#297 - 📑 Journal Club - The Complete Episode from April 6th 2025
- Mickael Guigui
- Apr 6
- 39 min read
Updated: Apr 11
Hello friends 👋
In this week’s Journal Club, Ben and Daphna dissect six critical studies reshaping our understanding of neonatal care. Kicking off with the Canadian Neonatal Network’s multicenter cohort on inhaled nitric oxide (iNO), they explore how early pulmonary hypertension responsiveness to iNO may predict survival in preterm infants. They then examine a poignant study on maternal decision regret following extremely preterm births, revealing elevated regret across all pathways—comfort, active care, or otherwise—underscoring the emotional complexity of shared decision-making. The duo dives into data from the NICHD Neonatal Research Network on how even a single dose of antenatal steroids improves outcomes incrementally by the hour, strengthening the case for early administration. Ben highlights an AI-driven TPN algorithm that may outperform clinicians in individualized nutrition planning, while Daphna introduces an unexpected intervention—xylitol chewing gum—to reduce preterm birth in low-resource settings. Rounding out the discussion is a provocative look at intermittent hypoxemia as a predictor of systemic hypertension and the landmark NICHD trial on therapeutic hypothermia for infants 33–35 weeks GA. With a striking 87% probability of harm, this trial challenges the current drift toward “cooling creep.” Essential listening for any neonatologist seeking evidence-based clarity in an ever-evolving field.
----
The articles covered on today’s episode of the podcast can be found here 👇
Phongpreecha T, Ghanem M, Reiss JD, Oskotsky T, Mataraso SJ, De Francesco D, Reincke SM, Espinosa C, Chung P, Ng T, Costello JM, Sequoia JA, Razdan S, Xie F, Berson E, Kim Y, Seong D, Szeto MY, Myers F, Gu H, Feister J, Verscaj CP, Rose LA, Sin LWY, Oskotsky B, Roger J, Shu CH, Shome S, Yang LK, Tan Y, Levitte S, Wong RJ, Gaudillière B, Angst MS, Montine TJ, Kerner JA, Keller RL, Shaw GM, Sylvester KG, Fuerch J, Chock V, Gaskari S, Stevenson DK, Sirota M, Prince LS, Aghaeepour N.Nat Med. 2025 Mar 25. doi: 10.1038/s41591-025-03601-1. Online ahead of print.PMID: 40133525
Baczynski M, Weisz D, Thomas L, Fevrier S, Castaldo M, Soraisham A, Hyderi A, Agarushi R, Bhattacharya S, Lalitha R, Sidhu A, Abdul Wahab MG, Altit G, Hébert A, Louis D, Elsayed Y, Mitra S, Deshpande P, Kharrat A, Zhu F, Ting J, Yoon E, Shah PS, Jain A; Canadian Neonatal Network Investigators.JAMA Netw Open. 2025 Feb 3;8(2):e2458843. doi: 10.1001/jamanetworkopen.2024.58843.PMID: 39928335 Free PMC article.
Martinez S, Chen Z, Di Fiore JM, Nguyen C, Minich NM, Hibbs AM.Pediatr Res. 2025 Jan 31. doi: 10.1038/s41390-025-03881-w. Online ahead of print.PMID: 39885241
Belden L, Kaempf J, Mackley A, Kernan-Schloss F, Chen C, Sturtz W, Tomlinson MW, Guillen U.Arch Dis Child Fetal Neonatal Ed. 2025 Feb 21;110(2):191-199. doi: 10.1136/archdischild-2024-327287.PMID: 39164062
Chawla S, Wyckoff MH, Lakshminrusimha S, Rysavy MA, Patel RM, Chowdhury D, Das A, Greenberg RG, Natarajan G, Shankaran S, Bell EF, Ambalavanan N, Younge NE, Laptook AR, Pavlek LR, Backes CH, Van Meurs KP, Werner EF, Carlo WA; National Institute of Child Health and Human Development (NICHD) Neonatal Research Network (NRN).JAMA Netw Open. 2025 Feb 3;8(2):e2461312. doi: 10.1001/jamanetworkopen.2024.61312.PMID: 39982720 Free PMC article.
Faix RG, Laptook AR, Shankaran S, Eggleston B, Chowdhury D, Heyne RJ, Das A, Pedroza C, Tyson JE, Wusthoff C, Bonifacio SL, Sánchez PJ, Yoder BA, Laughon MM, Vasil DM, Van Meurs KP, Crawford MM, Higgins RD, Poindexter BB, Colaizy TT, Hamrick SEG, Chalak LF, Ohls RK, Hartley-McAndrew ME, Dysart K, D'Angio CT, Guillet R, Kicklighter SD, Carlo WA, Sokol GM, DeMauro SB, Hibbs AM, Cotten CM, Merhar SL, Bapat RV, Harmon HM, Sewell E, Winter S, Natarajan G, Mosquera R, Hintz SR, Maitre NL, Benninger KL, Peralta-Carcelen M, Hines AC, Duncan AF, Wilson-Costello DE, Trembath A, Malcolm WF, Walsh MC; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network.JAMA Pediatr. 2025 Feb 24:e246613. doi: 10.1001/jamapediatrics.2024.6613. Online ahead of print.PMID: 39992674
Valentine GC, Antony KM, Sangi-Haghpeykar H, Wood AC, Chirwa R, Petro S, Dumba M, Nanthuru D, Shope C, Mlotha-Namarika J, Wilkinson J, Aagaard J, Aagaard EJ, Seferovic MD, Levison J, Kazembe P, Aagaard KM.Med. 2024 Nov 13:100539. doi: 10.1016/j.medj.2024.10.016. Online ahead of print.PMID: 39541971 Free article.
----
The transcript of today's episode can be found below 👇
Ben Courchia, MD (00:00.61)
Hello everybody, welcome back to the incubator podcast. We are back this Sunday for an episode of journal club. Daphna, how are you?
Daphna Yasova Barbeau, MD (00:06.585)
I'm doing well. I know you have some interesting papers planned, so I'm glad to get into it today.
Ben Courchia, MD (00:13.646)
Yeah, I'm glad to get into it as well. We're like a real podcast these days. I have printed notes of what I need to talk about.
Daphna Yasova Barbeau, MD (00:26.556)
If only we had a producer and they were printing our notes for us, you know? That's the next step.
Ben Courchia, MD (00:31.65)
None of that! We are definitely not there yet, but one step at a time. Rome wasn't built in a day.
