Predictors of prison based treatment outcomes a comparison of men

Predictors of prison-based treatment outcomes: a comparison of men
and women participants.(Clinical report).
Author(s):Nena Messina, William Burdon, Garo Hagopian and Michael
Prendergast.
Source:American Journal of Drug and Alcohol Abuse 32.1 (Feb
2006): p.7(22). (8013 words) From General OneFile.
Document Type:Magazine/Journal
Bookmark:Bookmark this Document
Library Links:

Abstract:
The purpose of this study was to examine differences between men
and women entering prison-based therapeutic community (TC)
treatment and to explore the relationship of those differences to
posttreatment outcomes (i.e., aftercare participation and
reincarceration rates). Extensive treatment-intake interview data for
4,386 women and 4,164 men from 16 prison-based TCs in California
were compared using chi-square analyses and t-tests. Logistic
regression analyses were then conducted separately for men and
women to identify gender-specific factors associated with posttreatment outcomes. Prison intake data and treatment participation
data come from a 5-year process and outcome evaluation of the
California Department of Corrections' (CDC) Prison Treatment
Expansion Initiative. The return-to-custody data came from the CDC's
Offender Based Information System. Bivariate results showed that
women were at a substantial disadvantage compared with their male
counterparts with regard to histories of employment, substance abuse,
psychological functioning, and sexual and physical abuse prior to
incarceration. In contrast, men had more serious criminal justice
involvement than women prior to incarceration. After controlling for
these and other factors related to outcomes, regression findings
showed that there were both similarities and differences with regard to
gender-specific predictors of posttreatment outcomes. Time in
treatment and motivation for treatment were similar predictors of
aftercare participation for men and women. Psychological impairment
was the strongest predictor of recidivism for both men and women.
Substantial differences in background characteristics and the limited
number of predictors related to posttreatment outcomes for women
suggests the plausibility of gender-specific paths in the recovery
process.
Keywords: Gender-specific pathways, prison-based treatment,
treatment outcome
Full Text :COPYRIGHT 2006 Taylor & Francis Ltd.
INTRODUCTION
From 1995 to 2002 the nation's state prison population increased by
27%, and the nation's federal prison population increased by 71% (1).
Moreover, annual increases in the number of incarcerated women have
been consistently larger than the increases in the number of
incarcerated men for the past two decades (1-3). The growth in the
nation's prison population largely has been due to the increased use of
incarceration for drug-related offenses (1), which also has created an
increased demand for appropriate drug treatment programs for men
and women within prison settings.
The therapeutic community (TC) treatment model has been shown to
be an effective method of substance abuse treatment (4) and many TC
programs have been incorporated into American prisons over the past
two decades. Despite the increase in prison-based TC treatment, little
is known about the individual characteristics and specific treatment
needs of drug-dependent men and women participating in prisonbased programs (5, 6). Moreover, many of the existing evaluations of
prison-based treatment focus specifically on the programs for men
(similar to community-based treatment evaluations). Only a handful of
studies have assessed outcomes of women in prison-based treatment
programs designed specifically for women, and even fewer studies
have compared and contrasted specific factors that are associated with
outcomes relative to men or women (7). Understanding possible
differences in the needs and recovery processes of drug-dependent
men and women offenders is important to help design appropriate
prison-based substance abuse programs.
This study begins to address the gap in the literature regarding gender
issues and treatment outcomes. We compare and contrast intake data
from a large sample of drug-dependent men and women offenders
who paroled from prison-based treatment programs within California.
We further assess posttreatment outcomes for men and women
separately to identify the correlates of success specific to gender, and
to examine the plausibility of "gender-specific" paths in the recovery
process. The following literature review pieces together the available
information on the background characteristics and treatment needs of
men and women in prison-based treatment, their postrelease
treatment outcomes, and gender-specific predictors of success.
Characteristics of Men and Women Entering Prison-Based Treatment
A review of the literature identified only 3 published studies that
directly compared the characteristics of incarcerated men and women
in drug treatment programs. Peters, Strozier, Murrin, and Kearns (8)
compared data from the intake assessments of 1,225 men and 430
women referred to a short-term cognitive behavioral jail treatment
program in Tampa, Florida. Langan and Pelissier (9) compared 1,332
men and 312 women who volunteered to participate in cognitive
behavioral drug treatment programs at 20 minimum, low, and medium
security federal prisons around the country. Messina, Burdon, and
Prendergast (10) replicated the design of the previous studies,
comparing the characteristics of 4,509 women and 3,595 men from 16
prison-based TCs throughout California.
The findings from these 3 studies were strikingly consistent and
paralleled findings on gender differences found in nonoffender
populations (11, 12), indicating that women were more likely than
men to present greater challenges to treatment practitioners. For
example, incarcerated women in treatment were significantly more
likely than incarcerated men to have severe substance abuse histories
(e.g., using hard drugs, using more frequently, using polysubstances,
or taking drugs intravenously), to have grown up in homes where drug
use was present, to have coexisting physical health and psychological
problems, to be taking prescribed medications for psychological
problems, and to have been sexually and physically abused as children.
However, men were found to have more serious criminal histories than
women (8-10). Although the findings from the above offender
populations are consistent, the extent to which the outlined differences
among men and women at treatment entry may affect posttreatment
outcomes largely is unknown.
Posttreatment Outcomes of Men and Women in Prison-Based
Treatment
Fairly consistent findings from prison-based treatment evaluations for
men have been reported: Treatment reduces posttreatment recidivism
and drug use; men who continue treatment in the community after
release from prison do better than men who do not continue treatment;
and length of time in treatment is positively correlated with greater
success on parole (13-16). The effect of treatment tends to disappear
at about 3 years (13, 14, 17).
However, findings from studies of men cannot be generalized to
women for several reasons. First, men and women have different
pathways to crime and addiction (18-20) and continue to use drugs for
different reasons (21, 22). Women's patterns of drug abuse have been
described as more socially embedded than men's and primarily revolve
around interpersonal relationships (23, 24). In fact, women are
frequently initiated to drug use by their male partners, and often
continue to use drugs to cope with abusive relationships (18, 22, 25).
