Background: Hyperventilation frequently is employed to reduce carbon dioxide partial pressure in patients in the operating room and intensive care unit. However the effect of hypocapnia on oxygenation is complex and may result in worsening in patients with preexisting intrapulmonary shunt. To better define the interplay between hypocapnia and oxygenation, the effects of hypocapnia and hypercapnia on the matching of ventilation (VA) and perfusion (Q) were studied in dogs with oleic acid-induced pulmonary edema, using the multiple inert gas elimination technique.
Methods: Eight pentobarbital-anesthetized, closed-chested dogs were administered 0.06 ml/kg of oleic acid at least 150 min prior to study. Ventilation was set with an FIO2 of 0.90, a tidal volume of 20 ml/kg, and a respiratory rate of 35 breaths/min. The arterial carbon dioxide tension (PaCO2) was varied in a randomized order to three levels (26, 38, and 50 mmHg) by altering the amount of CO2 in the inspired gas mixture. Gas exchange was assessed by true shunt, dead space, the log standard deviation of the perfusion (log SDQ) and the ventilation (log SDV) distributions, and the tracer inert gas arterial-alveolar difference ([a-A]D) area.
Results: Cardiac output (4.1 +/- 0.4 L/min), mean pulmonary artery pressure (25 +/- 1 mmHg), inert gas shunt (22 +/- 3%), and dead space (38 +/- 4%) during normocapnia were not different from that during hypocapnia and hypercapnia. Hypocapnia increased (P < .05) the alveolar-arterial oxygen tension difference (P[A-a]O2) and decreased (P < .05) the arterial blood oxygen tension (PaO2, 181 +/- 33 mmHg vs. 221 +/- 35 mmHg with normocapnia). P[A-a]O2 and PaO2 were unaffected by hypercapnia. During hypocapnia, VA/Q inequality increased, demonstrated by an increase (P < .05) in log SDQ (1.60 +/- 0.15 vs. 1.35 +/- 0.19 with normocapnia) and in the [a-A]D area (0.63 +/- 0.09 vs. 0.50 +/- 0.09 with normocapnia) indexes of VA/Q heterogeneity. During hypercapnia, the [a-A]D area (0.63 +/- 0.11) and log SDV (1.13 +/- 0.10 compared to 0.97 +/- 0.12 with normocapnia) also were increased (P < .05). With hypocapnia, there was a small but insignificant increase in blood flow to shunt and low VA/Q areas (29 +/- 4% compared to 26 +/- 4% with normocapnia). In the presence of a high FIO2, this small increase in shunt and low VA/Q may result in a significant decrease in PaO2.
Conclusions: Both hypocapnia and hypercapnia were associated with an increased VA/Q inequality. However, PaO2 decreased and P[A-a]O2 increased with only hypocapnia. These results suggest that hyperventilation to reduce PaCO2 may be detrimental to arterial PO2 in some patients with lung disease.