Elizabeth Kwong, Amy Cole, Amro Khasawneh, Carl Mhina, Lukasz Mazur, Karthik Adapa, Daniel R. Richardson
Abstract: Patients diagnosed with hematologic malignancies account for 10% of cancer related deaths. The growth of treatment options for hematologic malignancies has led to increased focus on treatment decision-making. However, little research has been done integrating patient-generated data and shared decision making to facilitate patient-clinician collaboration and understand patient preferences in cancer care. Our study aims to develop and evaluate data visualizations to support an electronic healthcare tool (EHT) to facilitate patient understanding of treatment outcomes using human-centered design methods. Data visualizations were developed and updated based on feedback from healthy volunteers, older adults with hematologic malignancies (patients), caregivers, and clinicians. We conducted a content analysis on the qualitative data gathered from participants. Our findings showed that users preferred easy to understand visualizations with simple, explanatory text compared to visualizations that were not immediately intuitive. Users also preferred visualizations that were more reflective of the individual's cancer treatment rather than a comparison to the patient population. Iterative improvements were made to the visualizations to reflect user feedback and will be used to inform the next iteration of visualizations for user testing in the clinic. This paper demonstrates the benefit of human- and user-centered design to iterate on data visualizations used to support a patient preference tool.