Evaluation of the application of a predictive model for red blood cell demand in surgical procedures
10.13303/j.cjbt.issn.1004-549x.2026.01.007
- VernacularTitle:外科手术红细胞需求预测模型应用评价
- Author:
Xiaoyu CAI
1
;
Yannan FENG
2
;
Chunya MA
1
;
Yuan ZHUANG
1
;
Yang YU
1
Author Information
1. Department of Blood Transfusion, The First Medical Center, Chinese PLA General Hospital, Beijing 100853, China
2. Department of Blood Transfusion, Sichuan Cancer Hospital & Affiliated Cancer Hospital of University of Electronic Science and Technology of China, Chengdu 610041, China
- Publication Type:Journal Article
- Keywords:
predictive models;
machine learning;
red blood cell transfusion;
application evaluation
- From:
Chinese Journal of Blood Transfusion
2026;39(1):51-55
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To assess the clinical application value of a prediction model for red blood cell (RBC) demand in surgical procedures. Methods: Demographic data, laboratory parameters, anesthesia and transfusion records, and model prediction data were retrospectively collected from surgical patients at the First Medical Center of Chinese PLA General Hospital between 2018 and 2024. Statistical analysis was performed using the Chi-square test, t-test, and Mann-Kendall trend test. Results: From 2018 to 2024, the predictive model for RBC demand in surgical procedures was used to evaluate a total of 112 293 surgeries. During this period, the model call rate (77.49%-98.91%, P<0.05), compliance rate (56.81%-84.92%, P<0.05), and prediction accuracy rate (66.82%-94.17%, P<0.05) all showed significant upward trends. The total blood usage across the hospital (13645.4-7723.5 units, P<0.05) and the average blood usage per surgery (0.21-0.1 units, P<0.05) exhibited overall downward trends. Postoperative average hemoglobin levels in the non-compliance group (112.1-105.3 g/L in the non-compliance group vs 106.9-92.7 g/L in the compliance group, P<0.05) and the intraoperative excessive transfusion rate (5.06%-6.05% in the non-compliance group vs 0.09%-0.04% in the compliance group, P<0.05) were significantly higher in the non-compliance group compared to the compliance group. Conclusion: The predictive model for RBC demand in surgical procedures has played a positive role in conserving blood resources, optimizing blood resource allocation, and reducing intraoperative risks.
