DEVELOPMENT OF A DECISION SUPPORT AID FOR GOALS OF CARE CONVERSATION FOR HOSPITALIZED PATIENTS WITH CANCER

Purpose: Goals of care (GOC) conversations between oncologists and patients are often ill-defined due to poor physician and patient preparation and unclear goals. Hospitalization may further increase patients’ fear and confusion about GOC. Creating a tool to facilitate shared decision making is critical to support patients and their health care team in this process. In this study, we have collected data as the initial step in the development of a decision aid (DA) for GOC conversations with hospitalized patients with advanced cancer.

Method: A group of 14 content experts including diverse health professionals, patients and caregivers responded to a needs assessment to determine key questions to elicit involved stakeholders’ perceptions on relevant content for the DA. Questions were used to conduct semi-structured interviews with patients, a clinician focus group and a patient and caregiver focus group. All responses were audio-recorded and transcribed.

Result: There were 30 participants: 10 patient interviews and 10 participants in each focus group. Patients generally wanted as much information as possible about prognosis and choices, but acknowledged they were often overwhelmed and frustrated at not receiving information sooner. Most were clear about who they would include in their GOC discussions. Clinicians’ concerns included not destroying patients’ hope and uncertainty about how to achieve the delicate balance of delivering bad news while preserving hope. Clinicians also felt challenged by the time needed to conduct GOC conversations and how to initiate them. Patients and caregivers emphasized the value of a multimodal approach to GOC conversations. Individually, they varied in their readiness for and the ideal timing of GOC conversations, noting these could be early or late in their illness trajectory.

Conclusion: We see a need for personalization in a decision support aid in order to have clear goals of care. Patients differ in readiness to activate different parts of the decisional tree at different times. Our data will be used to guide design of a decision support aid to help hospitalized patients with advanced cancer have meaningful goals of care conversations with their clinicians.