Transfusion therapy risk stratification in cardiac surgery

Abstract

Introduction. The use of allogeneic blood components in cardiac surgery is associated with the development of complications. One of the strategies for minimizing unjustified transfusions is the use of risk stratification models for transfusion therapy. Their goal is to predict the likelihood of using blood components in a particular patient based on clinical indicators.

Objective. To study the possibility of using a risk stratification model for transfusion therapy.

Materials and methods. The criteria for inclusion in the study were met by patients over 18 years of age who underwent emergency and planned open-heart surgery performed between January 01, 2024 and December 31, 2024. The endpoint was considered to be allogeneic blood transfusion, which was understood to mean the clinical use of one or more units of erythrocyte-containing components, any type of plasma or concentrate throughout the duration of hospitalization. Discrimination assessment was performed using the AUC-ROC method. The calibration was evaluated using the Hosmer-Lemeshov test. Descriptive analysis was performed using categorical variables expressed in absolute numbers and percentages. The quantitative variables were expressed as means and standard deviations. Statistical analysis was performed using Microsoft Excel 2010.

Results. A total of 218 patients were included in a single-center, observational, retrospective study. The accuracy of transfusion prediction was 0.67 95 % CI. The Hosmer-Lemeshov test demonstrated systematic calibration errors. In low–risk areas, the model overestimated the probability of transfusion, while in high-risk areas it underestimated it.

Conclusion. The use of this model makes it possible to optimize the appointment of donated blood and reduce the number of unjustified transfusions. For clinical use, adaptation to local conditions is required.