Objective: To determine whether vacuum-assisted venous return has clinical advantages over conventional gravity drainage apart from allowing the use of smaller cannulas and shorter tubing.
Methods: A total of 150 valve operations were performed at our institution between February and July 1999 using vacuum-assisted venous return with small venous cannulas connected to short tubing. These were compared with (1) 83 valve operations performed between April 1997 and January 1998 using the initial version of vacuum-assisted venous return, and (2) 124 valve operations performed between January and April of 1997 using conventional gravity drainage. Priming volume, hematocrit value, red blood cell usage, and total blood product usage were compared multivariably. These comparisons were covariate and propensity adjusted for dissimilarities between the groups and confirmed by propensity-matched pairs analysis.
Results: Priming volume was 1.4 +/- 0.4 L for small-cannula vacuum-assisted venous return, 1.7 +/- 0.4 L for initial vacuum-assisted venous return, and 2.0 +/- 0.4 L for gravity drainage (P <.0001). Smaller priming resulted in higher hematocrit values both at the beginning of cardiopulmonary bypass (27% +/- 5% compared with 26% +/- 4% and 25% +/- 4%, respectively, P <.0001) and at the end (30% +/- 4% compared with 28% +/- 4% and 27% +/- 4%, respectively, P <.0001). Red cell transfusions were used in 17% of the patients having small-cannula vacuum-assisted venous return, 27% of the initial patients having vacuum-assisted venous return, and 37% of the patients having gravity drainage (P =.001); total blood product usage was 19%, 27%, and 39%, respectively (P =.002). Although ministernotomy also was associated with reduced blood product usage (P <.004), propensity matching on type of sternotomy confirmed the association of vacuum-assisted venous return with lowered blood product usage.
Conclusions: Vacuum-assisted venous return results in (1) higher hematocrit values during cardiopulmonary bypass and (2) decreased red cell and total blood product usage.