JMIR Publications

ABSTRACT

Background:

Gastrointestinal bleeding (GIB) is a severe and potentially life-threatening complication in patients with acute myocardial infarction (AMI), significantly affecting prognosis during hospitalization. Early identification of high-risk patients is essential to reduce complications, improve outcomes, and guide clinical decision-making.

Objective:

This study aimed to develop and validate a machine learning-based model for predicting in-hospital GIB in AMI patients, identify key risk factors, and evaluate the clinical applicability of the model for risk stratification and decision support.

Methods:

A multicenter retrospective cohort study was conducted, including 1,910 AMI patients from the Affiliated Hospital of Guangdong Medical University (2005–2024). Patients were divided into training (n=1,575) and testing (n=335) cohorts based on admission dates. For external validation, 1,746 AMI patients were included from the publicly available MIMIC-IV database. Propensity score matching (PSM) adjusted for demographics, and the Boruta algorithm identified key predictors. Seven machine learning (ML) algorithms—logistic regression (LR), k-nearest neighbors (KNN), support vector machine (SVM), decision tree (DCtree), random forest (RF), XGBoost, and neural networks (NNET)—were trained using ten-fold cross-validation. The models were evaluated for area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, recall, F1score and decision curve analysis. SHAP analysis ranked variable importance. Kaplan-Meier survival analysis evaluated the impact of GIB on short-term survival. Multivariate logistic regression assessed the relationship between coronary heart disease(CHD) and in-hospital GIB after adjusting for clinical variables.

Results:

The RF model outperformed other machine learning models, achieving an AUC of 0.771 in the training cohort, 0.774 in the testing cohort, and 0.753 in the validation cohort. Key predictors included red blood cell count, hemoglobin, maximal myoglobin, hematocrit, CHD, and other variables, all of which were strongly associated with GIB risk. Decision curve analysis demonstrated the clinical utility of the RF model for early risk stratification. Kaplan-Meier survival analysis showed no significant differences in 7- and 15-day survival rates between AMI patients with and without GIB (P > 0.05). Multivariate logistic regression showed that CHD was an independent risk factor for in-hospital GIB (OR 2.79, 95% CI: 2.09–3.74). Stratified analyses by sex, age, occupation, marital status, and other subgroups consistently showed that the association between CHD and GIB remained robust across all subgroups.

Conclusions:

The machine learning-based RF model provides a robust and clinically applicable tool for predicting in-hospital GIB in AMI patients. By leveraging routinely available clinical and laboratory data, the model supports early risk stratification and personalized preventive strategies.