Transcript x - Kentucky IPS Supported Employment

Evidence for the
Effectiveness of
Individual Placement and
Support Model of
Supported Employment
Updated 2/12/14 by Gary Bond
Presentation Outline
 Role of work in recovery
 Core principles of Individual
Placement and Support (IPS)
 Evidence for effectiveness
 Implementation strategies
Why Focus on Work?
 Most clients want to work!
 Most clients see work as an essential
part of recovery
 Being productive = Basic human need
 In most societies, typical adult role
 Most clients live in poverty
 Working may prevent entry into
disability system
Most Clients Want Jobs in
Competitive (Open) Employment
 Regular community job
 Pays at least minimum wage
 Nondisabled coworkers
 Not temporary or “make work”
 Job belongs to the client, not to
the mental health or
rehabilitation agency
Expressed Interest in Employment
Reported in 8 Surveys
% Want
to Work
Survey Population
Rogers (1995)
71%
statewide survey of people with mental illness
Bedell (1998)
69%
sheltered workshop participants
Mueser (2001)
61%
study of family intervention
McQuilken (2003)
55%
clubhouse members
Drebing (2004)
53%
Veterans in a VA-sponsored VR program
Woltmann (2009)
70%
clients in psychiatric rehabilitation
Frounfelker (2011)
72%
clients with co-occurring substance use
Ramsay (2011)
78%
young adults experiencing early psychosis
Study
Mean
66%
The Gap Between Need and Access
 2/3 want to work
 1%-2% have access to evidencebased employment services
(SAMHSA, 2009; Brown, 2012, Twamley, 2013)
 15% employed at any time
(Lindamer, 2003; Pandiani, 2012; Perkins, 2002;
Rosenheck 2006; Salkever, 2007)
Core Principles of the
Individual Placement
and Support (IPS)
Model
Traditional  IPS Assumptions
Assumptions
Screen for readiness
Stabilize first
Prepare stepwise
Take available jobs
Short-term support
Keep voc services and
mental health separate
Zero exclusion
Focus on client goal
Rapid job search
Client choice
Ongoing support
Integrate voc services
with mental health
IPS Principles
1.
2.
3.
4.
5.
6.
7.
8.
Open to anyone who wants to work
Focus on competitive employment
Rapid job search
Systematic job development
Client preferences guide decisions
Individualized long-term supports
Integrated with treatment
Benefits counseling included
Research Evidence
for Effectiveness
6 Day Treatment Conversions
to Supported Employment:
Common Study Design
 Discontinued day treatment
 Reassigned day treatment staff to
new positions
 Implemented new supported
employment program
 Compared to 3 sites not converting
Sources: Drake and Becker
Competitive Employment Rates Before and
After Converting from Day Treatment to IPS
70%
65%
60%
57%
50%
44%
39%
40%
30%
30%
25%
23%
20%
19%
20%
16%
13%13%
13%
13%
10%
5%
7%
5%
0%
0%
(n = 71)
(n = 112)
(n = 112)
(n = 31)
(n = 31)
(n = 30)
(n = 43)
(n = 30)
(n = 41)
Convert
Control
Convert
Convert
Control
Convert
Convert-A
Convert-B
Control
Drake '94
Drake '96
Bailey '98
Pre
Gold '98
Post
Becker '01
Cumulative Employment Rates for
6 Sites Converting to IPS
vs. 4 Control Sites
40%
35%
30%
Percentage 25%
Competitively 20%
Employed
15%
10%
5%
0%
37.5%
13.4%
Sites Converting to
IPS (n = 317)
12.4%
15.2%
Control Sites
(n = 184)
Before
After
Similar Results in All
Day Treatment Conversions
 Large increase in employment rates
 No negative outcomes (e.g., relapses)
 Clients, families, staff liked change
 Overall, all former day treatment
clients got out into community more
 Resulted in cost savings
20 Randomized Controlled
Trials (RCTs) of Individual
Placement and Support (IPS)
 Best evidence available on
effectiveness
 RCTs are gold standard in medical
