Hyperoxia-induced hypocapnia: an underappreciated risk.

A dministration of supplementary O2 is considered to be safe, as exemplified by one editorial comment1: “Oxygen should be used as soon as possible, in as near 100% as possible in all resuscitation situations, and for the early management of injury and illness. Its use will never disadvantage [our emphasis] a patient under these circumstances.” We believe this claim merits examination. The rationale for administering O2 is that it increases the O2 content of blood and, therefore, O2 delivery to tissues. In a healthy person, hemoglobin is nearly saturated, and switching from air to pure O2 at sea level will increase O2 content by 10% due almost exclusively to the increase in O2 dissolved in the plasma. In these people, the more influential determinant of O2 delivery is tissue perfusion that is determined by perfusion pressure and local tissue vascular resistance. Vascular resistances in the brain, heart, and placenta are affected by the Pco2 in arterial blood. At issue is the existence of a strong link between hyperoxia and arterial Pco2. Oxygen has long been known, but seldom recognized, to be a respiratory stimulant resulting in hypocapnia in adults2–6 and infants,7–9 a result confirmed by more recent work.10 While oxygen may not have this effect in patients with a limited ability to increase ventilation because of disease,11 it can be expected to cause at least some hyperventilation in the vast majority of patients. This raises the possibility that hyperoxia-induced hypocapnia would cause vasoconstriction of CO2-resposive vascular beds and paradoxically exacerbate ischemia there, if present. Furthermore, the hypocapnia increases the affinity of hemoglobin for O2, reducing O2 unloading to tissues. These effects are known from basic physiologic principles but are seldom emphasized in clinical texts or taken into account when O2 administration is prescribed. The link between hyperoxia and hyperventilation can be explained by the Haldane effect. Oxygenated hemoglobin binds less CO2 (the Haldane effect); therefore, CO2 transport must be maintained by increases in bicarbonate and dissolved CO2, the latter increasing local tissue Pco2. In most tissues, this is of no consequence; but in the brainstem, the location of the central chemoreceptors responsible for most of the respiratory drive,12 the increased Pco2 and, more importantly, H 14 stimulate these receptors, increasing ventilation. This increase is greater if not blunted by the resulting arterial hypocapnia.10,15 Figure 1 illustrates this effect in a typical subject breathing sequentially air, O2, and O2 with Pco2 returned to and maintained at control values. We now briefly discuss several clinical situations in which hyperoxia-induced hypocapnia may paradoxically not improve—or even worsen—tissue oxygenation.