Previous studies conducted by the investigators have shown that the Hello game demonstrates successful advance care planning (ACP) engagement in general populations, but has yet to be tailored to meet the unique needs of patients with advanced cancer and their caregivers. Outlining their care preferences by engaging in ACP is an important aspect of care according to patients with advanced cancer. However, only 55% of patients with advanced cancer patients have participated in ACP. These patients have substantial bio-psycho-social stressors that distinguish their ACP needs from others. Tailoring established interventions that foster high quality conversations about medical treatment preferences and end-of-life issues (such as the Hello game) is critically important for this population given its unique needs. As evidenced by qualitative interviews with \>200 participants, the Hello game creates a safe environment for sensitive conversations about end-of-life issues and inspired sharing of rich perspectives, with no reported adverse events, excessive burden, or negative emotional effects. That said, the intervention must be adapted for patients with cancer, particularly those with advanced cancer and their caregivers.
Additionally, while several effective ACP interventions exist (including Hello), how best to disseminate these interventions has not been rigorously or systematically studied. In other ongoing and previous studies, the investigators have demonstrated success in both engaging individuals living in underrepresented communities in ACP and successfully enrolling them in interventional research about ACP. The investigators credit these successes to their unique intervention delivery approach called the Community Based Delivery Model (CBDM). The CBDM overcomes key barriers to ACP (such as healthcare distrust, resistance, and hesitancy to discuss end-of-life issues) by leveraging established community connections to recruit participants to participate in ACP interventions as well as research. In the CBDM, trusted community "hosts" (who are leaders from local hospice organizations, senior centers, health agencies) invite participants to attend an ACP event. They introduce the research team to the attendees who may choose to participate in the ACP activity, the research, or both. Hosts are provided with marketing materials and utilize their community network channels to advertise the event. This model allows for research to be conducted more easily within hard to reach and underserved communities such as Black, Hispanic and rural communities- much like the most remote communities across the Penn State Cancer Institute's 28-county catchment area.
Patients with cancer, however, are unique, and may require an alternative approach that involves partnering with their oncology care team to introduce the concept of ACP and encourage participation in ACP and research. Notably, there is evidence that patients are more likely to engage in ACP when recommended by their physician, so how best to approach ACP for cancer patients is unknown. A common approach to ACP intervention research is to use a Healthcare Based Delivery Model (HBDM). In contrast to the CBDM, the HBDM is positioned within the healthcare system (i.e., clinic-based recruitment) as the ACP intervention is recommended by the patient's clinician (rather than through community-based outreach groups). For this intervention delivery approach, research assistants support interactions between clinicians (providers or nurses) to find appropriate patients and garner interest in performing ACP. This model is commonly used to recruit patients for clinical trials, including ACP interventions. For patients with cancer, the HBDM may have some advantages over the CBDM, given the close bonds that form between a patient and clinical care team as they interact frequently during active treatments such as infusions and radiation that often span several hours and weeks. Leveraging these therapeutic relationships may support greater acceptance of opportunities to broach ACP than a community-based model, but this remains unknown.
