Impact of blood transfusion components on thrombosis risk and construction of a prediction model
10.13303/j.cjbt.issn.1004-549x.2026.07.008
- VernacularTitle:输血成分对患者血栓风险的影响及预测模型构建
- Author:
Tao CAO
1
;
Yuan ZHOU
1
;
Xianting KE
1
;
Jinbo FAN
1
Author Information
1. Department of Transfusion Medicine, Taihe Hospital, Hubei University of Medicine, Shiyan 442000, China
- Publication Type:Journal Article
- Keywords:
allogeneic transfusion;
blood components;
venous thromboembolism;
risk prediction model;
patient blood management
- From:
Chinese Journal of Blood Transfusion
2026;39(7):867-873
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To explore the association between transfusion of different allogeneic blood components and the risk of venous thromboembolism (VTE) in hospitalized patients, to clarify the differences in thrombotic risk among various blood components, and to develop a targeted VTE risk prediction model. Methods: A retrospective cohort study was conducted, enrolling 5 406 hospitalized patients who received allogeneic blood transfusion in our hospital from November 2024 to October 2025. Using 1∶1 nearest neighbor propensity score matching (PSM), 5 402 non-transfused patients were selected as controls. The matching variables included sex, age, VTE risk score, VTE prophylaxis, comorbid tumor, use of antibiotics, and invasive mechanical ventilation, with a caliper of 0.01 and sampling without replacement. Standardized mean difference (SMD) was used to assess matching quality, with SMD<0.1 indicating balanced baseline characteristics. The transfusion group was stratified by the type of blood component transfused. The chi-square test with Bonferroni correction and multivariate logistic regression (including forced adjustment for the unbalanced variables VTE prophylaxis and antibiotic use) were used to analyze risk factors. A prediction model was constructed and validated using the area under the receiver operating characteristic (ROC) curve and the Hosmer-Lemeshow test. The optimal cut-off value was determined by the maximum Youden index. Results: All variables showed SMD<0.1 after matching, indicating good baseline balance. The incidence of VTE was significantly higher in the transfusion group than in the control group (8.7% vs 4.3%, RR=2.02, 95% CI: 1.74-2.35, P<0.001). Significant differences in VTE incidence were observed among different transfusion subgroups, with a thrombotic risk gradient of combined transfusion > plasma > red blood cells ≈ platelets. The incidence of VTE in the combined transfusion subgroup was 11.7%, which was significantly higher than that in the red blood cell subgroup (RR=3.20, 95% CI: 2.41-4.24, P<0.001), plasma subgroup (RR=1.71, 95% CI: 1.24-2.35, P<0.001), and platelet subgroup (RR=6.81, 95% CI: 2.56-18.06, P<0.001). The prediction model showed good discrimination and calibration, with an area under the ROC curve of 0.800. The optimal cut-off value determined by the Youden index was 0.12, corresponding to a sensitivity of 62.3% and a negative predictive value of 91.4%. Conclusion: Allogeneic blood transfusion was an independent risk factor for VTE in hospitalized patients, with significant differences in thrombotic risk among different blood components. The prediction model developed in this study can achieve VTE risk stratification in transfused patients, providing evidence for precise transfusion therapy and patient blood management.
