Tracking Prescription Adherence in Electronic Medical Records

Over 50% of patients are non-adherent and do not take their medications, causing $100 billion in lost revenue and harming patient health. With the use of med-ALs, a machine learning algorithm, clinicians can view the likelihood of a patient's non-adherence and see the top contributing factors for their non-adherence. The purpose of this fifteen-week project is to create a proof-of-concept prototype that shows this information and appropriate patient education tools in an already-existing electronic medical record. The project workflow included two parts: the first part included user research, such as a literature review, comparative analysis, multiple modeling techniques, personas, and workflows. The second part pivoted to user design and rapid iterative test evaluation (RITE) testing, which allowed for an agile redesign process. The resulting prototype and report were handed off to the client at the culmination of the project.