Assessing fidelity of interventions to foster shared decision making in clinical encounters: Development of an implementation-focused fidelity checklist for use with electronic conversation aids.

Objectives: Conversation aids (CAs) support shared decision making (SDM) by presenting options and enabling the collaborative co-creation of care plans in clinical encounters. Fidelity assessment in implementation studies may require distinction between implementation of SDM and of the CA, as suboptimal CA implementation can disrupt SDM. We aimed to develop and pilot test an observer-based instrument for assessing implementation fidelity in a study of SDM using an electronic CA. Methods: Candidate items were guided by definitions of intervention components and constructs from (a) the Agency for Healthcare Research and Quality's SHARE Approach, and (b) a conceptual framework of implementation fidelity. The instrument and coding manual were developed using video-recorded encounters. We assessed response distributions and inter-observer agreement. Results: The final instrument included 10 items assessing clinician behaviors and 12 potential moderators (3 tool integration, 4 quality of delivery, and 5 participant responsiveness). Observer agreement was moderate to very good on all items. It was lowest for the item assessing whether the tool displayed risk with auto-populated patient data like age and cholesterol. Conclusion: An adapted conceptual framework for implementation fidelity guided checklist development in key domains related to CA implementation and how clinicians use CAs to foster SDM. Innovation: This study demonstrates the feasibility of observing implementation fidelity of SDM interventions and introduces an approach that distinguishes fidelity of intervention implementation from the clinical practice of SDM. This distinction is critical for interpreting effectiveness outcomes in hybrid effectiveness-implementation studies and targeting implementation strategies to bolster routine adoption of SDM.