|
The evaluation of the efficacy and impact of
case and disease management programs is one of the most
controversial topics within the care management field. One variable that limits the credibility of studies is that
members within any study and control groups vary with respect to the
stage and severity of their disease. A second confounding variable is the presence of
co-morbidities within the intervention and control populations. Most risk stratification algorithms fail to adequately
compensate for these two factors.
Employing Cohort Analysis
Cohort analysis is one of the most powerful
tools to determine the efficacy of program intervention over time. A cohort study is a one in which
subjects, who presently have a certain condition and/or receive a
particular treatment, are followed over time and compared with
another group who are not affected by the condition under
investigation. Because
RPNavigator categorizes members according to disease staging for
both single and co-morbid conditions, reliable and valid, matched
cohorts can be set up. In addition, due to its severity-adjustment capability,
RPNavigator can also easily measure changes in disease burden, case
mix and severity adjusted cost, hospitalization, and other outcomes,
on a longitudinal basis.
To provide valid comparisons between managed
and non-managed member groups, the two cohorts must be defined by
consistent characteristics that begin with the same disease, as well
as severity stage. For
example, one cannot compare a newly diagnosed diabetic with one that
has sustained associated degenerative processes such as kidney, eye
or circulatory damage. RPNavigator not only allows the end user to set up cohorts
based on disease, co-morbidities, and severity, but also additional
key variables, such as age, gender, product, care management and
provider specialty, etc.
Key Techniques and Findings
CareAdvantage, Inc. has used this technique
to compare and evaluate key outcomes associated with multi-variable managed and non-managed cohorts for its customers. The outcome measures that were tracked over time, using
T-test for significance analysis, include, but are not limited to,
the following:
Magnitude and rate of change in Burden of Illness
Incidence of hospital admissions
Incidence of emergency room visits
Complication rates
Actual cost performance against projected costs in
prior periods
Multiple cohort studies isolated
programs that had a statistically significant impact on the managed
population as opposed to the non-managed population.
Other programs demonstrated expenditures that did not result
in significant differences for the managed as opposed to the
matched, unmanaged population.
A finding associated with the latter cohort study also
suggested that significant resources may have been invested too
early in the disease process to have an impact on future resource
consumption and disease burden.
As a result of these studies, customers
can determine which programs are most cost-effective, whether the
programs are homegrown or outsourced to disease management vendors,
and at what stage of disease they are most efficacious.
Vendor Accountability
Cohort Analyses can be implemented with
RPNavigator so that managers can obtain a more objective evaluation
of performance than that provided by the Disease and Case management
vendors themselves, who may arrive at conclusions
which are influenced by selection bias, disease stage bias,
co-morbidity bias and a variety of other factors which skew
conclusions about how effective their programs have been.
RPNavigator returns control of program evaluation to the
purchaser, enhances vendor accountability, and assessment of return
on investment.
| 
