RPD302 - Patient and Clinician Strategies to Increase Goals of Care Conversations: A Cluster Randomized Clinical Trial

Key Message : Across three VA healthcare systems, in a sequential cluster randomized trial, mailed PREPARE materials and phone outreach did not increase primary care goals of care documentation beyond clinician training and team support. Documentation varied markedly by site. Organizational factors, such as social work and primary care engagement, likely drove differences.

Abstract:

Background: Goals of care conversations (GOCC) add value for patients with serious illnesses.

Objective: Test whether randomized patient-engagement strategies, delivered with clinician implementation strategies, increase documentation of GOCC in primary care.

Methods: Pragmatic, sequential, multiple assignment cluster randomized clinical trial (SMART) in 3 VA health systems. Eligible patients had serious illness and were in the top 10th percentile of risk of hospitalization or death. In Stage 1 (6 months), all clinicians received a brief GOCC training, and their patient panels were randomized as a unit to receive usual care versus mailed PREPARE materials. In Stage 2 (9 months), all clinicians received training to encourage team processes adapted to local context, leadership engagement, and priority patient lists. The patient panels of clinicians who documented < 4 GOCC were re-randomized to PREPARE materials versus PREPARE materials plus phone outreach. The primary endpoint was the rate of primary care documentation of a GOCC during Stage 2.

Results: Of 2,307 Stage 2 patients (mean [SD] age 72.5 [9.6] years; 96.4% male), 136 (5.9%) had goals of care documentation by primary care (site 1, 39/1434 [2.7%]; site 2, 87/404 [21.5%]; site 3, 10/469 [2.1%]. 410/2,307 patients (17.8%) had any clinician/setting documentation. Stage 2 documentation did not differ by level of patient engagement based on primary care (adjusted risk difference, 0.016, 95% CI -0.007, 0.053; p=0.25) or any clinician/setting (p=0.47). In Stage 1, 24/2786 (0.9%) had GOCC with primary care, and 278/2786 (10.0%) had any clinician/setting documentation. None of the analyses showed significant differences in documentation for patient-engagement strategies. Site was consistently associated with documentation (p < 0.001) as well as age.

Conclusions: In this pragmatic, multisite SMART, sequential patient engagement strategies did not significantly increase outpatient GOCC. Substantial site variation suggested that organizational context may be critical to increasing GOCC for clinician implementation strategies.

References: Bekelman DB, Giannitrapani K, Linn KA, et al. Increasing goals of care conversations in primary care: Study protocol for a cluster randomized, pragmatic, sequential multiple assignment randomized trial. Contemp Clin Trials. 2024;145:107643. doi:10.1016/j.cct.2024.107643.