Data mining applications has significant impact on all parties (such as doctors, business executives and patients) involved in the healthcare industry [7]. Applying data mining techniques to electronic health records will help in retrieving undiscovered knowledge. But mining medical data is not as easy as any other type of data because of the complexity of medical data. Medical data has a number of unique features. Cios and Moore have given a detailed analysis of the unique features (Heterogeneity, ethical legal and social issues, statistical boundaries, and special status) of medical data in [8]. Our research identifies how the complex medical data will pose a problem in applying data mining to the EHRs domain as shown in Table II.