Histories of sexual and/or physical abuse are also major preexisting
conditions in subsequent addiction and criminality for women (6, 21,
26). Second, men and women tend to enter treatment for different
reasons (27). Women cite family problems and health as motivation
for entering treatment, while men are more likely to cite employment
and legal issues (28). Third, the resources and services available in
women's correctional facilities may be different than those available to
men (21).
A small body of literature has evaluated posttreatment outcomes for
women in prison-based TC treatment. The existing research, however,
is limited, and findings are sometimes contradictory (for a full review,
see Messina & Prendergast (7)). For example, two studies found that
women in prison-based treatment had more success on parole
compared with a no-treatment group of inmates (16, 29); whereas
another study found there was no difference between women in a
prison treatment group versus women in a no-treatment group (30).
Two other studies found that women in prison-based treatment had
reductions in recidivism and drug use compared with a no-treatment
group of women (31, 32). However, Wexler's study (16) found no
difference in recidivism rates between treated and untreated women.
Moreover, Rhodes and associates (33) conducted a 3-year follow-up of
the women in Pelissier's study (32) and found that there was no
evidence of long-term treatment effectiveness (33). (For a
comprehensive review of community-based treatment outcomes for
women, see Ashley, Marsden, and Brady (11)). Additional limitations
of the existing research include reliance on bivariate comparisons
within relatively small sample sizes, which does not allow for the
control of pre-existing differences between groups or proper power to
detect treatment effects.
Predictors of Treatment Outcomes for Men and Women
Only two studies were found that directly explored gender-specific
predictors of treatment outcomes from men and women. Messina,
Wish, and Nemes (34) compared and contrasted the correlates of
success for 296 men and 116 women who were randomly assigned to
community-based residential TC treatment differing primarily in the
length of inpatient and outpatient phases. Pelissier and colleagues (35)
compared and contrasted the correlates of success for 1,842 men and
473 women who participated in prison-based cognitive behavioral
treatment in a multisite federal prison evaluation.
Both studies found that there were similarities and differences with
respect to predictors of outcomes for men and women. For example,
Messina et al. (34) found that men and women who completed
residential treatment followed by outpatient treatment had substantial
reductions in drug use and arrests and increased employment.
However, the longer residential treatment program had a particularly
beneficial impact on women. Number of prior arrests and a history of
physical abuse also were associated with negative postdischarge
outcomes for women, but not for men. Pelissier et al. (35) found that a
history of prior commitments and disciplinary actions during
incarceration increased the likelihood of post-treatment drug use and
recidivism for both men and women. In addition, being Black similarly
increased the likelihood of drug use for both men and women, and age
at release from prison (older) decreased the likelihood of drug use. A
variety of other factors predicted outcomes for men only. Positive
predictors of outcomes included living with a spouse at follow-up,
participating in treatment (versus the no-treatment group), and
employment. In contrast, only one additional factor was found to
decrease posttreatment drug use among women--participation in
mental health treatment.
The different pathways and patterns of drug abuse for men and
women, and the availability of gender-specific services, are all
considered to be directly related to the likelihood of treatment entry
and recovery (11, 27). However, the limited research on gender issues
does not provide enough information to fully understand how
treatment impacts men and women differently. Yet, the available
evidence does suggest that outcomes for men and women should be
assessed separately to identify specific factors that affect behavioral
outcomes following treatment.
The purpose of this study is to determine if the characteristics of
participants who paroled from the California prison-based TCs are
differentially correlated with treatment outcomes for men and women.
This study includes separate models for men and women assessing
treatment graduation, aftercare participation, and recidivism, and uses
multivariate analyses with appropriate independent variables selected
specifically for men and women. Based on the prior research
comparing treatment outcomes of men and women separately (34, 35),
we formulated the following research question: Do factors that predict
aftercare participation and reincarceration differ substantially for men
and women who participated in prison-based treatment?
METHODS
The CDC Prison Treatment Expansion Initiative
Based on previous research in California testifying to the effectiveness
of prison-based TCs (16, 29, 30), the legislature and the California
Department of Corrections (CDC) began an initiative in the mid 1990s
to expand treatment opportunities for inmates. As part of this initiative,
the CDC established TC treatment programs in designated housing
units within many of its prisons, including all of the institutions that
house women. The data for this study were collected as part of an
evaluation of the CDC prison treatment expansion initiative.
This initiative includes two 5-year evaluation studies of the drug
treatment programs within the California state prison system. The
University of California, Los Angeles (UCLA), Integrated Substance
Abuse Programs (ISAP; formerly known as the UCLA Drug Abuse
Research Center (DARC)) was contracted by CDC to evaluate these
programs, with contract management provided by CDC's Office of
Substance Abuse Programs. The two evaluation studies cover 16
substance abuse programs in 10 prisons, totaling approximately 3,300
beds (8 male programs totaling 1,600 beds and 7 female programs
totaling 1,700 beds). These programs became operational between
July 1998 and December 1999 and include participants at all levels of
security (Level I-Minimum through Level IV-Maximum).
Prison-Based TCs in California
CDC contracts with community-based organizations experienced in the
TC model to provide services in the prisons (i.e., Amity Foundation,
Center Point, Inc., Civigenics Inc., Mental Health Systems, Phoenix
House, and Walden House). Although all of the programs provide
treatment services using the TC model, each provider has adapted the
model somewhat to conform to its own treatment philosophy and the
specific needs of its population, including providing appropriate
programming for women. Characteristics of the prison-based TCs
include: (1) activities that embody positive values that start a process
of socialization; (2) treatment staff who provide positive role models
(and many of whom are recovering addicts themselves); (3) an
alternative concept of inmates that is usually much more positive than
prevailing beliefs and attitudes held by correctional staff; and (4) an
aftercare component for graduates from the prison-based TC programs
that provides funding for up to 6 months of continued treatment
(residential or outpatient services) in the community following release
to parole (36).
