A retrospective Cohort Study to Predict Acute Kidney Injury Based on the Coagulation and Inflammation in ICU Patients with Sepsis

Abstract Background. Sepsis is a major cause of morbidity and mortality worldwide. Sepsis with acute kidney injury (AKI) is associated with higher mortality risk when compared with those with sepsis and without AKI. Therefore, it is necessary to detect the predictors of sepsis-associated acute kidney injury (SA-AKI) in order to timely prevent, diagnose and treat this complication. Methods. From July 2016 to December 2019, 419 patients with sepsis admitted to the intensive care unit (ICU) were randomly divided into two groups: training group (n= 302) and validation group (n = 117). A least absolute shrinkage and selection operator (LASSO) regression was constructed to select variables within 24 h of admission, and then were included in a logistic regression model to find the independent risk factors of AKI. Hence, a nomogram for predicting SA-AKI with statistically significant covariates was constructed. Discrimination, calibration, and clinical utility of the nomogram performance were assessed and then validated.Results. The risk factors yielded by logistic regression were hypertension(HT), diabetes mellitus(DM), C-reactive protein(CRP), procalcitonin(PCT), activated partial thromboplastin time(APTT), platelet(PLT), and then were incorporated into the nomogram. The areas under the ROC curve of the nomogram in the training and validation groups were 0.856 and 0.885, respectively. The calibration curves demonstrated favorable consistency between the predictions of nomogram and the actual observations in both training as well as validation groups. Decision curve analysis (DCA) showed clinical usefulness of the proposed nomogram model.Conclusions. A risk prediction model by integrating variables can assist in identifying patients who are at high risk of developing SA-AKI. The nomogram had excellent predictive ability and might have significant clinical implications for early detection of AKI in patients undergoing sepsis.