Is oxygen good for premature infants or bad? The answer is…YES!

Lawrence Rhein, MD, is the director of the Center for Health Infant Lung Development and member of the divisions of Newborn Medicine and Pulmonology & Respiratory Diseases at Boston Children’s Hospital.

How much oxygen is too little, and how much is too much?

We all need reasonable oxygen levels to function appropriately, but this is especially true of premature infants, whose immature lungs and growing tissues are exquisitely sensitive to oxygen levels.

Some diseases or conditions, like lung disease of prematurity, can result in lower oxygen levels and potentially cause serious, even fatal, harm. We can provide supplemental oxygen, but this too can have significant side effects, especially in premature infants. If given in excess, oxygen can cause injury to an infant’s eyes (in the form of retinopathy of prematurity) and lungs.

As a result, finding the right target balance of oxygen delivery, while crucial, can be challenging. Too much oxygen is bad, but so is too little.

As noted recently in The New England Journal of Medicine, three large multicenter trials—the Benefits of Oxygen Saturation Targeting II (BOOST II) study, the Surfactant, Positive Pressure, and Pulse Oximetry Randomized Trial (SUPPORT) and the Canadian Oxygen Trial (COT)—have attempted to find the best oxygen saturation levels to maintain in premature infants. Each of these studies used similar methods, relying on the use of noninvasive sensors to display and record oxygen levels.

Foot of a premature infant in intensive care with an oxygen sensor taped onInfants in each of the trials were randomly assigned to have their oxygen levels tracked using one of two sets of special monitors. One set was calibrated such that they displayed oxygen levels that were several points higher than the actual values; the other set displayed oxygen levels that were several points lower.

Thus, while the bedside caregivers from both groups were told to target the same oxygen levels, 88 to 92 percent, there were two treatment groups, one with a higher target oxygen level and one with a lower level.

All three studies showed that infants with lower target oxygen levels were at an increased risk of death compared with the infants who had higher target oxygen levels, although the results were only statistically significant in two of the three studies. All studies also showed that the infants who had lower oxygen targets also had less eye disease.

What does this mean for the clinician working at premature child’s bedside? To answer that question, we must first recognize the limitations of these large studies:

  1. All of the studies used the same oxygen saturation targets throughout the infants’ stay in the neonatal intensive care unit (NICU). However, the appropriate target likely changes over the course of the few weeks and months that an infant is in the NICU.
  2. Maintaining an infant’s oxygen saturation within a target range is extremely difficult. Some of the studies were able to keep the infants in tighter ranges than others, which likely contributed to some of the slight differences in the results among the three studies.

For now, until new data become available and final analyses of the current trials are completed, it seems that neonatologists’ best approach would be to avoid extreme oxygen levels (below 90 percent or greater than 95 percent). The specific oxygen targets likely should depend on the capability of the NICU to stay within target levels, but the main underlying principle is to minimize high or low extremes of oxygen levels.

Researchers in our NICU at Boston Children’s Hospital are leading or participating in trials to find alternative, innovative ways of increasing oxygen levels without oxygen toxicity. Other studies are studying different examples of low-oxygen events in infants and children so as to distinguish those needing treatment from those that can be safely monitored without the risks of providing extra oxygen.

If too little oxygen is bad and too much oxygen is bad, one thing seems clear:  accurate oxygen level assessment is critically important. For many different diseases, correlation between oxygen and later outcomes will continue to be an important tool for determining when the benefits of oxygen therapy outweigh its risks.