Mechanical ventilation with permissive hypercapnia increases intrapulmonary shunt in septic and nonseptic patients with acute respiratory distress syndrome.

Objective: To compare the effects of conventional mechanical ventilation with low-volume, pressure-limited ventilation (LVPLV) and permissive hypercapnia on ventilation-perfusion (V/Q) distributions in patients with acute respiratory distress syndrome. We hypothesized that the advantageous cardiopulmonary effects of LVPLV would be greater in patients with sepsis than in those without sepsis.

Patients And Interventions: Twenty-two patients with acute respiratory distress syndrome were studied (group 1: 12 patients with hyperdynamic sepsis; group 2: 10 nonseptic patients). Intrapulmonary shunt (Qsp/Qt) (percentage of cardiac output), perfusion of "low" V/Q areas (percentage of cardiac output), ventilation of "high" V/Q areas (percentage of total ventilation [VE]), and deadspace ventilation (percentage of VE) were calculated from the retention/excretion data of six inert gases. Data were obtained during conventional mechanical ventilation and during LVPLV.

Measurements And Main Results: In group 1, LVPLV increased PaCO(0)rom 38 +/- 6 torr (5.1 +/- 0.8 kPa) to 61 +/- 12 torr (8.1 +/- 1.6 kPa). Qsp/Qt increased from 28 +/- 16% to 36 +/- 17%, whereas Pao2 (84 +/- 15 torr [11.1 +/- 2.0 kPa] vs. 86 +/- 21 torr [11.5 +/- 2.8 kPa]) and Qt (10.6 +/- 2.3 vs. 11.5 +/- 2.5 L x -1) remained unchanged and PVO(2) (40 +/- 4 [5.3 +/- 0.5 kPa] vs. 49 +/- 6 torr [6.5 +/- 0.3]) increased. In group 2, LVPLV increased PaCO(2) from 38 +/- 6 torr (5.1 +/- 0.8 kPa) to 63 +/- 11 torr (8.4 +/- 1.5 kPa). For Qsp/Qt (24 +/- 9% to 34 +/- 16%), the increase was not significant, whereas Qt (7.4 +/- 1.8 vs. 10.2 +/- 2.2 L x -1), PVO(2)(38 +/- 4 torr [5.1 +/- 0.5 kPa] vs. 50 +/- 6 mm Hg [6.7 +/- 0.8 kPa]), and PaO(2) (89 +/- 16 torr [11.9 +/- 2.1 kPa] vs. 98 +/- 19 torr [13.1 +/- 2.5 kPa]) increased. In both groups, the scatter of perfusion distribution (log SDQ) was greater than expected for normal subjects but was not different between the groups or altered by the treatments.

Conclusions: In patients with acute respiratory distress syndrome, LVPLV with permissive hypercapnia, tended to increase Qsp/Qt, without a concomitant decrease of PaO(2). This occurs because, although atelectasis and increased shunt result from the low ventilatory volume, the effects on PaO(2) are offset by increased PVO(2) resulting from the hypercapnic stimulation of cardiac output. This result was independent of the presence or absence of sepsis.