweden.

Ben Courchia, MD (39:18.677) Wow, that's not negligible, huh?

Daphna Yasova Barbeau, MD (39:45.278) That's the time period where they had home birth data. Of these infants, 0.09% were born at home. I actually thought it would be a higher percentage, but notably, most of the obstetric team in Swedish hospitals are midwives, so I think there's maybe a more balanced approach between home birth and hospital birth.

Among home-born infants, nearly 36% had documented intramuscular vitamin K administration, and 64% did not. Of the 254 infants without an IM injection, 7.5% received oral vitamin K — representing about 5% of all home-born infants. So that's roughly 35% plus about 5%, meaning less than half of home births received any vitamin K at all. Because the number of cases was so small — 396 home-born infants — no conclusions were drawn about bleeding risk specifically in that group.

I told you they did a number of sensitivity analyses: excluding infants who went to the NICU (Neonatal Intensive Care Unit) didn't change the results; excluding infants who died within the first two days of life didn't change the results; removing the three counties with the highest percentage of infants not receiving IM vitamin K didn't change the results; and restricting the analysis to infants with a low five-minute Apgar score slightly attenuated the results.

So, they conclude: in this nationwide cohort of more than two million live births of term gestation, there was a small but steadily increasing proportion of infants without intramuscular vitamin K. These infants had an increased risk of bleeding episodes, including intracranial hemorrhage, within the first six months of life. And there's increasing use of oral vitamin K, despite its elevated bleeding risk compared with intramuscular administration. Consistently, they found an almost doubled risk of bleeding episodes in infants without IM vitamin K compared with infants who did receive it — regardless of adjusting for numerous confounding variables.

Ben Courchia, MD (42:30.260) Yeah, so many things about this study. So the punchline, I think, everybody would have guessed — obviously, babies who don't receive vitamin K have a higher risk of developing vitamin K deficiency bleeding. The graph in Figure 2 is quite nice — you really see that linear progression of both refusal and bleeding diagnoses. A few takeaways: one, it's good in terms of counseling. It's hard to counsel families who really push for oral vitamin K — we can talk to them about the absorption, and about the unpredictability of understanding how much vitamin K their baby is actually going to receive, whether they do a single dose or repeated dosing. But it's nice to have this kind of data showing that it does rank well below IM vitamin K, and it still increases the risk of bleeding.

It seems to be at least a little better than no vitamin K, but it still doesn't match the benefits of the intramuscular injection. And then when it comes to the TikTok trend of not exposing your baby to "harmful" vitamin K — it looks like this is crossing borders. We tend, especially from a medical and sociologic standpoint, to look at Scandinavia as sort of a model society, almost. But it looks like no one's immune to this epidemic of ill-advised decisions. I was going to say something else.

Daphna Yasova Barbeau, MD (44:15.856) Misinformation. We have families asking all the time about oral — I mean, in the US, there's no FDA (Food and Drug Administration)-approved oral formulation, but parents are getting it, and I don't even know what they're bringing sometimes when they come to show it to us. Sometimes parents will say, "that's old data, they haven't done any recent studies" — but—

Ben Courchia, MD (44:42.401) We should ask people — do you remember that case we had, where a parent said no, then gave oral, regretted it, and said, "I want the IM"? And the question was, can we give an IM dose or not? It'd be interesting to see what people would do in that case — giving the dose in addition, or saying "now you're committed to this path." I won't say what we did, for HIPAA (Health Insurance Portability and Accountability Act) compliance purposes.

Daphna Yasova Barbeau, MD (45:08.815) Let us know what you all think. The counseling was so good that they wanted it again — they changed their mind. Well, we just keep going to work, and every time I'm in the nursery — today I did the nursery post-call, and I was like—

Ben Courchia, MD (45:29.997) So Daphna's email is on the website — I'm going to put you on the hook for that, okay? She recently redrafted our vitamin K refusal form, and it's perfect.

