Creating a tool for measuring preferences for treating blood cancers in older adults using implementation science

Abstract Background: Older adults with blood cancers now face complex treatment decisions that require weighing benefits and risks of various treatments. Despite continued calls for novel strategies to support decision-making for newly-diagnosed patients, especially among older adults, there is a paucity fit-for-purpose tools tailored for the decision context and adapted for key barriers like patient distress and perceived urgency of treatment decision-making. DISCOVERY is a theory-driven preference elicitation tool designed to support decision-making among older adults with blood cancers. DISCOVERY was developed using rigorous engagement and has demonstrated clinical feasibility and acceptability. We report here that DISCOVERY has been further optimized to enhance completion rates and facilitate longitudinal assessments to foster its implementation as part of a pragmatic RCT to evaluate its effectiveness. Methods: Human factors engineering using an iterative co-design process was used to adapt DISCOVERY. Data from usability testing with healthy volunteers, patients, caregivers, and clinicians were analyzed to update the method of preference elicitation, attributes included, and personalized feedback from DISCOVERY. The EPIS (Exploration, Preparation, Implementation, and Sustainment) implementation framework was used to guide implementation. Cognitive interviews with patients with newly diagnosed blood cancers were analyzed and used to understand barriers and facilitators to implementation. Iterative process improvement cycles were completed to develop a multi-phase implementation strategy for the tool. Post-implementation analyses were completed to improve feedback to patients and oncologists. Results: After 5 rounds of iterations involving a total of 121 interviews/usability sessions including patients (n=66), caregivers (n=20), healthy volunteers (n=30), and clinicians (n=5), several key refinements were made to DISCOVERY. First, the method of preference elicitation was changed from a multi-profile discrete choice experiment (DCE) to Best-Worst Scaling Case 1 based on usability testing demonstrating higher frustration scores and lower confidence in validity with the DCE. Second, the wording used for attribute descriptions were simplified and refined to improve understandability based on results from cognitive interviews. Third, a novel real-time evaluation of individual responses was developed to ensure validity of results. Feedback to patients was refined to include tailored question-prompts to facilitate shared decision-making and feedback to oncologists was updated to include results of validity testing. After several process improvement cycles, an implementation toolkit was created to target multi-level barriers to implementation. An electronic healthcare tool (EHT) including Discovery was developed containing several targeted improvements: an automated system was developed to allow participants to remotely consent, become randomized, and complete the intervention/control seamlessly to overcome barriers to recruitment and completion of assessments; a novel strategy was developed to provide results of the BWS to patients in real-time; and an interactive dashboard of all participants was created to facilitate real-time recruitment and retention to overcome challenges with data management and patient tracking. Implementation process maps were created to facilitate implementation. From June 2024 - June 2025, 625 older adults with blood cancers were invited to participate in the RCT (NCT06296368). 108 have consented to participate (17.3% recruitment rate; 43% leukemia/marrow failure, 33% Myeloma, 24% Lymphoma); 100% were randomized; 96% completed the baseline assessments which includes the intervention, if applicable; 83% have completed the primary outcome assessment. Conclusions: Implementation science and human factors engineering were used to refine a preference elicitation tool and to develop a strategy tailored to older adults with newly-diagnosed blood cancers. The implementation toolkit and EHT were successful at improving clinical implementation and overcoming known barriers to enrollment and completion of assessments. Future studies will assess the effectiveness and sustainability of using the tool in clinical practice.