| |
Hospice is
a defined set of services whereas palliative therapy is usually
defined as care with the intent of symptom control (or palliation).
Three broad options can be used to quantify palliative care using
administrative data:
- Consider
any care in the final months of life to be palliative. While
this is clearly a retrospective approach to end-of-life care,
there are situations where it is appropriate. A word of caution,
however: there will always be patients, even those with terminal
illnesses, who die unexpectedly. Some of these patients may
be receiving care that is considered to be curative in the most
basic sense. Defining treatment based on outcome may result
in some very nonsensical findings (e.g. aggressive surgery as
palliative treatment).
- Carefully
consider each treatment and classify it as: palliative, curative,
or "cannot be determined". This is probably the most
accurate in terms of connection to the practice of medicine.
As noted earlier, one cannot explicitly discern intent from
claims. It may be possible to take a 'rational doc' approach.
That is, a rational practitioner would only place a patient
on drug A if they believed the patient had a condition that
was treatable by drug A. From this, certain medications that
have a particular indication (e.g. pain control) may easily
be defined as palliative. Certain doses of chemotherapy or radiation
may follow a pattern that is more consistent with a goal of
palliation than cure. There will likely be a number of services
that cannot be classified and will need to be grouped as indeterminate.
The primary disadvantage to this approach is that it will need
to be repeated condition by condition, service by service. Thus,
this will not be a particularly efficient way to study palliative
care.
- Only consider
care provided in hospice programs to be palliative. This will
likely result in an under-ascertainment of end-of-life care
because much palliative and even hospice-like care is provided
outside of hospice care. This includes care prior to hospice
entry and actual end-of-life care for persons who, for a variety
of reasons, never formally enter a hospice program. This approach
would result in conclusions that may not match clinical intuition--e.g.
only 50% of persons dying of cancer receive any palliative care
(Virnig, 2002). A more
accurate conclusion might be that 50% of persons dying of cancer
receive care in a formal hospice program. Even defining hospice
is not as straightforward as one might expect. For hospice to
be reimbursed under Medicare, the care must be provided in a
Medicare certified hospice program. Alternate measures of hospice
availability (e.g. American Hospital Association annual survey)
provide estimates of hospice availability that do not line up
with hospice availability as estimated via presence of a Medicare
certified hospice. Medicare recipients receiving care in a non-certified
hospice program will have their care described in other ways
to allow for billing (e.g. routine home care through a home
health agency). This care may well fit our ad hoc definitions
of hospice but not fit our rigid definition of formal hospice
care. The bottom line is that this is the easiest definition
to implement, but probably under-represents the actual receipt
of services that can appropriately be defined as palliative.
Clearly, none
of these options is ideal. The correct choice necessarily involves
choosing among advantages and disadvantages and considering the
likely inference to be drawn from the study.
|