Transcript Study Overview

Are Family Treatment Drug Courts
Effective?
Results from a Four Site National Study
National Association of Drug Court
Professionals Meeting
June 12-15
Washington, D.C.
www.npcresearch.com
This study was funded by the Substance
Abuse and Mental Health Services
Administration, Grant No. 270-02-7107
NPC Project Team
Beth Green, Ph.D., Principal Investigator
 Michael Finigan, Ph.D., Co-Principal
Investigator
 Sonia Worcel, M.S., M.A., Project Director
 Carrie Furrer, Ph.D., Research Analyst
 Scott Burrus, M. A., Research Coordinator
 Jennifer Aborn, Data Specialist
 Becky Jones, Data Specialist

NPC Research
June 2007
2
Acknowledgements

NPC Site Research Coordinators and
Data Collection Staff

FTDC judges, coordinators, and other team
members at all four sites
 FTDC parents
 State and county child welfare and treatment
agencies in California, Nevada, and New
York
NPC Research
June 2007
3
The FTDC National Evaluation

A study conducted by NPC Research

A federally funded national evaluation
funded by the Center for Substance
Abuse Treatment, SAMHSA

Four FTDCs in the study: Santa Clara,
CA; San Diego, CA; Washoe, NV;
Suffolk, NY
NPC Research
June 2007
4
Primary Research Questions

I. Outcome Analysis
How do treatment, child welfare, and court outcomes
differ for families processed through FTDCs as
compared to traditional child welfare case
processing?

II. Black Box Analysis
How do FTDCs work? What factors influence
program outcomes?

III. Qualitative Analysis
What features of drug court most influence parents’
recovery and ability to make progress on the case
plan?
NPC Research
June 2007
5
General Conceptual Model
for FTDC Effects
Intermediate Parent
Outcomes
FTDC Experience
Intermediate Service
Outcomes
Mediating
Treatment &
Other
Outcomes
Child Outcomes
NPC Research – 1/26/07
6
Four Sites With Different
FTDC Models

San Diego: system-wide reform, the
“Substance Abuse Recovery Maintenance
System” (SARMS), with FTDC for noncompliant parents

Santa Clara: started as traditional FTDC
model; Made some systems changes later
in the study

Suffolk: neglect cases only, many children
not in out-of-home placements

Washoe: traditional FTDC model
NPC Research
June 2007
7
Study Samples

Complex design based on FTDC models and
availability of comparison groups
– Suffolk and Washoe relied on within-county
comparison groups of unserved eligible clients
– San Diego relied on comparison county
comparison group, matched at the individual
level
– Santa Clara relied on a combination of withincounty and comparison county comparison groups
NPC Research
June 2007
8
Final Study Samples
Site
Drug Court System
San Diego
438
(104 DC, 334 SARMS)
205
Santa Clara
100
553
Suffolk
117
239
Washoe
84
127
Total N=
739
1124
NPC Research
June 2007
Comparison
9
Data Collection Strategies

Administrative record review
– Treatment, court, and child welfare records
– Data were collected on both parents in two-parent
families, but data presented today are for mothers
only

Parent interviews
– A subset of 253 parents across the 4 sites were
interviewed up to 4 times during their case
– These data not presented here

Qualitative parent and key stakeholder
interviews and court observations
NPC Research
June 2007
10
Administrative Data Tool

Extensive data extraction tool captured the
following information:
– Family Background (e.g., number/ages of children,
marital status)
– Child Welfare Case (e.g., hearing dates, out-ofhome placements)
– FTDC Services (e.g., appearances, sanctions)
– Treatment Services (e.g., number & type of tx.)
– Permanency Outcomes (permanency decisions &
custody arrangements)
– Child Welfare Recidivism (e.g., new referrals,
petitions)
NPC Research
June 2007
11
Sample Demographics

Samples were well-matched, with very few
significant differences in demographic, risk, or
case characteristics

California sites had larger Hispanic
populations

Suffolk site had no meth users; this was the
most common drug at the other 3 sites
NPC Research
June 2007
12
Sample Demographics, cont’d

75% of parents were unemployed

60% of families were headed by single
mothers

25-50% had less than high school education

Families had an average of 2 children and
half had an infant
NPC Research
June 2007
13
Mother Risk Factors

