Phlebotomy with controlled hypovolemia is associated with decreased red blood cell transfusion in liver surgery

Background: Perioperative red blood cell transfusion is associated with poor outcomes in liver surgery. Phlebotomy with controlled hypovolemia, an intervention that is different from acute normovolemic hemodilution, is hypothesized to decrease transfusion requirements. The objective of this work was to determine whether phlebotomy is protective of transfusion, after accounting for other known confounders.

Methods: Consecutive patients who underwent liver resection at one institution (2010-2016) were examined. Factors predictive of perioperative blood transfusion, both on univariate analysis and previously published, were modelled using multivariate logistic regression.

Results: A total of 373 patients underwent liver resection (48% major). Phlebotomy was performed in 45 patients (12%) since 2013, for a median of 496 mL of removed blood (range 247-809). Phlebotomy patients were more likely to have a primary malignancy (31% vs 17%, p=0.0278) and to have a major resection (84% vs 43%, p<0.0001). A trend towards decreased blood loss was noted with phlebotomy (569 vs 748 mL, p=0.0563). On multivariate logistic regression, only having a phlebotomy (OR 0.231, 95% CI 0.081-0.655, p=0.0059), major liver resection (OR 2.858, 95% CI 1.622-5.036, p=0.0003), and preoperative hemoglobin <12.5 g/dL (OR 6.045, 95% CI 3.482-10.493, p<0.0001) were significantly associated with perioperative red blood cell transfusion. Gender, BMI, history of coronary disease, prior liver resection, having a primary malignancy, preoperative platelets, Pringle, and extrahepatic resection were not significantly associated with transfusion.

Conclusion: Phlebotomy with controlled hypovolemia appears to be significantly protective of red blood cell transfusion, independently from other known risk factors. This intervention warrants further study.