Oxygen and Resuscitation: Beyond the Myth

Abstract

Perhaps nothing in medicine is more steeped in myth than oxygen. From its initial success in treating pulmonary tuberculosis to its failure in treating cholera, oxygen has fallen in and out of style. Only gradually did oxygen gain favor as a treatment for respiratory distress with even the likes of Osler doubting its effectiveness. But by the early 1900s, treatment of patients with pulmonary parenchymal disease was so convincing that a lack of appreciation for oxygen’s potentially harmful effects led to its eventual widespread acceptance.a We started using oxygen for resuscitation because it seemed like a good idea. Now we use it because we always have.The use of 100% oxygen in resuscitation is intuitively appealing. Maximizing the gradient driving oxygen into hypoxic cells may speed recovery. However, it is well known that long-term exposure to 100% oxygen is toxic. The proponents of oxygen therapy argue that toxicity simply does not occur during the short period of resuscitation. Although this may be true with respect to effects like fibrosis and pulmonary edema, the real concern is for reperfusion injury. In reperfusion injury, hypoxic cells appear to undergo metabolic changes that prime them to create free radicals when oxygen is reintroduced.1–3 Experimental resuscitation with 100% oxygen has been associated with a variety of concerning physiologic changes when compared with room air resuscitation. They include increased generation of oxygen radicals,4 decreased central nervous system sodium-potassium ATPase function,5 and decreased dopamine metabolism.6 Although there is no data on the safety or danger of short-term exposure to 100% oxygen, the reperfusion injury literature suggests that resuscitation from hypoperfusion states (ie, sudden cardiac arrest, hypovolemic, or distributive shock) is precisely the time when oxygen toxicity from free radicals is likely to occur.Given the potential harm of supplemental oxygen use during resuscitation, it would seem that there must be some data to support its use. In fact, there is not. Recently, an argument was made supporting the use of supplemental oxygen during resuscitation,7 but the studies cited were not intended to specifically address that question. They do address other interesting aspects of resuscitation such as the importance of establishing circulation,8 the importance of gasping,9 and the potential harm of having carbon dioxide in the resuscitation gas.10 Unfortunately, none has an appropriate control group for making conclusions about the benefit of supplemental oxygen over room air for resuscitation.Animal studies have been conducted specifically to look at the appropriateness of using 100% oxygen rather than room air during resuscitation. Results show either no outcome advantage or the potentially harmful effects mentioned above. There is experimental animal evidence that, given a healthy pulmonary system, only 15% to 18% oxygen is needed for a successful resuscitation.11 This lends some scientific support to the successes of the age-old practice of mouth-to-mouth resuscitation, with 16% to 18% oxygen,12 and certainly to the practice of room air (21%) resuscitation.Although there is a paucity of animal data, there are actually 4 human studies comparing the use of room air to 80% or 100% oxygen.13–16 These are studies of the newborn, so the results do not clearly apply to the adult. Newborn resuscitation differs from adult resuscitation, in that the vast majority of newborn resuscitations involve reversal of vagal-mediated bradycardia, which is accomplished by pulmonary stretch.17 Some oxygen is needed to reverse the hypoxia-induced apnea and support the metabolic demands of the body once the vagal tone is lifted. The newborn resuscitation trials show that room air is completely sufficient for this purpose. In fact, compared with room air, resuscitation with 100% oxygen delays the onset of spontaneous respiration,15,16 and leads to a prolonged oxidative stress that, remarkably, is still apparent at 1 month of age.16Despite the short-term outcome differences, the newborn resuscitation trials all showed the same ultimate clinical outcome whether infants were resuscitated with room air or 80% to 100% oxygen. However, the authors of the trials had hypothesized that by avoiding exaggerated reperfusion injury, they could actually improve ultimate outcome. There are several possible reasons why they were unsuccessful. First, newborns may be protected from the harmful systemic effects of brief 100% oxygen exposure because of the residual right-to-left shunting of transition. In the RESAIR 2 study,15 there were no differences in oxygen saturation during the first 10 minutes of life whether infants were resuscitated with room air or 100% oxygen. In both cases, oxygen saturations started in the mid 60s at 1 minute and by 10 minutes were still only in the low 90s. However, it is important to note that while the rest of the body may be protected from hyperoxia by this mechanism, the lungs certainly are not. Furthermore, this possible protective mechanism is not present outside the immediate newborn period, so lack of adverse outcome in the newborn with 100% oxygen does not predict lack of adverse outcome for the older child or adult.Second, most term and near-term infants who require resuscitation in the delivery room probably do not undergo ischemia and reperfusion. Infants go through a series of protective steps in response to hypoxia designed to maintain oxygen delivery to the “essential” organs and decrease oxygen consumption elsewhere.18 Each of these steps can be reversed at birth by the pulmonary stretch and oxygen delivery of positive pressure ventilation.17 It is not until hypoxia is profound and these protective mechanisms are overwhelmed that there is circulatory collapse and shock.18 During neonatal resuscitation, failure to respond to adequate positive pressure ventilation is the indicator of cardiovascular compromise and the need to begin circulatory resuscitation.19 However, most term and near-term infants can be resuscitated with positive pressure alone, indicating that they are not in shock. Without serious circulatory compromise, there is probably not significant reperfusion, or reperfusion injury.Finally, these studies essentially excluded the group of infants that might have had the most benefit from room-air resuscitation. In these trials, there was a sub-population of infants that failed to respond to simple vagolytic ventilation. Some of these infants had probably progressed beyond protective mechanisms and into shock. Based on the concept of reperfusion injury, these are precisely the infants that might have had an improved outcome by avoiding excess oxygen when circulation was restored. But by design these infants were crossed over to receive 100% oxygen as resuscitation continued. Thus, in the subpopulation of infants that may have shown the biggest outcome difference between room air and 100% oxygen resuscitation, all the infants were resuscitated with 100% oxygen.At this point in time, there is no evidence of a role for oxygen concentration above 21% during resuscitation of the newborn, or for those older children and adults with lungs that were functioning adequately on room air before an arrest. In fact, there may be significant harm, based on the concept of reperfusion injury.Despite that, only now and only in the resuscitation of the newborn is this idea currently being challenged,20–22 20 years after the concept of room air resuscitation was introduced.23 The pediatric working group of the International Liaison Committee on Resuscitation states that although room air resuscitation makes sense, there is insufficient evidence of benefit over 100% oxygen resuscitation to change current guidelines.21 The World Health Organization, which developed guidelines for resuscitating newborns in areas with limited resources, states that room air is sufficient for resuscitation.22In the end, it may turn out that some supplemental oxygen during resuscitation has a beneficial role. However, we are still awaiting appropriate human studies that look at the role of supplemental oxygen as compared with room air for resuscitation outside the delivery room, where the newborn data are clear. There are physiologic differences between being resuscitated on room air or 100% oxygen, but to the limits of our ability to measure outcomes, neither is clearly superior,14–16,24 and the markers of prolonged oxidative stress are a significant concern.16The real question is this: in our age of evidence-based medicine, should we continue to use this historical, yet unsupported, therapy until it is proven harmful? Or, should we step back and recognize that supplemental oxygen is a medicine with potentially significant side effects that should be used only when there is an indication? Until we accept that the only reason we so strongly cling to supplemental oxygen as a therapy in resuscitation is because we were taught simplistically that “oxygen is good,” with the implication that “more must be better,” we are doomed to perpetuate the myth.I would like to thank Dr Russell R. Moores, Jr, for his invaluable editorial assistance during the preparation of this manuscript.