The integration of patient-specific computational models into clinical practice is a viable step forward in enhancing clinical treatment and improving the fundamental understanding of diseases. Clinically integrated models should be robust, provide meaningful, accurate, and timely feedback, and reduce costs associated with (a.o.) the time clinicians spend on diagnosis and treatment. One potential clinical application of such models involves the treatment of ventricular tachycardia (VT), a life-threatening heart rhythm disorder. To study mechanisms involved in VTs, new methods are required that enable high-fidelity analyses in a computationally efficient and accurate manner. Additionally, incorporating these methods into the clinical workflow should allow for the rapid output of indicators that support clinical decision-making. In this work, we aim to achieve these goals by first proposing a robust isogeometric cardiac analysis (IGA) model that leverages the advantageous properties of splines. The IGA cardiac model comprises several rigorously analyzed sub-modules: a multi-patch template-geometry module for single- and bi-ventricle geometries, a general geometry-fit module for sparse unstructured point clouds, and a cardiac-analysis module responsible for biomechanics. Special cases involving cardiac diseases, e.g., infarction, as well as treatment procedures, e.g., radio-frequency ablation, are also covered~. Second, we propose a reduced-order model (ROM) framework that utilizes the IGA model to efficiently generate training data. The ROM framework supports high-dimensional geometry input, similar to the IGA model, while providing instantaneous output results accompanied by confidence intervals. This ROM framework serves as an initial step toward model-based clinical decision-making.