AN E-PRESCRIPTION SYSTEM FRAMEWORK TO LOWER PRESCRIBING AND MEDICATION DISPENSING ERRORS USING BLOCKCHAIN, MACHINE LEARNING, AND NEAR FIELD COMMUNICATION (NFC)

Abstract

Medication errors are among the most significant risks facing the pharmaceutical industry. These errors can result from various issues such as a heavy workload, the misinterpretation of a prescriber’s handwriting, or dispensing the wrong medication to the wrong patient. Hence, many countries have implemented e-prescription systems trying to reduce medication errors. In addition, researchers have proposed several mobile apps that use near-field communication (NFC) to manage patients’ medication intake instructions and remind patients about intake times. We conducted a comparative review involving eight countries implementing e-prescription systems. One of the challenges and limitations of the reviewed systems is the availability of medication histories to the parties involved in the system. Moreover, the clinical decision support (CDS) systems are not part of the e-prescription system, and they do not provide quality, valuable alerts that would help avoid potential harm from the prescribed medication.

The objective of the thesis is to develop a framework for an e-prescription system that aims to enhance the security, privacy, availability, and reliability of the ePrescription information while prescribing and dispensing medication. Therefore, the framework benefits from the characteristics, features, and advantages of the technologies blockchain, Machine Learning (ML), Near Field Communication (NFC).

The framework will use blockchain technology to make the patient’s information secure, private, and available to the involved parties. Moreover, to enhance medication safety, we proposed using machine learning (ML) to detects any serious outcome caused by anomalies in the e-prescription before submitting it. Finally, using a mobile application enabled with NFC technology to transfer the patient’s token Id to the pharmacy management system will verify the patient’s identity and control the access to the patients’ ePrescription information. The application will help to manage and display the ePrescription information when needed to the patients. We developed a proof-of-concept and evaluated the reliability and performance of the blockchain and ML modules. Further, we conducted a user study of the NFC mobile application to evaluate its usability. Lastly, we conducted a survey study to understand better the strengths and shortcomings of the proposed features in the framework (i.e. blockchain and ML). The results are promising that the framework might help mitigate medication errors at different levels, starting from prescribing until dispensing the medications to the patients.