Predictive value of monocyte-to-high-density lipoprotein cholesterol ratio for acute kidney injury following coronary artery bypass grafting

Figures

• 

  Fig. 1. Comparison of clinical and laboratory variables according to AKI groups. Monocyte-to-high-density lipoprotein cholesterol ratio (MHR); White blood cell count (WBC); Blood glucose (BG); Triglycerides (TG); Left ventricular ejection fraction (LVEF); EuroSCORE. AKI = acute kidney injury; NS = not significant; **p < 0.01; ***p < 0.001

• 

  Fig. 2. ROC curves comparing the predictive performance of the combined model, MHR, triglycerides, and blood glucose for AKI following CABG.The combined model achieved the highest AUC. Pairwise comparisons using the DeLong test indicated statistically significant differences between the combined model and each individual predictor. (MHR = monocyte-to-high-density lipoprotein cholesterol ratio)

• 

  Fig. 3. ROC curves based on ten-fold cross-validation for prediction of postoperative AKI.The combined model incorporating MHR, triglycerides, and blood glucose showed the highest discriminative ability (AUC = 0.769), followed by MHR, triglycerides, and blood glucose. The results indicate improved performance of the combined model and good stability across validation folds. (MHR = monocyte-to-high-density lipoprotein cholesterol ratio, CABG = coronary artery bypass grafting, AKI = acute kidney injury)

• 

  Fig. 4. Calibration curve of the combined model for predicting postoperative AKI. The model shows good agreement between predicted and observed probabilities

• 

  Fig. 5. Decision curve analysis for prediction of postoperative AKI. The combined model incorporating MHR, triglycerides, and blood glucose demonstrated greater net benefit across a range of threshold probabilities compared with individual predictors. Among the individual predictors, MHR showed a relatively favorable net benefit

• 

  Fig. 6. Comparison between Group 1 (MHR < 0.658) and Group 2(MHR > 0.658) (AKI = acute kidney injury; WBC = white blood cell count; EF = left ventricular ejection fraction; PLT = platelet count). **p < 0.01; ***p < 0.001

• 

  Fig. 7. Temporal trajectory of serum creatinine by MHR groups (Group 1: MHR < 0.658, n = 182; Group 2: MHR ≥ 0.658, n = 98). Error bars represent SEM. CrH0–CrH72 denote creatinine at 0–72 h postoperatively. Longitudinal changes were assessed using linear mixed-effects models with time as a fixed effect and subject-level random intercepts (both groups: time effect p < 0.001). Between-group differences at each time point were evaluated using the Mann–Whitney U test with Holm correction (**p < 0.01; other p values indicated in figure)