Daphna Yasova Barbeau, MD (45:40.648) I don't know if it's perfect, but I am very proud of our oral vitamin K education-plus-refusal form.

Ben Courchia, MD (45:43.775) It is very good — a very informative refusal form. So if you're considering it, you can find Daphna's email on the website and ask her to send you her copy.

Daphna Yasova Barbeau, MD (45:57.500) Sure, I'll be happy to share. We'll let you know if it's effective first, but so far we're very excited about it. I had never had this experience before, where I pulled up and thought, "okay, I really have to check every baby," because now I'm not even sure somebody will tell me if a parent refuses vitamin K — it's becoming so—

Ben Courchia, MD (46:19.809) Yeah, it's become a new question I ask parents: "Has your baby received vitamin K at birth?" It used to be something taken for granted.

Daphna Yasova Barbeau, MD (46:28.336) That's right — especially for babies we're not admitting right after delivery, who come to the NICU later for an episode in the nursery, or a readmission.

Ben Courchia, MD (46:34.743) Right after delivery, I look for that little band-aid.

Daphna Yasova Barbeau, MD (46:45.163) That’s right

— — — —

Daphna Yasova Barbeau, MD (46:56.922) I'm very much looking forward to the paper you're presenting.

Ben Courchia, MD (47:08.441) One exciting paper. That's right. It's titled "Initial CPAP (Continuous Positive Airway Pressure) of 5 versus 8 Centimeters of Water During Delivery Room Stabilization of Very Preterm Infants — the PHILODENDRON Pilot Randomized Trial." Nate Sundgren made some jokes about the name of this trial.

Daphna Yasova Barbeau, MD (47:36.026) Do you know what a philodendron is? No? It's a plant — which is kind of funny, given that they're the ones studying the babies.

Ben Courchia, MD (47:38.690) No, I do not.

Daphna Yasova Barbeau, MD (47:52.295) If anybody's looking for a very easy houseplant to grow, a philodendron's a good pick.

Ben Courchia, MD (48:09.892) This is where my vocabulary ends.

Daphna Yasova Barbeau, MD (48:09.892) I feel like we were just talking about whether we should be using CPAP at all in delivery room stabilization, and now we're going to plus eight.

Ben Courchia, MD (48:16.848) Different topic — that was the study we reviewed about giving 20 minutes of prophylactic CPAP to late-preterm babies. That's a very different population from what we're talking about today. Here we're talking about babies born before 32 weeks. We know they have RDS (Respiratory Distress Syndrome), and that it's a major source of morbidity for them.

Immediate and appropriate respiratory support in the delivery room is critical to allow clearance of fetal lung fluid and to reach an early, adequate functional residual capacity (FRC), thereby mitigating complications such as atelectasis, overinflation, hypoxemia, and lung injury. Establishing adequate FRC in the first minutes after birth is one of the key determinants of a successful transition, particularly in preterm lungs, which are deficient in surfactant.

So the strategy we all use, based on the evidence, is delivering CPAP in the delivery room to provide continuous distending pressure that keeps the alveoli open during expiration, reducing the work of breathing and the need for mechanical ventilation. Now, there's a question of how much CPAP we should use — maybe a higher CPAP gives better alveolar recruitment. But the concern, obviously, is adverse events: pneumothoraces or other air leak syndromes, hemodynamic impairment.

At the time of this writing, the current European consensus guideline recommends initiating CPAP at approximately 6 cmH2O (centimeters of water) in spontaneously breathing preterm infants, with subsequent titration according to clinical response. However, this recommendation is largely based on limited evidence, and the optimal initial distending pressure in the delivery room isn't certain — which is interesting, because practice differs quite a bit. Most people here in the US, I'd say, start at a CPAP of 5.

The study aims to address this lack of evidence around the optimal initial CPAP at delivery. The authors compared a CPAP of 5 with a CPAP of 8, in babies born between 26 and 29 weeks of gestation — very preterm. I think it's a fascinating question, and that's why I wanted to present this paper: we put these kids on 5, and by the time they're in the NICU (Neonatal Intensive Care Unit), they're almost never still on 5 — by the end of day one, they're on 6, on 7.

