Operationalizing learning health system principles through PROMs implementation: A stakeholder-engaged decision-making process in cancer care.

Abstract

499 Background: Learning health systems (LHSs) prioritize stakeholder engagement to understand contextual factors that shape implementation processes and outcomes, but less attention is given to how engagement and co-design are operationalized in clinical settings. Patient reported outcome measures (PROMs) are being increasingly used to facilitate patient-provider communication on symptoms and functional status, necessitating engagement from patients and health care workers at all levels. This study describes the operationalization of stakeholder-engaged design and decision-making to advance PROMs implementation in complex oncology care settings. Methods: We used a multi-phase, stakeholder-engaged design embedded within a PROMs implementation initiative at a tertiary cancer center. The process was led by a PROMs-focused coordinating center with expertise in change management, clinical oncology, and implementation science. Phase 1 consisted of a mixed-methods formative evaluation of the program involving patients, providers, and clinic staff. In Phase 2, we synthesized Phase 1 findings with external evidence to identify implementation barriers and facilitators and generate candidate solutions. Phase 3 consisted of three focus groups with purposively sampled patient advisors, oncology providers, clinical staff, and system leaders. Solutions were prioritized and refined, considering patient experience and clinical workflows. Focus group data were examined using thematic analysis. Results: In Phase 1, 81 patients, 65 providers, and 31 clinic staff participated in surveys and interviews. Phase 2 identified five potential solutions that were perceived as high priority and impact within the health system. In Phase 3, 10-11 stakeholders participated in each focus group (2-3 patients, 3 system leaders, 2-3 clinical staff, 2-3 providers). Analysis identified six themes driving prioritization: 1) PROMs are insufficiently integrated into clinical workflow; 2) low completion rates limit their utility; 3) Uncertainty interpreting PROMs due to limited data regularity and context; 4) PROMs are more useful when meaning and trends are emphasized; 5) Low-burden supports are preferred to fit clinical workflows; and 6) PROMs discussions can support patient-provider trust. The two top-ranked solutions were: 1) increasing the visibility of PROMs within the EMR and 2) providing guidance to support providers in interpreting scores. Solution 1 was rated as acceptable, appropriate, and feasible; solution 2 was rated as feasible, but less acceptable and appropriate. Conclusions: Stakeholder-engaged co-design is essential for adapting PROMs implementation to oncology practice. Iterative, multi-stakeholder focus groups that address factors including feasibility, acceptability, and workflow fit can guide pragmatic adaptations within LHSs.