So the one thing I wanted to talk about is a very interesting project that one of our good friends of the show, Dr. Guilherme Sant'Anna made us aware of, the Smart Hospital Project. The Smart Hospital project is based at the Montreal Children's Hospital and it's transforming neonatal and pediatric care through cutting edge wireless technology, smart dashboards, and machine learning tools, all the while maintaining a family-centered approach. This initiative is a collaboration between medical professionals, biomedical engineers, and industry leaders led by Dr. Sant'Anna, Dr. Wissam Shalish, and Dr. Robert Kearney of McGill University. Their mission is to develop and rigorously test smart wireless technologies that have the potential to become the new standard of care in hospitals worldwide, which I think personally is really the future. Now their latest milestone, and they've asked us for some help on that, is that their project was selected to host a session at the International Conference on Clinical Neonatology this September in Turin, Italy. This prestigious opportunity allows the team to showcase their groundbreaking innovation to the global neonatal and critical care community. This session will feature several presentations on advancing wireless technologies led by Dr. Shalish, Dr. Sant'Anna, alongside with key collaborators, like Professor John Rogers, who is a pioneering expert in biomedical engineering from Northwestern University. Students from the Smart Hospital Project and Northwestern University will take center stage as well, presenting their research and leading an engaging poster session for over 300 expected attendees. So congratulations to the team there.
Daphna Yasova Barbeau, MD (02:27.324)
I have a conference shout out also. We didn't think we'd be able to air this before the conference, but we are. There's a conference April 9-10th, Monroe Carroll Junior Children's Hospital, if you're near Nashville, Tennessee. I like to highlight this one because it is a free event. I think that's really cool for trainees and early career neos about building a career in quality improvement and clinical leadership. You can look them up for this free conference on quality improvement, if you're close enough to get there in the next week.
Ben Courchia, MD (03:10.435)
Then the last piece of news that we have to cover is another upcoming conference that our good friend, Jen Canvasser is leading. This is the NEC Symposium. It will take place this September in Chicago. We'll have more coverage before the conference on the podcast. But if you are interested in attending, I thought that it would be a good opportunity to give our listeners a unique chance to secure lower rates for registration. So you can get a 10% promo code using NECINC25. So “NEC Inc. 25”, and you'll get 10% off your registration. So thank you to Jen and to the team at the NEC Society for hooking up the Incubator community with a discount code. These are always nice to have.
All right, buddy. Lots of papers to go through. I'll get started, if that's OK. Daphna is on service, and she is doing us the favor of recording this after a long day of work. So thank you, Daphna.
So the first paper I'm going to talk about is a study that was published recently in JAMA Network Open. it's called “Response to Inhaled Nitric Oxide and Mortality Among Very Preterm Neonates With Pulmonary Hypertension.” The first author is Michelle Baczynski, who is a respiratory therapist. It's cool that a respiratory therapist is the lead author of this type of publication. There's a lot of big names in this study and it comes to us from the Canadian Neonatal Network. It's a multi-center prospective observational cohort study that attempts to answer a longstanding question in neonatology which is, is there an identifiable benefit to inhaled nitric oxide among very preterm infants with acute pulmonary hypertension?
As we know, the use of iNO, inhaled nitric oxide, in preterm neonates has been a topic of debate for years. that's really due to the fact that there's inconsistent evidence from randomized clinical trials about this particular use. Despite this, in current clinical practice, iNO is predominantly used in the form of a rescue therapy in preterm neonates and in a population in which we estimate that they're suffering from pulmonary hypertension. Now, the authors point out that approximately 5% of neonates born before the age of 34 weeks gestation do receive iNO in the NICU. The rate of use is actually inversely related to gestational age. Now, sadly, the mortality among these infants remains extremely high, with estimates around 40%. There's recent registry data that suggests that oxygenation improvement patterns in response to iNO are similar in both preterm and term neonates when iNO is administered for pH-related hypoxic respiratory failure during the first week of life. this raises the question, can responsiveness to iNO predict survival? Like if you're one of these responders, does that mean you're going to survive? The second question they're asking is, are there any other factors that are also associated with mortality?
The objective of the study is to investigate the association between iNO responsiveness and other clinical factors with mortality in very preterm neonates who receive iNO specifically for acute pulmonary hypertension-associated hypoxic respiratory failure.
This is a multicenter prospective observational cohort study that was conducted between January 2018 and December 2022 in 12 tertiary NICUs in Canada. All very preterm neonates that were born before 32 weeks of gestation who received iNO for acute pulmonary hypertension or acute pH during the study were considered to be eligible. To be included in the study, you had to have a diagnosis of acute pulmonary hypertension assessed by the attending and hypoxic respiratory failure, which they defined as an FiO2 of 50% or greater before the initiation of iNO. They had some exclusion criteria involving congenital cardiac anomalies, known genetic anomalies, congenital diaphragmatic hernia, and/or chronic pulmonary hypertension.
Now they did something interesting, which is that they distinguished between early acute pH and late acute pH. They separated these entities into two different pathologies, almost like the pattern that we use for sepsis, where we have early onset and late onset. In this case, early acute pulmonary hypertension was defined as the use of iNO in the first three days of life, and late acute pulmonary hypertension was its use after three days of life. They also were able to use the patients double, because if you had pulmonary hypertension the first 72 hours of life, you could be included, and then you could be also included in the late acute pH cohort as well. They had some diagnostic criteria for the echocardiography-confirmed diagnosis of pulmonary hypertension. You can review them. They're quite thorough. the primary outcome of the study was pre-discharge mortality. Among the survivors, they had secondary outcomes that included chronic lung disease, stage 2A NEC or higher, grade 3 IVH or higher, and ROP requiring treatment.
Let's look at some of the results. A total of 371 neonates were included. 262 were in the early acute pulmonary hypertension group. 109 were in the late acute pulmonary hypertension group. In terms of baseline characteristics, the babies in the early pH group had a mean gestational age of 26.3 weeks. They were started on the iNO at about a day of life. 90% were on high frequency ventilation. About 68% had a pre-iNO FiO2 of 100%. Of these infants in that group, 72% had their pulmonary hypertension diagnosed via echocardiography. When we're talking about late acute pulmonary hypertension, these infants actually had a lower gestational age at birth. Their mean gestational age was 24.9 weeks. The median age at which iNO was started was 13 days. 82% were on high frequency [ventilation]. Only 47% were on 100% FiO2 before the initiation of the medication, as compared to 68% in the early acute phase. 63% had their pH diagnosed with echocardiography. In both groups, iNO was the primary pulmonary vasodilator that was used in 97% of the cases. I think that's important because obviously there's other ways to address pulmonary hypertension. In terms of mortality, the mortality rates were quite high. They were lower in the early acute pulmonary hypertension group with 34% mortality rate and 49% in the late acute pulmonary hypertension.
The question really is, what are some of the predictors of mortality? When we're looking at the predictors of mortality for the early pulmonary hypertension group, we note that the initial analysis identifies a lower gestational age at birth, a lower birth weight, less antenatal corticosteroid use, a higher incidence of prolonged rupture of membrane, a lower five-minute apgar, and greater use of vasoactive drugs before the initiation of iNO.
