To the Editor:-Cruz et al.  provide interesting data on the physiologic consequences of carbon dioxide (CO2) insufflation in pregnant ewes, but their conclusion, “capnography may be an inadequate guide to ventilation during carbon dioxide pneumoperitoneum in the pregnant patient,” merits reexamination.
Three issues we believe deserve attention. First, CO2pneumoperitoneum has been shown (two separate studies: see fig. 1, Cruz et al. ; Hunter et al. ) to increase arterial to end-tidal CO2gradient (PaCO2-PETCO2) in pregnant ewes by approximately 10 mmHg. Thus, it seems that capnographic data might be useful in adjusting minute ventilation during pneumoperitoneum, allowing for ETCO2to be maintained at a level 10 mmHg lower than preinsufflation ETCO2.
Second, Cruz et al. attributed increases in PaCO2-PETCO sub 2 (the gradient) to increases in alveolar deadspace without addressing other potentially important factors. The relation between PaCO2and PETCO2, for example, is altered by changes in the pattern of ventilation and by effects of pneumoperitoneum on the alveolar plateau (slope of phase III) of the capnogram. [3–5] Changes in the gradient correlate with changes in alveolar deadspace when phase III is nearly horizontal (minimal slope). [3,6] However, when phase III shows a steeper slope,  the gradient also depends on factors that influence the slope of phase III, such as ventilatory characteristics of alveoli. [3–5,7] For example, a lowering of total thoracic compliance (as in pregnancy and pneumoperitoneum) can alter the dynamics of alveolar emptying, causing a flattening of the initial portion of phase III and elevation of the terminal portion of the expiratory capnogram. [3,8] If, however, rate of ventilation is increased (method of Cruz et al. to prevent pneumoperitoneum-induced hypercarbia) when thoracic compliance is low,  there may be insufficient time for gases of alveoli with the highest CO2(responsible for terminal increase in alveolar plateau) to reach the CO2sensor at end-expiration. [3,8] This creates PETCO2samples with lower PCO2than end-expiratory alveolar gases. Cruz et al. could have tested the adequacy of PETCO2sampling by intermittently administering large volume breaths (squeeze PCO2) to determine whether values of PETCO2higher than those reported in their study could be obtained. 
Finally, one should be cautious about using capnographic data from gravid ewes to draw inferences about parturients. During general anesthesia, for example, the PaCO2-PETCO2in pregnant ewes ranges from 6 to 15 mmHg, [1,2] whereas the gradient in pregnant humans varies from -1 to 0.75 mmHg. [9,10] PETCO2often exceeds PaCO2in anesthetized pregnant women, [9,10] which may relate in part to a steeper slope of phase III in capnograms of pregnant versus nonpregnant subjects. [5,10] Thus, it seems conceivable that the consequences of pneumoperitoneum could differ in humans and sheep. In contrast to the recommendation of Cruz et al.,  Steinbrook et al.  used ETCO sub 2 (32–36 mmHg) to guide ventilation during laparoscopic operations in pregnant women (9–30 weeks gestation). In that study, there were no untoward effects on the mother or baby.  Thus, we believe that conclusions on the utility of capnography in pregnant patients undergoing laparoscopy should await appropriate clinical investigation, which include simultaneous measurements of PaCO2and PETCO2.
K. Bhavani Shankar, M.D.
Phillip S. Mushlin, M.D., Ph.D.
Department of Anaesthesia; Harvard Medical School; Brigham and Women's Hospital; 75 Francis Street; Boston, Massachusetts 02115
(Accepted for publication August 24, 1997.)