Team Relationships Shaped Hospice Use Across Palliative Care Programs
Hospice use for patients with metastatic cancer varied in ways that could not be explained by patient need alone. In six U.S. palliative care programs, the dividing line was the working relationship between palliative care and oncology teams. Programs with stronger hospice performance were marked by trust, familiarity, and a shared sense of purpose; programs with weaker performance showed distance, mistrust, and strained collaboration.
The findings came from a multicenter qualitative study that paired high-performing and low-performing sites through a positive-negative deviance approach. Investigators selected three programs with higher hospice use and three with lower hospice use, then embedded themselves in each site through interviews and direct observation. The design was built to answer a practical question: why do some palliative care teams have a greater effect on hospice transitions than others, even when the patient population is similar?
What Distinguished High-Performing Programs
Across the higher-performing sites, the relationship between palliative care and oncology was not left to chance. It was built deliberately. Clinicians described mutual trust and respect, along with warmth and familiarity that made discussions about prognosis and hospice easier to start and easier to sustain. Real-time communication between teams was common, and physical proximity appeared to matter because it reduced friction and made consultation feel routine rather than exceptional.
A single sentence captures the contrast: hospice discussion started earlier when the teams worked as collaborators rather than separate silos.
High-performing programs also had a service-oriented palliative care culture. The team’s reputation for being dependable seemed to shape how oncology clinicians used it. If palliative care was viewed as responsive, accessible, and willing to meet the clinical need in real time, oncology clinicians were more likely to involve the service early and to raise hospice sooner. In practice, that meant more coordinated care delivery, faster access to specialty support, and earlier incorporation of hospice into the care plan.
The bedside effect is straightforward. When oncology clinicians trust that palliative care will respond quickly and align with patient goals, they are less likely to delay difficult conversations. That timing can influence whether hospice enters the picture in a planned way rather than as a crisis decision near the end of life.
Friction Between Teams Slowed Hospice Transition
Low-performing programs showed a different pattern. The relationship between specialties was marked by lack of trust, lack of familiarity, and negative regard. Those are not abstract interpersonal details; in this setting, they functioned like clinical barriers. If teams do not know one another well, they communicate less freely, and if they do not respect one another’s role, they are less likely to coordinate care around a shared trajectory.
The study’s observation-based design made those differences visible in day-to-day practice rather than only in interview narratives. The investigators spent 424 hours observing care across sites and conducted 93 interviews with clinicians and patients. The consistency of the theme across multiple sites strengthens the impression that relationship quality was not incidental. It was part of the operational structure of hospice referral.
Some programs did not appear to rely on a single heroic clinician or an isolated communication fix. Instead, the stronger sites used repeated behaviors that reinforced the same goal: intentional relationship building, repeated contact, and communication in the moment. Those practices seem mundane until their absence becomes the reason a patient reaches hospice too late.
Why Trust Between Services May Change End-of-Life Care
The mechanism is plausible from a clinical standpoint. Hospice referral for patients with metastatic cancer often requires the oncology team to acknowledge when disease-directed treatment is no longer serving the patient’s goals. That conversation is easier when palliative care is seen as an ally rather than a competing voice. Mutual trust can lower the threshold for consultation, reduce hesitation around prognosis discussions, and make hospice feel like part of an integrated care plan rather than an admission of defeat.
Physical proximity likely amplifies the effect. Teams that share space or interact frequently can exchange information quickly, clarify confusion before it hardens into disagreement, and build familiarity through repeated contact. Real-time communication may also prevent the delays that accumulate when messages move slowly through disconnected systems. In that environment, hospice is introduced earlier because the whole team is operating from the same understanding of the patient’s trajectory.
Several elements in the data point to the same conclusion from different angles. The high-performing programs were not distinguished by a novel drug, a new device, or a rare patient subgroup. They were distinguished by team behavior, culture, and workflow. Those factors shaped how quickly palliative care reached the patient and how soon hospice became part of the conversation.
Limits of a Qualitative Multicenter Design
This was a qualitative study, so it was designed to explain variation, not to prove causation. The programs were selected because they differed in hospice performance, and the study did not measure a randomized intervention. The findings do not establish that any single organizational feature will raise hospice use on its own. They do, however, identify a set of conditions that appeared repeatedly in stronger programs and were absent or weaker in lower-performing ones.
The study also focused on six U.S. programs in the Midwest, Upper South, and Deep South, so the patterns reflect a specific health system context. The interviews and observations were rich, but the sample size and setting mean the findings should be used as a guide for practice improvement rather than as a universal model. Even so, the amount of direct observation gives the results unusual depth for a qualitative report. Forty-two hundred and twenty-four hours of observation is not a trivial window into how clinical teams actually work.
Hospice Access Appeared to Depend on Team Culture as Much as Clinical Need
The central message is practical: hospice use in cancer may depend as much on the relationship between services as on patient disease status. When palliative care and oncology teams trust each other, communicate in real time, and view one another as partners, hospice appears to enter care earlier and more smoothly. When that relationship is marked by distance or distrust, the transition may stall.
For hospitals and palliative care programs, the implication is not limited to etiquette or morale. It touches the structure of end-of-life care itself. If the goal is more consistent hospice use for patients with metastatic cancer, then the work may need to begin with how teams interact, where they sit, how often they communicate, and whether each service sees the other as reliable.
The study leaves an operational question in plain view: which of these relationship-building features can be strengthened quickly enough to change hospice access for patients who need it now?