Construction of blood transfusion prediction model for patients with severe traumatic brain injury
10.13303/j.cjbt.issn.1004-549x.2023.11.004
- VernacularTitle:重度颅脑创伤患者临床输血预测模型构建
- Author:
Chenggao WU
1
;
Qiang LIU
2
;
Wei XIONG
2
;
Aiping LE
1
;
Wei LIU
1
Author Information
1. Transfusion Medicine Department, The First Affiliated Hospital of Nanchang University, Nanchang 330006, China
2. Key Laboratory of Jiangxi Province for Transfusion Medicine, The First Affiliated Hospital of Nanchang University, Nanchang 330006, China
- Publication Type:Journal Article
- Keywords:
traumatic brain injury;
blood transfusion;
nomogram prediction model;
logistic regression
- From:
Chinese Journal of Blood Transfusion
2023;36(11):990-994
- CountryChina
- Language:Chinese
-
Abstract:
【Objective】 To construct a blood transfusion prediction model for patients with severe traumatic brain injury (TBI), in order to predict the risk of blood transfusion and guide blood transfusion decision-making. 【Methods】 The clinical data of 756 patients with severe TBI admitted to the hospital from January 1, 2015 to June 30, 2021 were retrospectively analyzed. According to whether the patients were transfused with red blood cells after admission, the patients were divided into blood transfusion group (n=354) and non-blood transfusion group (n=402). The basic clinical data and prognostic indicators of the two groups were compared. Logistic regression algorithm was used to screen the risk factors related to blood transfusion in hospital to establish a nomogram prediction model, and the performance of the model was evaluated. 【Results】 No significant differences were noticed in gender, age, body temperature, cause of injury, ABO blood group, Rh blood group, serum Na and K concentrations between the two groups (P>0.05). Significant differences were found in Glasgow coma score (GCS), heart rate (HR), systolic blood pressure (SP), diastolic blood pressure (DP), shock index (SI), respiratory rate (RR), clinical diagnosis, treatment, hemoglobin concentration (Hb), hematocrit (Hct), platelet count (Plt) and coagulation function between the two groups (P<0.05). The length of stay, complication rate, ventilator utilization rate, ICU admission rate and 90-day readmission rate in the blood transfusion group were significantly higher than those in the non-blood transfusion group (P<0.05). However, there was no significant difference in in-hospital mortality between the two groups (P>0.05). Multivariate logistic regression analysis showed that surgical treatment, skull fracture, hemorrhagic shock, decreased Plt, decreased Hct and increased INR were independent risk factors for blood transfusion. A nomogram prediction model was constructed and the area under the ROC curve of the training set and the test set was 0.819(95% CI: 0.784-0.854) and 0.866(95% CI: 0.818-0.910), respectively, which had good predictive performance. 【Conclusion】 Surgical treatment, skull fracture, hemorrhagic shock, decreased Plt, decreased Hct and increased INR are independent risk factors for blood transfusion in adult patients with severe traumatic brain injury. The nomogram prediction model can better predict the blood transfusion demand of TBI patients and has high application value.
