Background: Impaired gas exchange is a major complication after cardia
c surgery with the use of extracorporeal circulation. Blood gas analys
is gives little information on underlying mechanisms, in particular if
the impairment is multifactorial. In the current study we used the mu
ltiple inert gas technique with recordings of hemodynamics to analyze
the separate effects of intrapulmonary shunt (Q(S)/Q(T)), ventilation-
perfusion (V(A)/Q) mismatch, and low mixed venous oxygen tension on ar
terial oxygenation during cardiac surgery. Methods: V(A)/Q distributio
n was studied in nine patients undergoing coronary artery revasculariz
ation surgery. The obtained data related to V(A)/Q distribution were p
erfusion of lung regions with V(A)/Q < 0.005 (Q(S)/Q(T)) perfusion of
lung regions with 0.005 < V(A)/V(Q) < 0.1 (''low''-V(A)/Q regions), ve
ntilation of lung regions with 10 < V(A)/Q < 100 (''high''-V(A)/Q regi
ons), and ventilation of lung regions with V(A)/Q > 100 (dead space [V
(D)/V(T)]) In addition, arterial and mixed venous oxygen and carbon di
oxide tensions and systemic and pulmonary hemodynamics were analyzed.
Recordings were made before and after induction of anesthesia, after s
ternotomy, 45 min after separation from extracorporeal circulation, 4
h postoperatively during mechanical ventilation, and on the 1st postop
erative day during spontaneous breathing. Results: In the awake state,
Q(S)/Q(T) was 4 +/- 4%, and perfusion of low-V(A)/Q regions was 3 +/-
5%. The sum of Q(S)/Q(T) and low-(V(A)/Q units correlated with the al
veolar-arterial oxygen tension gradient (PA-a(O2)) (r = 0.63, P < 0.05
). After induction of anesthesia, Q(S)/Q(T) increased to 10 +/- 9% (P
= 0.069). Sternotomy had little effect on shunt, but Q(S)/Q(T) increas
ed to 22 +/- 8% (P < 0.01) after separation from extracorporeal circul
ation, which was correlated with a significantly higher PA-a(O2) (r =
0.77, P < 0.05). Postoperatively, gas exchange improved rapidly, as as
sessed by a decrease of PA-a(O2) from 341 +/- 77 to 97 +/- 36 mmHg (P
< 0.01) and a reduced Q(S)/Q(T) (5 +/- 4%, P < 0.05). On the 1st posto
perative day, arterial oxygen tension was significantly lower than pre
anesthesia values (58 +/- 6 vs. 68 +/- 8 mmHg, P < 0.05). Q(S)/Q(T) ha
d increased to 11 +/- 6% (P < 0.05), but little perfusion of low-V(A)/
Q units was observed. A correlation was found between PA-a(O2) and Q(S
)/Q(T) (r = 0. 2, P < 0.03). Conclusions. Q(S)/Q(T) is a major compone
nt of impaired gas exchange before, during, and after cardiac surgery.
Q(S)/Q(T) increases after induction of general anesthesia, probably b
ecause of development of atelectasis. After separation from extracorpo
real circulation, accumulation of extravascular lung water or further
collapse of lung tissue may aggravate Postoperatively, oxygenation imp
roves, possibly because of recruitment of previously nonventilated alv
eoli or resolution of extravascular lung water. During spontaneous bre
athing, additional mechanisms such as altered mechanics of the chest,
perfusion of low-V(A)/Q regions, and decreased mixed venous oxygen ten
sion may contribute to impaired gas exchange.