Data Collection
Client-level data were collected by the treatment providers upon
admission into the TC using the Intake Assessment (IA) instrument.
The IA is designed to assess a client's pretreatment/preincarceration
sociodemographic background, criminality, employment, and
substance use, abuse, or dependence. Adopted from the Initial
Assessment developed at the Institute of Behavioral Research at Texas
Christian University (37), the IA has been used extensively with
criminal justice populations and provides information that is useful for
both clinical and evaluation purposes. The programs provided the
intake data and treatment admission and discharge data to UCLA ISAP
through disclosure agreements under CFR 42 Part 2, Section 2.52. The
UCLA General Campus Institutional Review Board approved the study
protocol. Postrelease return-to-custody rates came from the CDC's
Offender Based Information System and include incarceration for both
parole violations and new charges.
Eligibility
Participation in these programs is open to inmates who have a
documented history of substance use or abuse (based on a review of
their criminal backgrounds as documented in their inmate central files),
and who have between 6 and 24 months left to serve on their current
sentence. Those who meet these eligibility requirements are mandated
into the treatment programs. There are, however, certain exclusionary
criteria that preclude otherwise eligible inmates from entering the
programs (e.g., gang-related enemy situations, documented
membership in a prison gang, time spent in administrative segregation
for violence or weapons charges within the last 12 months, and felony
and Immigration and Naturalization Service holds).
Participants
The current study focuses on 4,164 women and 4,386 men who
entered the participating programs between July 1998 and March 2001,
paroled prior to February 1, 2002 (i.e., in order to be at risk for one
year prior to our obtaining return to custody data), and for whom
intake data were available. Participants in the study were
predominately White (39%) or Black (31%), were 35-years-old on
average, and had completed approximately 11 years of education prior
to their current incarceration. About half (45%) had never been
married. Thirty-two percent were employed during the 30 days prior to
their current incarceration, and 60% were parents. Participants
reported an average of 15.5 arrests in their lifetime. Approximately
93% met DSM-IV criteria for alcohol or drug abuse or dependence,
and 57% were serving time for a drug-related offense at the time of
their treatment admission.
Data Analyses
The analyses for the current study are designed to examine the
relevance of pretreatment differences between men and women on
treatment outcomes (i.e., aftercare participation and reincarceration).
The distributions (shown in Table 1) of demographic characteristics,
substance use, criminal, and psychological histories prior to
incarceration by gender were evaluated using chi-square tests (for
categorical variables) and t-tests (for continuous variables).
Preliminary analyses (chi-square and t-tests results not shown)
examining correlates of aftercare participation and reincarceration
were conducted separately for men and women to identify genderspecific predictors of outcomes for inclusion in the logistic regression
models (see below).
Logistical regression analyses were conducted separately for men and
women for each dependent variable (aftercare participation and
return-to-custody). Aftercare participation was defined as any
participation in aftercare treatment (dummy coded; 0 = no and 1 =
yes) and return-to-custody was defined as any return-to-custody
during the 12 months since parole (dummy coded; 0 = no and 1 =
yes). All logistic regression models included demographic variables and
other correlates that were significantly related to the above outcomes
(i.e., age, race, education, employment, marital status, number of
years incarcerated, primary drug disorder, and the presence of
cooccurring disorders). Additional independent variables were included
depending on the appropriateness to gender and the dependent
variable. For example, histories of sexual and physical abuse were
included in the regression models for the women offenders only, as
they were not significantly correlated with outcomes for the men. The
logistic regression models examining predictors of participation in
aftercare included a motivation for treatment score (yielding a
distribution of motivation ranging from 0 to 6; see Burdon et al. (5)).
Length of time in treatment also has been shown to be a strong
predictor of treatment outcomes (15, 38). Therefore, we included a
variable in the aftercare and return-to-custody regression models that
defined total time in prison-based treatment. We also included a
variable that defined total time in aftercare treatment in the return-tocustody regression models.
Adjusted odds ratios were used to interpret the statistically significant
effect size at the p < .05 level: [Exp(Beta) - 1] x 100 = adjusted odds
ratio (e.g., the percentage increase or decrease in the odds of being
returned to custody).
RESULTS
Bivariate Comparisons of Men and Women at Admission
Table 1 displays the sample characteristics for men (n = 4,164) and
women (n = 4,386) entering the prison-based TCs during the selected
study period. Gender differences were found with regard to
demographic characteristics, sexual and physical abuse histories,
primary drug problem, substance abuse and criminal justice histories,
and the presence of cooccurring disorders. However, some of the
differences that reached significance are quite small (i.e., years of
education and motivation for treatment scores), as the large sample
size substantially increased the likelihood of finding statistically
significant differences between men and women.
Although the majority of men and women were White or Black, women
were significantly more likely than men to be of Asian or American
Indian decent (11 vs. 6%; p < .001) and to be divorced or separated
(34 vs. 26%; p < .001). Women were older than men (36.0 vs. 34.5
years; p < .001), had slightly less education (11.0 vs. 11.4 years; p
< .001), and were less likely to have been employed prior to
incarceration (33 vs. 53%; p < .001). Women also were much more
likely than men to report having been sexually and physically abused
as a child (27 vs. 13%; p < .001) and as an adult (27 vs. 3%; p
< .001). Women were more likely to report cocaine/crack as their
primary drug problem (34 vs. 19%; p < .001), while men were more
likely to report methamphetamine/amphetamine as their primary drug
problem (40 vs. 30%; p < .001). With the exception of
methamphetamine/amphetamine, women were more likely to report
daily drug use prior to incarceration and to report daily use of two or
more drugs (24 vs. 19%; p < .001). Compared with women, men
were younger at the age of first arrest (17.1 vs. 21.3 years; p < .001),
had been arrested more often in their lifetime (17.4 vs. 14.7 arrests; p
< .001), and had been incarcerated longer in their lifetime (5.5 vs. 3.2
years; p < .001). However, women were more likely than men to have
a cooccurring psychiatric disorder (27 vs. 14%; p < .001) and to have
a slightly higher motivation for treatment score (1.3 vs. 1.1; p < .001).