Collected information on the following
maternal risk factors identified in
administrative data sources:
– Mental illness
– Learning/developmental disorder
– Medical disability/condition
– History of domestic violence

Computed 0-4 risk index (one point for
each risk factor present)
NPC Research
June 2007
14
Average Number of
Mother Risk Factors
Drug Court
Comparison
3
2.5
# risk factors
2
2.3
2.3
2.2
2.4
2.6
2.7
2.2
1.8
1.5
1
0.5
0
San Diego*
* Statistically
NPC
Research
Santa Clara
significant at
p<.001.
June
2007
Suffolk
Washoe
15
Child Risk Factors

Collected information on the following child
risk factors:
– Developmental/educational issues
– Behavioral/emotional issues
– Child alcohol and drug issues
– Prenatal substance exposure
– Child sexually acting out
– Child was sexually abused

Computed 0-6 risk index (received one point
for each risk factor)
NPC Research
June 2007
16
Average Number of
Child Risk Factors
Drug Court
Comparison
3
2.5
# risk factors
2
1.5
1.5
1
1.2
1.3
1
1.1
1.3
0.9
1.1
0.5
0
San Diego*
NPC Research
Santa Clara
June 2007
* Statistically significant at p<.05.
Suffolk
Washoe
17
Summary: Sample Characteristics

Overall, samples were well-matched,
with very few significant differences in
demographic, risk, or case
characteristics

Some site differences in terms of
race/ethnicity, drug of choice,
treatment history & prior CPS
involvement
NPC Research
June 2007
18
Part I: Outcome Analysis
How do treatment, child welfare, and
court outcomes differ for families
processed through FTDCs as compared
to traditional child welfare case
processing?
NPC Research
June 2007
19
Outcomes: Analytic Approach


Propensity scoring is a method for reducing bias in
effect estimates associated with selection bias in nonrandomized designs (Rosenbaum & Rubin, 1983).
Propensity scores were modeled for each site using
the following characteristics:
– Race
– Marital status
– Education
– Employment status
– Type of abuse allegation
– Mother’s age
– Age 1st drug use
NPC Research
– Previous CPS involvement
– # mother risk factors
– # child risk factors
– # children on case
– Infant involved in case
– Frequency of drug use
– Previous TPR
June 2007
20
Outcomes: Analytic Approach





Outcomes were analyzed using weighted least
squares (WLS) regression
Propensity scores were used as site-specific weights
Data presented are adjusted means
Treatment effects were estimated within each site,
and then pooled for an overall study effect size
estimate
Effects of treatment group reflect FTDC vs.
comparison group except San Diego, where FTDC
and SARMS are combined and weighted, creating a
“FTDC system” treatment group
NPC Research
June 2007
21
Treatment Outcome Questions

Compared to Control Parents, Did
Parents in FTDC:
– Enter treatment at a higher rate?
– Enter treatment more quickly following
their child welfare petition?
– Spend more time in treatment?
– Complete treatment at a higher rate?
NPC Research
June 2007
22
Likelihood of Treatment Entry
Drug Court
Comparison
86%
83%
100%
80%
73%
67%
% Mothers
60%
60%
61%
61%
60%
40%
20%
0%
San Diego
Santa Clara*
Suffolk*
Washoe
* Statistically significant at p<.001.
NPC Research
June 2007
23
Days to Treatment Entry
Drug Court
Comparison
133
140
120
120
107
114
110
100
100
84
Days
80
58
60
40
20
0
San Diego
Santa Clara
Suffolk*
Washoe
* Statistically significant at p<.001.
NPC Research
June 2007
24
Days Spent in Treatment
Drug Court
Comparison
330
350
298
297
300
250
Days
200
179
154
150
172
135
132
100
50
0
San Diego
Santa Clara*
Suffolk*
Washoe*
* Statistically significant at p<.001.
NPC Research
June 2007
25
Percent Completing
at Least One Treatment
Drug Court
Comparison
80%
69%
62%
61%
% of mothers
60%
41%
40%
31%
32%
32%
37%
20%
0%
San Diego
NPC Research
Santa Clara**
June 2007
Suffolk**
Washoe*
* Statistically significant at p<.01. ** Statistically significant at p<.001.
26
Cross-site Effects on Substance
Abuse Treatment

