Carbon Dioxide Tensions During Anesthesia in the Prone Position

Abstract

The changes in lung function after placing patients in the prone position during general anesthesia have recently been carefully reexamined [1,2]. In these studies, the authors reported that the prone position markedly increased functional residual capacity (FRC) and improved oxygenation, whether the patients were lean [1] or obese [2]. They also found that PaCO2 values were unchanged. They did not, however, measure PCO2. The latter is, interalia, used as a noninvasive index of PaCO2[3] during anesthesia and intensive care. The purpose of our study was to provide the missing information regarding PCO2 and, hence, the PaCO2-PETCO2 gradient. Method We studied 20 ASA physical status I or II patients, 4 women and 16 men of average weight and height, scheduled for elective lumbar decompression and fusion. The study was approved by the clinical research committee of our hospital. Anesthesia was induced by IV anesthetics (thiopental/propofol) after an initial dose of narcotics (fentanyl/sufentanil). Tracheal intubation was facilitated by paralysis induced by succinyl choline. Anesthesia was then maintained by N2 O/O2 (without changes in inspired concentrations or fresh gas flow rates) plus narcotic infusion and either isoflurane or desflurane delivered by a partial rebreathing system with CO2 absorption. We used an Ohmeda 8000 anesthesia machine with an Ohmeda 7000 ventilator (Ohmeda, Madison, WI). Muscle paralysis was achieved by using pancuronium bromide. The ventilator settings were 10 mL/kg at a frequency of 10-12 bpm and was maintained throughout. Patients were positioned after achieving cardiovascular stability. The patients were placed on an orthopedic table (Andrews SST 3000; Orthopedic Systems Inc., Haywood, CA) that supports the trunk by two padded bolsters running from the shoulder tips to the anterior superior iliac spines, thus allowing the anterior thorax and abdomen to move unhampered during ventilation. The hips and knees are flexed at 90[degree sign]. The measurements were made just before and shortly after placing patients in the prone position, when the blood pressure and pulse rate were stable. The time lapse between induction or positioning and sampling was 10-12 min. In all instances, blood pressure and pulse rate were stable for 5-10 min. We excluded any patients with a rising phase III in the capnograph. The time interval between measurements was approximately 15 min. We did not, therefore, expect any change in patient temperature. We calculated the average PETCO2 of three to five consecutive breaths during quasi-simultaneous arterial blood sampling from an indwelling arterial catheter. Expired gases were analyzed for PCO2 by using an Ohmeda 5200 analyzer (side-arm) calibrated in the range of 0-55 mm Hg, and the arterial blood sample was analyzed by using a Mallinckrodt GEM Premier (Ann Arbor, MI) calibrated in the range of 35-79 mm Hg. The data were analyzed by using the two-tailed Student's t-tests for paired groups. Results Because of variations in the inspired oxygen concentration (FIO2), we present the oxygenation data as PaO2/FIO2 to standardize comparisons. The latter increased significantly from 413 +/- 25.2 to 483 +/- 19.5 (P = 0.002). PaCO2 did not change (36 +/- 1.4 vs 35.8 +/- 1.0 mm Hg). End-tidal PCO2 decreased significantly from 32.4 +/- 1.0 to 30.4 +/- 0.8 mm Hg (P = 0.03). Figure 1 gives the PaCO2-PETCO2 data. The gradient increased significantly from 3.7 +/- 0.9 to 5.9 +/- 0.6 mm Hg (P < 0.02).Figure 1: PaCO2-PETCO 2 data.Discussion Our report confirms that oxygenation is improved, and that PaCO2 is not affected, by placing patients prone. End-tidal PCO2 was significantly reduced, thus creating an enlarged PaCO2-PETCO2 gradient. The induction of general anesthesia results in an immediate reduction in FRC, which adversely affects oxygenation by two mechanisms [4]. The first is the creation of true shunt due to compression atelectasis. This compression atelectasis is not reversed by the prone posture. The second is an increase in ventilation to perfusion scatter and the increased influence of airway closure. Placing patients in the prone position restores much of the reduction in FRC [1,2] and improves oxygenation in surgical [1,2] and intensive care patients [5]. The most likely reason for the improvement is reduced ventilation-perfusion mismatch. The decreased PETCO2 and increased PCO2 gradient that we report can be either due to increased dead space ventilation or reduced cardiac output. The increased excursion of the chest wall can increase tidal volume and blow off CO2, thus reducing PETCO2. On the other hand, reduced cardiac output also can increase the dead space effect. The reduced cardiac output is the result of anesthesia and the prone position with the lower limbs below the level of the heart. Our data do not allow us to differentiate between the two mechanisms. Our study shows that PaCO2-PETCO2 enlarges in the prone position, but does not indicate the possible changes with time thereafter. In summary, we report that PaCO2-PETCO (2) enlarges after placing anesthetized patients in the prone position. We thank Drs. R. Covert and S. Vilderman for their help with the measurements and Mrs. S. Scholl for her work on the manuscript.