Survivors on the other hand had a larger post-iNO change in their FiO2, which I think is very interesting, and they had more frequent use of echocardiography. They did a multivariate analysis, and after running the multivariate analysis, some of these factors actually did not reach statistical significance anymore. so what we were left with is that the independent predictors for survival included a higher gestational age at birth (so the more mature the better), a greater reduction in FiO2 after iNO (so that response in the early acute pulmonary hypertension does make a big difference on whether survival will be in the cards for these patients), and the more frequent use of echocardiography.
Higher pre-iNO vasopressor use was associated with lower survival. So if you were on a bunch of pressors, that was not good. They did look at the response in FiO2. They found that the optimal threshold for FiO2 change to discriminate survival was 16%. So if you were like about 20% change in FiO2, that had a sensitivity of 87% and a specificity of 49%. So looking at the other group, the late pulmonary hypertension group, they found that the demographic factors were mostly similar between survivors and non-survivors. Non-survivors had higher severity illness, greater prevalence of late onset sepsis, a higher pre-iNO FiO2, mean airway pressure, vasoactive drug use, and lower pre-iNO blood pH. Unlike the early group, there were no significant differences in the change in FiO2. So the response in iNO in the form of reduced FiO2 was no longer seen as a factor that potentially impacted survival in the late group. The multivariate analysis shows that of all these different parameters, the only two that really matter are a lower pre-iNO FiO2 and a higher pre-iNO pH. So these two are really the factors that are associated with survival.
A few more interesting analyses: they looked at responders versus the non-responders, and the Kaplan-Meier analysis showed lower mortality among iNO responders; the difference was statistically significant only in the early pH group, as we just said. Among the survivors, prematurity-related morbidities were common in both groups. In the responders to iNO in the early pulmonary hypertension group, the rates of BPD were actually lower in the non-responders. Everything else didn't really show much of a difference.
Across the entire cohort, the strongest predictor of positive iNO response was the fact that you had pulmonary hypertension early on and that you had a greater post-menstrual age at treatment. Importantly, among neonates weighing less than 1,000 grams or less than 750 grams, iNO response remained associated with lower mortality without any increased risk of severe IVH.
So it's a very interesting paper. The discussion is quite interesting. They mentioned how the findings support the nuanced clinical impression that iNO may provide benefit in select very preterm neonates, particularly those with early onset acute pulmonary hypertension. They mention that there's a lot of dissonance in terms of what the use of iNO should be. The Pediatric Pulmonary Hypertension Network currently recommends iNO for preterm neonates with confirmed acute pulmonary hypertension physiology, citing that it has established safety, there's lack of alternatives, and there's rapid resolution of symptoms. But the American Academy of Pediatrics surveyed about 300 neonatologists, which showed that only 36% supported iNO use for acute pulmonary hypertension, 22% were opposed, and 42% said that this had to be a shared decision with the parents.
So to wrap up, this is a large multi-center prospective cohort study that found that very preterm neonates with acute pulmonary hypertension associated hypoxic respiratory failure during post-natal transition frequently showed a positive response to iNO, and that this iNO responsiveness was associated with improved survival independent of pretreatment illness severity. On the other hand, among neonates who developed acute pH later in the NICU course, oxygenation improvement with iNO was modest and mortality was more strongly associated with baseline illness severity, rather than with the physiologic response to iNO. Obviously, they're mentioning that this needs to be researched a bit further.
I think that they're right on the money when they discriminate between early and late pulmonary hypertension, because as they say, while they are the same symptom, they are very different from a pathophysiologic standpoint, where early really may have to do with transition to extrauterine life, versus late, where this might be a completely different process altogether. So a very interesting study, and congratulations to the authors.
Daphna Yasova Barbeau, MD (16:11.546)
I think separating out the two types was critical. We're a unit that uses a lot of nitric, we're willing to give it a try pretty quick, and if there's not any improvement, to turn it back off. But we like to see who's a responder and who's not a responder. I think this is neat data. It makes sense, right? If you're a responder and you get better, you can make it through that period. if you're not a responder and you can't make it, this makes sense. It's so true about so many medications that we use - we can't use the same thing for everybody. We don't know who's going to respond well and who's not going to respond well (yet). I hope more and more people will study it. I think for the perivariable babies, the data never showed improvement for nitric. So I think some people won't try it, when in the N of one trials, as they say, the individual babies, for some it really seems to help.
Ben Courchia, MD (17:26.178)
It's very interesting that sometimes we may give up and say, my God, this gets so sick, is it even worth it as a Hail Mary? But remember that baseline illness severity, aside from the use of pressers, was not as much of a factor in the early pH compared to the late pH. So something maybe to think about.
Daphna Yasova Barbeau, MD (17:43.356)
Interesting. You got another one for us? I know you have a whole stack.
Ben Courchia, MD (17:47.458)
I have a whole stack, I can keep going. My next paper is called, “Evaluating Decision Regret After Extremely Preterm Birth,” and it's published in the Archives of Diseases in Childhood, Fetal and Neonatal Edition by Belden and colleagues. This was a fascinating study that I really enjoyed reading. It dives into the often overlooked and complex concept of decision regret, specifically in the context of parents, more precisely mothers, who delivered extremely premature infants between 22 and 25 weeks of gestation. The context here is critical. While survival rates for extremely preterm infants have improved, the neurodevelopmental outcomes and risk of chronic health conditions remain beyond significant. So despite efforts to support shared decision-making through consultation guidelines, decision aids, and social-cultural sensitivity, institutional practices do remain very variable. The long term impact of these decisions, particularly the emotional aftermath in the form of regret remains underexplored. The literature is sparse. Most prior reports on decision regret regarding children do focus on other areas of pediatrics like oncology, surgery, but not neonatology.
The objective of the study was twofold. Number one, to measure decision regret in women who delivered an extremely preterm infant at 22 to 25 weeks of gestation. Number two, to compare decision regret scores based on the resuscitation decision, whether active care, comfort care, or another path was chosen. This was a survey-based study conducted at two unaffiliated high-risk obstetric neonatology hospitals, one of them in Delaware and the other one was in Oregon. Both institutions had longstanding perivariability consultation guidelines based on gestational age. At “hospital one” - I'm not sure which one of the two is which by the way, but if you work at any of these places, maybe you can let us know - at hospital one, comfort care was advised for any patient that delivered before 22 weeks. There was individualized parent-guided care between 22 and 24+6. then there was active neonatal care at 25 weeks and above. At hospital number two, it was slightly different. They offered comfort care for anything less than 23 weeks, individualized between 23 and 25+6. then they offered active care for anybody above 26 weeks.
