A Brain-Oxygen Sensor at Birth: What a Two-Year Study of Premature Babies Found

Following the children of the COSGOD III trial to age two — where an early-life brain monitor did, and did not, seem to help

When a baby is born many weeks too early, the first few minutes of life are among the most important they will ever have. A new study followed a large group of premature babies to the age of two to answer a hopeful question: if doctors add a small sensor that watches the baby's brain-oxygen level during those first minutes, do the children grow up healthier? The answer, for most babies, was no clear difference — but among a group of somewhat more mature premature babies, those who had the brain sensor had noticeably fewer problems with movement, vision, and hearing. Because that promising finding came from a small slice of the study, the researchers say it is encouraging but not yet proven [1].

Why This Research Matters

Every premature baby who needs help breathing at birth needs oxygen — but the right amount is a delicate balance. Too little oxygen can injure the brain and other organs; too much can harm the eyes and lungs. For many years, delivery-room teams around the world have judged that balance using two signals: the oxygen level in the blood, measured by a sensor on the hand or wrist, and the heart rate. These signals are built into the resuscitation guidelines that neonatal teams are trained to follow [2], [3]. They save lives, but they are indirect: a number from the wrist does not tell you exactly how much oxygen is reaching the brain [4]. Doctors have tried adjusting the blood-oxygen targets up and down to find the safest setting, and those studies showed just how much the early results can shift depending on the exact target chosen [5].

This uncertainty is what led researchers to look directly at the brain. Using a technology called near-infrared spectroscopy — a soft sensor placed on the baby's forehead — they can estimate how much oxygen the brain tissue is actually getting. Earlier research found that premature babies whose brain-oxygen readings dipped very low in the first minutes were more likely to suffer bleeding in the brain and worse outcomes later [6], [7], [8], and one study linked those early readings to how children did in the long run [9]. A first, small trial then showed that when teams watched the brain sensor and followed a plan to respond to it, they could cut the time the brain spent low on oxygen by more than half [10]. That success is what made a large trial worth doing.

The Study

That large trial was called COSGOD III [11]. It enrolled babies born before 32 weeks of pregnancy at 11 hospitals across Europe and Canada. By the flip of a coin (random assignment, the fairest way to compare two approaches), each baby was given either the usual delivery-room care or the usual care plus the brain-oxygen sensor with a plan for how to respond to it during the first 15 minutes of life. The first results, published in 2023, found that adding the sensor led to a small improvement in the number of babies who survived without early brain injury, but the difference was small enough that it could have been due to chance [11].

The new study picked up where that one left off. Doctors tracked down as many of those same children as they could and assessed them at about two years of age — checking how they moved, thought, saw, and heard [1]. Of the children who could be included, 417 were assessed, about eight in ten of those eligible. The people doing the developmental check-ups did not know which babies had received the brain sensor, so their assessments could not be swayed by that knowledge [1].

What They Found

For the whole group, the brain sensor did not change how the children were doing at two years. About 61 out of every 100 children in the sensor group were alive and free of moderate-to-severe developmental problems, compared with about 62 out of 100 in the standard-care group — essentially the same. Survival was the same, and scores on standard tests of thinking and language were the same [1].

The interesting part came from a planned look at a subgroup: babies born between 28 and 31 weeks — premature, but not as extremely early as some. Among these children, those who had received the brain sensor had strikingly fewer specific problems. Cerebral palsy (a movement disorder) affected about 1 in 100 in the sensor group versus about 6 in 100 with standard care; serious vision problems and hearing problems followed the same pattern. Put together, this trio of impairments affected fewer than 1 in 100 sensor-group children versus about 8 in 100 in the standard-care group [1]. These are exactly the kinds of differences that matter enormously to a child and family.

But the researchers are careful, and so should we be. This encouraging result came from a small subgroup with very few affected children in total, and it was one of several comparisons the team looked at — the kind of finding that sometimes appears by chance and does not hold up when tested again. Tellingly, the babies born earliest of all (before 28 weeks), the ones many people would expect to benefit most, showed no difference at all [1]. There was also one cautionary note: among the 28-to-31-week babies, a serious eye condition of prematurity was actually a little more common in the sensor group early on, even though vision problems at age two went the other way. That kind of mixed signal is one more reason to treat the hopeful finding as a lead to follow rather than a settled fact [1].

What This Means for Families

If your baby is premature and needs help breathing at birth, this study does not change what should happen in the delivery room. The trusted approach — carefully guiding oxygen and support using the blood-oxygen sensor and heart rate — remains the standard, because it works and because the brain sensor has not yet been shown to improve how children do years later [2], [3]. The brain-oxygen sensor appears to be safe and is a genuinely promising idea, but for now it belongs in research rather than as routine care.

There is also reassuring context in the numbers. Across both groups, very few babies died and most scored close to typical ranges on development tests — better than many older studies of extremely premature infants would predict [12], [13], [14]. Part of that reflects real improvements in newborn care over the years, and part reflects the way this study enrolled babies, which tended to include somewhat lower-risk pregnancies. Either way, families can take some comfort that outcomes for premature babies continue to improve.

