Practice Variability in the SHARING Choices Pragmatic Trial of Primary Care Advance Care Planning

Barriers to advance care planning intervention implementation and impact on outcomes at the patient, clinician, and health system levels are well-documented. Understanding practice-level variation in implementation and outcomes could elucidate relevant contextual factors and potential strategies for improving future implementation. We therefore examined practice variation and associations between processes and outcomes in the pragmatic trial of SHARING Choices, a primary care advance care planning (ACP) and communication intervention for older adults.

We conducted an explanatory sequential mixed-methods analysis of quantitative variation among intervention practices in trial processes and outcomes, and qualitative interview analysis of ACP facilitators' perceptions of variation in implementation. We evaluated variation in key processes: (1) reach (phone contact between ACP facilitator and patient/family) and (2) uptake (facilitator-led ACP conversations) and outcomes: (1) new electronic health record (EHR)-documented advance directives (ADs) at 12 months and (2) receipt of potentially burdensome care within 6 months of death for Maryland residents with serious illness who died. We examined practice-level correlations among processes and outcomes.

Variation among practices was substantial for key processes (ACP facilitator reach and uptake) and outcomes (new EHR AD documentation and potentially burdensome care at end of life; all p < 0.01). Processes of reach and uptake were significantly correlated with the outcome of new EHR AD documentation but not with potentially burdensome care at end of life. ACP facilitators perceived variation in practice engagement with the intervention related to practice relationships and teams, relative priority of ACP, and resources such as space.

Practice variation in processes and mixed associations with outcomes highlight pragmatic trial implementation challenges and the impact and complexity of ACP. Future ACP trials should consider accounting for and evaluating practice variation in study design, implementation, and analysis.