We appreciate the interest shown by Drs Garite, Nageotte, and Parer in our Clinical Opinion, particularly because individually and collectively they have done much to advance our understanding of fetal monitoring. We agree with some of the excellent points they raise but disagree with others. Rather than recapitulating the data, observations, and suggestions from our Clinical Opinion, we will respond to their Viewpoint in a point-by-point manner.

We agree that the practice of administering oxygen to laboring women, as they pointed out, is “vastly overdone” and may add “to other unnecessary and unwarranted interventions” but were somewhat perplexed by their reference to the paper by Clark et al as support for the contention that certain fetal heart rate patterns may benefit from oxygen administration. In the body of that paper (of which Dr Rouse is a coauthor), the only reference to maternal oxygen administration is negative: “no evidence exists to support the efficacy of maternal oxygen administration in commonly achievable concentrations in increasing fetal tissue oxygenation or in improving newborn outcomes, regardless of oxygen concentration. …”

Garite et al argue that the literature we cite suggesting maternal oxygen administration may lower pH is “weak,” “counterintuitive,” and not consistent with data from nonrandomized trials. We agree that the literature is not robust (and in fact point this out in our paper [page 125, paragraph 2]). However, the nonrandomized trials to which they refer, both of which we cited in our Opinion, are hardly compelling.

In the paper by Newman et al, the effect of administering 50-100% oxygen to women during the first stage of labor had a negligible effect on fetal scalp pH, which went from 7.28 ± 0.013 at baseline to 7.30 ± 0.01 with hyperoxygenation; in contrast, the fetal scalp base deficit actually worsened, from 2.5 ± 0.8 at baseline to 4.0 ± 1.0 with hyperoxygenation.

Khazin et al reported that maternal hyperoxia during labor was associated with an increase in fetal scalp pO ₂ . However, to the extent to which one can tell from the report (which is difficult), maternal oxygen administration seemed to have no effect on fetal scalp pH. To our mind, these 2 small, nonrandomized, and somewhat obscurely analyzed studies provide little support for maternal oxygen administration.

Garite et al take issue with our discussion of the study by Morishima et al, in which oxygen was administered for 30 minutes to nonhuman primates laboring with acidotic fetuses. In that study, the effect on fetal arterial pH was with minimal and went in the wrong direction with oxygen administration, from an average of 7.10 ± 0.029 to 7.07 ± 0.034. Garite et al argue that 30 minutes is not an adequate duration of time to reverse metabolic acidosis and likely explains why pH did not improve with oxygen supplementation. Although this may be true, it also may not, and optimism that longer periods of oxygenation will raise fetal pH, if unfounded, clearly has the potential to delay a delivery that should be expedited.

Although it may not be physiologically possible for fetal pO ₂ to exceed maternal venous pO ₂ , maternal hyperoxia clearly raises markers of free radical activity in not only maternal but also in fetal blood as well. In our Clinical Opinion, we discuss (and only speculatively) the potential negative effects of free radical exposure on the fetus. Garite et al question whether this is a realistic concern, citing the use of oxygen in the setting of surgery or trauma and the lack of apparent harms from this practice. However, neonates, particularly those born preterm, may be especially vulnerable to free radical damage because their antioxidant defense systems have yet to fully mature and thus may have a decreased capability to counter the effects of free radical activity. We merely raised the issue that oxygen-induced free radical generation may not be salutary for the fetus. We are not alone in this concern.

At some level, we do not disagree that maternal oxygen administration to improve the status of the fetus might make sense. But so too did home uterine activity monitoring and fetal pulse oximetry, to name but 2 of many examples in obstetrics in which theoretically sound interventions were found ineffective, and in some cases harmful, when finally properly evaluated.

And we should not forget that oxygen is being administered in response to a technology (electronic fetal heart rate monitoring) to which the US Preventive Services Task Force previously assigned a grade of D (ie, one more likely to do harm than good). It is not implausible that injudicious oxygen use might exacerbate this potential harm.

Given that we called for a randomized trial of maternal oxygen administration, we naturally agree that it would be prudent for practitioners to be more selective in identifying candidates for maternal oxygen administration. However, we are not sanguine that this will occur. Despite Association of Women’s Health, Obstetric and Neonatal Nurses recommendations, in our large-volume obstetric hospital, laboring women are frequently placed on oxygen for an abnormal fetal heart rate tracing without oxytocin being discontinued, and prolonged use of oxygen in labor is common. Thus, it does not seem at all far-fetched to us that faith in maternal oxygen therapy may delay an indicated intervention or result in continued labor stimulation when neither is in the best interests of the fetus. Out ultimate point was simply that a properly done clinical trial would, to paraphrase Winston Churchill, help us all to “pass with relief from the tossing sea of cause and theory to the firm ground of result and fact.”