This research compares decision-making techniques for diagnosing diabetes, including Bayesian decision theory, rough set theory, and fuzzy soft set theory. It discusses applying Bayesian decision theory and rough set theory, utilizing the latter to approximate three-way regions for patient classification. The study aims to provide effective medication plans by considering uncertainty. It concludes that the Bayesian rough set theory-based method identifies boundary regions for further investigation. It suggests decision-making based on minimum overall cost and provides a framework for more precise medical diagnoses for diabetic patients.