Strong cross-site treatment effects

Compared to comparison parents, drug court
parents:
– Were more likely to enter treatment
– Entered treatment more quickly than comparison
parents
– Stayed in treatment longer than comparison
parents
– Completed treatment more often than comparison
parents
NPC Research
June 2007
27
Child Welfare & Court System
Outcome Questions

Did children of FTDC parents:
–
–
–
–

Receive more ancillary services?
Have fewer placement changes?
Spend less time in out-of-home care?
Have more kinship placements?
Were children of FTDC parents reunified at a
higher rate?
 Were FTDC parents less likely to become
involved with the CWS subsequent to their
case?
 Were court cases shorter and less often
NPC Research
June 2007
28
contested?
Child Welfare: Levels of Analysis

Analysis of child welfare outcomes is
complicated by the fact that multiple
children may have outcomes for each
parent

Two levels of analysis: children “nested”
within parents
P1
C1
NPC Research
C2
P2
C3
June 2007
C1
C2
29
Child Welfare: Levels of Analysis

Outcomes for children within a family
are likely to be similar

Analyzing each child’s outcome can
result in bias in significance testing

SPSS linear mixed models used to
adjust the error terms to reduce
possible bias caused by the “nesting” of
children within families
NPC Research
June 2007
30
Children’s Experiences
During the Case

Number of services for children (medical,
early intervention, mental health, education,
substance abuse, and “other” services)

Children’s living situations during case
– # of living situation changes
– Days & % of case in parental care
– Days & % of case in out of home placements
– Days & % of case in kinship care
NPC Research
June 2007
31
Number of Services for Children
Site
Drug Court
Comparison
San Diego*
mean
N=788
1.4
N=475
1.2
Santa Clara
mean
N=199
1.3
N=1,132
1.2
Suffolk
mean
N=259
0.5
N=491
0.5
Washoe
mean
N=148
1.7
N=226
1.7
* Statistically significant at p<.05.
NPC Research
June 2007
32
Number of Living Situations
During CW Case
Site
Drug Court
Comparison
San Diego
mean
N=482
3.3
N=844
3.3
Santa Clara*
mean
N=201
4.0
N=1,150
3.4
Suffolk
mean
N=262
1.9
N=497
1.9
Washoe
mean
N=166
3.3
N=246
3.2
* Statistically significant at p<.001., controlling for length of case
NPC Research
June 2007
33
Time in Parental Care
Site
Drug Court
Comparison
San Diego
mean days
% of case
N=788
150
26%
N=456
143
28%
Santa Clara
mean days**
% of case*
N=174
207
31%
N=1,100
128
23%
Suffolk
mean days
% of case
N=262
284
46%
N=495
269
47%
Washoe
mean days**
% of case**
N=164
286
50%
N=244
90
20%
232 days
38%
158 days
29%
Overall Average Days
NPC Research
June 2007
* Statistically significant at p<.05. ** Statistically significant at p<.001.
34
Time in Out-of-Home Placements
Site
Drug Court
Comparison
San Diego
mean days
% of case
N=824
226
33%
N=463
232
34%
Santa Clara
mean days
% of case
N=194
190
28%
N=1,112
218
33%
Suffolk
mean days
% of case
N=262
114
18%
N=496
82
14%
Washoe
mean days*
% of case*
N=165
199
33%
N=245
336
58%
182 days
28%
217 days
35%
Overall Average
NPC Research
June 2007
* Statistically significant at p<.001.
35
Time in Kinship Care
Site
Drug Court
Comparison
San Diego
mean days
% of case
N=819
252
39%
N=459
244
37%
Santa Clara
mean days
% of case
N=198
247
37%
N=1,115
287
43%
Suffolk
mean days
% of case
N=262
198
36%
N=495
228
39%
Washoe
mean days
% of case
N=164
102
17%
N=245
129
23%
NPC Research
June 2007
36
Cross-site Children’s
Experiences Effects

No significant difference between groups in
number of services children received

Drug court children had significantly more
living situation changes than comparison
children

