A Small Sensor on the Forehead, and What It Can Tell a NICU Team About a Baby's Brain

Explaining the NIRTURE trial (Jani et al., JAMA Network Open 2026), which tested whether watching brain oxygen levels helps keep them steady in babies born very early

A research team in Australia, New Zealand and the United States tested whether a small sensor placed on a premature baby's forehead — one that continuously measures how much oxygen is reaching the brain — helps the medical team keep that oxygen level steady. In 100 babies born before 29 weeks of pregnancy, the babies whose monitors the team could actually see spent far less time outside the healthy oxygen range than babies whose identical monitors were covered up. Whether that steadier oxygen level leads to healthier brains later on is still unknown, and larger studies are being planned to find out.

Why This Question Has Been So Hard to Answer

Babies born extremely early — before 29 weeks of a 40-week pregnancy — face a set of risks that come from being born before the body is finished getting ready. One of the most serious is injury to the developing brain. Doctors use the term "encephalopathy of prematurity" for a mix of problems that can affect the brain's wiring and the way it matures, and these problems are a major reason some children born very early go on to have learning difficulties, movement difficulties, or both [1][2].

Oxygen sits right at the centre of this problem, and in a way that surprises many families. Too little oxygen is obviously dangerous — that is easy to understand. But too much oxygen is also harmful to a very immature brain, because excess oxygen generates unstable molecules that damage delicate developing tissue. Studies that followed extremely premature babies have found that oxygen levels in the first days of life, in both directions, are linked to later evidence of brain injury [3]. The goal, in other words, is not "more oxygen" — it is "the right amount, steadily."

For years, neonatal units have tried to achieve this using a pulse oximeter: the small light-up probe wrapped around a baby's hand or foot. It is a genuinely valuable device, but it measures oxygen in the bloodstream at the edge of the body, not oxygen actually arriving at the brain. A very large research effort combining five clinical trials compared aiming for slightly lower blood-oxygen readings with aiming for slightly higher ones, and found real trade-offs — the higher target was linked with fewer deaths and less severe bowel disease, but more eye disease requiring treatment — while the rate of death or major disability at around two years of age was similar either way [4]. That result told clinicians they had probably reached the limit of what a finger-and-toe measurement could do.

So attention turned to a technology that measures the brain directly. It is called near-infrared spectroscopy, usually shortened to NIRS, and it works using harmless infrared light — the same principle as a pulse oximeter, but with a soft adhesive sensor on the forehead. It shines light through the skin and skull and reads how much comes back, and from that it estimates how much oxygen the brain tissue underneath is holding. It causes no pain, delivers no radiation, and can run continuously for days.

The first careful test of this idea, a study called SafeBoosC-II, showed that when clinical teams could see the brain-oxygen reading and had a written plan for responding to it, babies spent less time with low brain oxygen over the first three days of life [5]. That was encouraging, but an independent review of all the available evidence concluded there was simply not enough of it to recommend the technology for routine care [6]. A much larger study, SafeBoosC-III, then enrolled 1,601 babies across 70 hospitals in 17 countries — and found that using the monitors for the first 72 hours did not reduce death or severe brain injury [7]. A follow-up of those same children at two years of age, published in 2026, also found no benefit for development or thinking skills [8].

That could mean the idea is simply wrong. But there was another possibility that kept nagging at researchers. In that large study, different hospitals used monitors from different manufacturers, which do not all report the same number for the same amount of oxygen. Some babies wore sensors designed for adults, which read lower than sensors designed for newborns. Alarm settings were adjusted. Most staff had little experience with the technology. It was, in short, possible that the study had tested a blurry version of the intervention rather than the intervention itself.

What the Researchers Did

The NIRTURE trial was designed to settle that narrower question first: if you remove all that variation, can a team actually keep a premature baby's brain-oxygen level steady? [1]

Between October 2021 and July 2024, researchers enrolled 100 babies born before 29 weeks at five large neonatal intensive care units — three in Australia, one in New Zealand, and one in the United States. Babies were enrolled within six hours of birth, and a computer randomly assigned each one to one of two groups. Random assignment is the heart of a fair trial: the two groups end up similar in every way except the treatment being tested, so any difference at the end can be credited to the treatment rather than to some hidden difference between the babies.