research:
Random assignment = Participants
assigned by a flip of a coin
Principal
Investigator/
Year Pub
Study Site Location
Control Condition
Months
N
N
Follow(IPS) (Ctl)
up
Drake 1996
Mancester & Concord, NH Skills training, nonintegrated
18
73
67
Drake 1999
Washington, DC
Sheltered workshop
18
74
76
Lehman 2002
Baltimore, MD
PSR
24
113
106
Mueser 2004
Hartford, CT
(1) Brokered SE; (2) PSR
24
68
136
Gold 2006
Rural SC
Sheltered workshop
24
66
77
Latimer 2006
Montréal, Québec
Traditional vocational services
12
75
74
Bond 2007
Chicago, IL
Diversified placement approach
24
92
95
Burns 2007
6 cities in Europe
Traditional vocational services
18
156
156
Wong 2008
Hong Kong
VR referral
12
46
46
Killackey 2008
Melbourne, Australia
Traditional vocational services
6
20
21
Nuechterlein 2010 Los Angeles, CA
VR referral
18
46
23
Heslin 2011
London, UK
Usual Care
24
93
95
Twamley 2012
San Diego, CA
VR referral
12
28
22
Hoffmann 2012
Bern, Switzerland
Traditional Vocational Rehab
24
46
54
Davis 2012
Tuscaloosa, Alabama
Standard VA vocational rehab
12
36
35
Killackey 2012
Australia
Traditional vocational services
6
73
73
Drake 2013
23 sites in US
No additional services
24
Oshima 2014
Tokyo, Japan
Usual care
6
18
19
Michon 2014
4 cities in the Netherlands
Traditional vocational services
30
71
80
Bond 2014
Chicago, IL
Job club adaptation
12
43
44
1004 1051
Competitive Employment Rates in 20
Randomized Controlled Trials of
Individual Placement and Support
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
96
NH
12
AL
07
IL
04 12
08
CT AUST HK
12
08
06
CA AUST SC
IPS
99
DC
12
12
07
SWZ CA EUR
Control
Control 2
13
US
06
14
14
QUE JAP HOL
14
IL
02
MD
11
UK
Overall Findings for 20 RCTs
 All 20 studies showed a significant
advantage for IPS
 Mean competitive employment rates
for the 20 studies:
–57% for IPS (Median = 60%)
–24% for controls (Median = 26%)
(Each study weighted equally in
calculating mean rates)
Four Measurement Domains
of Employment Outcomes
 Job acquisition (% employed)
 Job duration (weeks worked)
 Hours worked per week
 Total hours worked/earnings
(Bond,Campbell, & Drake, 2012)
18-Month Competitive Employment
Outcomes in 4 Controlled Trials of IPS
IPS
Control
N = 307
N = 374
Job acquisition
216 (70.4%) 91 (24.3%)
Work ≥20 hrs/wk 128 (41.7%) 50 (13.4%)
Days to first job
140
212
Total hours
417.0
105.8
Total wage
$3,704
$1,001
p
d
<.001
<.001
<.001
<.001
<.001
0.96
0.67
-0.58
0.62
0.51
(Bond, Campbell, & Drake, 2012)
Competitive Employment Duration
2-Year Follow-up After IPS Job Start
N = 142
Total months worked 12.9
Months worked at 1st job 10.0
Number of jobs
1.9
(Bond & Kukla, 2011)
IPS for SSDI Beneficiaries
Mental Health Treatment Study
(MHTS) (Drake et al. submitted)
 Large multi-site controlled trial
 Recruited SSDI beneficiaries with
mental illness
–Most were not receiving community
mental health services
–Unemployed at baseline
MHTS Design
 Study population:
Cold calls to SSDI
beneficiaries on SSA rolls
 23 communities throughout US
 Randomized controlled trial
–Intervention group: IPS + other
mental health services
–Control group: Usual services
 Two-year follow-up
MHTS Sample
 14% of nearly 16,000 beneficiaries
contacted agreed to participate
 Final sample:
Intervention:
Control:
2,055
1,004
1,051
Monthly Paid Employment Rates in MHTS
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Intervention
Control
Overall Employment Rates:
Intervention: 61% Control: 40%
***
4
Month of Follow-up and Significance
Level
***
***
***
***
***
***
***
***
***
***
***
***
***
***
***
***
***
***
3
***
2
***
1
*
35%
30%
25%
20%
15%
10%
5%
0%
What About
Long-Term Outcomes?