Daphna Yasova Barbeau, MD (50:49.378) Nope — they always need more than 5.

Ben Courchia, MD (50:59.675) So that's for you and your practice to know.

Daphna Yasova Barbeau, MD (51:00.502) Maybe I give more than you do.

Daphna Yasova Barbeau, MD (51:10.561) I think I'd start at 5, at least.

Ben Courchia, MD (51:13.615) Start at 5, sure, but even then it doesn't fully make sense, given we know babies typically need more — and apparently the Europeans now start at 6. So it's a very good question. This is an investigator-initiated pilot randomized controlled trial conducted at a level 3 NICU — the Buzzi Children's Hospital in Milan, Italy — between 2020 and 2023.

The primary objective was to evaluate the feasibility and safety of eventually conducting a multicenter randomized controlled trial comparing two initial CPAP strategies during delivery room stabilization of very preterm infants. Secondary, exploratory objectives were to compare early physiological respiratory adaptation and short-term respiratory outcomes between the two groups. They included babies born between 26 weeks 0 days and 29 weeks 6 days who required non-invasive respiratory support immediately after birth.

Infants with major congenital malformations, diagnosed either antenatally or postnatally, or whose parents declined participation, were excluded. Participants were randomized 1:1 to receive CPAP at a level of 5 or 8. Respiratory support was delivered using a T-piece resuscitator with an appropriately sized face mask. CPAP was maintained at the assigned level for the first 15 minutes after birth, unless escalation to positive pressure ventilation or intubation was required. Initial FiO2 (fraction of inspired oxygen) was set at 30%, as most people do, and subsequently adjusted to maintain oxygen saturation within target range.

After the first 15 minutes, CPAP was standardized to a level of 6 in both groups — so the CPAP-5 group went up to 6, and the CPAP-8 group came down to 6, until NICU admission. Subsequent respiratory management followed local NICU protocol, and surfactant was given at a dose of 200 mg/kg once the infant reached an FiO2 threshold of 30% or greater on CPAP. That threshold is an important detail to understand, since it becomes one of the outcomes they measured — essentially, how many of these babies did better than others.

The primary objective of the pilot trial was feasibility. Safety was evaluated by monitoring predefined adverse events, including pneumothorax, PIE (pulmonary interstitial emphysema), and need for endotracheal intubation during the early postnatal period. Predefined exploratory physiological outcomes included preductal oxygen saturation, FiO2, heart rate, and the preductal SpO2-to-FiO2 ratio — the S/F ratio — during the first 15 minutes after birth. Have you ever used the S/F ratio?

Daphna Yasova Barbeau, MD (54:20.961) No, I have not.

Ben Courchia, MD (54:22.929) It's interesting — makes sense once you see it. I'll walk you through it when we get to the numbers. They use the S/F ratio as a non-invasive surrogate marker of oxygenation efficiency during the transition, along with some other outcomes we'll get to.

During the study period, 56 preterm infants were successfully enrolled and randomized — 28 to each arm. It's a pilot trial, small, as expected. The two groups were well balanced for gestational age, birth weight, sex, Apgar scores, and other prenatal characteristics: mean gestational age of 28 weeks in both groups, mean birth weight of about 1 kilo in the CPAP-5 group and about 1,185 grams in the CPAP-8 group, similar rates of SGA (small for gestational age), similar rates of antenatal corticosteroid exposure.

In terms of safety outcomes, both interventions were well tolerated, with no statistically significant differences between the two groups. Pneumothorax occurred in 0 of the infants in the CPAP-5 group versus 3 out of 28 in the CPAP-8 group — higher in the CPAP-8 group, but with a p-value of 0.23, not statistically significant. PIE occurred in none of the babies started on CPAP of 5, and in 2 out of 28 in the CPAP-8 group — p-value of 0.49, again not statistically significant. Delivery room intubation was uncommon: 2 out of 28 in the CPAP-5 group, 1 out of 28 in the CPAP-8 group — also not significant.