They identified the deliveries that took place between January 2004 and December 2019. They had some exclusion criteria about life-limiting diagnoses, parents who did not speak English (they justify that exclusion based on the limited research support so that's understandable), or if there was a maternal death, unfortunately. They were contacted via mailed letter and they were invited to complete the survey, which included the decision regret scale. It's a validated five item instrument where you can interpret the score of zero (meaning you have no regret), 1 to 25 (mild regret), and 25 and above (moderate to strong regret). The secondary outcomes were the perinatal consultation recall, the decision-making process characteristics, the importance of influencing factors such as faith, survival chances, and infant outcome.
Out of 787 eligible women, 56% (442) were successfully contacted. 242 mothers completed the survey. There were no significant differences in gestational age at delivery or the infant survival between the people who responded to the survey versus the one who did not respond. So there's no attrition bias there. However, they do note that the mothers who did respond were more likely to be of white race. So in terms of maternal and infant characteristics, of the 242 respondents, 84% indicated that they made a decision before delivery: 68% of those chose active care, 21% chose comfort care, 11% chose other (this was a wait and see approach, maybe a trial of life, trial of support, or a limited care pathway). 16%, and that's staggering, reported no option given/no decision made, and were excluded from further analysis. Women who chose comfort care had higher education levels and delivered at earlier gestational ages. A higher proportion in the active care group reported being the primary decision makers. Notably, 12 infants in the comfort care group survived the delivery room. Five of them died in a NICU and seven survived to discharge.
So let's look at the primary outcome: did they have decision regret? Decision regret scores were elevated across all groups, but it was lower in the group that decided to proceed with active care. Active care group had a decision regret score of 14.5, compared to 39.1 in the comfort care group and 37.4 in the other category. Multi-variable analysis was done and it confirmed the significantly higher regret in the comfort group and other group, even after adjusting for potential co-founders.
Some additional findings were that prenatal consultation was associated with lower regret, so please, do your prenatal consults. Do a good job because they matter. Those who self-identified as the primary decision maker had lower regrets as compared with when the doctor had a major role or made the decision outright. Let's just think about that for a second. Infant survival at the time of the survey was linked with lower regret compared with delivery room or NICU death, or death after discharge. We're comparing survival to death, not survival to survival without neurological impairment or disability, so just keep that in mind.
In terms of secondary outcomes, among reported factors influencing decision-making, hope was mentioned in 84% and babies’ chances of survival in 82%, as factors that influenced their decision. Having a support system, the chance of problems later in life, the doctor's advice, and faith/religion were also considered important. I was very interested to see that financial worries were of low importance for 56% of respondents. Kind of an impressive number if you ask me, considering this was done in the US where everything is outrageously expensive from a medical standpoint. Interestingly, hope was more important among those who chose active care, while chance of problems later in life was rated higher among the people who chose comfort care and other groups, which is not really surprising.
I think these are all very interesting findings. This is a study that is the largest of its kind that reveals that decision regret after extremely preterm birth is quite common, even years later. It really varies based on the decision that was made, the extent of shared decision-making, and the outcome of the infant. Some of the key findings that we can highlight again are that all groups reported an elevated score, that women who received a prenatal consultation, who were the primary decision-maker, who chose active care, and/or who had a surviving infant had lower regrets. The people who chose comfort or other groups had a higher regret score. They're saying that one possible explanation is that comfort care inherently leads to infant death, which even if accepted, may lead to long-term emotional consequences. Alternatively, some women may have preferred active care, but chose differently due to limited options or perceived lack of autonomy. Despite elevated regret, most women described counseling as honest, understandable, and tailored, which suggests that regret is not necessarily a criticism of the medical communication nor of the comfort care pathway itself, rather the authors are saying it may reflect the emotional weight of the decision, the trauma of preterm birth, or simply the passage of time. Notably, even women whose infants survive without disability sometimes expressed high levels of regret and vice versa. The most extreme regret scores, up to a hundred, came from women whose infants had received active care and survived. That is something to really keep in mind that underscores again that decision regret may not reflect objective outcome. The study did not find significant effects from maternal age, education, religion, marital status, or mental health. They also had free text responses, but they said that they were too limited for thematic analysis.
The study reinforces the importance of high quality prenatal consultation, the importance of shared decision-making where pregnant women are affirmed as the primary decision makers. The key takeaways are that we should ensure that all eligible women receive prenatal consults. We reaffirm their autonomy and that we offer individualized compassionate discussion around all care options, including comfort care. Now, as the limits of viability shift and our understanding of outcomes evolve, ongoing research and larger more diverse studies will definitely be needed, as the author suggests.
Daphna Yasova Barbeau, MD (28:14.044)
On face value, [this one is] not so unexpected, but it's quite interesting if you really get down to the details, which you did such a nice job doing, that everybody carries regret. It's hard to say which family should we target with interventions for their guilt or regret (the answer is all of them). I commend the centers where this was done because obviously they're doing a really good job with the prenatal counseling rate. For those scores to be so high, I think there has been a change in the community probably in the last 5-10 years where we're really taking that seriously and trying to meet parents where they are and connect with them in shared decision-making. I think this information about paternalism was really quite striking. I think we've definitely moved away from paternalism and it shows that I guess that's what families want and it helps them with their decisions when they are the decision-maker. I think that's what I'm taking away from this.
Ben Courchia, MD (29:41.87)
Awesome. Are you going to make me keep going or are you going to present?
Daphna Yasova Barbeau, MD (29:47.118)
Okay, I'll do one. It's my turn, I suppose. I was so enjoying listening to you talk. You haven't used your inhaler yet, so I feel like we're okay! I am gonna do “Short Duration of Antenatal Corticosteroid Exposure and Outcomes in Extremely Preterm Infants” from JAMA Network Open Pediatrics. Lead author is Sanjay Chawla, senior author is Wally Carlo, and this is coming from the NICHD Neonatal Research Network. The question really is, what is the minimum duration of exposure to a single dose of antenatal steroids associated with improved neonatal survival and reduced morbidities?
We know antenatal steroids in general, good. We know the guidelines. There are a few different ways to give steroids, but in general, the two dose, 24 hours apart, regimen is probably the most common here in the States. They looked at infants born at 22+0 to 27+6 weeks of gestation between January 2016 and February 2021at 15 US academic centers. They excluded babies who were outborn, had major congenital anomalies, had a birth weight more than 1,500 grams, were exposed to antenatal dexamethasone instead of or in addition to betamethasone, or who received more than one dose of antenatal steroids because again, the goal of the study was to look at hours from that first dose where the baby doesn't make it long enough to get subsequent doses. They also excluded infants who did not receive active treatment (endotracheal intubation, surfactant, CPAP, bag and mask ventilation or mechanical ventilation, chest compressions, epinephrine, or parental nutrition). So they were looking at babies who got the full critical care package, and looking at their outcomes.
