The effect of preexisting pulmonary vascular disease on the response to mechanical ventilation with PEEP following open heart surgery

Abstract

The effects of mechanical ventilation with and without positive end expiratory pressure (PEEP) on hemodynamic performance and blood gas exchange were studied in 10 patients following open heart surgery. Ventilation at constant tidal volume (15 ml/kg body weight) with 10 cm H2O PEEP following aortic valve replacement (AVR) in 5 patients without pulmonary vascular disease, was associated with the following significant changes: a rise in arterial P(O2), a fall in the alveolar arterial P(O2) gradient when FI(O2) = 1.0, decreases in calculated Q(s)/Q(t) (pulmonary right to left shunt) and cardiac index. Using a similar pattern of ventilation following mitral valve replacement (MVR) in patients with elevated pulmonary vascular resistance, the authors found a significant decrease in cardiac index (but less than in the AVR group), a significant elevation of calculated physiologic deadspace (V(D)/V(T) and no change in Q(s)/Q(t). An hour after removal of PEEP, intravascular pressures, blood flow and blood gas exchange values of all patients with AVR had returned to control levels; patients with MVR had persistently significantly low cardiac indices, while V(D)/V(T) returned to pre PEEP values. These findings suggest that evaluation of responses to different ventilation patterns must take into account preexisting V/Q abnormalities secondary to pulmonary vascular disease, particularly when these are secondary to chronic congestive heart failure. Following AVR, Q(s)/Q(t) changed in the same direction as cardiac index (CI) irrespective of ventilatory patterns. The authors conclude that with increasing severity of pulmonary vascular disease, changes in airway pressure will have an unpredictable effect on cardiac index unless the level of myocardial competence is taken into account. In the presence of ventricular failure, changes in pleural (and therefore transmural) pressures will be minimal compared with the high filling pressures and exert no influence on stroke volume. Although pulmonary venous hypertension was more pronounced in the MVR than in the AVR group, there was no significant difference between the postoperative values for Q(s)/Q(t) FI(O2) = 1.0), a condition probably fostered by marked differences in preexisting V/Q.