For exploratory physiological outcomes, the preductal S/F ratio during the first 15 minutes was similar between groups. At 15 minutes, it was 271 ± 100 in the CPAP-5 group and 269 in the CPAP-8 group. To put that in context: if you're saturating 97% on 21% oxygen, that's 97/0.21 — about 460. If you're saturating 90%, the bare minimum, on 60% oxygen, your S/F ratio is about 150. So these babies, with ratios around 270, were probably saturating around 92%, on FiO2s in the 30s to 40s. The best possible score is close to 500; a poor score would be down in the 150s.

Ten infants — 35.7% — in each group required positive pressure ventilation during delivery room stabilization. Duration of PPV (positive pressure ventilation) was shorter in the CPAP-8 group, though not statistically significant. Overall proportion of infants receiving surfactant during the NICU stay was similar between groups — 22 out of 28 in CPAP-5, 20 out of 28 in CPAP-8, p-value of 0.9. However, infants in the lower CPAP group reached the predefined surfactant treatment threshold significantly earlier than those in the CPAP-8 group: 2.2 ± 1.7 hours versus 3.5 ± 2.1 hours. Statistically significant — clinically, you'll be the judge.

For intubation in the first 72 hours of life: 9 out of 28 in the CPAP-5 group, 7 out of 28 in the CPAP-8 group — not statistically significant. Duration of mechanical ventilation, BPD (bronchopulmonary dysplasia), and the composite outcome of BPD or death showed no significant difference between groups — although they note a trend toward shorter mechanical ventilation and lower BPD rates in the CPAP-8 group. I don't think I'd put too much stock in a trend from such a small cohort.

Their conclusion: this pilot RCT shows that a trial comparing CPAP-5 versus CPAP-8 is feasible and not associated with major short-term safety concerns. Exploratory physiological findings suggest a higher CPAP may influence early respiratory adaptation and timing of surfactant administration, though these observations should be considered hypothesis-generating, not confirmatory. A larger trial is obviously needed.

That's really it. My honest take: I look at this paper and get excited, because physiologically it makes sense to me that maybe we should give a higher PEEP (Positive End-Expiratory Pressure). Then I read the pilot data and think — maybe not so fast. If you haven't listened to our episode with Brett Manley about clinical trials, I think it's a great follow-up to this conversation — it really raises the question: you're designing a trial comparing CPAP-5 versus CPAP-8, but what if CPAP-8 isn't the right number, and it should really be CPAP-7? The time and resources these trials take really make you wonder whether we need new trial designs for these kinds of questions. Something to think about — we'll see what comes out of the larger trial they're planning.

Any thoughts on this?

Daphna Yasova Barbeau, MD (60:39.035) No, I totally agree with you. That's the mainstay of our understanding of the transition — fetal lung fluid moving out of the lungs, and establishing functional residual capacity. It feels like it should work physiologically. But the trial was small, so we'll see.

Ben Courchia, MD (61:01.914) I know — but you do worry a bit about that air leak signal being a little larger in the CPAP-8 group. You wonder whether, with a bigger sample, that difference might become more pronounced.

Daphna Yasova Barbeau, MD (61:06.679) Absolutely — it might become significant.

Ben Courchia, MD (61:06.679) But I'm sure there'll be a safety board, and if they start seeing negative outcomes, they'll halt that kind of trial sooner rather than later.

— — — —

Daphna Yasova Barbeau, MD (61:13.373) I was looking forward to this paper you found on birth defects and IVF (In Vitro Fertilization). It just so happens that in the last week, a number of people have asked me about comorbidities associated with IVF treatment — we know those babies are more likely to end up in the NICU.

Ben Courchia, MD (61:25.116) Yeah, there's a lot we don't fully understand about the effects of IVF.

