Watson CARE Insights
Watson Care Insights applies natural language processing and data visualization to clinical workflows and electronic health record (EHR) data. This allows primary care physicians to view their patient’s condition-specific treatment history temporally and access previous encounter notes, outcome readings, meditations, and insights. I worked to create an experience that offers the ability for users to see how treatments are affecting their patient's health outcomes that are relevant to a specific condition. Our solution allowed doctors to get the full story of why clinical decisions were made and how those decisions changed the course of their patient's treatment in a matter of seconds. This allows users to take in large amounts of historic summarized treatment information without having to forage through their electronic medical record. Additionally, I created a precision cohort Sankey diagram visualization that used outcomes from similar patients (based on condition-specific acceptance criteria) to provide evidence-based insight into how their patients might do on a certain medication.
This was my primary project at IBM. My team was made up of a design lead and a senior design researcher. I crafted UX wireframes, high fidelity visuals, created prototypes and handed off designs to front end developers primarily by utilizing Sketch and Invision. Our team received feedback from sponsor users by utilizing user testing, surveys, record reviews, observations and in-clinic rounding and behavioral analytics. I worked with the design researcher to find insights from our feedback and updated designs accordingly.