Here is the elegant part of the design. Every baby in the study — in both groups — wore exactly the same forehead sensor, connected to exactly the same monitor, recording continuously for the first five days of life. The only difference was whether anyone could see the screen. For half the babies, the reading was visible, and the team followed a written step-by-step plan whenever the brain-oxygen level drifted below 65% or above 90%, acting within one minute and checking the result within half an hour. For the other half, the monitor kept recording silently behind an opaque cover, and those babies received the unit's normal care, which includes responding to the usual pulse oximeter as always.

This means the researchers could compare not "monitored versus unmonitored," but "acted upon versus not acted upon" — a much cleaner comparison. They also used a single brand of monitor with a sensor made specifically for newborns, kept the alarms at their factory settings, checked the skin under the sensor every four hours, and required every staff member to complete training before their hospital joined. The full plan was published in advance, before any results were known [9]. Parents who had themselves been through a premature birth helped design the study, including advising on how families should be approached for consent.

The main thing measured was the total amount of time each baby's brain oxygen spent outside the 65%–90% range across those five days, weighted by how far outside it went — a measure the researchers call the "burden."

What They Found

The difference was striking. Babies whose monitors were visible had a typical burden of 5.7% hours outside the target range. Babies whose monitors were covered had a typical burden of 39.6% hours — roughly seven times as much [1].

Perhaps the most interesting detail is which direction the problem was in. Most of the time spent outside the range was time with brain oxygen too high, not too low. In the covered-monitor group, high-oxygen time accounted for a typical 23.7% hours, while low-oxygen time accounted for 2.5% hours. When teams could see the reading, the high-oxygen time dropped to 3.5% hours and the low-oxygen time to 0.7% hours. This is a genuinely useful piece of information about how oxygen is being given during a premature baby's first days, quite apart from whether the monitor itself is worth using — and it fits with what laboratory research suggests about excess oxygen being harmful to an immature brain [10].

The effect was also larger in the smallest, most fragile babies — those born before 26 weeks — than in those born between 26 and 29 weeks.

Now the important caution. When the researchers looked at the outcomes families care about most — survival, bleeding in the brain, chronic lung disease, bowel problems, eye disease — the two groups were essentially the same. Six percent of babies in each group died before going home. Around a third of babies in each group had some degree of bleeding in the brain. Around 60% in each group developed some eye disease of prematurity, though fewer babies in the monitored group needed treatment for it — a difference too small, in a study this size, to mean anything reliable.

This is not a disappointing result so much as an expected one. A study of 100 babies is far too small to detect differences in outcomes that happen to only a handful of infants. The trial was designed to answer a physiological question, and it answered it.

On safety, the news was good: not one baby in the monitored group developed any skin injury from wearing the sensor for five days. One baby in the other group had a temporary mark from the sensor that faded once it was repositioned.

What This Means for Families, and What Comes Next

If your baby is in a unit that uses one of these forehead sensors, this study is genuinely reassuring on two counts: it is safe over five days of continuous use, and when clinical teams can see and act on the readings, they really can hold a baby's brain oxygen much steadier.

But it is important to be clear about what has not been shown. There is no evidence yet that a steadier brain-oxygen level leads to a healthier brain later on. The largest study that looked for that benefit did not find it, at either 36 weeks or at two years of age [7][8]. If a member of your baby's team explains that they are using this monitor, it is entirely reasonable to ask what they will do differently because of it — and to understand that this technology currently tells the team more about your baby's physiology than about your baby's future. It is a promising tool, not a proven protection, and honest clinicians will say so.