Drug court children spent significantly less
time in out-of-home placements and more
time with parents than comparison children,
especially in Santa Clara and Washoe sites
NPC Research
June 2007
37
Permanency Outcomes

Time to permanent placement
 Permanency decisions:
– % reunification
– % terminations of parental rights
– % another permanency outcome

About one-fourth (24%) of children had not
yet reached permanency at the end of the 2year window:
–
–
–
–
San Diego:
20%
Santa Clara: 12%
Suffolk:
57%
Washoe:
13%
NPC Research
June 2007
38
Days to Permanent Placement
Drug Court
Comparison
400
347
350
300
286
255
277
243
250
262
Days
216
200
163
150
100
50
0
San Diego
Santa Clara*
Suffolk
Washoe
* Statistically significant at p<.05.
NPC Research
June 2007
39
San Diego Permanency Decisions
Drug Court
Comparison
100%
% of children
80%
60%
56%
45%
40%
24%
28%
27%
20%
20%
0%
Reunified*
TPR
Other**
* Statistically significant at p<.05. ** Statistically significant at p<.001.
NPC Research
June 2007
40
Santa Clara Permanency Decisions
Drug Court
Comparison
100%
80%
76%
% of children
60%
44%
34%
40%
22%
11%
20%
13%
0%
Reunified*
TPR
Other
* Statistically significant at p<.001.
NPC Research
June 2007
41
Suffolk Permanency Decisions
Drug Court
Comparison
100%
% of children
80%
60%
57%
55%
35% 35%
40%
20%
8%
11%
0%
Reunified
NPC Research
TPR
June 2007
Other
42
Washoe Permanency Decisions
Drug Court
100%
Comparison
91%
80%
% of children
60%
45%
34%
40%
20%
20%
3%
5%
TPR**
Other*
0%
Reunified**
* Statistically significant at p<.01. ** Statistically significant at p<.001.
NPC Research
June 2007
43
Cross-site Permanency Effects

Drug court cases took significantly
longer than comparison cases to reach
permanent placement

Drug court children were significantly
more likely to be reunified and less
likely to have TPRs than comparison
children, in all sites except for Suffolk
NPC Research
June 2007
44
Parents’ Subsequent Involvement
with Child Welfare
% With second petition on original case
 % With a new CPS referral
 % With a new CPS petition (new case)
 % With subsequent out-of-home
placements
 % With subsequent TPR
 % With a new drug-exposed baby

NPC Research
June 2007
45
Cross-site Effects

When we pooled results across sites,
there were no significant differences
between drug court and comparison
families on any of the child welfare
recidivism outcomes

Time frame is likely too short to
adequately assess subsequent
involvement with child welfare
NPC Research
June 2007
46
Court Outcomes
Contested hearings
 Indication of noncompliance with case
plan
 Time to case closure

NPC Research
June 2007
47
Cross-site Court Effects

No significant differences in number of
contested hearings between FTDC and
comparison cases

FTDC parents had significantly fewer
incidents of noncompliance with court
orders than comparison parents

FTDC cases took significantly longer to
reach case closure than comparison
cases
NPC Research
June 2007
48
Summary: Outcomes for FTDCs

Strong treatment outcomes: FTDC parents
more likely to enter treatment, spend more
time in treatment, and complete treatment

Longer time to permanent placement and
case closure for FTDC parents could be
explained by the longer treatment stays

Cases took longer to reach permanency and
closure, but FTDC children spent more of this
time with their parents
NPC Research
June 2007
49
Outcomes for FTDCs, cont’d
FTDC children were more likely to be
reunified with their parents at the end of
the case
 No differences in child welfare
recidivism, but follow-up period was
short

NPC Research
June 2007
50
Part II: Black Box Analysis
How do FTDCs work? What factors
influence program outcomes?
NPC Research
June 2007
51
Unpacking the “Black Box” of
Family Treatment Drug Court
Outcome analysis tells us whether
FTDCs work
 Analysis of parent characteristics and
experiences with services can begin to
tell us about how, why, and for whom
FTDCs work
 A preliminary look within the FTDC
sample

NPC Research
June 2007
52
Conceptual Model for Understanding
How FTDC Works
FTDC
Treatment
Child Welfare
Outcomes
Parent
Characteristics
NPC Research
June 2007
53
How Do FTDC Experiences
Influence Outcomes?
FTDC
Treatment
Child Welfare
Outcomes
Parent
Characteristics
NPC Research
June 2007
54
Key Questions About FTDC

Key Questions:
– Does how long it takes parents to enter
FTDC relate to key outcomes?
– Does time spent in FTDC relate to
outcomes?
– Do the number of FTDC appearances
relate to outcomes?
– Does FTDC graduation relate to
outcomes?

