Background/Objectives: The aim of this study was to investigate the association between intraoperative blood transfusion (BT) and the short-term outcomes of pancreatoduodenectomy (PD) for patients with periampullary malignancies. Methods: In a retrospective two-center cohort analysis, we utilized a logistic and mixed-effects ordinal regression, nonparametric partial correlation, and mediation analysis, complemented by propensity score matching (PSM) and weighting. Results: A total of 491 patients were included. Of these, 18 (3.7%) received an intraoperative BT. An intraoperative BT was associated with blood loss (odds ratio (OR) per 100 mL = 1.42; 95% CI 1.27 to 1.62; p < 0.001) and relatively high ASA classes (OR = 3.75; 95% CI 1.05 to 17.74; p = 0.041). Intraoperative blood loss (r = 0.27; p < 0.001) but not intraoperative BT (r = 0.015; p = 0.698) was associated with postoperative complications. Intraoperative BT was associated with postoperative complications according to the unadjusted regression (OR = 1.95; 95% CI 1.38–2.42, p < 0.001) but not the multivariable ordinal regression. In the mediation analysis for relative risk (RR), intraoperative BT was beneficial (RR = 0.51; 95% CI: 0.01–0.78), and blood loss (RR = 2.49; 95% CI: 1.75–177.34) contributed to the occurrence of major postoperative complications. After PSM, analyses revealed that an intraoperative BT did not have a significant impact on the rates of postoperative major complications (OR = 1.048; p = 0.919), clinically relevant postoperative pancreatic fistula (OR = 0.573; p = 0.439) or postoperative 90-day mortality (OR = 0.714; p = 0.439). Conclusions: When adjusting for intraoperative blood loss, intraoperative BT is not associated with postoperative complications.
Read full abstract