They wanted to look at a number of morbidities and outcomes. They were looking for severe intracranial hemorrhage defined as grade three or four. They looked at advanced resuscitation in the delivery room, including endotracheal intubation, chest compressions, or epinephrine. They looked at NEC Bell stage 2A or greater. They defined severe BPD as invasive mechanical ventilation at 36 weeks post-menstrual age. They looked at severe retinopathy of prematurity, so ROP requiring medical or surgical treatment, which included laser, cryo, or vitrectomy.
The primary outcome though was survival at discharge. They had secondary outcomes of survival without major neonatal morbidities, which included any of the following: severe intracranial hemorrhage, cystic PVL, surgical NEC, severe BPD, and severe ROP.
Initially what they did is take this time interval from antenatal steroid administration. They used one hour increments in time from that first steroid dose. They looked at adjusted relative risks per one hour increase. Then they took the total number of babies, split them into four equal groups by time (not like 12 hours divided by four, but all the babies divided by four and they fell into whichever quartile that they fell into by time). They have this beautiful chart (figure two) that looks at association of interval from antenatal steroid administration to birth interval with survival at hospital discharge. So this was just the time between antenatal steroid and birth and looking at this scatter plot smoothing curve. It's really beautiful. It's quite striking how this upward linear graph shows improvement in survival to hospital discharge by hour from that first antenatal dose.
I'll get into the details of the results. They had 7,400 infants born during the study period. 2,000 infants were eligible and 1,806 infants were included in the cohort. There were 470 infants in the no antenatal steroid group. They compared the quartiles to the no antenatal steroid group. They had 1,331 infants with exposure to a single dose of betamethasone within 24 hours before birth. The median time from antenatal steroid administration to birth for neonates who only got one dose was actually 3.8 hours with a range of 1.4 to 9.5 hours in their cohort. Among the 475 infants with no antenatal steroid exposure, 65% survived. In the group born after exposure to a single dose of antenatal steroids, 74% survived. I already told you they saw this incremental increase. The actual numerical data showed that among infants born after exposure to a single dose of antenatal steroid, each one hour increase in administration to birth interval was associated with a 1% higher rate of survival to hospital discharge and a 1% higher rate of survival without major morbidity. They also saw continuous reductions of adverse outcomes observed for each one hour increment of betamethasone exposure. So they saw improvements in severe intracranial hemorrhage, severe intracranial hemorrhage or death, cystic PVL, cystic PVL or death, grade 3 BPD or death, NEC or death, surgical NEC or death, and severe ROP or death.
They also had a post hoc analysis controlling for chorioamnionitis, in addition to the previous factors. Among infants born after exposure to a single dose of antenatal steroids, each six hour increase in the interval from administration to birth was associated with a 4% higher rate of survival to hospital discharge and a 9% higher rate of survival without major morbidity. So every one hour counted, but certainly getting to increments of six hours counted even more.
They looked at quartiles of babies compared to no antenatal steroids. The reference group was no steroid exposure, quartile one is 1.4 hours from antenatal exposure, quartile two is 1.5 to 3.8 hours, quartile three is 3.9 to 9.5 hours, and quartile four is more than 9.5 hours. What they saw is that there were differences for time-dependent increases in survival without major morbidities and reductions in severe intracranial hemorrhage, severe intracranial (39:50) hemorrhage or death, cystic PVL or death, grade three BPD or death, NEC or death, and severe ROP or death. Again, that's when they compared them to no antenatal steroids. You can look at these tables and really see the increases by the quartiles for almost all of these outcomes.
The take-home point is obviously that for individuals at risk of imminent preterm birth, even a few hours of exposure to a single dose of antenatal betamethasone has beneficial associations, and this benefit increases with greater duration of exposure. We can take this back to our teams and say as soon as we identify these moms on labor and delivery, every hour matters. So how can we get steroids in faster?
Ben Courchia, MD (40:23.136)
It's becoming clear with antenatal steroids that you should try to avoid giving them if you don't think that there's going to be a delivery.
Daphna Yasova Barbeau, MD (40:34.31)
Which is tricky for our obstetric colleagues.
Ben Courchia, MD (40:52.718)
The OBs have to figure out whether the delivery is imminent or not. That's not our specialty. They will need to have a better granular understanding of the actual risk to make that decision, because it's sometimes a bit subjective and that can make it tough for them to decide. Very interesting.
Daphna Yasova Barbeau, MD (41:06.117)
Your turn.
Ben Courchia, MD (41:07.768)
For sure. Do you have another paper? Because we're cruising. It's almost like 40 minutes. I just want to make sure that we cover the topics that we have to cover. Do you have another paper in the bag for today?
Daphna Yasova Barbeau, MD (41:25.506)
I do, but you go for it. I know you had some you really wanted to share, so go for it.
Ben Courchia, MD (41:28.984)
Are you going to be reviewing that cooling paper later on? Yes? Okay, great. So then let me just talk a little bit about some articles that came through my desk. To be honest with you, I don't think it would be valuable or compelling for the audience to have a thorough, thorough review of the paper. But I just want to bring some attention to them.
The first one is one that you will definitely enjoy because I saw this article and I was like, this was written for me. But I also realized that probably it's been written for every neonatologist. It's in Nature Medicine, called “AI-guided precision parenteral nutrition for neonatal intensive care units.”
Daphna Yasova Barbeau, MD (42:17.021)
That sounds delightful.
Ben Courchia, MD (42:27.392)
The article that highlights the bane of our existence: neonatologists [trying to] order TPN. It's very variable, and they often have to rely on their pharmacist to make some changes. There's lack of consistency and sometimes things get missed. So [the authors ask], what if we could make this AI model that could actually order TPN for you and follow based on a pretty large training data set, how to do TPN? They created this thing called TPN 2.0. it maps patients to a fixed library of formulas. It's enhancing reproducibility, allowing for potential mass production of pre-compounded TPN bags. They also designed some physics, like transformer models. It's very interesting because it has a pretty large data set. The figures of the paper are quite good. They did some validation you can take a look at, where they showed it to blinded clinicians and they asked them to rate the quality of the TPNs. The ones that were made through this AI algorithm were ranked higher than the ones made by physicians. They had a random TPN as a control. The score that the AI one got was 56 and the one from the actual physician was 35. They also are showing that this really can improve variability, and it has a high correlation with expert prescription. They do some of that analysis trying to tell you how using their model could reduce mortality, cholestasis, NEC, sepsis. I have to say, the paper is quite daunting to read because I don't have the knowledge to understand every intricacy, but I don't know how that calculation was done. So I would be careful about how they actually do that, because these are models that they did some projections. They did not really test it on babies and actually measure that. I thought that was quite interesting and maybe potentially a solution to a very vexing problem for a lot of us in the NICU…you don't seem sold. I was expecting giggles and fireworks.