Outcomes: days in treatment, treatment
completion, & reunification
NPC Research
June 2007
55
FTDC Processing Variables
San
Diego
Santa
Clara
Suffolk
Washoe
Days from petition to
drug court entry
227A
101
N.A.
105
Days spent in drug
court
209C
355
449B
353
17
(approx.
2 per
month)
15
(approx.
1 per
month)
31
(approx.
2 per
month)
25
(approx.
2 per
month)
Number of drug court
appearances
A
San Diego is significantly different from Santa Clara and Washoe.
B Suffolk is significantly higher than all other sites.
C San Diego is significantly lower than all other sites.
NPC Research
June 2007
56
Drug Court Graduation
San Diego
Santa Clara
Suffolk
Washoe
100%
84%
77%
80%
58%
60%
44%
40%
29%
30%
20%
26%
18%
13%
13%
3%
5%
0%
Graduated
NPC Research
Terminated
June 2007
Still Enrolled
57
FTDC Experiences and Substance
Abuse Treatment Outcomes
Variable
Statistically
Significant?
Nature of Relationship to
Treatment Outcomes
Time to FTDC entry
(petition to entry)
No
No relationship
Time spent in FTDC
Yes
Longer stays in FTDC are
related to longer stays in tx
and more tx completion
Number of FTDC
appearances
Yes
More FTDC appearances are
related to longer stays in tx
and more tx completion
FTDC graduation
Yes
Graduation is related to
longer stays in tx and more tx
completion
NPC Research
June 2007
58
FTDC Experiences
and Reunification
Variable
Statistically
Significant?
Nature of Relationship to
Reunification
Time to FTDC
entry
No
No relationship
Time spent in
FTDC
Yes
The more time spent in FTDC,
the greater the likelihood of
reunification
Number of FTDC
appearances
Yes
The greater number of FTDC
appearances, the greater
likelihood of reunification
FTDC graduation
Yes
Parents who graduate from
family treatment drug court are
more likely to be reunified
NPC Research
June 2007
59
How Do Treatment Experiences
Relate to Outcomes?
FTDC
Treatment
Child Welfare
Outcomes
Parent
Characteristics
NPC Research
June 2007
60
Key Questions

Does time to treatment entry relate to
outcomes:
– time spent in treatment
– treatment completion
– reunification

Does likelihood of treatment entry relate to
reunification?
 Does time spent in treatment relate to
treatment completion or reunification?
 Does treatment completion relate to
reunification?
NPC Research
June 2007
61
Treatment Experiences and
Treatment Completion
Variable
(administrative
data)
Statistically
Significant?
Nature of Relationship to
Treatment Outcomes
Time to treatment
entry
Yes
Faster time to treatment is
related to longer treatment
stays and higher rates of
treatment completion
Time in treatment
Yes
Longer treatment stays are
related to higher rates of
treatment completion
NPC Research
June 2007
62
Treatment Experiences
and Reunification
Variable
Statistically
Significant?
Nature of Relationship to
Reunification
Likelihood of
treatment entry
No
No relationship
Time to treatment
No
No relationship
Time spent in
treatment
No
No relationship
Treatment
completion
Yes
Mothers who completed at
least one treatment
episode are more likely to
be reunified
NPC Research
June 2007
63
Do Parent Characteristics Influence
Outcomes?
FTDC
Treatment
Child Welfare
Outcomes
Parent
Characteristics
NPC Research
June 2007
64
Do Parent Characteristics Influence
Outcomes?

