Construction and validation of an early predictive model for intraoperative massive transfusion of red blood cells in patients with Stanford type A aortic coarctation

Abstract

Objective To analyze the risk factors for intraoperative massive red blood cell (RBC) transfusion in patients with Stanford type A aortic dissection (TAAD), in order to develop a risk-prediction model and validate its predictive effect. Methods The clinical data of 233 patients with TAAD admitted to our hospital from July 2018 to June 2021 (modeling set) were retrospectively analyzed. They were divided into routine transfusion group (n=128, RBC≤8 U) and massive transfusion group (n=105, RBC>8 U). Risk factors for intraoperative massive RBC transfusion in TAAD patients were analyzed by multivariate logistic regression and a risk prediction model was developed. Calibration curve and receiver operating characteristic (ROC) curve were used to assess the accuracy and discrimination of the model. In addition, 61 TAAD patients admitted to our hospital from July 2021 to May 2022 (validation set) were used for external validation. Results The rate of intraoperative massive RBC transfusion in 233 TAAD patients was 45.06% (95% CI: 38.59%-51.69%). Logistic analysis showed that women, age >50 years, preoperative Hb≤131.50 g/L, intraoperative bleeding >720 mL, and CPB time >155 min were independent risk factors for massive intraoperative RBC transfusion (P<0.05). The intraoperative risk prediction model formula for massive RBC infusion was: -4.427+ 0.925×gender+ 1.461×age+ 2.081×preoperative Hb+ 1.573×bleeding volume+ 2.823×CPB time. The area under the ROC curve of the modeling set and validation set were 0.904 (95% CI: 0.865-0.943) vs 0.868 (95%CI: 0.779-0.958), and the slopes of the calibration curves all converged to 1, indicating that the model predicted the risk of intraoperative massive RBC infusion in TAAD patients in good consistency with the actual risk of massive infusion. The decision curve shows that the model exhibits a positive net benefit with a threshold probability of 0.15-0.67 and has a high clinical application value. Conclusion The prediction model constructed based on the risk factors of intraoperative massive RBC infusion in TAAD patients can effectively predict the risk of intraoperative massive RBC infusion with high clinical predictive efficacy.

Keywords

• stanford type a aortic coarctation
• risk factors
• transfusion of red blood cells
• predictive model