Daphna Yasova Barbeau, MD (61:46.333) We don't totally understand why, but we know we're going to be seeing them. That's why I thought this was a neat paper — it's titled "In Vitro Fertilization and VACTERL (Vertebral defects, Anal atresia, Cardiac defects, Tracheoesophageal fistula, Renal anomalies, Limb abnormalities) Birth Defects." I like looking for VACTERL, I like teaching about it — and I feel like we've been seeing a lot of it recently. Don't you think?

Ben Courchia, MD (62:24.419) No.

Ben Courchia, MD (62:29.861) I reviewed the data and I don't think there's been a spike, but that's okay — that's your reality.

Daphna Yasova Barbeau, MD (62:35.645) Well, that's what it feels like to me.

This is published in JAMA Network Open.

Ben Courchia, MD (62:46.875) What did they call this today — your lived experience?

Daphna Yasova Barbeau, MD (62:50.031) This is my lived experience. Lead author Ji Yun

Tark, senior author Barbara Luke. The objective was to characterize the association between IVF and VACTERL. This comes to us from a reporting system called SART — the Society for Assisted Reproductive Technology's Clinical Outcome Reporting System (SART CORS). I hadn't heard of that before, and I learned quite a bit.

The study linked IVF cycles reported to SART CORS with birth records and birth defect registry data from four states — New York, Texas, Massachusetts, and North Carolina, for reasons I'll explain — between January 2004 and December 2018 for New York, Massachusetts, and North Carolina, and between 2004 and 2017 for Texas. During that period, these states recorded 12,456,523 births, representing approximately 20.5% of all US live births for that time. These states also have a high proportion of IVF births: New York ranks third, Texas 15th, Massachusetts second, and North Carolina sixth nationally in annual IVF births as of 2018. IVF births comprised 3.5% of all births in New York, 1.7% in Texas, 5.4% in Massachusetts, and 1.5% in North Carolina.

More than 90% of all ART (Assisted Reproductive Technology) cycles in the US are performed at SART member clinics. SART also conducts clinic evaluations — validation visits — and if there are reporting problems, clinics can go through a probationary period. In 2003, the US birth certificate added an infertility treatment indicator with treatment type. Births were classified as IVF if linked to a SART CORS cycle. Births with no birth certificate indication of infertility and no SART CORS linkage were classified as naturally conceived. Any birth with an indication of infertility treatment on the birth certificate but no SART CORS linkage was excluded, to keep the SART CORS cohort clean.

Each state linked SART CORS records to birth certificates to identify IVF-conceived births, achieving greater than 90% linkage. They also linked mothers to other births to identify siblings — any sibling conceived via IVF was included in the IVF group, and those conceived without fertility treatment were classified as naturally conceived. They also looked at IVF subtype: embryo state, ICSI (Intracytoplasmic Sperm Injection) use, male infertility, and thawed versus fresh embryo transfer — and then specifically at VACTERL.

Daphna Yasova Barbeau, MD (67:38.320) The four participating states were either current or former CDC Centers for Birth Defects Research and Prevention, so they had enhanced birth defects surveillance and good-quality data on babies with birth defects. Texas, North Carolina, and Massachusetts used active birth defects surveillance; New York used passive surveillance.

To promote consistency, they used standard case definitions established by the National Birth Defects Prevention Study and the National Birth Defects Prevention Network. Birth defects were classified using British Pediatric Association (BPA) codes — not ICD-9 or ICD-10 codes. They're often similar to ICD codes, but not synonymous.

Children with chromosomal defects or known genetic syndromes were excluded before birth defect classification — this removed about 3,500 children. Children with VACTERL were identified by having codes for the individual phenotypes or components — you couldn't just be coded as "VACTERL," you had to have proof of at least three of the VACTERL criteria: vertebral defects, anal atresia, cardiac defects, esophageal atresia with or without TEF (tracheoesophageal fistula), renal defects, or limb defects. They also excluded babies with VACTERL plus hydrocephalus, considering that a distinct pathology, and analyzed babies with multiple anomalies that weren't VACTERL separately as well.

