Does the timing of initiating community-based palliative care affect the time spent in hospital at the end of life for cancer patients? This was the question that spurred research my colleagues David Youens and Rachael Moorin and I recently published in the Journal of Pain and Symptom Management.
The research used whole-of-population individual level linked data, which allows records from mortality and cancer registries, hospitals and, in this case, a Perth-based palliative care service provider to be linked.
For this study we looked at people dying due to cancer in Western Australia between 2001 and 2011. This gave us a population-level view of how people used community-based palliative care, and how often and for how long the same people were unexpectedly admitted to hospital in the final few months of life.
Access to community-based palliative care – an umbrella term for multi-disciplinary palliative care delivered in a person’s usual place of residence – is not restricted in WA in terms of time before death.
Transitioning to palliative care does represent a change in treatment goals, from curative to patient comfort and symptom management.
Thus, it is unsurprising that of the 16,439 people in our study, 64% accessed community-based palliative care in the last three months of life. Because some people accessed earlier than this, we had an opportunity to explore if earlier access was associated with fewer hospital admissions at the end-of-life.
This is valuable because this is not possible to evaluate across all jurisdictions. For example, the US Medicare program restricts access to hospice care to an expected six months of life remaining. We found that initiation more than six months prior to death (i.e. earlier initiation) was associated with fewer unplanned hospitalisations.
While the difference was small at an individual level, it was important when viewed at the population level. Because we were observing what happened in the past, there are other reasons apart from use of community-based palliative care that hospital use might be different. To account for this, we adjusted for these other factors as much as the available data allowed.
This increases our confidence that the difference is linked with timing of initiation. Overall, estimated hospital costs were less among those who accessed the service early and the trends for hospital admissions mirrored emergency department presentations. However, people initiating community-based palliative care earlier tended to spend longer in hospital when they were admitted.
This could be for reasons including more complicated admission reasons that could not be managed by community providers Our findings are broadly consistent with emerging literature from Australia, Canada and Taiwan. They also build on previous work from WA, which found that community-based palliative care was associated with fewer hospitalisations and a greater likelihood of dying outside of hospital.
A review of community-based palliative care, which included 23 studies, found that its availability reduced symptoms at the end-of-life. Our single study should be interpreted in the context of evidence available and considering relevant community attitudes towards palliative care.
In jurisdictions where community-based palliative care is restricted to a certain expected time to death (usually for cost-saving reasons), our results argue that this may not always yield a net cost saving.
ED: Cameron Wright, David Youens and Rachael Moorin are a members of the School of Public Health at Curtin University.