#447 - [Journal Club] - 📌 Is phototherapy doing more harm than good in very preterm infants?

In this Journal Club episode, Ben and Daphna review a nationwide Swedish cohort study examining the association between phototherapy duration and neonatal outcomes in very preterm infants (22 to 31 weeks). The study’s primary outcome, late neonatal mortality on days 8 to 27, was not significantly associated with phototherapy duration. However, longer phototherapy exposure was associated with increased odds of severe neonatal morbidity, including IVH and BPD, in infants born at 26 to 31 weeks. The findings prompt an important conversation about the near-universal use of phototherapy in preterm neonates and whether current practice warrants reassessment.

----

Phototherapy, Morbidity, and Mortality in Very Preterm Newborns. Deschmann E, Håkansson S, Söderling J, Norman M.JAMA Netw Open. 2026 May 1;9(5):e2614107. doi: 10.1001/jamanetworkopen.2026.14107.PMID: 42166159 Free PMC article.

Transcript

[00:00] Ben Courchia, MD: Hello everybody, welcome back to The Incubator Podcast. We're back today for an episode of Journal Club. Daphna, good morning. How are you?

[00:08] Daphna Barbeau: I'm doing so well. I was actually, just —

[00:11] Ben Courchia, MD: Liar.

[00:14] Daphna Barbeau: Okay, fine. I've been disgruntled about a number of logistical things in the last week, so that's true. God, you don't have to air all of our dirty laundry.

[00:18] Ben Courchia, MD: You're grumpy today.

[00:25] Ben Courchia, MD: There we go. We come as we are.

[00:32] Daphna Barbeau: All right, sure. I mean, don't we all have those small logistical things — you feel like they stack up over the course of a week and you're like, gosh, I just can't do anything today. But here we are. We're going to do Journal Club regardless.

[00:46] Ben Courchia, MD: I found that this actually happens when you're quite busy. When you get quite busy, if things go well, you manage. But as soon as something doesn't go quite well, the toll it takes is exponentially higher than it truly is. Sometimes you deal with things that are not objectively a big deal, but because you're tired and managing a million different things, that small thing is going to be like the grain of sand in the watch mechanism — it's going to throw the whole thing off.

[01:13] Daphna Barbeau: Yeah, it really just feels like it puts you over the edge. But nothing will get in the way of doing Journal Club. I think it's good for us. It centers us, and it's good for the community. So here we are.

[01:29] Ben Courchia, MD: While we're venting, we just want to remind people that we're going to let the survey run for a couple more weeks. We're celebrating the five-year anniversary of the podcast and we're going to have our usual feedback survey. Feel free to go and fill it out. We take the survey results very seriously — all the changes you've seen happen on the podcast over the years are driven by survey results. And we have lots of gifts we're hoping to send to people just for thanking them for their time in sharing their thoughts, because we believe that is not taken for granted.

[02:07] Daphna Barbeau: Reading the survey is something that is the opposite of disgruntling — and that doesn't mean people can't be constructive. We've changed a lot of things based on that feedback.

[02:18] Ben Courchia, MD: Yeah, a lot of things. For example, people say things like, "I have trouble finding the episodes." So it's not all flowers.

[02:24] Daphna Barbeau: Please do be constructive. It actually makes us feel quite good when the community is engaged in making the platform better — that's what we've always wanted.

[02:35] Ben Courchia, MD: When the feedback is constructive, it also gives us a hint that people have been listening and engaging with the platform. All right.

[02:41] Daphna Barbeau: We'll take all comers. Please leave your feedback, and thank you to everyone who has been leaving feedback on their podcast streaming app — that's always nice.

[02:53] Ben Courchia, MD: We have a new comment on the Apple Podcasts platform. If you have the time and want to leave us a five-star review or a little feedback, it helps the platform a lot. This one is from Junior Neo Attending — I'm assuming somebody who recently graduated. They said: "I've been listening to The Incubator in various forms for several years now and continue to find it helpful with staying up to date with current topics and controversies, and learning the historical context behind disease processes." This person mentions they particularly loved the episodes on hypoxic-ischemic encephalopathy (HIE) and therapeutic hypothermia reviewing physiology. "Thank you, Daphna and Ben, for providing such an easily accessible forum that covers so many layers of neonatology, inclusive of not just neonatal expert opinion, but also parental values. It is what I listen to on my daily drive to and from the unit." Lovely.

[03:49] Daphna Barbeau: We love that. Thank you for writing.

[03:51] Ben Courchia, MD: Thank you, Junior Neo Attending. Okay. The first paper we're reviewing this week is from JAMA Network Open. Very interesting subject: phototherapy, morbidity, and mortality in very preterm infants. The first author is Imoki Deshman — I'm going to mispronounce that — and this is a paper out of Sweden.

