Increasing Early Outpatient Goals of Care Conversations: A Sequential Multiple Assignment Randomized Trial (SMART)

AMANDA GLICKMAN, MD
University of Colorado Anschutz Medical Campus

Early goals of care conversations are recommended by the National Academy of Medicine, preferred by patients and families, and associated with less aggressive medical care at end of life, earlier hospice referral, and better caregiver bereavement adjustment. Conversations aim to explore and establish patients’ underlying values and priorities to guide decisions about specific medical interventions in the present and future. Despite the VA’s intent to increase early conversations in the outpatient setting by implementing the National Center for Ethics in Health Care’s Life Sustaining Treatment Decisions Initiative (LSTDI) in 2017, the majority of conversations still take place in the inpatient setting near end of life, when patients may be critically ill and unable to fully participate. Thus, further implementation strategies are needed. Using a sequential multiple assignment randomized trial (SMART) design, various adaptive strategies will be compared to better understand whether low- or high-intensity patient-facing and/or clinician-facing strategies improve conversations. This will help health care systems decide which approaches to use.

Study sites include the VA Eastern Colorado, Greater Los Angeles, and Palo Alto Health Care Systems. Advance practice outpatient clinicians with low rates of documented conversations who care for ?15 veterans with cancer, heart failure, COPD, interstitial lung disease, dementia, or end-stage renal or liver disease in the top 10th percentile for risk of hospitalization or death will be eligible for randomization.

This study employs a cluster (clinician-level randomization) SMART design. The primary outcome is number of documented conversations, measured 6 and 12 months after initial randomization. In stage 1, clinicians will receive written materials and one-time training. Their patients will be randomized to no patient strategy or a letter directing them to an interactive website to prepare for conversations. After 6 months, in stage 2, clinicians in either group who do not respond (<4 notes completed) will receive in-person team facilitation and lists of patients potentially appropriate for a conversation. Clinicians’ patients will be randomized to receive the letter or the letter with follow-up phone calls. Clinician and patient implementation strategy success or failure will be evaluated through surveys and semi-structured interviews. Data will be analyzed using quantitative and qualitative methods.

This study will help determine whether an implementation strategy requiring fewer resources could be effective, and to learn what sequence of strategies is effective overall and for specific patients, clinicians, or sites. Increasing goals of care conversations in the outpatient setting, earlier in the course of serious illness while the patient has decision making capacity and prior to a health crisis, will better align medical care with Veterans’ values.