Decision Support Training for Advanced Cancer Family Caregivers: The CASCADE Factorial Trial

A priority focus in oncology and palliative care is preparing the 3.2 million U.S. family caregivers of persons with cancer to effectively partner with patients in health-related decision-making from diagnosis to the end of life, particularly in underserved settings. Over 70% of patients with cancer involve relatives, friends, and partners in healthcare decisions, including choices about cancer treatments, surgery, transitions and location of care, accessing palliative and hospice care, and many others. Patients making healthcare decisions with unprepared family caregivers may experience inadequate family decision support leading to heightened distress and receipt of care/treatments inconsistent with their values and preferences. This in turn may increase distress for family caregivers. Hence, there is a critical need to train cancer family caregivers to effectively support patient decision-making; however, few interventions exist that enhance caregiver skills in providing decision support.

We have developed and successfully pilot tested CASCADE (CAre Supporters Coached to be Adept DEcision partners), a lay navigator-led, telehealth early palliative care intervention to train advanced cancer caregivers how to effectively partner with patients in health-related decision-making. Evolving out of our prior early palliative care caregiving interventions, decision partnering relevant content for family caregivers includes principles of effective social support in decision-making, decision support communication, and Ottawa Decision Guide training; however we do not know which of these components and component interactions influences patient and caregiver decision-making outcomes. Traditional research approaches typically treat interventions as "bundled" treatment packages, making it difficult to assess definitively which aspects of an intervention can be reduced, eliminated, or replaced to improve efficiency. Using traditional research methods (e.g., two-arm randomized controlled trials that test new features one at a time) requires conducting multiple studies, which is an exorbitantly expensive and time consuming process. This paradox prompts us to consider methodologies that offer a more efficient way to test multiple intervention components simultaneously. The research question we raise is: "What set of decision support training components best optimizes family caregiver support of patient healthcare decision-making?"

Using a highly innovative methodology, the Multiphase Optimization Strategy (MOST), the purpose of this study is to conduct an optimization trial to develop and refine the decision support skills of family caregivers of persons with newly-diagnosed advanced cancer (CASCADE: CAre Supporters Coached to be Adept DEcision partners). Using a 2x2x2x2 factorial design, 256 family caregivers of persons with newly-diagnosed advanced cancer will be randomized to receive one or more lay coach-delivered decision support training components, based on the Ottawa Decision Support Framework and Social Support Effectiveness Theory: 1) psychoeducation on effective decision support and social support principles (1 vs. 3 sessions); 2) decision support communication training (yes vs. no); 3) Ottawa Decision Guide training (yes vs. no); and 4) monthly follow-up (1 monthly follow-up call vs. monthly follow-up calls for 24 weeks).

The specific aims of this study are to:

Aim 1: Identify CASCADE decision support training components (main effects/interactions) that contribute meaningfully to improvement over 24 weeks in the primary outcome, patient-reported decisional conflict, measured by the Decisional Conflict Scale and secondary outcomes, including:

Caregivers: a) distress (Hospital Anxiety and Depression Scale),42 b) quality of life (PROMIS Global 10) Patients: a) distress (Hospital Anxiety and Depression Scale), b) quality of life (PROMIS Global 10), c) healthcare utilization, d) advance directive completion. A component will be considered effective if its presence produces a statistically significant main effect or two-way interaction of Cohen's d ≥.30.

Aim 2: Apply results obtained in Aim 1 to build: (1) an intervention made up of only active components and (2) a second intervention that is optimized for scalability and cost. Intervention (1) will be made up of the best set of component and component levels, based on Aim 1 results. Intervention (2) will be comprised of the set of components for the smallest cost that still yields a clinically meaningful effect in outcomes (d≥.30).

Exploratory Aim: Explore mediators and moderators (e.g., sociodemographics, decision self-efficacy, social support) of the relationship between intervention components and patient and caregiver outcomes.