An Analysis of Perioperative Factors Associated with Hemorrhage in Laparoscopic Sleeve Gastrectomy

Jonathan E Miller, Pei-Wen Lim, MD, Richard Perugini, MD, John J Kelly, MD, Donald Czerniach, MD, Philip Cohen, MD. UMass Memorial Medical Center

Background Laparoscopic sleeve gastrectomy (LSG) has rapidly gained popularity as a primary bariatric procedure in the United States. The most frequent complication of LSG in our series, after perioperative nausea, was hemorrhage. We studied our inception series to identify factors associated with hemorrhage following LSG.

Methods A registry was conducted of patients who underwent LSG at UMass Memorial Health Care between July 2010 and June 2014. Our criteria for hemorrhage included hematocrit drop requiring extended postoperative stay, need for transfusion, or return to the operating room to control bleeding. Data included demographics and comorbidities, metabolic status, length of stay, operative technique, rate of complication, and excess bodyweight loss at postoperative follow-up. Univariate analysis was used to identify perioperative factors most strongly associated with hemorrhage.

Results In this series 20 patients out of 345 met our definition of hemorrhage. Mean hematocrit drop (11.2% vs. 2.0%, p < 0.001), cumulative length of stay (5.2 days vs. 1.8, p < 0.001), mean rate of transfusion (1.4 units/person vs. 0.0 units/person, p < 0.001), and rate of readmission (20% vs. 4.9%, p = 0.022) were significantly increased among patients who bled postoperatively. Only preoperative DVT prophylaxis was identified as significantly associated with hemorrhage in this study. Of 189 patients who received DVT prophylaxis hemorrhage occurred in 9.25%, compared to 2.2% of the 156 patients who did not (p = 0.016). Age, gender, BMI, and use of Seamguards or Evicel were not found to significantly impact hemorrhage.

Conclusion Postoperative hemorrhage has important effects that can be presumed to lead to higher costs. As preoperative anticoagulation therapy is the only factor we identified as having an association with hemorrhage, we must weigh the benefit of prophylaxis to reduce thromboembolic events against the increased risk of postoperative hemorrhage. We plan to look at selective administration of Lovenox versus methods of staple line reinforcement (tissue sealant, bioprosthetic reinforcement) to decrease rate of hemorrhage.

« Return to SAGES 2015 abstract archive