Development and validation of a nomogram for predicting advanced chronic kidney disease

Figure 1. Flow chart of the study design NHANES: national health and nutrition examination survey; LASSO: least absolute shrinkage and selection operator; AUC: area under the receiver operating characteristic curve; CKD: chronic kidney disease; DCA: decision curve analysis.

Figure 2. LASSO regression analysis was employed to identify the key variables associated with advanced CKD (A) Path diagram of LASSO regression coefficients. (B) LASSO regression cross-validation curve. (C) Importance ranking diagram of LASSO variables. CKD: chronic kidney disease; LASSO refers to the least absolute shrinkage and selection operator; ALT: alanine aminotransferase; HDLC: high-density lipoprotein cholesterol; LDH: lactate dehydrogenase; TBIL: total bilirubin; ALP: alkaline phosphatase.

Figure 3. Univariate and multivariate analyses were conducted to identify variables associated with advanced CKD (A) Forest plot of univariate logistic regression analysis. (B) Forest plot for multivariate logistic regression analysis. CKD: chronic kidney disease; LDH: lactate dehydrogenase; HDLC: high-density lipoprotein cholesterol; TBIL: total bilirubin; ALP: alkaline phosphatase; ALT: alanine aminotransferase.

Figure 4. A nomogram for predicting the risk of advanced CKD was constructed using logistic regression CKD: chronic kidney disease; ALT: alanine aminotransferase; LDH: lactate dehydrogenase.

Figure 5. Comprehensive evaluation of advanced CKD models derived from training set (A) ROC curve illustrating discriminatory ability. (B and C) Bootstrap distribution and scatterplot of the AUC, demonstrating the model’s stability. (D) Calibration curves comparing predicted probabilities with observed outcomes. (E) DCA to assess net clinical benefit. CKD: chronic kidney disease; ROC: receiver operating characteristic; AUC: area under the curve; DCA: decision curve analysis.

Figure 6. Comprehensive evaluation of advanced CKD models derived from test set and external dataset (A) The discrimination power of the ROC curves in the test set. (B) The calibration curve in the test set was used to compare the consistency between the predicted probabilities and the observed outcomes. (C) DCA to assess the net clinical benefit of the test set. (D) The discrimination power of the ROC curves in the external dataset. (E) The calibration curve in the external dataset was used to compare the consistency between the predicted probabilities and the observed outcomes. (F) DCA to assess the net clinical benefit of the external dataset. CKD: chronic kidney disease; ROC: receiver operating characteristic; DCA: decision curve analysis.