JMIR Publications

ABSTRACT

Background:

Major adverse renal and cardiovascular events (MARCE) have been proposed as a comprehensive endpoint reflecting both cardiac and renal outcomes in patients with acute myocardial infarction (AMI) and chronic kidney disease (CKD). However, existing prediction tools lack accuracy and generalizability for this high-risk population.

Objective:

This study aimed to develop and externally validate explainable machine learning models and a nomogram for predicting in-hospital MARCE among patients with AMI and CKD.

Methods:

We enrolled patients with AMI and CKD who underwent emergency treatment at the First Affiliated Hospital of Wenzhou Medical University as the discovery cohort and at Wenzhou TCM Hospital of Zhejiang Chinese Medical University eternal validation cohort. Clinical features, including demographic details, comorbidities, laboratory results, echocardiography measurements, procedural details, and medication parameters, were available for the enrolled patients. The number of important variables was determined using the least absolute shrinkage and selection operator (LASSO) regression. The following model development methods were introduced: extreme gradient boosting (XGBoost), random forest (RF), decision tree (DT), and logistic regression (LR). Model performance was accessed using receiver operating characteristic curve (AUC) scores, the accuracy, precision, recall, and F1-score. We additionally applied SHapley Additive exPlanation (SHAP) values to explain the XGBoost model. A nomogram model based on logistic regression to facilitate the prediction of MARCE is freely available at https://abcde-l.shinyapps.io/DynNomapp/.

Results:

In the training set, we identified 6 important variables. The 6 variables were Killip class, lactate, estimated glomerular filtration rate (eGFR), use of diuretics, use of beta blockers, and use of renin-angiotensin system inhibitors (RASI). In the test set, the XGboost model exhibited the highest AUC (0.882, 95% confidence interval [CI] 0.816–0.947), followed by the LR model (0.875, 95% CI 0.799–0.950), RF model (0.870, 95% CI 0.800–0.940), and DT model (0.814, 95% CI 0.720–0.886). The SHAP values successfully illustrated the positive or negative contribution of the 6 variables attributed to the output of the XGBoost model and the individual variable’s effect on model prediction. A nomogram model was constructed based on logistic regression and was further externally validated with an AUC of 0.873 (95% CI 0.714–0.926) in the validation set. 

Conclusions:

Our study identified 6 important predictors and provided a series of prediction models to enhance the risk stratification of MARCE for patients with AMI and CKD. These models might aid in risk stratifying and choosing proper therapeutic strategies in the management of patients with AMI and CKD.