A Prediction Model of Sepsis-Associated Acute Kidney Injury Based on Antithrombin III

Xie Yun,Zhang Yi,Tian Rui,Jin Wei,Du Jiang,Zhou Zhigang,Wang Ruilan
DOI: https://doi.org/10.1007/s10238-020-00656-x
2020-01-01
Abstract:The incidence of sepsis-associated acute kidney injury (AKI) is on the rise. Recent studies have found a correlation between antithrombin III and AKI. We established a predictive model for sepsis-associated AKI based on plasma ATIII levels. A prospective study (March 2018–January 2020) was conducted in sepsis patients admitted to the Critical Care Medicine Department at Shanghai General Hospital. ATIII levels were obtained within 48 h after admission to the ICU and before the diagnosis of sepsis-associated AKI was recorded. Renal function was assessed by measuring serum creatinine levels and urine volume. Male sex, other cardiovascular disease, and low ATIII levels were identified as independent risk factors for AKI. Age, immune disease, and low ATIII levels were identified as independent risk factors for death. Plasma ATIII levels in the non-AKI group were higher than those in the AKI group, plasma ATIII levels were higher in the survival group than in the non-survival group, plasma ATIII levels in the non-CRRT group were higher than those in the CRRT group, and plasma ATIII levels in the non-CKD group were higher than those in the CKD group. ATIII was significantly higher in the group with pulmonary infection than in the group without pulmonary infection. ATIII was significantly lower in the celiac infection group than in the nonceliac infection group. There was no statistically significant difference between the ATIII in the gram-positive group and the gram-negative group. ATIII was significantly higher in medical patients than in surgical patients. The predictive model of sepsis-associated AKI established based on ATIII was ln[P/(1 − p)] = −1.211 × sex − 0.017 × ATIII + 0.022 × Cr + 0.004 × BUN − 2.8192. The model goodness-of-fit test (p = 0.000) and the area under the ROC curve of the model (0.9862) suggested that the model has a high degree of discrimination and calibration. ATIII reduction was closely related to the prognosis of patients with sepsis. ATIII reduction was an independent risk factor for sepsis-associated AKI and an independent risk factor for mortality in patients with sepsis. ATIII reduction could predict sepsis-associated AKI. Low ATIII predicted a poor prognosis.
What problem does this paper attempt to address?