Electronic medical records form a major source of information regarding a patient's health history. Governments in the present take necessary steps to gather the patient's health history to carry out research and be prepared for any disease outbreaks at large to the citizens. Research has shown that the disease outbreaks are due to the lifestyle, the living conditions and the treatment undergone during the past. Medical literature states that many drugs whose complete safety profile unknown have been approved. Some drugs have shown serious adverse events (SAE), and subsequently withdrawn. There may be some drugs which still show adverse effects on the patients. This work makes an attempt to extract details regarding the drug administration from the electronic medical records (EMR) and employ the Bayesian classifier to find any SAE. It also analyses the various data mining techniques to find adverse events. The main advantage in using EMR is that they can be enhanced with powerful classification algorithms that can deal with images also. © 2015 Elsevier B.V., All rights reserved.