Is Addiction Treatment Enough?

1
Is Addiction Treatment Enough? Long-term
Employment and Housing Outcomes Among Treated
Homeless Cocaine Abusers
Stefan G. Kertesz1 Ashley N. Mullins1 Dennis
Wallace2 Joseph E. Schumacher1 Jesse B.
Milby1 1University of Alabama at Birmingham
2Rho Federal Systems, Inc Support NIDA
K23-DA15487 NIDA R01-DA089475
2
Background

• Substance abuse (SA) prevalent among urban
  homeless (43)
• Treatment, typically residential, reduces drug
  use.1,2,3
• Whether treatments produce long-term housing and
  employment rehabilitation is not well-studied in
  clinical trials

Stahler, J Addict Dis. 1995,14. Schumacher et al.
JSAT. 200081-88. Kertesz et al.
JSAT.2003197-207.
3
Threats to Long-term Success

• Chronic nature of addiction1
• Complex socioeconomic conditions that produce
  homelessness
• Criminal history. Laws permit1
• refusal of employment, housing and benefits

1. McLellan et al. JAMA. 20002841689-95.
2.Legal Action Center, 2004 http//www.lac.org
4
Objectives

• Aim 1 Quantify proportion of subjects with
  housing or employment success 12 months after
  treatment entry1
• Aim 2 Identify factors associated with housing
  or employment success
• Original Hypothesis Criminal history will
  predict worse housing employment outcomes.

1. Milby, Schumacher et al. In Press.
5
Treatment Data Source

• Secondary analysis of RCT of behavioral Rx for
  homeless cocaine-dependent adults, comparing
  6-month interventions (Homeless 3)
• Exclusion Criteria for Current Analysis
• incomplete data for main predictors at baseline
  (N2)
• unavailable at 12 months (N76)----------gt117
  analyzed

Abstinence-Contingent Hsg (N63)
N195
Non-Abstinence-Contingent Hsg (N66)
No Housing (N66)
1. Milby, Schumacher et al. In Press.
6
Treatment Phase 1

• Phase I (months 1-2)
• All Behavioral Day Treatment individual
  counseling, groups, goal management, case
  management
• ACH abstinence-contingent housing after 2
  consecutive drug free urine tests.
• NACH housing after 2 consecutive urine tests,
  regardless of results.
• NH subjects no housing.

7
Treatment Phase 2 Aftercare

• Phase II (months 3-6)
• ACH NACH housing _at_ rent of 161/mo.
• All Non-Abstinent Contingent Work Therapy
  (paid)
• Daily therapy groups individual counseling
• Aftercare (months 7-12)
• Some ACH NACH subjects used program housing if
  slots avail, all subjects eligible for case
  management
• Aftercare group available 1x/week

8
RCT Abstinence Outcomes
Weekly Proportion of Abstinent Participants
During Months 1-6
Overall treatment effect p0.0020
9
Variables

• 12 month Outcomes
• Good employment 45 days in last 60
• Good housing 45 days housed in last 60
• Dual Success on employment housing
• Predictors
• CATEGORICAL
• Gender
• Axis I (non-SUD) or Axis II Disorder (none
  psychotic)
• Baseline status of employment or housing
• Criminal history (gt1 conviction of any kind,
  lifetime)
• CONTINUOUS
• Age
• Education (years)
• Treatment Attendance Days
• Consecutive weeks abstinent on urine tests
  (longest)

10
Subjects Available vs. Unavailable at 1 Year8
8 For continuous variables, median values are
shown, and Wilcoxon Rank Sums were used.
Longest consecutive weeks abstinent during months
1-6 of treatment. Attendance during months 1-6
of treatment.
11
(No Transcript)
12
Multivariable Models
13
Multivariable Models
14
Employment Housing Dual Success
Employed gt 45 days
Dual Non-success 57 (49)
19
Dual Success 18 (15)
Housed gt 45 days
23
15
Dual Success vs Dual Non-success
16
Limitations

• Small numbers
• Local cohort, mainly African-American
• Potential follow-up bias (60 at 12 mos.)
• Subject preferences unknown
• Limited power to measure the effect of criminal
  history 68 had convictions.

17
Conclusions

• 6 months after program completion
• 35 good employment (13 prior)
• 32 good housing (12 prior)
• 15 both
• Female--gt ? employment
• Education, attendance--gt? employment
• Dual success associated with
• trial-provided housing
• abstinence, attendance
• male

18
Implications

• Treatment housing for 6 months helps only some
  persons
• Addictions chronicity and complex socioeconomic
  challenges reduce benefits
• Consider
• long-term housing SA interventions
• addressing housing/employment market forces

19
Thank-you
20
Bivariate Models
21
Bivariate Models