Title: Measuring Outcomes of a Psychiatric Rehabilitation Intervention for Medicaid Beneficiaries using Pro
1Measuring Outcomes of a Psychiatric
Rehabilitation Intervention for Medicaid
Beneficiaries using Propensity Sores
- Presented by Marsha Langer Ellison, Ph.D.
- Center for Health Policy and Research
- Univ. of Mass. Medical School
- Shrewsbury MA.
- Marsha.Ellison_at_umassmed.edu
- and
- Asya Lyass, Ph.D. (Cand).
- Center for Psychiatric Rehabilitation
- Boston University
- Boston, MA 02215
- 617-353-3549
-
- Presented at American Public Health Association
Annual Meeting. - Boston, MA, November 2006
2Background
- The Intensive Psychiatric Rehabilitation (IPR)
initiative evolved as a response to consumers of
mental health services who wanted additional
psychiatric rehabilitation services that would
assist them in overcoming disabilities resulting
from SPMI. Other stakeholders that promoted
implementation of IPR included rehabilitation
providers, the Iowa Department of Human Services,
and families. - The occasion of writing a renewed contract for a
behavioral health care carve-out allowed DHS to
include IPR as a performance indicator. At the
same time Iowa mental health services regulatory
code was amended to support implementation of the
new service. - The IPR initiative was developed and implemented
in 1997 by Magellan Behavioral Care of Iowa (MBC
of Iowa), a subsidiary of Magellan Behavioral
Health. MBC of Iowa is the contracted
administrator of state behavioral health services
described in the Iowa Plan for Behavioral Health.
3The Guiding Vision of IPRRecovery is Possible
for Mental Health Consumers
- Core elements of the IPR initiative
- The belief that people with SPMI can recover.
- A mission of enhanced role functioning through
personal goal development that aims to improve
success and satisfaction in key areas of life. - The empowering value of self-determination.
- The availability of readiness, skill, and support
development intervention strategies. - IPR is an aid to recovery and is one of the
essential components of a comprehensive community
system.
4IPR Service Description and Components
5IPR Implementation
- MBC of Iowa was the first managed behavioral
healthcare company to design and implement a
formal IPR initiative that incorporates
recovery-oriented principles as part of a public
sector carve-out. - The IPR initiative has worked directly with more
than 30 agencies in Iowa. Over 700 people have
been referred to the IPR program from 1998 to
2002. - Demographics of baseline data (738) show a group
that is predominantly white, of divided gender,
mean age 38, who have either schizophrenia or
major affective disorders. Most are not married,
live in supervised facilities, have earned a high
school diploma, and are either not working or
earn very little money.
6Service Delivery Requirements and Supports
- IPR services are provided by bachelors or
masters-level clinicians who meet State of Iowa
accreditation standards for Psychiatric
Rehabilitation Practitioners. - MBC of Iowa/DHS Sponsored Training including an
initial 120 hours of rigorous IPR training for
practitioners as well as ongoing training focused
on the needs of the IPR provider sites. - Joint Site Visits and case consultation to
providers by representatives from DHS and MBC of
Iowa for technical assistance and specific
implementation planning. - Formal Outcomes Evaluation of the IPR Initiative
co-sponsored by DHS and MBC of Iowa and conducted
by the Boston University Center for Psychiatric
Rehabilitation.
7IPR Outcomes Study Design
- Baseline to end-point comparisons were made for
IPR participants and sub-groups, in 3
rehabilitation outcomes (IAPSRS Toolkit, 1995). - Cross group comparisons were made with the
drop-out group. The drop-out group did not
differ from IPRs for most demographics, except
that they were 4 years younger, and had lower
monthly earnings than IPRs at baseline. - A quasi-experimental comparison was made of
changes in mental health service utilization and
related costs for IPR recipients and a control
group of matched treatment as usual mental
health service recipients. Control group members
were matched to experimentals on age, gender,
diagnosis and number and cost of in-patient
mental health services using propensity scoring
techniques. - Analysis used Medicaid claims for mental health
services made available to BU from MBC of Iowa.
8Pre and Post Changes in Employment Status Among
IPR Participants
- Significant increases in average employment
status from baseline to endpoint were found for
graduates and 18 month completers. There was a
significant decrease for drop-outs and no change
for the Intent to Treat group.
