A nine-variable nomogram successfully predicted 30-day mortality in cardiorenal syndrome patients undergoing continuous renal replacement therapy with an AUC of 0.7.
Cohort (n=8,055)
No
Does a nomogram based on 9 clinical variables accurately predict 30-day mortality in patients with cardiorenal syndrome treated with continuous renal replacement therapy?
A novel nomogram based on 9 readily available clinical variables can predict 30-day mortality with moderate accuracy (AUC 0.7) in critically ill patients with cardiorenal syndrome requiring continuous renal replacement therapy.
Effect estimate: AUC 0.7
Studies have shown that most cardiorenal syndrome (CRS) patients treated with continuous renal replacement therapy (CRRT) have a poor prognosis. This study aimed to develop a predictive model for the risk of death within 30 days in patients with CRS treated with CRRT. Data from CRS patients treated with CRRT were chosen for this investigation from the MIMIC-IV database. The training and validation sets were formed on screened CRS patients. Also, based on the survival status of CRS, survival and death groups were generated in training and validation sets, respectively. Afterwards, in the training set, key variables were identified by univariate Cox, LASSO, as well as multivariate Cox regression analyses. Next, to evaluate the key variables’ capacity for prediction, the nomogram was established. Meanwhile, decision curves, calibration curves, and the receiver operating characteristics (ROC) were drawn. Following the removal of ineligible people, 8,055 CRS patients treated with CRRT were recruited and randomly assigned in a 7:3 ratio to a training set (5,640) and a validation set (2,415). In the training set, a survival group was created, and in the validation set, a death group. Nine factors were then determined to be important predictors of 30-day mortality through the multivariate Cox model: temperature, respiratory rate, white blood cells (WBC), hemoglobin, creatinine, prothrombin time (PT), charlson comorbidity index (CCI), systemic inflammatory response syndrome (SIRS), and systolic blood pressure (SBP). Following that, in both the training and validation sets, the nomogram’s AUC value was 0.7. The choice curve and the calibration curve both showed that the nomograms that were created had strong predictive power. The 9 key variables were identified and the nomogram model was established to forecast the risk of death within 30 days after CRRT treatment of patients with CRS, which might provide useful predictive information for clinical decision making.
Wang et al. (Sat,) conducted a cohort in Cardiorenal syndrome (n=8,055). Continuous renal replacement therapy (CRRT) was evaluated on 30-day mortality prediction accuracy (Area Under the Curve) (AUC 0.7). A nine-variable nomogram successfully predicted 30-day mortality in cardiorenal syndrome patients undergoing continuous renal replacement therapy with an AUC of 0.7.