[04:20] Ben Courchia, MD: It's a very interesting topic, obviously, because we're talking about phototherapy in very preterm infants — something we're still not entirely sure about. The background notes that almost all preterm neonates develop hyperbilirubinemia within the first two weeks of life, and that excessive bilirubin concentrations can lead to neurological dysfunction, sometimes even permanent brain injury in the form of kernicterus. The introduction of phototherapy has substantially lowered the incidence of excessive bilirubin levels and kernicterus. Phototherapy has also been proven to reduce the need for exchange transfusion for hyperbilirubinemia — I think that's the classic board question where they ask what the effect of phototherapy is, and we all tend to think it lowers the bilirubin, but the data was actually driven to find out whether it reduced the need for exchange transfusion. And that's obviously a very big deal. The striking statistic in the background is that more than 80% of extremely preterm newborns have been reported to be treated with phototherapy, so most very preterm babies do get some form of phototherapy at one point or another.

[05:52] Ben Courchia, MD: Now, there is emerging evidence suggesting that phototherapy may be harmful and overused in preterm infants. A paper I pulled from the Journal of Perinatology, dating from 2012, from the Neonatal Research Network, is called "Does Aggressive Phototherapy Increase Mortality While Decreasing Profound Impairment Among the Smallest and Sickest Newborns?" — first author John Tyson. It's always very interesting to see what the supposed effect of phototherapy can be on neonates. Phototherapy has been associated with increased dermal water losses, delayed circulatory adaptation, and feeding intolerance. More importantly, a secondary analysis of a randomized clinical trial indicated that aggressive phototherapy among newborns with birth weights between 500 and 750 grams reduced neurodevelopmental impairment at the cost of higher mortality. In an earlier US study, newborns with a birth weight of less than 1,000 grams who received phototherapy had a 19% increased mortality rate.

[06:44] Ben Courchia, MD: The primary aim of the study we're reviewing today is this nationwide population-based cohort study testing the association between the duration of phototherapy and neonatal mortality in very preterm newborns, with major neonatal morbidities as secondary outcomes. To evaluate any confounding, they also tested the association between bilirubin levels and neonatal mortality and morbidity, with and without restriction for conditions known to contribute to phototherapy use and mortality.

[07:20] Ben Courchia, MD: The study was approved by Swedish authorities in 2024 and used the SNQ — the Swedish Neonatal Quality Register — a database we've referenced on the podcast through various studies and a very robust one. Between November 2015 and December 2024, almost 9,000 neonates were born between 22 weeks and 31 weeks and 6 days and admitted to the NICU. Of those, 2,647 were ineligible because they were cared for in units that did not report daily exposure data on phototherapy duration or peak bilirubin. Among the remaining newborns, they excluded babies who died within the first few days after birth and babies with congenital anomalies, leaving them with almost 664 newborns with complete data.

[08:50] Ben Courchia, MD: The primary exposure was the duration of phototherapy in days during the first week of postnatal life. Phototherapy within 24-hour increments was reported as either started, ongoing, or stopped on the day of charting — and any of those registrations on a given day was categorized as a day with phototherapy. Three categories were defined: zero to three days, four to five days, or six to seven days. Other exposures included peak bilirubin, defined as the highest serum value registered for each newborn, and the area under the curve (AUC) in the first week after birth. The analysis included a total of about 22,000 bilirubin values. Bilirubin percentiles were calculated across two gestational age strata and four categories: 25th percentile or below, 26th to 50th, 51st to 75th, and above the 75th percentile.

[09:35] Ben Courchia, MD: Now, how did they manage phototherapy in preterm newborns in Sweden? The Swedish guideline for treatment of neonatal hyperbilirubinemia recommends phototherapy in cycles of two to four hours, followed by a break for feeding and nursing care, after which phototherapy was resumed. They did not have a specific protocol for when to start phototherapy for hyperbilirubinemia.

[10:52] Ben Courchia, MD: The primary outcome was late neonatal mortality on postnatal days 8 to 27. Secondary outcomes included a composite of severe neonatal morbidity: intraventricular hemorrhage (IVH) grade 3 to 4, patent ductus arteriosus (PDA) treated pharmacologically or surgically, necrotizing enterocolitis (NEC) stage 2A or higher, severe bronchopulmonary dysplasia (BPD) defined as greater than or equal to 30% supplemental oxygen at 36 weeks, and retinopathy of prematurity (ROP) treated with laser, anti-vascular endothelial growth factor (anti-VEGF), or surgery.

[11:26] Ben Courchia, MD: Let's get into the results. In this cohort study of 4,917 neonates — 1,800 born extremely preterm at less than 28 weeks, and 3,162 born very preterm at 28 to 31 weeks — the median gestational age was 29 weeks and the median birth weight was 1,180 grams. Importantly, 36% of the cohort weighed less than one kilogram. Overall, 95.5% were treated with phototherapy — even higher than what they reported in their methods. The number of days on phototherapy was higher for babies born at lower gestational ages: the smaller and more immature, the more likely to have extended days on phototherapy.