Daphna Yasova Barbeau, MD (44:49.318)
Well, I was just thinking that this makes me feel like all of my tinkering doesn't really matter!
Ben Courchia, MD (45:01.262)
That's not what they're saying. I think that the TPN does do the tinkering based on EMR data and make some adjustments. But it also suggests that people will forget to turn down the zinc, or they will miss a parameter and adjust that accordingly.
Daphna Yasova Barbeau, MD (45:25.594)
But it sounds like it might do a better job than we do.
Ben Courchia, MD (45:28.748)
That's what they're selling it as. I think if it could do as good of a job, it would have been great. Some of the validation stuff has to be reviewed, but I thought that was interesting. It's published in a very reputable journal. So I thought I would just mention it.
The other paper I wanted to mention is this paper that's published in this journal that I don't frequently [read], it's called Med. It's an open access paper, and it's called “A cluster randomized trial of xylitol chewing gum for prevention of preterm birth: The PPaX trial.” This was very interesting because it looked at the rates of preterm birth in Malawi and looked at if chewing gum to try to improve periodontal disease could reduce preterm birth. I'm going to read you a little bit of the abstract: “while periodontal disease in pregnancy is associated with the delivery of preterm and low birth weight newborns, trial of routine treatment strategies have not shown significant benefits here in this study. The investigators evaluated the use of xylitol containing chewing gum among 10,069 pregnant participants in an eight-site cluster randomized trial in Malawi, a country with one of the highest preterm birth rates.” We actually had on the podcast the chief pediatrician of Malawi. So check that out if you're interested to learn more about pediatrics and neonatology over there. Now, what it's showing is that, compared to an active control group receiving education and Malawi Ministry of Health standard prenatal care, early pregnancy initiation and ongoing use of the xylitol containing chewing gum twice daily reduced the occurrence of preterm and low birth weight deliveries. In order to prevent one such occurrence, fewer than 26 pregnancies needed xylitol chewing gum use, making it efficacious and affordable in her low resource settings. I thought that it was quite interesting that they're tackling preterm birth through using chewing gums, and how that's actually having a significant effect is quite interesting. It's a relevant thing for countries that have issues with high mortality rates, high preterm birth, and difficulty accessing care.
Daphna Yasova Barbeau, MD (47:50.052)
I'd love to see if we paired it with what do we currently do to prevent preterm birth. I'd be happy to put myself out of a job. I mean, there are babies other than preterm babies in the NICU, but it'd be lovely to see something this simple [work].
Ben Courchia, MD (48:08.994)
When you speak to dentists, they talk about the importance of oral health and how the oral health [affects-]
Daphna Yasova Barbeau, MD (48:18.358)
[-affects] all of your physical health. It makes sense.
Ben Courchia, MD (48:22.84)
I'm going to do one more paper, and then I can let you finish. The last paper I'll review today was published in Pediatric Research, and it's called “Neonatal Intermittent Hypoxemia Events Are Associated with Later Systemic Hypertension.” I thought that was a very catchy title. It's a very interesting study. So neonatal hypertension is a relatively understudied issue. This paper explores actually a potentially modifiable risk factor being intermittent hypoxemia and how it could provide new insight into how systemic hypertension develops in preterm babies.
We know from previous work that about 5% of preterm infants born before 30 weeks of gestation develop systemic hypertension. The etiology remains unclear. In about 50% of these cases, commonly what's implicated as a risk factor is an umbilical catheter, a history of acute kidney injury, or other comorbidities like bronchopulmonary dysplasia. Still, there's a large number of cases that are labeled as idiopathic or unexplained hypertension. Intermittent hypoxemia is extremely common in the preterm population, especially between the second and the fourth weeks of life with evidence linking them to adverse outcomes such as ROP, BPD, poor neurodevelopmental outcomes, prolonged hospitalization, and so on. Little is known about their association with hypertension in this particular population.
In adult humans and in animal models, there is a strong growing body of evidence that shows that intermittent hypoxia can contribute to systemic hypertension, through mechanisms like oxidative stress, inflammation, reactive oxygen species, sympathetic activation, hypoxia-inducible factors, and so on. The hypothesis of this study is that maybe these factors are at play in preterm infants and that early intermittent hypoxia events, specifically within the first three weeks of life, are associated with elevated systolic pressures at 34 to 36 weeks post-menstrual age. So a very interesting question. Let's dive in.
This is a secondary analysis of the large Pre-Vent study (Prematurity-Related Ventilatory Control study) that was conducted at Case Western Reserve University. They included from that study babies who were born before 31 weeks of gestation, who were less than seven days old on admission to the NICU at Rainbow Babies, that survived to 36 weeks, and that had intermittent hypoxia data recording during the first three weeks of life. They had some exclusion criteria so from 175 enrolled infants, 164 were included in the final study. The data on intermittent hypoxemia is collected using the continuous pulse ox, and they're averaging them over eight seconds. The intermittent hypoxia events were defined as having an O2 sat of less than 90% for 10 seconds to 5 minutes, which is scary. The data analysis focused on day 8 to day 21, which according to the authors, is a period that is known to be associated with a linear increase in frequency of intermittent hypoxia. They looked at the frequency of these events, the looked at the duration of time spent below 90%, and the mean duration of the individual events. Now, there were three main other parameters that were measured. So we have blood pressure data. The one that they're interested in is the one collected between 34 and 36 weeks of gestation. They did at least one measurement per day. They were classified as having hypertension if their median systolic or mean arterial pressure was above the 95th percentile (they have a reference with a table that shows them what that looks like), or if at least eight out of 15 daily BP measurements were above the 95th percentile. The other data that they collected was the presence of a UAC, the serum creatinine, some renal ultrasound finding, the history of a PDA, etc.
I know we're running short on time, so I'm going to go over the results. 664 infants were analyzed. 23% met the criteria for systemic hypertension. Despite that, only 13% had clinically recognized hypertension in the documentation…I'm not laughing at the clinicians, because I think we're all in that bucket, but man, that's a big drop off!
They didn't find any significant difference in antenatal or early hospital characteristics between the hypertensive and the normotensive group. [They did find] that intermittent hypoxia events during day eight to 14 of life were significantly associated with hypertension at 34 to 36 weeks post-menstrual age. When they looked at intermittent hypoxemia beyond that window of 8 to 14 days, specifically day 15 to 21, they no longer saw that association with systemic hypertension down the road. The number of intermittent hypoxia events during the week, during day 8 to 14, was significantly associated with an elevated systolic BP, measured as a continuous variable. The percent of time that babies spent below 90% during day 8 to 14 was also associated with an elevated systolic BP. The mean duration of the events however was not associated with either hypertension or an elevated systolic blood pressure. Now, they're saying that none of the traditional hypertension risk factors, such as SGA, having a UAC, a PDA, or AKI, were associated with hypertension in this cohort. Do with that what you wish.
