#453 - [Journal Club] - 📌 Should We Cool 35 Week Infants with Encephalopathy?

Cooling works at 36 weeks. At 35 weeks, nobody is sure. Ben brings a new Journal of Perinatology analysis of the National Inpatient Sample, covering 1.4 million infants from 2016 to 2022, asking what happens when therapeutic hypothermia is offered just below the evidence line. Cooled 35-weekers died at higher rates than cooled 36-weekers, but within the 35-week group, cooling changed nothing either way. Coagulopathy rose with cooling. Mediation analysis says it wasn't the cause. Ben and Daphna work through what that leaves us, and why shared decision making and careful documentation carry the weight here

----

Therapeutic hypothermia and in-hospital mortality in 35-week infants with encephalopathy. Aly H, Eltaly H, Mohamed FA, Saker F, Acun C, Mohamed MA.J Perinatol. 2026 Jun 3. doi: 10.1038/s41372-026-02738-2. Online ahead of print.PMID: 42236997

Transcript

Ben Courchia, MD (00:00.408)
Hello, everybody. Welcome back to the Incubator Podcast Journal Club. We're back today for another episode of Journal Club. I have a very interesting paper to review. Daphna, good morning. How are you?

Daphna Yasova Barbeau (00:08.196)
Good morning. I'm excited for your very interesting paper.

Ben Courchia, MD (00:10.432)
All right. Yeah, just a reminder, today's the last day to submit your feedback form on our website for the Incubator anniversary to get access to potential special prizes. It takes two minutes. We have lots of things to give.

Daphna Yasova Barbeau (00:25.999)
Two minutes, and we've already read some of the responses. We are making changes in real time. So people should know that it doesn't just go to the cloud. We read them and we use them.

Ben Courchia, MD (00:32.321)
Yeah.

Ben Courchia, MD (00:35.986)
That's true. That is true. Okay, so the paper I'm reviewing next was published in the Journal of Perinatology. It's called "Therapeutic Hypothermia and In-Hospital Mortality in 35-Week Infants with Encephalopathy." First author is Hany Aly from Cleveland, Ohio. Lots of interesting people on that paper as well.

Daphna Yasova Barbeau (00:45.231)
Mm-hmm. Oof.

Ben Courchia, MD (01:00.982)
Obviously, we know the devastating effect of HIE on our neonates. And the criteria for therapeutic hypothermia, which is, to be honest with you, the only evidence-based intervention that we currently have available for these babies, is indicated for infants who are 36 weeks of gestation or more and who have moderate to severe HIE. The safety and efficacy of hypothermia in infants who are born before 36 weeks is uncertain.

Recently, we reviewed a paper that was published in JAMA Pediatrics on a Bayesian probability approach to babies receiving hypothermia between 33 and 35 weeks, which showed that there was an increased signal for mortality in that population. And the question really lingered as to maybe 33, 34 weeks, okay, but what about 35 weeks? There were a lot of questions about whether the data

was really strong for that specific gestational age. Importantly, as the authors mention, current evidence remains inconclusive regarding the safety of therapeutic hypothermia in infants who are born at 35 weeks. And a recent international survey of 88 centers demonstrated substantial practice variation, with many centers continuing to offer cooling at 34 to 35 weeks despite findings from the randomized trial. Notably, 22 centers reported in-hospital mortality rates lower

than those observed in the trial, and therefore did not find sufficient justification to alter their practice, citing more favorable outcomes in their multicenter experience. In addition, several small single-center studies have reported mixed results. Consequently, robust population-level data remains limited. The purpose of this study is to conduct an epidemiological analysis using the National Inpatient Sample, the NIS

database, covering 2016 to 2022, to examine hypothermia utilization and outcomes among infants born between 35 and 36 weeks with HIE. So basically, they took the data from the NIS looking at the infants specifically during that 35th week from 2016 to 2022. I'm not going to get into what the NIS is. It's a very well known database from the Healthcare Cost and Utilization Project, HCUP,

Ben Courchia, MD (03:28.014)
and the Agency for Healthcare Research and Quality, the AHRQ. Infants that were included in the study basically had a gestational age of 35 weeks. They compared them to infants who were 36 weeks and more, and they had to have a diagnosis of HIE and receive therapeutic hypothermia. So whenever we have these database studies, I think the methods are pretty straightforward. I'm going to let you review that,

and I want to get into the results. So the sample of patients who met inclusion and exclusion criteria was 1.4 million infants. Of these, there were 493,000 infants in the 35-week group. Of them, 2,354 had a diagnosis of any degree of HIE. They had about 900,000 babies in the 36-week group, and of them, 3,900 had any degree of HIE.

In terms of hypothermia utilization, fewer infants in the 35-week group received hypothermia. In the 35-week group, 19.8% received cooling, compared to the 36-week group, where 889 infants, 22.4%, received therapeutic hypothermia. This yielded an adjusted odds ratio of 0.84 and a

significant p-value of 0.01. Respiratory distress syndrome was significantly associated with the 35-week group, which is not so surprising, while persistent pulmonary hypertension of the newborn and thrombocytopenia were more associated with the 36-week group. Coagulopathy was more strongly encountered in the 35-week group compared to the 36-week group. The median length of hospitalization did not differ between the two gestational ages in infants with HIE, and that was about 15

days. So let's talk a little bit about in-hospital mortality. The overall in-hospital mortality among 35-week infants with HIE, regardless of whether they were cooled or not, was comparable to that of 36-week infants. That was about 10%, 9.98 versus 9.99. So very, very much the same. Infants in the 35-week group with HIE who received cooling had significantly higher in-hospital mortality

Ben Courchia, MD (05:51.865)
compared to the 36-week infants with HIE who received cooling. So mortality was seen to be higher. That was 10.3% mortality in the 35-week infants compared to 6.8% in the 36-week infant group. P-value 0.04. In-hospital mortality among the 35-week infants did not change with versus without therapeutic hypothermia. 10.3%

versus 9.9%, which shows that it doesn't seem like cooling really made a difference on that mortality ratio. The use of therapeutic hypothermia was associated with a significant decrease in in-hospital mortality among 36-week infants compared to those who did not receive cooling, 6.8% versus 11%, with a p-value of less than 0.001. The p-values are obviously very strong because we're talking about a lot of babies.

