Perioperative red blood cell infusion and deep vein thrombosis in patients with femoral and pelvic fractures: A propensity score matching

Abstract Background:The relationship between perioperative red blood cell（RBC）infusion and deep vein thrombosis(DVT) has not been determined.Objectives:To analyze the time-event relationship between perioperative RBC infusion and DVT in patients with femoral and pelvic fractures after adjusting for confounding factors and to provide reference for optimizing DVT risk factors.Methods: The clinical data of 569 patients with femoral and pelvic fractures who received surgical treatment from May 2018 to December 2019 were retrospectively analyzed.Propensity score matching (PSM) was performed on 20 covariates of DVT. With the formation or progression of DVT after RBC infusion as the end point, the time-event relationship between perioperative RBC infusion and DVT in patients was analyzed by binary Logistic regression.Results: After 1:1 PSM of 569 patients included in this study, 126 patients were in the transfusion group and the non-transfusion group, respectively.Before PSM (P=0.023, OR=1.496 [95%CI,1.058-2.115]), perioperative RBC infusion was associated with DVT formation for femoral and pelvic fractures.This conclusion was still obtained after PSM (P=0.038,OR=1.728, 95%CI=(1.031,2.896)). The risk of DVT in patients with RBC infusion of 2-4U and >4U is 1.833 and 2.667 times that of ≤2U, respectively. Excluding patients who received preoperative RBC infusion and had DVT formation or progression prior to transfusion, there were still differences in perioperative RBC infusion between non-DVT group and DVT group (P=0.037, OR=2.231 [95% CI ,1.049-4.745]).Conclusion: Perioperative RBC infusion is the cause of DVT in patients with femoral and pelvic fractures, and the risk of DVT is positively correlated with the amount of RBC infusion.