The comparison groups were: children with other birth defect syndromes and/or associations sharing the VACTERL BPA code — about 200 babies with multiple anomaly syndromes; children with two birth defects, about 13,000 infants; children with at least three birth defects not meeting VACTERL criteria, about 14,000 babies; and children with isolated structural defects defined by other BPA codes.

Daphna Yasova Barbeau, MD (70:02.600) The study included 1,555,936 live births, about 48% female, including 175,161 IVF-conceived births. Ethnic background: 8.3% Asian or Pacific Islander, 14% Black, 70% white, 6.8% other; 26.1% of mothers were of Hispanic ethnicity.

Among the 200 births meeting VACTERL criteria, 159 were conceived without IVF and 41 with IVF. Compared with naturally conceived children, those conceived through IVF were more likely to be born to older, non-Hispanic white women with higher educational level, higher plurality, lower parity, higher rates of gestational diabetes or hypertension, younger gestational age at delivery, and more frequent cesarean delivery — findings that have been replicated numerous times in IVF populations.

Among the 200 children meeting VACTERL criteria, 64% were male, 36% female. The most common phenotype combination was cardiac, renal, and anal atresia — 20.5% of cases. Distribution of VACTERL component phenotypes: vertebral defects in 51%, anal atresia in nearly 60%, cardiac defects in 80%, esophageal atresia with or without TEF in 23%, renal defects in 84%, and limb defects in 27.5% of cases. That's actually much higher for vertebral defects and anal atresia than what we tend to see clinically in our group.

Daphna Yasova Barbeau, MD (72:28.073) After adjustment, IVF conception was associated with higher odds of anal atresia — adjusted odds ratio (aOR) 2.28 — higher odds of cardiac defects, aOR 1.25, higher odds of renal anomalies, aOR 1.31, and higher odds of limb anomalies, aOR 1.52. The odds of vertebral defects and esophageal atresia with or without TEF were elevated but lost statistical significance.

For the overall IVF–VACTERL association: after adjusting for demographic and pregnancy-related factors, IVF-conceived children were more likely to meet VACTERL criteria compared with naturally conceived children, with an adjusted odds ratio of 1.86. Estimates were similar for multiple anomaly syndromes — aOR 1.82 for IVF-conceived babies. Associations for other birth defect groups outside VACTERL had lower odds than the VACTERL group.

Associations were generally consistent across IVF treatment types and infertility categories, though all remained significant. When restricted to singleton live births, the association between IVF and VACTERL remained consistent. They found no evidence of interaction between IVF conception and child sex — though the VACTERL rate was higher in male than female children among natural births, unlike the pattern seen with IVF conception.

In conclusion, they observed a VACTERL prevalence of 1.2 per 10,000 births among natural conceptions. The prevalence of VACTERL among IVF-conceived births was higher than among naturally conceived infants, and roughly equal between males and females. The highest odds of birth defect patterns in IVF conception was for the VACTERL association, followed by multiple anomaly syndromes. IVF was also associated with patterns of multiple birth defects and isolated defects, but with smaller odds than for VACTERL and multiple anomaly syndromes. Taken together, the authors conclude that conception via IVF is associated with complex patterns of birth defects.

Ben Courchia, MD (75:31.101) Very interesting.

Daphna Yasova Barbeau, MD (75:32.035) Very interesting — how does this change our practice? I like knowing this. I think if it crosses your mind that a baby was conceived via IVF, you'd better think twice. If they have some of the soft signs — anal atresia, you're going to work them up; TEF, you're going to work them up — but babies can have other combinations of VACTERL that are less obvious. Since the most common phenotypes are renal-cardiac-anal atresia and renal-cardiac-vertebral, we may be missing some of those renal-cardiac-vertebral babies.

Ben Courchia, MD (76:20.809) Is that going to change your practice? Are you going to start looking more?

Daphna Yasova Barbeau, MD (76:23.689) You'd get a renal ultrasound.