And again, just to show that the data is pretty strong, in-hospital mortality for babies who were 37 weeks with HIE who received cooling was significantly lower than for those who did not receive cooling. Because again, if you had seen that cooling did not make a difference across all gestational ages, then you say, okay, well, maybe it's not the gestational age that's the issue, maybe it's the data. The median length of hospital stay did not differ significantly, as we said.

And interestingly enough, they looked at coagulopathy. Since 35-week infants who received cooling had higher rates of coagulopathy and in-hospital mortality when compared to 36-week infants, and we know coagulopathy and coagulation disturbance is a risk factor for babies who are born at 35 weeks, the authors repeated their regression analysis, adding coagulopathy as a confounding variable. After controlling for coagulopathy in the regression model, TH was not associated

with in-hospital mortality in 35-week infants, meaning that it doesn't seem like it's coagulopathy that's leading to an increased mortality in that particular group. It was then important to identify whether coagulopathy existed in HIE infants regardless of their exposure to therapeutic hypothermia. And coagulopathy in that 35-week infant group was significantly increased in those who received therapeutic hypothermia, 28.7%,

Ben Courchia, MD (08:17.651)
28.7% versus only 18.4% coagulopathy in those who did not receive therapeutic hypothermia. So clearly that side effect that we know from cooling was seen in that group. They did a mediation analysis, which didn't show any correlation between coagulopathy and inpatient mortality. And I said I would do this one relatively quick, but that's really it.

So the conclusions of the article are that although the role of therapeutic hypothermia in 35-week infants remains uncertain, its use was not associated with increased in-hospital mortality. In this national cohort, therapeutic hypothermia was associated with an increased risk of coagulopathy. However, mediation analysis did not support a significant role for coagulopathy in in-hospital mortality.

These findings underscore the need for careful risk stratification and heightened safety surveillance when considering therapeutic hypothermia in this population. Prospective studies are needed to define eligibility thresholds and monitoring strategies before routine use can be supported. Until more definitive evidence is available, close monitoring and early management of coagulopathy remains prudent when TH is administered in this gestational age group. I thought this was very interesting. I know you have a lot of thoughts on that.

Daphna Yasova Barbeau (09:39.706)
Yeah, very interesting. And just to clarify, because I think it can be really confusing if you don't have the tables in front of you, there was a significant difference in mortality between the 35-weekers and the 36-weekers who were treated. So that was different, 10.3 versus 6.76. But if you look at the 35-weekers as a group themselves, and look at mortality for those who were treated and untreated, that's where there was no difference.

Ben Courchia, MD (09:56.046)
Yeah.

Ben Courchia, MD (10:08.427)
No difference. And that's what the first line of the conclusion really mentions when they say, although the role of therapeutic hypothermia in 35-week infants remains uncertain, its use was not associated with increased in-hospital mortality. It was also not associated with a reduction in in-hospital mortality. And so when you're 35 weeks, whether you get cooled or not, based on the data from this database, it doesn't really impact your ultimate outcome, which is survival.

Daphna Yasova Barbeau (10:09.123)
And mortality. Yeah. Yeah.

Daphna Yasova Barbeau (10:22.863)
That's right.

Daphna Yasova Barbeau (10:32.291)
Yeah, and it was hard to get into this, but degree of HIE in babies who were treated and untreated, fortunately or unfortunately, there's a difference there too. In those who were treated, moderate HIE was 34.1% versus 25%. That was not statistically significant. But severe HIE was 7.28% in those treated, 12.8% in untreated, and that was statistically significant.

Same thing, unspecified HIE, which I guess is unspecified, 37.7% in the treated, 45% in untreated, and that was statistically significant. So it looks like there was potentially some improvement in the HIE diagnosis, even in the 35-week group. So it adds definitely another layer to our conversation

about 35-weekers and therapeutic hypothermia, for sure.

Ben Courchia, MD (11:33.367)
Yeah. And I think that what this paper is actually opening the door to is that we potentially can't address HIE in the 35-week infant, and maybe even younger than that, the same way we address it in a 38-weeker. For a 38-week infant, you're going to say moderate to severe, I'm treating, or I'm doing this, I'm doing that. Maybe for 35-weekers there are certain classifications that will matter more.

Maybe mild, moderate, severe will not be treated the same way. And I think that this matters tremendously. So.

Daphna Yasova Barbeau (12:10.883)
Yeah, and I think the last thing, people in units will say, well, now what are we supposed to do? I think that's when we really have to have good shared decision making with families, good documentation of those discussions, especially in light of the current AAP recommendations and guidelines.

Ben Courchia, MD (12:30.103)
Yeah, I think this data comes to support what the AAP has published, in the sense that it's giving you caution about cooling below 36 weeks. But it's not definitive by any means. And I think that this is where, if you're going to deviate from that, like you said, shared decision making, review of the data with the family, and deciding together is going to be key.

Daphna Yasova Barbeau (12:56.699)
All right, buddy.

Ben Courchia, MD (12:57.817)
All right, I'll see you tomorrow.

Daphna Yasova Barbeau (12:59.728)
Right.