The Results
|
Published: October, 2001
|
This report summarizes results for the ALERT outcomes management
system and examines the validity of the assumptions underlying
the design of the system. The Outcome Index results presented
are for the PBH commercially insured population, and breaks
out the results by age group and severity levels. Note that
one of the severity ranges is designated "Normal."
The target outcome (average expected effect size using the
case mix formulas) is .33. Since the actual effect size is
.36, the Outcome Index is .03 (actual change – expected change).
This result is above the target at the 95% confidence level.
This indicates that PBH results are equivalent if not better
than the benchmark
The results for children exceed the benchmark to a greater
degree than the results for adults (Change Index of .09 as
opposed to .02). The children’s outcomes are above the benchmark
at the 95% confidence level, and this pulls up the overall
results for all members. The adult results do not exceed the
benchmark at the 90% confidence level, and so they can be
interpreted as being essentially consistent with expected
change.
The LSQ and YLSQ displayed the expected psychometric properties.
The case mix adjustment formulas used to establish the benchmark
for outcomes have been found to be valid for the PBH commercially
insured treatment population. Encouragingly, all indications
are that these norms are also appropriate for PBH’s public
sector population.
Overall results for PBH’s treatment population are at least
on par, and possibly superior to, the national benchmark established
by the data repository for the outcome instruments. None of
the PGPs have outcomes below the target (at a 95% confidence
level), while four of the groups have results that are above
the target (95% confidence). Members treated by one of these
groups average 33% more improvement than the target.
Back to Top
 
|