Ben Courchia, MD (76:28.681) Yeah, this is quite impressive.

Daphna Yasova Barbeau, MD (76:37.533) I also think we're finally getting better information — in residency and fellowship, I didn't always even know if a baby was IVF-conceived.

Ben Courchia, MD (76:49.957) I ask — specifically for twins and such — but for a singleton, not always.

Daphna Yasova Barbeau, MD (76:58.173) I feel like we're getting better information about the IVF process for families now — it's following them through to delivery, whereas before it felt like it was getting lost somewhere along the birth or perinatal history.

Ben Courchia, MD (77:12.177) Yeah — I guess if you suspect it... you mentioned it already in the review, but it's about an 86% increase in the odds. The odds were 86% higher — 1.86 was the odds ratio.

Daphna Yasova Barbeau, MD (77:20.573) Does that mean we should work up every IVF pregnancy? No — but it's not insignificant.

Ben Courchia, MD (77:41.095) And it's not like this is a small sample — 1.5 million births. That's substantial.

Daphna Yasova Barbeau, MD (77:48.177) Yeah, and it crossed several different states — a broad population — so I think it was a good cohort to look at.

Ben Courchia, MD (77:59.089) Sure, broad, broad, broad. Alright, now you know.

Daphna Yasova Barbeau, MD (78:02.853) Now you know. You heard it here.

— — — —

Eli (78:04.853) We're moving on to our next article. Our last article talked about work-life balance and the importance of having that balance — existentially, morally, spiritually, as well as some ruggedly pragmatic outcomes around patient safety, satisfaction, burnout, and workforce turnover.

Today I think we should talk about the work side of the equation. This comes from a really stimulating article published in the Journal of Perinatology, which gets at a fundamental question: are NICU (Neonatal Intensive Care Unit) admissions being accurately reported? Or is the utilization of NICU services actually much higher than some of the data suggests? And if that's true, which data should we be relying on to assess how many babies actually need NICU care — and how should policymakers at the state and federal level decide what resources to devote to NICUs, based on data that shows them the true burden of utilization?

This study, published in the Journal of Perinatology, used revenue codes in discharge data to assess whether infants had actually used NICU services. The top-line finding: almost 40% of babies admitted to a NICU did not have a NICU admission recorded on their birth certificates. That gap in birth certificate data was even more pronounced in certain subgroups — for term babies, birth certificates missed more than half of NICU admissions,

Eli (80:14.224) including term infants with conditions categorized as severe. The gaps in reporting were also pronounced in certain subpopulations. For extremely premature babies — those under 28 weeks — infants born to Black mothers were 20% more likely to have a missed NICU admission on their birth certificate than infants born to white mothers. Likewise, missed NICU admission rates were 31% higher in very preterm infants born to Black mothers compared to white counterparts, and 39% higher in moderately preterm infants.

The main takeaway is something the authors put really well in two separate quotes, in the conclusion and discussion sections. Their first quote: "We found that birth certificates dramatically misrepresent true NICU admissions across gestational ages, even among very preterm infants and infants with serious medical conditions" — exactly the kinds of infants you'd think NICU utilization data would most reliably reflect. Their second quote: "Misclassification of NICU admission status on birth certificates can lead to substantial error and bias in studies of newborn outcomes and disparities, NICU utilization, and will result in serious underestimation of newborn healthcare needs."

Ben, what did you think of this study? The findings, and the conclusions they draw?

Ben Courchia, MD (81:53.620) First of all, I don't know about you, Eli, but I didn't realize NICU admission was even a box on the birth certificate. When I went to get my daughter's birth certificate, I don't remember that being a question. So I was pleasantly surprised just by the premise of the study — I didn't realize birth certificates captured NICU admission information at all. That alone surprised me.

There are two things I want people to focus on. First, the basic comparison between birth certificate data and actual NICU admissions. This was clever methodologically — they had the NICU discharge list, so they knew every baby who had actually been in the NICU, and correlating that with birth certificates made the data quite robust.

