The essential need of database management systems to improve the quality of healthcare delivery makes the use of data mining techniques a phenomenon that cannot be ignored. Today, many healthcare providers are in the business of capturing and storing patients' personalized health information such as demographics, family history, allergies, medications, and diagnosis. This information is generally collected not only to make the healthcare practitioner well-informed about the health status of patients but also to improve the efficiency of care delivery and reduce waiting times. This paper aims to discover the applicability of data mining algorithms on clinical datasets. An experimental study was conducted to compare the performance of four different learning algorithms across four clinical datasets using 10 fold cross-validations.