Establishment and verification of acute kidney injury model in sepsis patients based on albumin-corrected anion gap
10.3760/cma.j.cn114656-20250407-00256
- VernacularTitle:基于白蛋白校正阴离子间隙的脓毒症患者并发急性肾损伤模型建立及验证
- Author:
Lingquan ZENG
1
;
Shuping GUO
Author Information
1. 赣州市人民医院急诊科，赣州 341000
- Keywords:
Albumin corrected anion gap;
Sepsis;
Acute kidney injury;
Prediction model
- From:
Chinese Journal of Emergency Medicine
2025;34(11):1579-1585
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct and validate a model of acute kidney injury ( AKI ) in sepsis patients based on albumin-corrected anion gap ( ACAG ).Methods:The clinical data of patients with sepsis admitted to our hospital from September 2020 to January 2025 were retrospectively analyzed. They were divided into training set and validation set according to the ratio of 7 : 3. The patients in the training set were divided into AKI group and non-AKI group according to whether AKI occurred during hospitalization. The basic information, disease treatment data and laboratory data of the patients were collected, and ACAG was calculated. Multivariate Logistic regression analysis was used to analyze the risk factors of AKI, and R language was used to construct a nomogram model to predict the risk of AKI. The receiver operating characteristic ( ROC ) curve was drawn to measure the distinguishing ability of the model, and the calibration curve was drawn to test the consistency between the predicted results of the model and the actual situation. Decision curve analysis ( DCA ) was used to judge the practicability of the model in clinical practice.Results:382 patients with sepsis were included and divided into training set ( n = 267 ) and validation set ( n = 115 ) according to the ratio of 7 : 3. Among the 382 patients, 124 cases of AKI occurred during hospitalization, accounting for 32.46%( 124 / 382 ). Among them, 85 cases of AKI occurred in the training set, accounting for 31.84%( 85 / 267 ), and 39 cases of AKI occurred in the validation set, accounting for 33.91%( 39 / 115 ). Multivariate Logistic regression analysis showed that age ( OR=2.815, 95% CI:1.316~6.022), white blood cell count ( WBC ) ( OR = 1.926, 95% CI : 1.330-2.790 ), lactic acid ( Lac ) ( OR = 2.189,95%CI : 1.300 ~ 3.687 ), serum creatinine ( Scr ) ( OR = 3.156,95% CI : 1.702 -5.852 ), blood urea nitrogen ( BUN ) ( OR = 2.951,95% CI : 1.652-5.271 ), uric acid ( OR = 3.122,95% CI: 1.588-6.139 ), C-reactive protein ( CRP ) ( OR = 2.847,95% CI: 1.384-5.856 ), ACAG ( OR = 2.953,95% CI: 1.669-5.224 ) were risk factors for AKI in sepsis patients with intensive training ( P < 0.05 ). According to the results of multivariate logistic regression analysis, a nomogram model was established. ROC analysis showed that the area under the curve ( AUC ) of the nomogram model for predicting postoperative AKI in patients with sepsis in the training set and the validation set were 0.95 ( 95% CI: 0.90-0.98 ) and 0.92 ( 95% CI: 0.89-0.97 ), respectively. The statistical values after Hosmer-Lemeshow goodness-of-fit test were 0.295 and 0.264 ( P = 0.563 and 0.488 ), respectively. The calibration curve and DCA results showed. The calibration and clinical practicability of the nomogram prediction model are good. Conclusions:Age, WBC, Lac, Scr, BUN, uric acid, CRP and ACAG are all associated with AKI in patients with sepsis. Based on these risk factors, a nomogram model was constructed to predict the risk of AKI in patients with sepsis. The model has good discrimination ability, calibration ability and clinical practical value.
