Impact of Bleeding Revision on Outcomes After Left Ventricular Assist Device Implantation

Abstract

Few reports have described associations between a bleeding revision and outcomes after implantation of a continuous-flow left ventricular assist device (LVAD). We retrospectively analyzed all adult recipients of a continuous-flow LVAD who were operated on between January 2006 and December 2016 at the Division of Cardiac Surgery, Medical University of Vienna (N=257). LVAD types were HeartWare HVAD (Medtronic, Minneapolis, MN), HeartMate II (Abbott [Thoratec], Abbott Park, IL), and HeartMate 3 (Abbott [Thoratec]). The primary outcome variable was survival during LVAD support. Secondary outcome variables were rates of prolonged ventilation, intensive care unit stay, hospital stay, renal replacement therapy, and stroke. We additionally investigated risk factors for a bleeding revision. Thirty-six patients (14%) underwent surgical revision for bleeding. In a multivariable Cox proportional hazards model, bleeding revision was associated with significantly reduced survival during LVAD support (3months: hazard ratio [HR], 2.04; 95% confidence interval [CI], 1.20 to 3.48; 6 months: HR, 1.63; 95% CI, 0.95 to 2.82; 12 months: HR, 1.30; 95% CI, 0.69 to 2.46; 24 months: HR, 1.04; 95% CI 0.48 to 2.26; overall p= 0.007). Rates of secondary outcome variables were substantially higher in the revision group compared with patients who survived the hospital stay without a bleeding revision. Concomitant implantation of a temporary right ventricular assist device (p= 0.002) and age (p= 0.028) were independent risk factors for a bleeding revision. A bleeding revision is associated with significantly reduced survival and substantially increased morbidity during LVAD support. Preoperative, intraoperative, and postoperative prophylactic strategies may help to prevent this life-threatening complication.