k-anonymity is one of the most important techniques to preserve privacy when we build anonymous data from medical data source. The k-anonymity prevent the identity disclosure of a specific individual from a dataset. However, k-anonymity does not assume situations where an adversary has prior knowledge of the source. On the other hand, the differential privacy is a strong concept of noise addition technique that guarantees privacy even if the adversary has any kind of prior information. But it does not protect the identity that the k-anonymity protects. Hence, there is a strong need to develop methods that satisfies both k-anonymity and differential privacy. There have been several studies to this end, but most of them are based on random sampling and none of them are considering publishing the whole data. We propose a method of publishing the whole anonymized data that satisfies both k-anonymity and differential privacy, and evaluate it on actual medical data.