Gender and Posttreatment Outcomes
Bivariate comparisons between men and women were conducted for
the two posttreatment outcomes: aftercare participation and a returnto-custody within 12 months from parole. Men were significantly more
likely to participate in aftercare than women (46% of men vs. 41% of
women; p < .001); yet, men who participated in aftercare did not stay
as long as women (4.6 months for men vs. 5.1 months for women; p
< .001). In addition, men were significantly more likely than women to
be returned to custody within 12 months from parole (40% of men vs.
31% of women; p < .001), and were returned sooner than women
(8.4 months for men vs. 9.2 months for women; p < .001).
Logistic regression analyses assessing the relationship of gender to
aftercare participation and 12-month return-to-custody rates were
conducted previously with this CDC sample (see Burdon et al. (5)).
These results (not shown) indicated that when other factors (e.g.,
demographics, number of years in prison, substance abuse disorder,
time in prison treatment, and motivation) were controlled, gender was
no longer a significant predictor of aftercare participation. The most
important predictor of aftercare participation was motivation for
treatment regardless of gender. However, gender remained associated
with a return-to-custody within 12 months from parole. After
controlling for various other factors, men still were significantly more
likely to be returned to custody compared with women.
Gender-Specific Paths (Logistic Regression Models for Men and Women)
Aftercare Participation
Separate logistic regression models assessing predictors of aftercare
participation for men and for women are shown in Table 2.
Men: Seven out of 10 variables significantly predicted aftercare
participation among the men (length of prison-based TC treatment,
age, race, motivation for treatment, length of lifetime incarceration,
employment, and primary drug problem). For each additional day in
prison-based treatment, the odds of participating in aftercare
increased by .4% (p < .001). For each additional year of age, the odds
of aftercare participation increased by 5% (p < .001). Compared with
Hispanic male participants, the odds of White males participating in
aftercare were increased by 79% (p < .001), and the odds of Black
males participating in aftercare were increased by 45% (p < .05). For
each additional score on the motivation for treatment scale, the odds
of participating in aftercare were increased by 26% (p < .001). For
each additional year of lifetime incarceration, the odds of aftercare
participation decreased by 3% (p < .01). Compared with men who
were unemployed prior to incarceration, the odds of aftercare
participation for those who were employed were reduced by 26% (p
< .01). Compared with men who reported opiates as their primary
drug problem, the odds of aftercare participation for men who reported
alcohol/other as their primary drug problem were reduced by 52% (p
< .001). The presence of cooccurring disorders, prior education, and
prior marital status were not significantly related to aftercare
participation among the men.
Women: Three out of 11 variables significantly predicted aftercare
participation among the women (length of prison-based TC treatment,
motivation for treatment, and primary drug problem). For each
additional day of prison-based treatment, the odds of participating in
aftercare increased by .2% (p < .001). For each additional point on
the motivation scale, the odds of participating in aftercare were
increased by 13% (p < .01). Compared with women who reported
opiates as their primary drug problem, the odds of participating in
aftercare for women who reported amphetamine/methamphetamine as
their primary drug problem were reduced by 44% (p < .01). The
presence of cooccurring disorders, sexual/physical abuse histories,
prior education, prior employment, age, race, prior marital status and
length of lifetime incarceration were not significantly related to
aftercare participation among the women.
Return-to-Custody
Separate logistic regression models assessing predictors of a 12-month
return-to-custody for men and for women are shown in Table 3.
Men: Six out of 10 variables significantly predicted a 12-month returnto-custody among the men (the presence of cooccurring disorders,
race, length of lifetime incarceration, age, employment, and total days
in aftercare). Compared with men with only a substance abuse
disorder, the odds of men with cooccurring psychiatric disorders being
returned to custody within 12 months were increased by 40%, (p
< .05). Compared with Hispanic men, the odds of white men being
returned to custody were increased by 55% (p < .01), and the odds of
black men being returned to custody were increased by 63% (p
< .001). For each additional year of lifetime incarceration, the odds of
returning to custody within 12 months were increased by 8% (p
< .001). For each additional year of age, the odds of men returning to
custody within 12 months were reduced by 4% (p < .001). Compared
with men who were unemployed prior to incarceration, the odds for
those who were employed of returning to custody within 12 months
were reduced by 29% (p < .001). For each additional day of aftercare,
the odds of returning to custody within 12 months were reduced by
1% (p < .001). Time in prison-based treatment, education, primary
drug disorder, and prior marital status were not significantly related to
a return-to-custody within 12 months.
Women: Six of the 11 variables significantly predicted a 12-month
return-to-custody among the women (length of prison-based TC
treatment, age, education, total days in aftercare, the presence of a
cooccurring psychiatric disorder, and length of lifetime incarceration).
For each additional day in prison-based treatment, the odds of
returning to custody within 12 months were reduced by .1% (p < .02).
For each additional year of age and education, the odds of returning to
custody within 12 months were reduced by 4% (p < .001) and 6% (p
< .05), respectively. For each additional day spent in aftercare, the
odds of returning to custody within 12 months were reduced by 1% (p
< .001). Compared with women with only a substance abuse disorder,
the odds of codisordered women returning to custody within 12
months were increased by 58%, (p < .001). For each additional year
of lifetime incarceration, the odds of returning to custody within 12
months were increased by 4% (p < .05). Sexual/physical abuse
histories, prior employment, race, primary drug problem, and prior
marital status were not significantly associated with a return-tocustody within 12 months.
