Construction and performance evaluation of a prediction model for risk factors of acute kidney injury in patients with multiple trauma
10.3760/cma.j.cn501098-20241023-00617
- VernacularTitle:多发伤患者合并急性肾损伤的危险因素及其预测模型构建与效能评估
- Author:
Dengkui ZHANG
1
;
Zhenjun MIAO
1
;
Yapeng LIANG
1
;
Feng ZHOU
1
;
Qixiang YIN
1
;
Huazhong CAI
1
Author Information
1. 江苏大学附属医院急诊中心，镇江 212001
- Publication Type:Journal Article
- Keywords:
Multiple trauma;
Lactic acid;
Risk factors;
Nomograms;
Acute kidney injury
- From:
Chinese Journal of Trauma
2025;41(2):177-187
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To screen the risk factors of acute kidney injury (AKI) in patients with multiple trauma, construct a prediction model accordingly, and evaluate its predictive value.Methods:A retrospective cohort study was performed to analyze the clinical data of 560 multiple trauma patients who were admitted to while Affiliated Hospital of Jiangsu University from January 2017 to June 2023, including 424 males and 136 females, aged 18-91 years [(55.5±15.0)years]. The patients were randomly divided into a training set ( n=392) and validation set ( n=168) with a ratio of 7∶3. Of all, 77 patients were combined with AKI in the training set, while 33 patients combined with AKI in the validation set. The AKI group and non-AKI group in the training set were compared in terms of gender, age, hypertension, diabetes, cause of injury, abbreviated injury scale (AIS) score of head and neck injury, AIS score of maxillofacial injury, AIS score of chest injury, AIS score of abdominal injury, AIS score of extremities and pelvic injury, AIS score of body surface injury, systolic blood pressure, diastolic blood pressure, heart rate, respiratory rate, body temperature, red blood cell and plasma transfusion volume within 24 hours following admission, emergency surgery, mechanical ventilation, vasoactive drug therapy, Glasgow coma score (GCS) on admission, revised trauma score (RTS) on admission, acute physiology and chronic health assessment II (APACHE II) on admission, injury severity score (ISS) on admission, and laboratory test results on admission including white blood cell count, neutrophil count, lymphocyte count, C-reactive protein, hemoglobin, platelet count, activated partial thromboplastin time (APTT), prothrombin (PT), fibrinogen (FIB), thrombin time (TT), international normalized ratio (INR), D-dimer, blood lactate, base excess, total bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), albumin, globulin, urea nitrogen, serum creatinine, blood glucose, potassium, sodium and chloronium. In the training set, univariate analysis and Lasso regression analysis were used to screen the risk factors of AKI in patients with multiple trauma, which were then included into multivariate logistic regression analysis to identify the independent risk factors. A nomogram prediction model was constructed using the R software based on the above independent risk factors. Hosmer-Lemeshow (H-L) goodness-of-fit test was performed to evaluate the fitting degree of the prediction model in the training set and the validation set, and the receiver operating characteristic (ROC) curve, calibration curve and clinical decision curve (DCA) were plotted in the training set and the validation set to evaluate the predictive performance of the prediction model. Results:There were statistically significant differences in AIS score of abdominal injury, heart rate, body temperature, red blood cell and plasma transfusion volume within 24 hours following admission, emergency surgery, mechanical ventilation, vasoactive drug therapy, GCS on admission, RTS on admission, APACHE II on admission, ISS on admission as well as hemoglobin, platelet count, APTT, PT, FIB, TT, INR, blood lactate, base excess, AST, albumin, globulin, urea nitrogen, serum creatinine, blood glucose and sodium on admission between the AKI group and the non-AKI group ( P<0.05 or 0.01). The characteristic variables screened by Lasso regression analysis included AIS score of abdominal injury, red blood cell transfusion volume within 24 hours following admission, mechanical ventilation, vasoactive drugs therapy, blood lactate on admission, blood creatinine on admission, AST on admission, and blood sodium on admission. Multivariate logistic regression analysis showed that red blood cell transfusion volume within 24 hour following admission ( OR=1.09, 95% CI 1.01, 1.18), mechanical ventilation ( OR=2.49, 95% CI 1.06, 5.85), vasoactive drug therapy ( OR=2.04, 95% CI 1.03, 4.03), blood lactate on admission ( OR=1.10, 95% CI 1.01, 1.21) and serum creatinine on admission ( OR=1.02, 95% CI 1.01, 1.03) were independent risk factors for AKI in patients with multiple trauma ( P<0.05). The regression equation was constructed: Logit[ P/(1- P)]=0.086 2×"red blood cell transfusion volume within 24 hour following admission"+0.912 7×"mechanical ventilation"+0.713 2×"vasoactive drug therapy"+0.098 9×"blood lactate on admission"+0.019 2×"serum creatinine on admission" -4.822 3. H-L goodness-of-fit test showed χ2 value of 9.50 in the training set ( P>0.05) and 6.43 in the validation set ( P>0.05). The results of the ROC curve indicated that the area under the curve (AUC) was 0.84 (95% CI 0.78, 0.89) in the training set and 0.80 (95% CI 0.72, 0.88) in the validation set. The calibration curves showed good agreement with the actual curves, with the predicted probability consistent with the actual probability in both training set and validation set. DCA analysis showed that the threshold probability ranged from 2% to 70% with the net benefit rate of the prediction model greater than 0 in the training set, while the threshold probability ranged from 3% to 69% with the net benefit rate of the prediction model greater than 0 in the validation set. Conclusions:Red blood cell transfusion volume within 24 hours following admission, mechanical ventilation, vasoactive drug therapy, lactate and serum creatinine on admission are independent risk factors for AKI in patients with multiple trauma. The nomogram prediction model based on the above 5 predictive variables of AKI in patients with multiple trauma shows good predictive efficacy and clinical application value.
