<p>Advancements in quantum computing have introduced new opportunities for enhancing machine learning models in healthcare applications such as heart disease diagnosis. In this study, we propose a quantum-enhanced machine learning framework (QuEML) that advances beyond prior works by integrating four distinct quantum classifiers, including Quantum Support Vector Classifier (QSVC), Quantum Neural Network (QNN), Quanvolutional Neural Network (QVC), and an ensemble Bagging-QSVC into a unified diagnostic pipeline. Unlike previous studies that typically evaluate a single quantum model in isolation, our framework demonstrates that a hybrid integration improves robustness, reduces computational time, and enhances predictive performance. QSVC achieved up to 90% classification accuracy on the publicly available dataset of 1,190 patient records. Notably, the ensemble Bagging-QSVC exhibited the most consistent performance, providing stability across folds. These results underscore the potential of multi-model quantum integration for advancing time-sensitive medical diagnostics and highlight QuEML as a novel, reproducible, and computationally efficient contribution to quantum machine learning in healthcare.</p>