The Oxygen Number on Your Baby's Monitor: How Decades of Research Settled One of Neonatology's Hardest Questions

For Families and General Readers

That Number on the Screen

If your baby is in the neonatal intensive care unit, you have almost certainly fixed your eyes on the oxygen saturation reading — the percentage number on the monitor that rises and falls with every breath. Nurses watch it constantly. Alarms sound when it drops below a threshold. It may have become one of the most anxiety-producing numbers in your life.

What you may not know is that for most of the history of neonatal medicine, doctors genuinely did not agree on which number to aim for. Too high carried one set of risks. Too low carried another. Finding the right range required decades of research — including two major clinical studies, published in 2010 and 2017, that finally settled the question with enough confidence to set a global standard.

Why Oxygen Is Both Medicine and Risk

Premature babies need supplemental oxygen because their lungs are not yet developed enough to extract sufficient oxygen from room air on their own. But oxygen is not a neutral support; it is a biologically active substance that, in excess, damages immature tissue. Two organ systems are at particular risk.

The first is the retina — the light-sensing layer at the back of the eye. Premature infants exposed to high oxygen levels can develop abnormal blood vessel growth in the retina, a condition called retinopathy of prematurity (ROP). In severe cases, ROP can lead to retinal detachment and blindness. This risk was first recognised in the 1940s, when a surge in premature infant survival was accompanied by an epidemic of infant blindness.

The second is the brain and body. Too little oxygen causes the kind of harm you would expect: cells die, organs fail, babies deteriorate and can die. Restricting oxygen too aggressively in the 1950s to reduce blindness caused a measurable increase in infant deaths. The field had swung too far in the opposite direction [1].

The challenge for neonatologists has always been finding the narrow range where the baby gets enough oxygen to survive and develop, without tipping into the zone where oxygen itself becomes harmful to the eyes and lungs. Excess oxygen also damages the premature lung itself — contributing to a chronic lung disease called bronchopulmonary dysplasia (BPD), which is now understood as a failure of normal lung growth caused in part by oxygen toxicity [8]. This is why how babies are supported at the moment of birth matters too: moving to gentle mask-based breathing support (CPAP) rather than inserting a breathing tube changes how much supplemental oxygen the baby needs from the very first breath [9].

The Long Road to a Trial

Pulse oximetry — the clip-on sensor that reads oxygen saturation through the skin — became standard in NICUs in the 1980s. For the first time, nurses could monitor saturation continuously without taking blood. But the monitors revealed something uncomfortable: different hospitals were targeting completely different saturation ranges. Some aimed for 85-92%. Others kept babies at 94-99%. There was no consensus, and no randomised trial to create one.

By the early 2000s, a group of researchers across the UK, Australia, New Zealand, Canada, and the United States decided this had to change. They formed a collaboration called NeOProM — the Neonatal Oxygenation Prospective Meta-analysis — and planned five parallel clinical trials that would all answer the same question: is it safer to target 85-89% or 91-95% saturation in the most premature babies? The plan was to combine all five trials' data at the end for a definitive answer [2].

The SUPPORT Trial: The First Answer (2010)

The first of the five trials to report was SUPPORT, led by Dr Waldemar Carlo and published in the New England Journal of Medicine in 2010 [1]. It enrolled 1,316 babies born between 24 and 27+6 weeks of pregnancy at 20 major US hospitals. Each baby was randomly assigned to a target saturation range of either 85-89% or 91-95%.

The results were striking, and they produced a clear warning:

Babies in the lower saturation group died more often. About 1 in 5 babies in the 85-89% group died before leaving hospital (19.9%), compared to about 1 in 6 in the 91-95% group (16.2%). That 3.7 percentage point difference — representing roughly 1 additional death for every 27 babies assigned to the lower range — was statistically significant [1,9].

Babies in the lower saturation group had much less severe eye disease. The rate of severe ROP was nearly halved in the lower group (8.6% vs 17.9%). This confirmed the biological intuition: less oxygen exposure protects the retina.

So the lower target produced fewer babies with severe eye disease — but more babies died. The SUPPORT investigators urged caution about using the lower range as a clinical standard.

The NeOProM Synthesis: The Definitive Answer (2017)

By 2017, all five NeOProM trials had completed, and the combined data on 4,965 extremely premature babies — the largest dataset ever assembled for this question — was analysed and published in JAMA [2].

The headline finding was carefully phrased: there was no statistically significant difference in the combined outcome of death or serious disability at 18-24 months between the two groups. Some headlines reported this as "no difference found." That is a misreading.

What the data actually showed was this:

Lower target (85-89%): More deaths. Significantly higher risk of dying — about 17% more deaths compared to the higher group [2]. And significantly more cases of necrotising enterocolitis (NEC) — a devastating bowel disease unique to premature babies — about 33% more cases [2].