Researchers are working on several things next. The children who took part in NIRTURE are being followed up to see how they develop, and those results are not yet available. Earlier work hinted that low brain-oxygen levels in the first days might be linked to development at two years of age [11], though the same group found no change in early blood markers of brain injury [12] — findings that raise a question rather than answer it. The next major step is a larger trial that keeps everything NIRTURE standardised, uses detailed brain imaging rather than bedside ultrasound alone, and follows children long enough to see whether steadier oxygen in five early days makes a difference to the years that follow. Until then, the honest answer is that we know how to steady the signal, and not yet what steadying it is worth.

References

  1. Jani PR, Goyen T, Balegar KK, et al. Cerebral Oximetry–Guided Treatment and Cerebral Oxygenation in Extremely Preterm Infants: A Randomized Clinical Trial. JAMA Netw Open. 2026;9(2):e2557620. doi:10.1001/jamanetworkopen.2025.57620 ↩
  2. Volpe JJ. Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. Lancet Neurol. 2009;8(1):110-124. doi:10.1016/S1474-4422(08)70294-170294-1) ↩
  3. Rantakari K, Rinta-Koski OP, Metsäranta M, et al. Early oxygen levels contribute to brain injury in extremely preterm infants. Pediatr Res. 2021;90(1):131-139. doi:10.1038/s41390-021-01460-3 ↩
  4. Askie LM, Darlow BA, Finer N, et al; Neonatal Oxygenation Prospective Meta-analysis (NeOProM) Collaboration. Association Between Oxygen Saturation Targeting and Death or Disability in Extremely Preterm Infants in the Neonatal Oxygenation Prospective Meta-analysis Collaboration. JAMA. 2018;319(21):2190-2201. doi:10.1001/jama.2018.5725 ↩
  5. Hyttel-Sørensen S, Pellicer A, Alderliesten T, et al. Cerebral near infrared spectroscopy oximetry in extremely preterm infants: phase II randomised clinical trial. BMJ. 2015;350:g7635. doi:10.1136/bmj.g7635 ↩
  6. Hyttel-Sørensen S, Greisen G, Als-Nielsen B, Gluud C. Cerebral near-infrared spectroscopy monitoring for prevention of brain injury in very preterm infants. Cochrane Database Syst Rev. 2017;9(9):CD011506. doi:10.1002/14651858.CD011506.pub2 ↩
  7. Hansen ML, Pellicer A, Hyttel-Sørensen S, et al. Cerebral Oximetry Monitoring in Extremely Preterm Infants. N Engl J Med. 2023;388(16):1501-1511. doi:10.1056/NEJMoa2207554 ↩
  8. Rasmussen MIS, Hansen ML, Pellicer A, et al. Cerebral Oximetry in Extremely Preterm Infants: 2-Year Follow-Up of the SafeBoosC-III Randomized Clinical Trial. JAMA Pediatr. 2026;180(6):619-627. doi:10.1001/jamapediatrics.2026.1066 ↩
  9. Jani P, et al. Targeted Cerebral Oxygenation Using Dedicated Treatment Versus Usual Care in Extremely Preterm Infants: Protocol for a Multicentre International Phase II Randomised Controlled Trial. J Paediatr Child Health. 2025;61(7):1020-1029. doi:10.1111/jpc.70066 ↩
  10. Yates N, Gunn AJ, Bennet L, Dhillon SK, Davidson JO. Preventing Brain Injury in the Preterm Infant—Current Controversies and Potential Therapies. Int J Mol Sci. 2021;22(4):1671. doi:10.3390/ijms22041671 ↩
  11. Plomgaard AM, et al. Early cerebral hypoxia in extremely preterm infants and neurodevelopmental impairment at 2 year of age: A post hoc analysis of the SafeBoosC II trial. PLoS One. 2022;17(1):e0262640. doi:10.1371/journal.pone.0262640 ↩
  12. Plomgaard AM, van Oeveren W, Petersen TH, et al. The SafeBoosC II randomized trial: treatment guided by near-infrared spectroscopy reduces cerebral hypoxia without changing early biomarkers of brain injury. Pediatr Res. 2016;79(4):528-535. doi:10.1038/pr.2015.266 ↩