[12:20] Ben Courchia, MD: Late neonatal mortality — the primary outcome — occurred in 1.7% of the cohort treated for zero to three days, 2.9% for those treated four to five days, and 4.6% for those treated six to seven days. I invite you to look at Figure 1, because you might think there's a clear progression — and it seems like the longer you're on phototherapy, the more likely you are to have late neonatal mortality. But when you look at the confidence intervals, it's not statistically significant. Excluding newborns with IVH grade 3 to 4, isoimmunization, or Apgar scores below 4 at five minutes did not change the results, and neither did extending the mortality endpoint from day 27 to hospital discharge.

[13:15] Ben Courchia, MD: What I wanted to focus on is the finding that was actually significant — unfortunately not the primary outcome, but severe neonatal morbidity (SNM). It occurred in 10.7% of infants treated for zero to three days, 17% in those treated four to five days, and 28.3% in those treated six to seven days — and that was statistically significant. So the longer on phototherapy, the more likely to have a severe neonatal morbidity. These associations were limited to newborns between 26 and 31 weeks — when you look at babies between 22 and 25 weeks, the association doesn't hold. And this finding could be largely attributed to higher prevalence of IVH grade 3 to 4 and severe BPD in newborns receiving phototherapy for six to seven days. Interestingly, phototherapy for six to seven days was associated with decreased odds of treated PDA in newborns born at 22 to 25 weeks — always a subject of controversy. Phototherapy duration was not associated with NEC or ROP in any gestational age stratum.

[15:33] Ben Courchia, MD: Going back to late neonatal mortality — it was not related to peak bilirubin levels and occurred in 2.5% with a peak bilirubin below the 25th percentile and 3% above the 75th percentile, so not much difference across those thresholds. A peak bilirubin above the 75th percentile was associated with increased odds of the composite severe neonatal morbidity compared to below the 25th percentile, and after stratification this association remained significant only in babies at 30 to 31 weeks gestational age.

[16:17] Ben Courchia, MD: The conclusion: in this cohort of very preterm newborns, the duration of phototherapy was not associated with neonatal mortality. However, a longer duration was associated with increased odds of some severe neonatal morbidities. Even if this association may have resulted from some confounding, the findings underscore recommendations to reduce unwarranted phototherapy exposure in very preterm neonates. And I think that's really the point of discussion — we use a lot of phototherapy assuming it's very safe with very little side effects, but this is the kind of data that should give us pause to rethink how long we keep a baby under phototherapy.

[17:03] Daphna Barbeau: Yeah. For me it highlighted two things. One — and I think we forget this — bilirubin is probably also kind of an inflammatory marker. The sickest babies have the most prolonged needs. They tried to account for some of that, but there's some of it we just can't account for without a really, really good severity score. And second, I think even if it were just the water losses or the temperature changes from the lights, we should be cognizant of that — especially in the smallest babies. Rather than just leaving a baby on continuously for a week, which I think some colleagues have done — I'm not sure that's the right practice given the data we do have.

[18:02] Ben Courchia, MD: And we have very arbitrary values for phototherapy thresholds. For some of the smaller babies, some people say if it's above 5, start it. And then if the bilirubin is 5.6, 5.2, 5.1 — they remain on phototherapy even if it's relatively stable. I'm not blaming the practice, I'm just saying we really have no idea what we're doing at that point. So I think this is something we'll have to rethink. It's interesting also to see what is considered safe for a preterm. We had a very interesting episode series on the podcast a while back about rethinking phototherapy — it mostly focused on full-term infants. We spoke with Dr. Alex Kemper, one of the authors of the AAP guidelines, and Dr. Dan Rauch. We also had a very interesting discussion with Dr. Deepak Mehta about how we think about phototherapy in preemies — he wrote a nice NeoReviews paper on the subject, which I'd refer you to.

[19:36] Daphna Barbeau: The other thing that complicates all of this — related to that series — is that we can talk about duration of phototherapy, but we don't always capture the intensity. How much phototherapy are we giving? It may be many days at low intensity, or just a few days at high intensity. I'm not sure we've looked at this in a granular enough way.

[20:09] Ben Courchia, MD: Right — are we measuring the intensity of phototherapy every time we touch the baby, every time we adjust the light? These are important considerations. But I am always fascinated by interventions we consider super safe. 95% of that cohort was exposed to phototherapy — what else do we do to babies at a 95% rate? And for us to do that with so little evidence behind it is kind of remarkable. We're going to have a discussion soon about erythropoiesis-stimulating agents (ESAs) — there's so much data and yet the debate continues about whether to use them — and yet phototherapy, with far less data, is essentially universal. It's always interesting how we agonize over starting something with robust supporting data, but something with very little data becomes standard of care.

[21:15] Daphna Barbeau: Yeah. The standard of care.

[21:18] Ben Courchia, MD: 95% for very preterm babies — and we're not talking about bilirubins of 12. We're starting phototherapy for levels below 10 without knowing what the real target should be. It's fascinating. But my fascination may not be shared by the rest of the audience, and so for that reason we're going to move on. All right, Daphna, I'll see you tomorrow for another episode.

[21:42] Daphna Barbeau: Sounds good. Thanks.