DISCUSSION
The purpose of this study was to outline the differences between men
and women offenders entering TC treatment and to explore the
relationship of those differences to post-treatment outcomes. The
differences found between men and women entering prison-based
treatment in California mirrored the findings of previous studies of
incarcerated treatment populations; whereby women were at a
substantial disadvantage compared to their male counterparts in all
areas of life history except for their past criminal involvement (8, 9,
35). The degree to which these differences affected posttreatment
outcomes was less clear, as very little research had previously
examined gender-specific paths of recovery.
Table 4 provides a summary of the significant predictors of the two
outcome measures in the regression models, by gender. One of the
most notable findings from this table is that there were fewer
predictors of outcomes for women than for men (similar to findings
reported in Pelissier et al. (35)). Even though women were
significantly more likely than men to have more severe substance
abuse histories, sexual and physical abuse histories, and cooccurring
psychiatric disorders prior to incarceration, these variables were not
associated with the outcome measures. For example, after controlling
for other related factors, a history of sexual and/or physical abuse as a
child was no longer related to aftercare participation or reincarceration
among the women. However, a fairly small percentage of women and
men in our sample reported histories of sexual and physical abuse.
Previous studies have shown percentages ranging from 19% to 55%
among offender populations (8, 9). The low percentages reported here
could be due to the fact that these particular questions were asked by
treatment personal prior to treatment entry and not experienced
researchers. Since these questions were only asked from a partial
sample of participants, the regression analyses might not have
detected variability due to histories of sexual/physical abuse.
Similarities Between Men and Women
Consistent with previous findings from incarcerated populations, age,
motivation for treatment, time in prison treatment, and time in
aftercare were positively associated with at least one (and in some
cases both) of the outcomes for men and women (13-16, 29, 31). In
addition, both men and women who reported opiates as their primary
drug problem were most likely to go to aftercare treatment compared
to those who reported other drug dependencies. The direction of the
correlation for age and motivation for treatment to outcomes are often
replicated in offender populations, as older parolees often "age-out" of
criminal behavior and those with substantial internal motivation are
highly likely to succeed upon release from prison. Moreover, selection
bias was not likely to be an important factor in producing the "time in
treatment" finding since length of time in prison treatment was not
within the control of the inmate (which would create selection bias),
but rather was determined by when CDC placed the inmate in the
treatment program and the length of their sentence. Since the mean
time spent in prison treatment for men was 7.1 months (SD = 4.8)
and 6.9 months (SD = 4.5 months) for women, it appears that a
substantial number of inmates in this sample were assigned to these
TCs with approximately 7 months remaining on their sentence.
Predictors that negatively affected outcomes for both men and women
included the cooccurrence of psychiatric disorders during treatment
and total years in prison over one's lifetime. Codisordered inmates and
those with more total years in prison in one's lifetime were more likely
to be reincarcerated regardless of gender, a finding consistent with
previous research indicating that those with the most severe
psychological problems and those with serious criminal histories are
more likely to recidivate (34, 39).
Differences Between Men and Women
Time in treatment warrants further clarification (discussed above), as
it was also differentially associated with reincarceration among men
and women. Total time in prison-based treatment and aftercare were
related to a reduced likelihood of reincarceration for women only,
indicating the positive effect of continuous and long-term treatment
episodes (similar to that reported in Messina et al. (34)). In contrast,
only time in aftercare was related to a reduced likelihood of
reincarceration for men. This finding creates an interesting dilemma,
as it also appears that men were more likely to go to aftercare
treatment than women, but did not stay as long as the women. The
lower aftercare participation rate reported for women may have been
due to gaps in reporting procedures. The Female Offender Treatment
and Employment Program (FOTEP) aftercare treatment program was
an additional option for women paroling from prison in California.
Women entering the FOTEP program did not access aftercare
treatment using the same procedure as men or women parolees opting
for other forms of aftercare (e.g., outpatient, sober living, residential).
As a result, data on the number of women entering the FOTEP
aftercare program may not have been complete.
Another difference between men and women was that race was a
significant predictor of both outcomes for men, but not at all for
women. Among the men, Whites and Blacks were more likely than
Hispanics to participate in aftercare treatment and to be returned to
custody. Previous findings from this sample (see Burdon et al. (5))
have shown that a greater proportion of Hispanics reported living with
family/relatives prior to their current incarceration. In addition, a
recent review of studies that examined drug use behaviors among
Hispanics found that social support systems (including familial factors)
are important factors in preventing drug abuse among this population
(40). These findings suggest that the familial support systems for
Hispanics are stronger, and that they may tend to rely on these
support systems to a greater degree and with greater success
following release to parole than Whites or Blacks.
Finally, being employed prior to incarceration decreased the likelihood
of aftercare participation and a return to custody for men (similar to
that reported in Pelissier et al. (35)), but was not related to outcomes
for women. In contrast, prior education decreased the likelihood of
reincarceration for women, but was not related to outcomes for men.
Limitations
It should be noted that the current study relied on general intake data
collected by treatment personnel for a large sample of men and
women entering prison-based treatment. Due to the overall scope of
the CDC initiative, the questions available on the IA instrument were
limited in both range and depth. The IA instrument was not originally
designed to capture detailed differences between men and women
entering prison-based treatment, and, therefore, many factors that
may be predictive of posttreatment outcomes for women were absent.
For example, the questions on the IA regarding histories of sexual and
physical abuse were dichotomous (yes/no) questions, which did not
inquire about the specific type of abuse, the perpetrator of the abuse,
the age at which it occurred, or the duration of the abuse, which have
previously been shown to be related to women's recovery (6, 21). In
addition, the findings generated by this study are limited to inmates
who were selected for treatment participation and thus, cannot be
generalized to general inmate populations in state prison.
Implications and Conclusion
The results of this study have highlighted relevant individual-level
factors that serve as predictors of participation in aftercare treatment
and 12-month return to custody rates among men and women
parolees of prison-based TC treatment programs. One of the most
consistent findings, relative to previous research, is the success for
both men and women associated with aftercare participation. The
consistency of this finding indicates the need to place greater
emphasis on promoting appropriate aftercare treatment for inmates in
the prison-based treatment programs, regardless of gender.
