Mapping cancer patient experience across the care journey through a Bayesian Belief network approach

Cancer patients face a demanding journey spanning diagnosis, hospital and GP visits, treatment, and post-care recovery. Ensuring high-quality care across these phases is essential for a positive patient experience. Although prior studies have identified factors influencing cancer patient experience (CPE), few have examined their interactions or relative importance in an integrated framework. To address this gap, this study developed a Bayesian Belief Network (BBN) using multi-year National Health Service (NHS) survey data to predict patient satisfaction. Among four structural learning algorithms, the Augmented Naive Bayes (ANB) model achieved superior predictive accuracy and ROC performance. Sensitivity analysis identified Nurse and Staff Support and Homecare Support as the most influential factors of overall satisfaction, exhibiting the highest structural leverage. These results provide healthcare managers with a prioritized framework for intervention, demonstrating that targeting staff engagement and home-based support yields the highest return on patient well-being.