Preferred Group Practice Results
|
Published: October, 2001
|
The ALERT system provides monthly reports comparing results
for PGPs, and a graph is generated that depicts the results
for all groups with at least 50 cases with two data points.
The following graph presents the results for 13 groups that
meet this criterion of at least 50 cases. The chart also plots
the combined results for all groups and for all solo practitioners.
The Outcomes Benchmark is the overall expected effect size
for the entire sample (.33 effect size). The individual group
results are plotted by adding their individual Change Index
to this benchmark value. While the actual effect size for
each group will vary slightly from the results plotted here,
this method provides a fair means of comparison while conveying
information about the general magnitude of the effect size.
The results are plotted with a 90% confidence interval, meaning
that there is a 5% chance that the results are above the interval
and 5% below the interval. Based on this criterion, four of
the 13 groups exceed the benchmark at the 95% confidence level.
None of the groups are significantly below the benchmark.
These four groups are responsible for elevating the overall
results. If these groups are removed from the calculation
of aggregated effect size, the Outcome Index score for the
remaining groups and individual providers is .00, meaning
that the PBH outcomes are virtually identical to those in
the data repository.
A patient seen in one of these four groups averages over
33% more improvement per episode of care than patients treated
elsewhere in the network. The reasons for the superior results
are not apparent. These results present an opportunity to
determine how care is delivered in these groups and then promote
"best practices" within other groups.
Back to Top
 
|