In contrast, consistent findings regarding the increased likelihood of
reincarceration among inmates with cooccurring psychiatric disorders
highlights the difficulty involved in delivering effective treatment
services to codisordered men and women in correctional settings.
Because treatment staff may not be adequately trained to treat certain
psychiatric disorders that offenders present upon entry into prisonbased programs, perhaps there should be separate treatment tracks
for codisordered offenders to provide effective treatment to this highrisk population. At the very least, referrals to appropriate aftercare
treatment should be in place upon release from prison programs,
which would require increased communication, coordination, and
collaboration between substance abuse and psychiatric treatment
systems (41). The high prevalence rates of psychiatric disorders
among incarcerated drug offenders across the nation suggest that
these issues are probably not unique to California (39).
In conclusion, the limited number of identified predictors of outcomes
for women is notable. It would appear that we know more about what
leads to successful outcomes for men than for women. Indeed, the
limitations of the available data of our study may have drawn further
attention to the gap in our knowledge of gender-specific paths to
recovery. Future studies will need to explore and incorporate additional
predictors of posttreatment outcomes that more appropriately reflect
paths of recovery for women. Future studies also may need to explore
additional posttreatment outcomes for women such as improved
relationships with children, living situations, and psychological status.
REFERENCES
(1.) Ashley O, Marsden ME, Brady TM. Effectiveness of substance
abuse treatment programming for women: a review. Am J Drug
Alcohol Abuse 2003; 29(1):19-53.
(2.) Beck A. Prisoners in 1999. Bureau of Justice Statistics Bulletin.
Washington, DC: U.S. Department of Justice, Bureau of Justice
Statistics, 2000.
(3.) Beck A. Prison and jail inmates at midyear 2000. Bureau of Justice
Statistics Bulletin. Washington, DC: U.S. Department of Justice,
Bureau of Justice Statistics, 2001.
(4.) Bloom B, Owen B, Covington S. Gender-responsive strategies:
research, practice, and guiding principles for women offenders.
Available at http://nicic.org/ pubs/2003/018017.pdf. Accessed October
2, 2003.
(5.) Blume S. Women, alcohol and drugs. In Substance Abuse. A
Comprehensive Textbook. Lowinson J, Ruiz P, Millman R, Langrod J,
eds. Baltimore: Williams, & Wilkins, 1992; 794-807.
(6.) Brecht ML, O'Brien A, von Mayrhauser C, Anglin MD.
Methamphetamine use behaviors and gender differences. Addictive
Behaviors 2003; 29:89-106.
(7.) Broome K, Knight K, Hiller M, Simpson D. Drug treatment process
indicators for probationers and prediction of recidivism. J Subst Abuse
Treat 1996; 13(6):487-491.
(8.) Burdon W, Farabee D, Prendergast M, Messina N, Cartier J.
Evaluating prison-based therapeutic community substance abuse
programs: the California Initiative. Federal Probation 2002; 66(3):3-8.
(9.) Burdon W, Messina N, Prendergast M. The California treatment
expansion initiative: Participant characteristics, aftercare participation,
recidivism, and predictors of outcome. The Prison Journal 2004;
84(1):61-80.
(10.) Covington S, Surrey J. The relational theory of women's
psychological development: Implications for substance abuse. In
Gender and Alcohol: Individual and Social Perspectives. Wilsnak S,
Wilsnak R, eds. Piscataway, NJ: Rutgers University Press, 1997; 335351.
(11.) Covington S. Women in prison: Approaches in the treatment of
our most invisible population. Binghampton, NY: Haworth Press, Inc,
1998.
(12.) De La Rosa MR, White MS. A review of the role of social support
systems in the drug use behavior of Hispanics. J Psychoactive Drugs
2001; 33(3): 233-240.
(13.) DeLeon G. The Therapeutic Community: Theory, Model, and
Method. New York: Springer Publishing Company, 2000.
(14.) DeLeon G, Jainchill N. Male and female drug abusers: social and
psychological status 2 years after treatment in a therapeutic
community. Am J Drug Alcohol Abuse 1981; 8(4):465-497.
(15.) Grella C, Gillmore J. Improving service delivery to the dually
diagnosed in Los Angeles County. J Subst Abuse Treat 2002; 23:115122.
(16.) Grella C, Joshi V. Gender differences in drug treatment careers
among clients in the national drug abuse treatment outcome study.
Am J Drug Alcohol Abuse 1999; 25(3):385-406.
(17.) Grella CE, Stein JA, Greenwell L. Associations among childhood
trauma, adolescent problem behaviors, and adverse adult outcomes in
substance-abusing women offenders. Psychology of Addicted Behavior
2005; 19:43-53.
(18.) Hall E, Prendergast M, Wellisch J, Patten M. Treating Drug
Abusing Women Prisoners: An outcomes evaluation of the Forever Free
substance abuse treatment program. The Prison Journal 2004;
84(1):81-105.
(19.) Harrison P, Beck A. Prisoners in 2002. Bureau of Justice Statistics
Bulletin. Washington, DC: U.S. Department of Justice, Bureau of
Justice Statistics, 2003.
(20.) Henderson D. Drug abuse and incarcerated women. J Subst
Abuse Treat 1998; 15(6):579-587.
(21.) Hodgins D, El-Guebaly N, Addington J. Treatment of substance
abusers: single or mixed-gender programs? Addiction 1997;
92(7):805-812.
(22.) Hser Y, Anglin M, Booth M. Sex differences in addict careers. 3.
addiction. Am J Drug Alcohol Abuse 1987; 13:33-57.
(23.) Jarman E. An Evaluation of Program Effectiveness for the Forever
Free Substance Abuse Program at the California Institute for Women,
Frontera, California, Sacramento: California Department of Correction,
Office of Substance Abuse Programs, 1993.
(24.) Knight K, Simpson D, Hiller M. Three-year incarceration
outcomes for in-prison therapeutic community treatment in Texas. The
Prison Journal 1999; 79(3):337-351.
