2024 FSA Podium and Poster Abstracts

P024: PREDICTIVE MODELING OF MASSIVE TRANSFUSION IN LIVER TRANSPLANTATION
Sagar Jolly, MD; Sathish Kottaisamy, MD; Nicolás P Caram, MD; Eslam Fouda, MD; Fouad Souki, MD; Marina Gitman, MD; Ramona Nicolau-Raducu, MD, PhD; University of Miami/Jackson Memorial Hospital

Introduction: Massive transfusion (MT) during liver transplant is an important concern due to its association with increased morbidity, mortality, and healthcare costs. Anticipating the need for MT is critical for effective preoperative preparation and for the efficient use of blood bank resources. Identifying predictors of MT allows blood banks to strategically plan blood supply, improving patient care and reducing wait times. The purpose of our study was to identify risk factors associated with MT during liver transplant surgery and to develop a probability model for MT.

Methods: With institutional board review approval, 532 liver transplant recipients data from 2019 to 2022 was analyzed retrospectively, including 74 (14%) combined liver-kidney transplants. Sequential kidney transplant was done in all combined liver-kidney cases after kidneys were placed on a hypothermic machine perfusion pump. MT was defined as the intraoperative administration of ≥10 packed red blood cells (pRBC) units.

Results: In our cohort of 532 liver transplant recipients, 38% required massive transfusions, using a median of 17 units (IQR: 12-25), compared to 62% who did not, receiving a median of 5 units (IQR: 3-7). When comparing the 2 groups (≥ 10 vs <10 units pRBC), the following significant pre-transplant variables were identified: type of organ transplanted (liver versus combined liver-kidney), MELD score, portal vein thrombosis, presence of trans jugular intrahepatic portosystemic shunt, presence of ascites, pre-transplant hospitalization, pre-transplant diagnosis of hepatocellular carcinoma, viral hepatitis and alcoholic cirrhosis, baseline hemoglobin and platelets count. After adjusting for confounders, logistic regression identified four pretransplant risk factors associated with MT: combined liver-kidney transplant vs. liver, a higher MELD score, lower baseline platelet count, and lower hemoglobin level (χ2 = 118; P < 0.0001). A C-statistic of 0.76 was calculated for these risk factors. Recursive partitioning identified a MELD score ≥23, hemoglobin <9.7 g/dL, and platelet count <84 x 103/mcL as critical thresholds for increased MT risk. A probability model for MT is presented in Table 1.

Conclusion: Our study is consistent with previous risk factors reported, such as preoperative hemoglobin, platelet count and MELD score. The novelty of our study validates the fact that, for the same risk factors, combined liver-kidney transplant had a higher probability for MT than liver transplant alone (88% vs 61%), even if the kidney transplant followed the next day.