So these are the key findings. Just to recap, the study finds a 23% incidence of hypertension, much higher than the previously reported 5-10%. I think that this has to do with an issue of definition. Now, again, the critical finding is that intermittent hypoxia during the second week of life, and not later, was seen to be associated with both binary hypertension classification and a continuous elevation in systolic blood pressure. It raises the possibility, the study authors are suggesting, of a critical window where intermittent hypoxia may initiate vascular remodeling, maybe sympathetic activation leading to hypertension. This would be supported by animal data that is demonstrating changes in peripheral baro- and chemoreceptor function, organ remodeling, reactive oxygen speciation generation, inflammation, and endothelial dysfunction. Despite the high prevalence of AKI in that cohort that was close to 50%, it was not associated with hypertension, mirroring the finding that intermittent hypoxia may be an independent and under-recognized contributor to neonatal hypertension. Another important observation, again, was that only 3% of the infants had both clinically recognized hypertension and hypertension by normative values, highlighting the potential under recognition of hypertension in the NICU settings.
So to sum up, intermittent hypoxia and time spent below 90% during the second week of life is a very important deal and can translate to potentially high systemic blood pressures at 34 or 36 weeks. Later events and median duration of these events are not associated with these problems. It supports this hypothesis of having a critical window of vulnerability that exists, having potentially long-term hemodynamic consequences. This does not answer the ultimate question of the study, but does need to open the door for more studies needed in the future.
Daphna Yasova Barbeau, MD (57:04.38)
Well, I'm not sure what to say. I mean, I think a lot of people are like, oh, the baby desats but spends most of the time and we're in our happy range, but we really can't ignore them anymore.
Ben Courchia, MD (57:14.766)
While it is not a good idea to let therapies trickle down from the adult world to the neonatology world or from older children, and [babies] have to be studied, we still can learn from the pathophysiologic mechanisms that we know from adults. Clearly, if you talk to adult cardiologists and adult pulmonologists, having chronic hypoxemia or intermittent hypoxemia does raise your blood pressure. It is basic adult cardiovascular health. I don't know if we can truly ignore that. We had a discussion in our unit recently where it was brought up that when somebody [undergoing chemotherapy] was told that they should not ingest caffeine and iron together, because it could reduce the absorption of iron. We've never looked at that as a field. How often are our babies getting caffeine and iron close together? I think these pathophysiologic mechanisms where they've been well-established in adults or older children, maybe we should take a look at. I thought this study was very interesting for that specific purpose.
Daphna Yasova Barbeau, MD (58:29.39)
Yeah, and I think even more so it's underscored that this brand new, still-developing kidney, is taking an even greater hit in that first week that we have to be careful about.
Ben Courchia, MD (58:45.76)
Even though they showed that AKI did not really correlate with their pulmonary hypertension, remember they're only measuring it between 34 and 36 weeks. It's been pretty well established that AKI and prematurity lead to pulmonary hypertension, systemic hypertension down the road. Maybe too early to detect it at that point based on that particular causality.
Daphna Yasova Barbeau, MD (59:13.03)
Very interesting. Well, no pressure saving this paper for last, but it's definitely making the rounds. I think it’s to your point about treatment creep or practice creep, I think that's what people are really talking about. This is a JAMA Pediatrics article, “Whole-Body Hypothermia for Neonatal Encephalopathy in Preterm Infants 33 to 35 Weeks' Gestation: A Randomized Clinical Trial.” Lead author is Roger Faix, senior author is Michelle Walsh. This is another NICHD neonatal research network study. We know that therapeutic hypothermia less than six hours after birth in babies with this hypoxic encephalopathy at 36 or more weeks gestation has been helpful in reducing death or disability. More and more people are moving the needle of cooling later, cooling earlier gestation, cooling smaller babies. So that's really what they were looking to study. They wanted to look at therapeutic hypothermia less than six hours after birth and how it affects death or disability in infants 33 to 35 weeks gestation with moderate or severe hypoxic ischemic encephalopathy.
This is a neonatal research network, so all infants admitted to NICUs of those centers at 33+0 to 35+6 weeks gestational age less than six hours of age with a diagnosis of encephalopathy, perinatal asphyxia, neurologic depression, or a similar condition were screened for eligibility. The inclusion criteria should really be familiar to everybody. This is the same inclusion criteria for other cooling studies, except for the gestational age criteria. They needed to have a blood pH of 7.0 or less, a base deficit greater than or equal to 16, an acute perinatal event with a 10 minute apgar score of five or less, ventilation initiated at birth and continued for 10 minutes or longer, moderate or severe encephalopathy using the Sarnat score, or clinical seizures at less than six hours. Exclusion criteria included core temperatures were less than 34 degrees for greater than one hour before screening (if they were cooled before randomization), receipt of paralytic or sedative agents that may complicate your Sarnat examination, encephalopathy that was unlikely to be due to hypoxic ischemic encephalopathy, a major anomaly, a baby who's not intended to get intensive care, a birth weight less than 1,500 grams.
Ben Courchia, MD (01:02:50.126)
That's a big deal.
Daphna Yasova Barbeau, MD (01:02:52.828)
I agree, I wanted to highlight that one. The last one was clinician-declined enrollment. Again, 1,500 grams was the cutoff that they used, regardless of the gestational age. That is moving the needle from the cooling trials.
Ben Courchia, MD (01:03:09.826)
You could have a very growth-restricted smaller infant that would not have met this trial [inclusion criteria].
Daphna Yasova Barbeau, MD (01:03:16.24)
Okay, so there's a lot of data here and I know we're running out of time, but I want to get into the primary outcomes. The primary outcome was death or disability, which included severe or moderate at 18 to 22 months corrected age. They used a post-discharge neurologic exam, the Bayley-III, the Gross Motor Function Classification System level and vision and hearing status.
Ben Courchia, MD (01:03:49.016)
Good for you. I would not have been able to say that. The GMFCS, I could never.