(25.) Langan N, Pelissier B. Gender differences among prisoners in
drug treatment. J Subst Abuse 2001; 13(3):291-301.
(26.) Martin S, Butzin C, Saum C, Inciardi J. Three-year outcomes of
therapeutic community treatment for drug-involved offenders in
Delaware: From prison to work release aftercare. The Prison Journal
1999; 79(3):294-320.
(27.) Messina N, Burdon W, Hagopian G, Prendergast M. One-year
return to custody rates among co-disordered offenders. Behavioral
Sciences and the Law 22:503-518.
(28.) Messina N, Burdon W, Prendergast M. Assessing the needs of
women in institutional therapeutic communities. Journal of Offender
Rehabilitation 2003; 37(2):89-106.
(29.) Messina N, Prendergast M. Therapeutic community treatment for
women in prison: Assessing outcomes and needs. In Treating Addicted
Offenders: A Continuum of Opportunities. Knight K, Farabee D, eds.
New York: Civic Research Institute; 18, 1-12.
(30.) Messina N, Wish E, Nemes S. Predictors of treatment outcomes
in men and women admitted to a therapeutic community. Am J Drug
Alcohol Abuse 2000; 26(2):207-228.
(31.) Owen B. "In the mix": Struggle and Survival in a Women's Prison.
Albany: State University Press of New York, 1998.
(32.) Pelissier B, Wallace S, O'Neil J, Gaes G, Camp S, Rhodes W,
Saylor W. Federal prison residential drug treatment reduces substance
use and arrests after release. Am J Drug Alcohol Abuse 2001; 42:315337.
(33.) Pelissier B, Camp S, Gaes G, Saylor W, Rhodes W. Gender
differences in outcomes from prison-based residential treatment. J
Subst Abuse Treat 2003; 24:149-160.
(34.) Peters R, Strozier A, Murrin M, Kearns W. Treatment of
substance-abusing jail inmates: Examination of gender differences. J
Subst Abuse Treat 1997; 14:339-349.
(35.) Pollock J. Counseling women offenders. Thousand Oaks, CA:
Sage Publications, 1998.
(36.) Prendergast M, Wellisch J, Wong M. Residential treatment for
women parolees following prison-based drug treatment: Treatment
experiences, needs, and services, outcomes. The Prison Journal 1996;
76(3):253-274.
(37.) Rhodes W, Pelissier B, Gaes G, Saylor W, Camp S, Wallace S.
Alternative solutions to the problem of selection bias in an analysis of
federal residential drug treatment programs. Evaluation Review 2001;
25(3):331-369.
(38.) Wasilow-Mueller S, Erickson C. Drug abuse and dependency:
Understanding gender differences in etiology and management. J Am
Pharmacol Assoc 2001; 42(1):78-90.
(39.) Wexler HK, De Leon G, Kressel D, Peters J. The amity prison TC
evaluation: Reincarceration outcomes. Criminal Justice and Behavior
1999; 26(2): 147-167.
(40.) Wexler H, Falkin G, Lipton D, Rosenblum A. Outcome evaluation
of a prison therapeutic community for substance abuse treatment. In
Drug Abuse Treatment and Jails. National Institute on Drug Abuse
Research Monograph Series 118. U.S. Department of Health and
Human Services. Alcohol, Drug Abuse, and Mental Health
Administration. Rockville, MD: U.S. Government Printing Office, 1990;
156-175.
(41.) Wexler H, Melnick G, Lowe L, Peters J. Three-year
reincarceration outcomes for Amity in-prison therapeutic community
and aftercare in California. The Prison Journal 1999; 79(3):321-336.
This study is supported by an Interagency Agreement (Contracts
#C97.355 and #C98.346) between the California Department of
Corrections (CDC), the Office of Substance Abuse Programs (OSAP),
and UCLA Integrated Substance Abuse Programs (ISAP). The findings
and conclusions of this article are those of the authors and do not
necessarily represent the official policies of the Department.
We would like to give a special thanks to the contracted substance
abuse programs for providing prison intake data, to Bubpha Chen for
providing us with reincarceration data for our sample.
Nena Messina, Ph.D., William Burdon, Ph.D., Garo Hagopian, J.D., and
Michael Prendergast, Ph.D. UCLA Integrated Substance Abuse
Programs, Criminal Justice Research Group, Los Angeles, California,
USA
Address correspondence to Nena Messina, Ph.D., UCLA Integrated
Substance Abuse Programs, Criminal Justice Research Group, 1640 S.
Sepulveda Blvd., Suite 200, Los Angeles, CA 90025. E-mail:
nmessina@ucla.edu
Table 1. Sample characteristics at treatment admission, by
gender *
Men
Women
(n = 4,164)
(n = 4,386)
Characteristics
M(SD)
% M(SD)
Race/ethnicity
White
37
Black
31
Hispanic
21
Other
11
Marital status
Never married
41
Married
24
Previously married
34
Age at admission
36.0 (7.6)
Number of years of education
11.0 (2.2)
Employed 30 days prior to incarceration
33
Ever sexually/physically
27
abused as a child (b)
Ever sexually/ physically
27
%
42
32
20
6
48
26
26
34.5 (8.9)
11.4 (2.0)
53
13
3
abused as an adult (b)
Primary drug problem (self-report)
Methamphetamine/amphetamines
30
Cocaine/crack
34
Heroin
21
Other (c)
10
None
5
Drug use 60 days prior to incarceration:
Amphetamines/meth. daily
37
Cocaine/crack daily
38
Opiates daily
50
Alcohol/marijuana daily
42
2 or more drugs daily
24
Age first arrested
21.3 (7.5)
Number of arrests in lifetime 15.8 (26.2)
14.7 (24.6)
Number of years incarcerated in lifetime
3.2 (4.1)
Cooccurring disorders during treatment
27
Motivation for treatment (0 to 6)
1.3 (1.3)
Characteristics
Race/ethnicity
White
Black
Hispanic
Other
Marital status
Never married
Married
40
19
14
16
11
42
19
33
37
19
17.1 (6.1)
17.4 (28.1)
5.5 (5.8)
14
1.1 (1.2)
Total
(N = 8,550)
(a)
% M(SD)
39
31
21
9
45
25
Previously married
Age at admission
Number of years of education
Employed 30 days prior to incarceration
Ever sexually/physically
abused as a child (b)
Ever sexually/ physically
abused as an adult (b)
Primary drug problem (self-report)
Methamphetamine/amphetamines
Cocaine/crack
Heroin
Other (c)
None
Drug use 60 days prior to incarceration:
Amphetamines/meth. daily
Cocaine/crack daily
Opiates daily
Alcohol/marijuana daily
2 or more drugs daily
Age first arrested
Number of arrests in lifetime 15.8 (26.2)
Number of years incarcerated in lifetime
Cooccurring disorders during treatment
Motivation for treatment (0 to 6)
30
35.4 (8.2)
11.2 (2.1)
41
22
18
35
28
18
12
7
40
29
42
39
22
19.5 (7.3)
4.2 (5.0)
21
1.2 (l.2)
* All bivariate analyses are statistically significant at p
< .001.