Higher target (91-95%): More severe ROP — but importantly, a treatment now exists. An injection of a drug called bevacizumab directly into the eye is highly effective for severe ROP [3]. When severe eye disease became treatable, the balance of the argument shifted decisively toward the higher oxygen target, which prevents deaths and NEC without condemning babies with eye disease to blindness [3].

The reason the combined "death or disability" figure was not statistically different is a technical one: babies who died in the lower-target group did not count as having a disability, and babies who survived with less ROP in the lower-target group had lower measured disability. Mathematically, these effects partially cancelled. But from the perspective of a family — and from the perspective of every neonatologist making a real-time clinical decision — the choice between more deaths and more treatable eye disease has a clear answer [2].

What Changed in Hospitals Around the World

The SUPPORT trial's mortality signal in 2010, confirmed and amplified by the BOOST II trials in 2013 [4] and the full NeOProM synthesis in 2017, drove a global convergence on the 91-95% target — one piece of a broader improvement in outcomes for extremely premature babies documented over the 1993-2012 period [7]. Two of the BOOST II trials were stopped early by their safety committees — an unusual and serious step — when interim data confirmed that the lower target was causing more deaths [4].

By 2020, international resuscitation guidelines formally endorsed 91-95% as the standard saturation range for premature infants receiving supplemental oxygen [5]. They also changed the approach at birth itself: premature babies should now be resuscitated starting with low concentrations of supplemental oxygen, not pure oxygen, because even in the first minutes of life, very high oxygen concentrations cause measurable oxidative damage to immature lungs [6].

Today, when a nurse adjusts the FiO2 on your baby's ventilator or oxygen supply to keep the saturation between 91% and 95%, that decision is backed by the most rigorous evidence in neonatal medicine — evidence assembled through decades of work, five international clinical trials, and the lives of nearly 5,000 enrolled infants [2].

What This Means for Your Baby's Care

When you see the saturation alarm sound — because the number has dropped below 90% or risen above 96% — the nurse is managing a normal event in a medically intentional way. The target range is set where it is because evidence shows it saves the most lives and causes the least preventable harm. Not every moment will be within range; premature babies have variable breathing patterns. The goal is keeping the average in the right place over time [1,9].

If your baby is diagnosed with ROP and requires treatment, intravitreal bevacizumab is now the most common treatment for severe disease and is highly effective [3]. The oxygen question and the eye disease question now have evidence-based answers for both.

Ongoing research is exploring whether computer-controlled oxygen adjustment — where an algorithm adjusts the FiO2 automatically, faster than a nurse can respond — might keep babies more consistently within the target range, potentially improving outcomes further. The story of neonatal oxygen targeting is not finished; it is evolving toward more precision with every generation of evidence.

References

  1. Carlo WA, Finer NN, Walsh MC, et al. Target Ranges of Oxygen Saturation in Extremely Preterm Infants. N Engl J Med. 2010;362(21):1959-1969. doi:10.1056/NEJMoa0911781
  2. Askie LM, Darlow BA, Finer N, et al. Association between Oxygen Saturation Targeting and Death or Disability in Extremely Preterm Infants. JAMA. 2017;318(24):2345-2357. doi:10.1001/jama.2017.19218
  3. MintzHittner HA, Kennedy KA, Chuang AZ; BEAT-ROP Cooperative Group. Efficacy of Intravitreal Bevacizumab for Stage 3+ Retinopathy of Prematurity. N Engl J Med. 2011;364(7):603-615. doi:10.1056/NEJMoa1007374
  4. Tarnow-Mordi W, Stenson B, Kirby A, et al. Outcomes of Two Trials of Oxygen-Saturation Targets in Preterm Infants. N Engl J Med. 2016;374(8):749-760. doi:10.1056/NEJMoa1514212
  5. Wyckoff MH, Wyllie J, Aziz K, et al. Neonatal Life Support: 2020 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science. Circulation. 2020;142(Suppl 1):S185-S221. doi:10.1161/CIR.0000000000000895
  6. Vento M, Moro M, Escrig R, et al. Preterm Resuscitation with Low Oxygen Causes Less Oxidative Stress, Inflammation, and Chronic Lung Disease. Pediatrics. 2009;124(3):e439-e449. doi:10.1542/peds.2009-0434
  7. 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):1039-1051. doi:10.1001/jama.2015.10244
  8. Jobe AH. The New BPD: An Arrest of Lung Development. Pediatr Res. 1999;46(6):641-643. doi:10.1203/00006450-199912000-00007
  9. Morley CJ, Davis PG, Doyle LW, et al. Nasal CPAP or Intubation at Birth for Very Preterm Infants. N Engl J Med. 2008;358(7):700-708. doi:10.1056/NEJMoa072788