Effective Palliative Care for Hospice Transitions.

Abstract

Importance: Variation in hospice use for patients with cancer exists between palliative care programs, but factors underlying this heterogeneity are unknown.

Objective: To identify hospital and palliative care program factors that may impact the ability of palliative care teams to affect hospice use in cancer.

Design, setting, and participants: This multicenter qualitative study included 6 palliative care programs in the United States. Using a cohort of patients with metastatic cancer who received specialist palliative care, a positive-negative deviance approach was used to select 3 programs with high and 3 programs with low performance on hospice use. In-depth site visits were conducted and included semistructured interviews with palliative care and oncology clinicians and patients as well as in-person observation. Interviews were conducted from February 2025 to January 2026. Site visits were conducted from May 2025 to January 2026.

Main outcomes and measures: Program-specific practices and their potential effect on hospice use for patients with cancer.

Results: A total of 93 interviews and 424 hours of in-person observation were conducted across 6 palliative care programs in the Midwest, Upper South, and Deep South regions of the United States. Interview participants had a median (IQR) age of 43 (37-47) years; 68 (73.1%) were female; clinicians had a median (IQR) of 11 (5-16) years of experience. There were notable differences in the relationships between palliative care and oncology teams between high- and low-performing programs. At high-performing programs, the relationship was characterized by mutual trust and respect, warmth and familiarity, and empathy for each other, whereas at low-performing programs, the relationship was characterized by a lack of trust and familiarity and negative regard for each other. Factors observed in programs with positive relationships included intentionality in relationship building, physical presence and proximity between teams, use of real-time communication, and the palliative care team having a service-oriented mentality and a reputation for reliable service. At high-performing programs, observable differences in care included collaborative care delivery, increased access and timeliness of palliative care, and earlier introduction of hospice by oncology clinicians.

Conclusions and relevance: In this multicenter qualitative study, differences in the relationship between palliative care and oncology teams were identified that may mechanistically affect end-of-life patient outcomes. These findings are important to ensure the delivery of high-quality care for all patients with cancer.