What's interesting is how that comparison plays out across gestational age groups. For the preemies, it's not that bad — still significant, but not huge. For 22 to 27 weeks, it's an 8% discrepancy. 28 to 31 weeks, 7%. 32 to 33 weeks, 12%. But for late preterm and full-term babies, the discrepancies are huge — 34 to 36 weeks shows about a 30% difference between birth certificate data and discharge data, and for full-term infants, it's a 54% discrepancy.

What's interesting to me is that this late-preterm-to-full-term population represents a large proportion of our NICU admissions. As the authors point out, if we misrepresent how much NICU utilization is actually happening, we're less likely to see resources directed toward neonatology, and less likely to recognize the need to support families in the NICU. That's a tremendously important factor.

Ben Courchia, MD (84:10.119) We try to wave our arms and say, "Hey, look at us, we should care more about babies" — but this is a very low-hanging fruit, in my opinion, that we could address to accurately represent the actual healthcare utilization needs of our society. For context, this data came from Pennsylvania and South Carolina.

And then looking at the data by race, I was surprised there too — for more mature infants, the discrepancy tended to happen more often in babies born to white mothers. But when we're talking about preterm infants, it skews the other way — the discrepancy tends to happen more for babies born to Black mothers. I thought that was very interesting. The authors don't venture to explain why, and I don't think we can know why just from this data, but it's definitely something worth pursuing for anyone interested in this subject. Something I hadn't thought about before — I'm actually planning to ask our birth certificate team at the hospital about this on Monday.

Eli (85:50.823) Yeah — in the Neo News tradition of giving people ideas for their tenure-track research career: NICU utilization. Certainly Scott Lorch and the folks at CHOP (Children's Hospital of Philadelphia) have been doing amazing health services research in neonatology for years, but there seems to be a lot of room to run here.

I think people might be wondering, why are we covering this on Neo News? I think this is a good example, first, of Ben and I being pretty wonky — we love studies with interesting methodologies. This is a great example of using imagination and creative methodology to answer an important research question that maybe gets overlooked or taken for granted. The other thing we think about is what's in the zeitgeist of the broader conversation. The extent of NICU utilization is squarely in the zeitgeist right now, in a moment of widespread hospital closures, including maternity care closures. And the second thing is, in an era of budgetary constraints, with a lot of conversation especially at the state level — since states have to balance their budgets — the question becomes: where should they put their money?

Eli (87:16.368) The first thing anyone listening to Neo News could do overnight is write to their local representatives and ask: how are you making decisions about how much money to allocate to NICUs? Are you using billing codes? Birth certificate data? What does that data actually look like, and does it reflect the true burden of utilization? Because I think you'd assume they're using the right data, and that it's accurate — but this study shows that, and if there's been any consistent trend we've seen, especially during this administration, it's that the data being used may only capture a subset of the full picture. So is the data reliable, and does it accurately reflect the needs of NICUs? Or is there an opportunity to rethink how we evaluate the kind of support NICUs need, so that we all have access to the resources required to best care for our babies and their families?

Ben Courchia, MD (88:31.702) Exactly right. I think sometimes even the discussion at the state or national level can feel overwhelming, and we end up having to make our case with individual stories. But this is solid, grounded epidemiologic data that can really help turn the tide. I think this is all very interesting.

Eli (89:04.198) And I think the magic is in pairing epidemiologic data like this with individual stories — that's where our listeners have such amazing insight. All of us remember a story where we were strapped, where there weren't enough kits, we were missing equipment, or there was some other resource gap that forced us to pivot our plan for caring for a baby. The question is: are those resource gaps preventable? Are they related to funding decisions at the state and federal level that could be solved for? And if so, is the reason those gaps exist because the data decision-makers are using to think about resource allocation is incompletely reflective of both the care needs of our patients and the resource needs of our facilities?

Ben Courchia, MD (90:03.576) Thank you, Eli — I'll see you next week. Bye.

Eli (90:06.632) See you next week then.