Daphna Yasova Barbeau, MD (01:03:55.44)
It needs a cute name. Severe disability was deemed by any of the following: a Bayley-III cognitive score less than 70, a GMFCS level 3-5, blindness or hearing loss with inability to hear commands despite amplification. A moderate disability was defined by a cognitive score of 70 to 84, and any of the GMFCS level 2, a treated seizure disorder, hearing loss requiring amplification, or an implant to understand commands. Those infants with cognitive scores greater than or equal to 85 and no deficits were considered “normal.” The pre-specified secondary outcomes included death alone, severe or moderate disability alone, death or profound disability, which was defined as severe disability with assignment of lowest possible cognitive score because the infant was untestable due to impairment, survival with a normal outcome, and each component of the severe and moderate disability and cause of death. They looked at a number of adverse events, which have been previously studied in the other cooling trials. For the sake of time, I'm not going to go to the individual adverse events. I’ll get to the meat of the study here.
I will highlight they're using a Bayesian approach to look at probability of treatment, benefit and harm. We won't totally describe that here, but I think it helps people to understand the statistics. From July, 2015 to September, 2020, 168 infants were randomized. There were 88 in the hypothermia cohort with a mean age of 34 weeks gestation, and 80 in the normothermia cohort with a mean age of 34+1 weeks gestation. There were no major differences between the demographic and clinical characteristics of the mother or the infant before randomization.
They do highlight these concerns about hyperthermia, so I'll tell you about that in a second. The mean temperature of the two groups, which obviously are expected to be different during the intervention period, hypothermia (33.4 degrees Celsius) versus normothermia (37.2 degrees Celsius). In that normothermia group, 16 infants met the threshold of greater than 37.5 degrees Celsius. So we know that also getting babies too warm or too cold is a risk factor for worse outcomes, regardless of cooling intervention. Randomization occurred at a mean of 4.5 hours for both groups.
The primary outcome of death or moderate to severe disability occurred in 29/83 infants or 35% in the hypothermia group, versus 20/69 or 29% in the normothermia group. The adjusted ratios using the neutral prior probability was 1.1. This yielded a probability of benefit of 26% and a harm of 74%. Death occurred in 20% of the hypothermia group and 12% of the normothermia group, with a probability of benefit 13% and harm 87%.
Ben Courchia, MD (01:08:05.038)
How impressive are those numbers?
Daphna Yasova Barbeau, MD (01:08:07.672)
Impressive numbers, for sure.
Ben Courchia, MD (01:08:11.534)
13% probability of benefit in the case of death alone and 87% probability of harm. That is crazy. Before you said 74% probability of harm for the primary outcome of death or moderate to severe disability, versus 26% probability of benefit. Crazy.
Daphna Yasova Barbeau, MD (01:08:25.574)
That's right. Then they stratified by gestational age. This is actually in the supplemental table, but I think it bears discussion. So I'm going to go over this. The 33+0 to 33+6 weeks, let's look at death or moderate to severe disability. It was 42% in the hypothermic group versus 35% in the normothermic group. For death, 16% in the hypothermic group and 21% in the normothermic group. In the 34+0 to 34+6 weeks for death or moderate to severe disability, 38% in the hypothermic group, 31% in the normothermic group. For death, 31% in the hypothermic group, 6% percent in the normothermic group. So this was the biggest difference of the group. In the 35+0 to 35+6 weeks, 25% in the hypothermic group for death to moderate severe disability and 20% in the normothermic group. For death, 18% in the hypothermic and 15% in the normothermic.
Ben Courchia, MD (01:10:01.198)
By gestational age, when they break it down, you can see that for the primary outcome, as you mentioned, the hypothermic group did worse at every level of stratification. It's not so clear when it comes to death alone, because at 33 weeks, it seems that the normothermic group may have had a little bit less death, but for every other gestational group, 34 to 35+6, it's still higher.
Daphna Yasova Barbeau, MD (01:10:29.946)
Yeah. I think this is especially important because based on the early cooling trials, a lot of centers are routinely cooling 35-weekers and considering these 34-weekers. They do discuss in the discussion that in those hallmark trials, the number of 35-weekers in those trials were really, really, really low.
It was noted that 32 infants randomized to hypothermia and one to normothermia attained temperatures less than 32 degrees Celsius during the intervention. So this was a lot of infants. They did an analysis which excluded all of those infants, and they didn't reveal any clinically important difference between groups for the primary outcome or death alone. I think that's another important feature. We know that a baby getting too cold or too hot is problematic. They did see some of that in the initial cooling trials as well.
Death was attributed by site investigators to brain injury in 15 of 18 infants, so 83% in the hypothermia cohort and 56% in the normothermia cohort. Multi-organ failure was considered the cause of death in 11% of the hypothermia cohort and 22% in the normothermia cohort. The cause of death in two remaining infants in the normothermia cohort were pulmonary hypoplasia with chronic pulmonary hypertension and severe BPD, whereas the cause in the remaining infant in the hypothermia cohort was cardiomyopathy. Three deaths occurred, two hypothermia and one normothermia, after NICU discharge. The frequency of pre-specified non-death safety events during intervention was comparable between the two groups.
So, the highlight of the discussion is that our findings indicate that therapeutic hypothermia initiated by six hours postnatal age did not reduce the primary outcome of death or moderate to severe disability or death alone at 18 or more months in infants born at 33 to 35 weeks gestation with moderate to severe neonatal encephalopathy. A 74% probability of treatment harm for the primary outcome, an 87% probability for death alone with hypothermia was observed. Although survival with moderate to severe disability appeared to be slightly better in the hypothermia group overall as a total cohort, the absolute difference was small, and this potential small benefit was accompanied by much higher incidence of death.
So this was, I think, a major shift in the community in terms of cooling. I think we were all very hopeful that cooling these younger babies would be as good as cooling the older babies.
Ben Courchia, MD (01:13:42.092)
We're always rooting to have more tools at our disposal to help these infants and these families. The outcomes are unfortunate, but I think that it's important for people, especially the people listening to us internationally, that they realize how litigation has really influenced the therapeutic creep of the management of HIE in the US. There is so much litigation, so many lawsuits, that a lot of people feel they don't want to take the risk and end up cooling more babies that maybe may not completely meet criteria and that contributes to the therapeutic creep. To me, it's almost better that the outcomes of this study were so stark that it clearly shows it's not good. It actually will cause harm. If it had shown no harm or if it had shown that they had similar outcomes, I wouldn't have been surprised if that would have been translated into [cooling] this category of patients, even though the benefits may not be there. So I think it paints a very clear picture that it should not be done. Maybe this will start pausing some of the therapeutic creep.
Daphna Yasova Barbeau, MD (01:14:55.194)
It should be noted this is not the first paper to look at this group of babies. Those other studies considered it in very small groups or single centers, and considered it to be “safe” with not a lot of differences in outcomes. But this is the biggest trial for this group of babies.
Ben Courchia, MD (01:15:18.35)
Definitely something you can take to the bedside tomorrow. Thank you for reviewing that impressive study. All right, I think that wraps up Journal Club for us today. Thank you for reviewing the papers. I'll see you next time.