9Pre and Post Changes in Earnings Among IPR
Participants
- Significant increases in gross monthly earnings
were found for graduates and 18 month
completers. Intent to Treat group had a
significant but smaller increase. Drop-outs had
a non significant change in earnings. (All tests
were adjusted for baseline differences)
10Pre and Post Changes in Residential Status
- Significant increases in average residential
status from baseline to endpoint were found for
graduates, 18 month completers, the ITT groups
and for drop-outs, although the increase is
larger for IPRs.
11Comparison of Change Values Between Participants
and Drop-Outs
- There are significant differences between change
values in employment status and in monthly gross
earnings comparing Graduates and 18 Month
Completers with Drop-Outs (with adjustments made
for significant differences in baseline
earnings).
12Changes in Mental Health Services Units Used
Costs
- On the whole IPRs used more services at baseline
than controls (despite matching techniques).
Mean change values were adjusted for these
differences when comparing groups for change. - Combining all mental health services units used
(excluding IPR), control group subjects showed
significantly larger decreases in mental health
service units used costs compared with IPR
participants (Graduates, 18 month completers and
ITT).
13Change in All Mental Health Service Units Used
Baseline to Endpoint (over 2 months)
14Changes in Units Used Costs Selected Mental
Health Services
- For day treatment this pattern continues when use
and costs of selected services are compared
between IPR participants and controls. -
- The decreases found for in-patient use among IPR
and control sub-groups are not significantly
different. - There are increases in use of out-patient
services by IPR participants that are
significantly different from the decreases found
for controls. - For community services the decreases for controls
are not always significantly larger than
decreases for experimentals.
15Changes in Selected MH Services IPR Graduates
and Controls
Statistically significant difference in values
16Discussion
- Data demonstrate appreciable evidence for the
success of ICM on improved rehabilitation
outcomes especially for employment status and
gross monthly earnings. - Given that findings show an improvement in
residential status for all groups including
drop-outs, this suggests that there may be
changes in state housing policy affecting all
groups. However, findings for the larger
improvement for IPR participants support
continuation of the program. - A sizeable number of people were lost to the
study for a variety of reasons (e.g., missing
data, changing service providers). This
indicates some bumps in the road with
implementation of IPR as well as difficulties
with study data collection. The result is that
the outcomes for a large number of people are
unknown. - Further, there is a large group of formal
drop-outs (their dropout status was communicated
to the study). This result (common to many
mental health interventions) indicates that
further study is needed to learn when, for whom,
and under what circumstances IPR works.
17Discussion (continued)
- Findings for service utilization pose several
questions - Decreases in mental health service use and cost
that are seen for both IPR and controls suggest
that an underlying goal of managed care i.e., to
reduce unnecessary services or costs, is being
realized. - While psychiatric rehabilitation is often assumed
or hoped to promote integration and thereby
reduce use of acute mental health services, this
relationship has received little formal testing.
A theory of the impact of improved role
functioning on symptoms and service use is still
to be developed. - The relative shallower decrease or increase of
service use among IPRs suggests that IPR actually
improves service access or the acquisition of
services needed. Anecdotally providers suggest
that IPRs become service savvy through the
program.
18Implications for future policy and study
- Continued piloting of IPR in other states is
warranted in concert with testing using a more
rigorous design. - A recovery mission and rehabilitation focus is a
successful strategy to promote employment.
Services such as these can be integrated into
behavioral healthcare contracts. - A full-blown costeffectiveness study of IPR is
still outstanding. - While reduction of acute services use can be
interpreted as a good thing for individuals, a
mission of improved access or increased
penetration to an underserved population appears
unsuccessful. - Different strokes for different folks -- findings
suggest that a wide array of community-based
rehabilitation services with multiple entry
points is needed to promote outcomes.
19Citations
- Ellison, M.L. Anthony, W.A., Sheets, J.L., Dodds
W., Barker, W.J., Massaro, J.M. Wewiorski, N.J.
(2002) The integration of psychiatric
rehabilitation services in behavioral health care
structures A state example. Journal of
Behavioral Health Services and Research, 29(4),
381-393. - Ellison, M.L. Lyass, A., Anthiny, W.A. Massaro,
J. (November 2005) Outcomes Study of the
Intensive Psychiatric Rehabilitation Program in
Iowa 4th Interim Report. Center for Psychiatric
Rehabilitation, Boston University - Sheets, J. Yamin, Z. (1998) Intensive
psychiatric rehabilitation services A best
practice design for managed care. In 1999
Medicaid managed behavioral care sourcebook. New
York Faulkner Gray. - International Association of Psychosocial
Rehabilitation Services Research Committee
(1995). Toolkit for Measuring Psychosocial
Outcomes, Columbia, MD Author.