 Many psychosocial interventions
produce improvements in
short-term outcomes
 BUT improvements often
disappear over the long term
 True for IPS?
2 Long-Term IPS Follow-up Studies
(Salyers 2004; Becker, 2007)
Steady Workers
(> 50% Follow-up)
52%
Length of Follow-up
10 years
Sample Size
74
Clients also reported: Greater selfconfidence and hopefulness, more energy,
less loneliness and boredom
Conclusion:
IPS Has…
Strong Evidence for
Effectiveness Across a Range
of Employment Measures,
Including Long-Term
Outcomes
Impact of IPS on
Recovery
Is Work Too Stressful?
 As compared to what?
 Joe Marrone: If you
think work is stressful,
try unemployment
Benefits of
Steady Competitive Employment
 Improved self esteem
 Improved social networks
 Increased quality of life
 Reduced psychiatric symptoms
 Reduced substance use
 Less use of disability system
Arns, 1993, 1995; Barreira, 2011; Bond, 2001; Burns, 2009;Drake,
in press: Fabian, 1992; Krupa, 2012; Kukla, 2012; McHugo, 2012;
Mueser, 1997; Van Dongen, 1996, 1998; Xie, 2005
Conclusion:
Steady Employment
Contributes to
Long-Term Recovery
IPS Implementation
 National EBP Project
(McHugo, 2007)
 IPS Learning Collaborative
(Becker. 2011)
Fidelity
 Fidelity defined as: the degree to
which a service model is
implemented as intended
 Working hypothesis: Better
implemented programs (with
higher fidelity) have better
outcomes
Implementation Factors in
National EBP Project
 After 2 years, 29 (55%) of 53 sites
implemented at high fidelity
–Training, staff attitudes, and
barriers not predictive
 Strongly predictive:
–Leadership at all levels
–Integration of new practices into
work flow
(Torrey, Bond, et al., 2011)
8 Keys to Implementing IPS
 End old ways of doing things
(e.g., day programs)
 Use fidelity reviews as guide
 Exercise leadership:
–Agency directors provide
resources
–Supervisors set firm behavioral
expectations
8 Keys to Implementing IPS
(continued)
 Hire and retain staff suited for IPS
 Provide field mentoring
 Collaborate closely with treatment
teams
 Track indicators you want to
improve
(Bond et al., 2008)
IPS Learning Collaborative
 Launched in 2002
 Grown to 130 programs in 14 states
 83% of programs joining since 2002
have been sustained
(Becker et al., 2011)
.
Features of IPS Learning
Collaborative
 Provide time-limited seed money
 Create position of IPS state leader
 State agency buy-in (MH and VR)
 Start small, grow gradually
 Commit to fidelity and outcome
reporting
 Nurture mechanisms for sharing of
experiences
IPS Fidelity and Outcomes
in 88 Sites in Learning Collaborative
IPS Fidelity
Category
N (%)
Mean Competitive
Employment Rate
Exemplary Fidelity
8 (9%)
44%
Good Fidelity
45 (51%)
39%
Fair Fidelity
30 (34%)
32%
Not Supported
Employment
5 (6%)
29%
(Bond et al., 2012)
Conclusions
 Employment is a key to
recovery
 IPS is an effective program to
achieve this goal
 Learning collaborative best
way to implement IPS widely