ICES | Primary Care Models in Ontario English - page 33

Comparison of Primary Care Models in Ontario by Demographics, Case Mix and Emergency Department Use, 2008/09 to 2009/10
Networks. CHCs serve disadvantaged
populations as a consequence of their
community mandate, but the reasons why
they are associated with lower than expected
ED visits is not known. Possible factors
include health promoting services,
community engagement, longer appointment
duration, the presence of long-established
interdisciplinary teams, extended hours,
client preferences, provider practice styles,
practice location in relation to existing
services and the nature of appointment
scheduling. The mechanisms responsible for
lower than expected ED visits are important
for health policy decisions and require further
investigation, as does the efficiency of CHCs
in relation to outcomes.
12
24
ICES
LIMITATIONS
This report has a number of limitations that
should be considered when interpreting its
findings.
Among CHC clients, 11.5% did not have a
health card number and could not be linked to
Ontario databases. The CHC profile is
therefore not representative of all CHC
clients, but it does include close to 90% of
clients seen in the previous two years. The
characteristics and patterns of ED use of
those lacking health coverage requires
further investigation.
Nurse practitioners often see patients who do
not see physicians. Nurse practitioner
encounter data were available for CHCs but
not FHTs and for that reason, nurse
practitioner encounters (representing 22% of
clients in CHCs) were excluded. Inclusion of
nurse practitioner data may have resulted in
lower levels of morbidity and comorbidity for
CHCs and FHTs if nurse practitioners had
practices that were less complex than those
of physicians. In keeping with that
assumption, inclusion of nurse practitioner
data for CHCs resulted in a SAMI value of 1.67,
lower than that for physician visits only (1.84)
but still considerably higher than that in other
models (data not shown).
Patients and clients who died before April 1,
2010 were excluded from the analysis. This
may have underestimated the complexity
within all of the models because those who
died may have had complex problems and
high resource utilization needs during the
period prior to death.
Income quintiles represent area-level income
and may not accurately reflect income levels
of individuals. They are very commonly used
in health services research, however, and do
correlate with individual-level income.
The completeness of data may have been an
issue at CHCs, especially those that more
recently began to use electronic records, and
it may also have been an issue in capitation
models (FHN, FHO, FHT) that shadow bill, as
the completeness of shadow billing is not
known.
These analyses are cross-sectional and do
not help to distinguish whether physicians
altered their practices or mix of patients as a
result of joining a model of care or whether
the patterns seen here were pre-existing. An
earlier comparison of ED visits in FHNs and
FHGs found that higher rates of ED visits
were pre-existing in FHNs
31
and were not
likely the result of changing practice.
1...,23,24,25,26,27,28,29,30,31,32 34,35,36,37,38,39,40,41
Powered by FlippingBook