Parent characteristics examined:
– Demographic variables
– History of substance abuse, mental health
– Child welfare history
– Maternal risk factors
– Child risk factors
– Psychosocial characteristics (perceived
stress, perceptions of control, social
support)
NPC Research
June 2007
65
Summary: Influences of Parent
Characteristics

FTDC Experience
– More appearances for white, older, lower risk moms
– Less graduation for African American moms

Treatment Completion/Retention Better For:
– Less educated moms
– More years drug use
– More treatment motivation

Reunification More Likely For:
– Lower risk moms
– Moms with no CPS history

No strong, consistent pattern of differences for
different
“types” of parents
NPC Research
June 2007
66
How Do FTDCs Work? Summary of
Findings from Quantitative Data
Speed of Tx entry
Duration of Tx
FTDC:
Time spent
Appearances
Graduation
Treatment
Completion
Parent
Characteristics
Child Welfare
Outcomes
Does FTDC Influence Reunification
“Above and Beyond” its Effect on
Treatment Completion?
FTDC
(TX vs Control)
Treatment
Completion
.14***
.28***
Reunification
Parent
Characteristics
NPC Research
June 2007
68
Part III: Qualitative Parent
Interviews
What features of drug court most
influence parents’ recovery and ability to
make progress on the case plan?
NPC Research
June 2007
69
Qualitative Parent Interviews
 Brief, open-ended questions asked
of 219 parents, and in-depth
qualitative interviews with 30 parents
Interviews provide contextual data
to facilitate the interpretation of the
quantitative data

Parents’ perspective of what
facilitates the drug court process

NPC Research
June 2007
70
Emotional Support

Parents talked about how the drug court
team, and in particular the judge and the drug
court-dedicated case workers, provide a
support system.
“The drug court team and the drug court case
worker have helped me a lot. My first case worker,
that wasn’t the drug court one, didn’t spend much
time with me, but my drug court case worker
always knew what was going on with me, and
helped me get what I needed to get my kids back.”
NPC Research
June 2007
71
Accountability and Collaboration

Parents also explained how frequent hearings
and attendance in drug court provided
accountability for their behavior because:
– “the team knows what’s going on with you and you
get immediate support for whatever is going on as
soon as you need it.”
– “it’s helpful going every two weeks because things
can come up during that time, and in drug court
these problems are addressed quickly.”
NPC Research
June 2007
72
Accountability and
Collaboration, cont’d

Frequent court attendance means that the
judge and others are well informed about the
parents’ cases and able to provide
appropriate support for recovery and other
issues facing the parent.
“(attending drug court regularly) helps you
feel less alone, that someone knows what’s
going on in your life and the all the issues that
you face, they know how to support you and
what you need.”
NPC Research
June 2007
73
Practical Support

Participants in drug court receive practical
assistance. Parents talked about:
– how the drug court helped get them housing and
employment,
– helped with life improvement needs such as tattoo
removal, dentures and obtaining birth control.

These practical and external supports helped
to increase parents’ sense of confidence and
ability to make improvements in their lives.
NPC Research
June 2007
74
Sense of Accomplishment

Parents who graduated from drug court
spoke eloquently about the significance of
graduation. Parents discussed how
graduation from drug court gave them a
sense of accomplishment, some for the first
time in their life.
“It (graduation) was great. Everyone applauded for
me, I got a hug from the Judge, and they gave me
flowers. I felt like a beauty queen. I also felt that my
graduated meant that I finished something I started,
and this is the first time I ever accomplished
something like this in my life. Now I feel like I can
succeed in life.”
NPC Research
June 2007
75
Putting It All Together—
What Have We Learned?




Data strongly support the effectiveness of the
FTDC model in improving both treatment and
child welfare outcomes; “traditional” FTDC
models may be most effective
FTDC influence on outcomes goes beyond its
positive influence on treatment retention and
completion – but what accounts for this remains
largely unknown
Retention of families in FTDC programs is
important to success
FTDC influence on child welfare recidivism needs
additional data and research
NPC Research
June 2007
76
For More Information:
The final report is posted on NPC’s website:
www.npcresearch.com
 E-mail Sonia Worcel at
[email protected]
 Article on an earlier phase of the FTDC study:

Green, Furrer, Worcel, Burrus & Finigan (2007). How
Effective Are Family Treatment Drug Courts?
Outcomes from a Four-Site National Study, Child
Maltreatment 12(1), 43-50.
NPC Research
June 2007
77