For decades, medical records have been under the ownership of hospitals. The process of transferring a record between two hospitals often takes days, and sometimes it is impossible to transfer records across different countries. Traditional EHR systems rely on a central authority to store, manage and control access to medical records. This centralized approach has several weaknesses, including single points of failure, data breaches, and data manipulation. Therefore, a need for a system that leverages deep learning algorithms to analyze patient health data and generate insights to improve patient outcomes where distributed ledger technology also ensures the immutability and transparency of the system, allowing for secure and auditable storage and sharing of medical records across multiple parties arises. The proposed framework DEHRSys utilizes IPFS to store patients’ medical records in a distributed file system, independent of any central entity. With the help of distributed ledger technology, true ownership of the medical records can be obtained. On the Hedera network, patients can access their records by interacting with smart contracts, forming a digital identity of the patient. DEHRSys operates in two phases, in the first phase, patients are assisted in securing their critical data over an off-chain storage system such as IPFS and the second phase trains and tunes the neural network model in order to predict a patient’s eligibility for a particular drug administration at a 99.4% F1 score.