<p>This research enhances the healthcare IS literature by presenting a dual-level interpretable decision support framework tailored for organ transplantation. It incorporates a synergistic combination of elastic net, Bayesian belief networks, and Shapley additive explanations to construct a decision support system (DSS) that combines predictive capabilities regarding organ graft survival with elucidating contributing factors extracted using techniques from the explainable artificial intelligence (XAI) domain. This methodological innovation merges the computational prowess of machine learning with the clarity and ethical considerations of XAI to foster trust and transparency in healthcare AI applications. Based on a real-life dataset of 31,207 transplant records from the United Network for Organ Sharing (UNOS), the proposed DSS empowers healthcare professionals by offering a tool for refined feature selection, nuanced risk assessment, and personalized explanations, which enhance decision-making in organ transplantation. Crucially, the framework provides powerful dual-level interpretability: Bayesian Belief Networks offer transparent global insights into variable dependencies and population-level risks. At the same time, SHAP quantifies individual feature contributions to specific patient predictions, revealing why a particular outcome is predicted. We demonstrate the framework’s capability to support informed decision-making through simulated clinical scenarios and an interactive online tool, showcasing how its dual-level insights can be leveraged for practical use. This study contributes by introducing an interpretable DSS capable of dissecting and interpreting risk factors at both global and individual levels, advancing knowledge of determinants for graft longevity, and critically examining the pitfalls of traditional global explanation methods in ML, thus setting new standards for explainable analytics in healthcare.</p>