It can also help to understand why a result like the one in the 28-to-31-week group needs a second look before anyone can rely on it. When researchers divide a study into smaller groups and examine many possible outcomes, chance alone will occasionally produce a difference that looks convincing but disappears when the same question is asked again in a fresh study. That is not a flaw in this research — the team planned the subgroup in advance and reported it openly, including the puzzling eye finding that pointed the other way. It is simply how careful science works: a promising clue is announced honestly, along with all the reasons it might not hold, so that the next study can settle the matter rather than a headline. For parents, the practical takeaway is that trusting well-established care, while researchers test new ideas thoroughly, is exactly the right instinct.

If you read a headline claiming a "brain monitor that prevents cerebral palsy," the fuller truth is gentler: the overall study found no clear benefit, a hopeful signal appeared in one group of more mature premature babies, and researchers now need to test whether it is real.

What Researchers Are Working On Next

The natural next step is a new trial focused specifically on babies born at 28 to 31 weeks — the group where the signal appeared — designed from the start to measure movement, vision, and hearing rather than a broad combined score, and to keep a close eye on the eye condition seen here. If that trial confirms the benefit, brain-oxygen monitoring could become a valuable addition to delivery-room care. Until then, the honest and hopeful message is that this is science in progress: a promising idea being tested with the care it deserves, so that one day families can be told, with confidence, exactly what helps [1].

References

  1. Pichler G, Wolfsberger CH, Goeral K, et al. Two-year neurodevelopmental outcome in preterm neonates with cerebral oxygenation monitoring after birth: a multinational, multicenter retrospective follow-up study of the COSGOD III trial. Front Pediatr. 2026;14:1754084. doi:10.3389/fped.2026.1754084 ↩
  2. Wyllie J, Bruinenberg J, Roehr CC, et al. European Resuscitation Council Guidelines for Resuscitation 2015: section 7. Resuscitation and support of transition of babies at birth. Resuscitation. 2015;95:249–63. doi:10.1016/j.resuscitation.2015.07.029 ↩
  3. Madar J, Roehr CC, Ainsworth S, et al. European Resuscitation Council Guidelines 2021: newborn resuscitation and support of transition of infants at birth. Resuscitation. 2021;161:291–326. doi:10.1016/j.resuscitation.2021.02.014 ↩
  4. Kim E, Nguyen M. Oxygen therapy for neonatal resuscitation in the delivery room. Neoreviews. 2019;20(9):e500–12. doi:10.1542/neo.20-9-e500 ↩
  5. Oei JL, Finer NN, Saugstad OD, et al. Outcomes of oxygen saturation targeting during delivery room stabilisation of preterm infants. Arch Dis Child Fetal Neonatal Ed. 2018;103(5):F446–54. doi:10.1136/archdischild-2016-312366 ↩
  6. Baik N, Urlesberger B, Schwaberger B, et al. Cerebral haemorrhage in preterm neonates: does cerebral regional oxygen saturation during the immediate transition matter? Arch Dis Child Fetal Neonatal Ed. 2015;100(5):F422–7. doi:10.1136/archdischild-2014-307590 ↩
  7. Fuchs H, Lindner W, Buschko A, et al. Brain oxygenation monitoring during neonatal resuscitation of very low birth weight infants. J Perinatol. 2012;32(5):356–62. doi:10.1038/jp.2011.110 ↩
  8. Katheria AC, Harbert MJ, Nagaraj SB, et al. The Neu-Prem trial: neuromonitoring of brains of infants born preterm during resuscitation. J Pediatr. 2018;198:209–13. doi:10.1016/j.jpeds.2018.02.065 ↩
  9. Wolfsberger CH, Pichler-Stachl E, Höller N, et al. Cerebral oxygenation immediately after birth and long-term outcome in preterm neonates — a retrospective analysis. BMC Pediatr. 2023;23(1):145. doi:10.1186/s12887-023-03960-z ↩
  10. Pichler G, Urlesberger B, Baik N, et al. Cerebral oxygen saturation to guide oxygen delivery in preterm neonates for the immediate transition after birth: a 2-center randomized controlled pilot feasibility trial. J Pediatr. 2016;170:73–8. doi:10.1016/j.jpeds.2015.11.053 ↩
  11. Pichler G, Goeral K, Hammerl M, et al. Cerebral regional tissue oxygen saturation to guide oxygen delivery in preterm neonates during immediate transition after birth (COSGOD III): multicentre randomised phase 3 clinical trial. BMJ. 2023;380:e072313. doi:10.1136/bmj-2022-072313 ↩
  12. Stoll BJ, Hansen NI, Bell EF, et al. Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993–2012. JAMA. 2015;314(10):1039–51. doi:10.1001/jama.2015.10244 ↩
  13. Manuck TA, Rice MM, Bailit JL, et al. Preterm neonatal morbidity and mortality by gestational age: a contemporary cohort. Am J Obstet Gynecol. 2016;215(1):103.e1–103.e14. doi:10.1016/j.ajog.2016.01.004 ↩
  14. Norman M, Nilsson D, Trygg J, Håkansson S. Perinatal risk factors for mortality in very preterm infants — a nationwide, population-based discriminant analysis. Acta Paediatr. 2022;111(8):1526–35. doi:10.1111/apa.16356 ↩