What does this research mean for the field?

An XGBoost machine learning model incorporating routine clinical features, most notably the uric acid-to-albumin ratio, accurately identifies concomitant chronic kidney disease in patients with coronary heart disease. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To construct and validate machine learning models based on routine clinical data to detect undetected chronic kidney disease in patients with coronary heart disease.

December 1, 2025International Journal of General MedicineOpen Access

Machine Learning Models for Identifying the Risk of Chronic Kidney Disease in Patients with Coronary Heart Disease: A Retrospective Study

View Full Paper

Ask AI

Bookmark

Why the study?

CHD has a significant comorbid association with CKD, but identification tools for the risk of concomitant CKD in CHD patients are lacking.

Can machine learning models using routine clinical data accurately identify the risk of undiagnosed chronic kidney disease in patients with coronary heart disease?

Key Points

To construct and validate machine learning models based on routine clinical data to detect undetected chronic kidney disease in patients with coronary heart disease.
Retrospectively analyzed 1786 coronary heart disease patients undergoing coronary intervention, randomly allocating them into training (70%) and validation (30%) cohorts.
Screened variables using LASSO regression and multifactor logistic regression to build five machine learning models: logistic regression, support vector machine, random forest, gradient boosting machine, and extreme gradient boosting (XGBoost).
Evaluated model performance using area under the receiver operating characteristic curve, calibration plots, decision curve analysis, and SHapley Additive explanation analysis.
Chronic kidney disease was present in 21.8% (390/1786) of participants, with male sex (OR 3.89, 95% CI 2.24-6.84, p<0.001), hypertension (OR 3.01, 95% CI 2.10-4.36, p<0.001), diabetes (OR 2.11, 95% CI 1.43-3.11, p<0.001), and atrial fibrillation (OR 4.83, 95% CI 2.31-10.2, p<0.001) identified as significant independent risk factors.
The XGBoost model achieved the highest predictive performance in the validation set with an AUC of 0.909 (95% CI 0.881-0.937), outperforming logistic regression (AUC=0.905), gradient boosting machine (AUC=0.897), support vector machine (AUC=0.878), and random forest (AUC=0.873).
SHapley Additive explanation analysis identified the uric acid-to-albumin ratio, hypertension, Gensini score, age, and systemic inflammation response index as the top five key predictive features.

Structured PICO

Can machine learning models using routine clinical data accurately identify the risk of undiagnosed chronic kidney disease in patients with coronary heart disease?

Population

1,786 adults (≥18 years) with clinically evident coronary heart disease (CHD) undergoing coronary intervention, without previously diagnosed chronic kidney disease (CKD) or renal insufficiency. Median age 65.0, 69.5% male, based in a tertiary care center in China. Key exclusions: non-atherosclerotic coronary artery disease, contrast agent use within 1 month before admission, and presence of other systemic diseases.

Intervention

Five machine learning models (logistic regression, support vector machine, random forest, gradient boosting machine, and extreme gradient boosting [XGBoost]) constructed using routine clinical and laboratory data to identify undetected CKD.

Outcome

Presence of undiagnosed chronic kidney disease (CKD) on admission, defined as an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73m2.

An XGBoost machine learning model utilizing routine clinical data, particularly the uric acid to albumin ratio (UAR), can accurately identify the risk of undiagnosed chronic kidney disease in patients with coronary heart disease.

Implication

ML models may aid undetected CKD screening in CHD; hypothesis-generating pending prospective validation.

Abstract

Purpose: Coronary heart disease (CHD) has a significant co-morbid association with chronic kidney disease (CKD), but identification tools for the risk of concomitant CKD in patients with CHD are still lacking. The purpose of this research was to construct machine learning (ML) models for identifying undetected CKD in CHD patients. Methods: 1786 CHD patients undergoing coronary intervention were retrospectively included. Lasso regression and multifactor logistic regression were used to screen feature variables. Five ML models, ie, logistic regression (LR), support vector machine (SVM), random forest (RF), gradient boosting machine (GBM), and extreme gradient boosting (XGBoost), were constructed. Participants were divided into the training set and validation set in a 7:3 ratio. The evaluation metrics included the area under the curve, calibration curve, and decision curve. Results: Totally, 1786 CHD patients were enrolled and split into training (70%) and validation (30%) sets. The prevalence of CKD was 21.8% (390/1786). Multivariate logistic regression analysis showed that men, advanced age, hypertension, diabetes mellitus, history of atrial fibrillation (AF), high Gensini, low hemoglobin, low plateletcrit (PCT), high triglycerides (TG), high lipoprotein(a) (Lp(a)), hyperkalemia, high uric acid to albumin ratio (UAR), high systemic inflammation response index (SIRI), low lymphocyte to monocyte ratio (LMR), and high apolipoprotein B to apolipoprotein A1 (ApoB/ApoA1) ratio were the key clinical and laboratory test indicators of CKD. The XGBoost model performed optimally in the validation set (AUC=0.909, 95% CI 0.881 -0.937). SHapley Additive explanation analysis identified UAR, hypertension, Gensini score, age, and SIRI as the top 5 key features. Conclusion: The XGBoost model constructed on routine clinical data was effective in identifying CKD risk in CHD patients, with UAR as a novel strong predictor. Decision curve analysis confirmed the clinical utility of the model, indicating that it may be used to guide decisions for enhanced monitoring and early intervention over a wide range of risk thresholds.