(a) N's vary slightly due to missing data.
(b) Data collected from partial sample due to survey
revisions
(men = 801; women = 1,933).
(c) Includes marijuana/hashish, PCP, hallucinogens,
barbiturates,
sedatives, alcohol, and inhalants.
Table 2. Logistic regression predicting aftercare
participation
Variable
Sig
Men (N = 1,743)
B
df
Time in prison treatment
.000 **
Cooccurring disorders
1
.516
Age
1
.000 **
Education (years)
1
.599
Recent employment history
1
.011 **
Number of years incarcerated
1
.003 **
Race [Hispanic] (a)
3
.005 **
White
1
.000 **
Other
1
.180
Black
1
.037 *
Marital status [never married] (a)
2
.106
Married/living together
1
.190
Divorced /separated
1
.327
Primary drug problem [Opiates] (a)
3
.000 **
Cocaine
1
.306
Alcohol/other
1
.000 **
Methamphetamine
1
.749
Motivation
1
.000 **
Constant
1
.000 **
Women (N = 1,192)
Time in prison treatment
1
.000 **
Cooccurring disorders
1
.685
Age
1
.818
Education (years)
1
.150
.004
1
.106
.046
.015
-.296
-.032
.584
.386
.370
-.192
.147
.208
-.743
.059
.230
-3.695
.002
.056
.002
-.039
1
1
3
1
1
1
2
1
1
3
1
1
1
1
1
1
Recent employment history
.792
Number of years incarcerated
.261
Race [Hispanic] (a)
.282
White
.106
Other
.900
Black
.164
Marital status [Never Married] (a)
.432
Married/living together
.589
Divorced/separated
.197
Primary drug problem [Opiates] (a)
.003 **
Cocaine
.275
Alcohol/other
.424
Methamphetamine
.006 **
Abused as child
.924
Motivation
.012 **
Constant
.020 *
Variable
Men (N = 1,743)
Time in prison treatment
Cooccurring disorders
Age
Education (years)
Recent employment history
Number of years incarcerated
Race [Hispanic] (a)
White
Other
.036
-.019
.286
.032
.277
.091
.209
-.209
.153
-.582
-.014
.125
-1.020
Exp(B)
1.004
1.114
1.047
1.015
.743
.969
1.793
1.471
Effect
size
(%)
0.4
5
26
3
79
Black
Marital status [never married]
Married/living together
Divorced /separated
Primary drug problem [Opiates]
Cocaine
Alcohol/other
Methamphetamine
Motivation
Constant
Women (N = 1,192)
Time in prison treatment
Cooccurring disorders
Age
Education (years)
Recent employment history
Number of years incarcerated
Race [Hispanic] (a)
White
Other
Black
Marital status [Never Married]
Married/living together
Divorced/separated
Primary drug problem [Opiates]
Cocaine
Alcohol/other
Methamphetamine
Abused as child
Motivation
Constant
1.447
45
(a)
.825
1.159
(a)
1.232
.476
1.060
1.259
.025
1.002
1.057
1.002
.961
1.037
.981
52
26
75
0.2
1.331
1.032
1.320
(a)
1.096
1.233
(a)
.811
1.165
.559
.986
1.133
.361
44
13
64
(a) Brackets indicate reference category.
* p < .05.
** p < .01.
Table 3. Logistic regression predicting return to custody
within 12 months
Variable
Sig
Wen (N = 1,744)
Time in prison treatment
1
.373
B
.000
df
Cooccurring disorders
.022 *
Age
1
.000 **
Education (years)
1
.250
Recent employment history
1
.001 **
Number of years incarcerated
1
.000 **
Race [Hispanic]
3
.004 **
White
1
.002 **
Other
1
.267
Black
1
.001 **
Marital status [Never married] (a)
2
.784
Married/living together
1
.541
Divorced/separated
1
.935
Primary drug problem [Opiates] (a)
3
.953
Cocaine
1
.674
Alcohol/other
1
.916
Methamphetamine
1
.703
Aftercare participation
1
.000 **
Constant
1
.005 **
Women (N = 1,192)
Time in prison treatment
1
.019 *
Cooccurring disorders
1
.001 *
Age
1
.000 **
Education (years)
1
.020 *
Recent employment history
1
.124
.338
1
-.037
-.030
-.338
.074
.437
.286
.487
-.078
.012
-.082
-.018
-.066
-.007
1.055
-.001
.455
-.038
-.067
-.220
1
3
1
1
1
2
1
1
3
1
1
1
1
1
1
Number of years incarcerated
.024 *
Race [Hispanic] (a)
.860
White
.540
Other
.592
Black
.415
Marital status [Never married] (a)
.364
Married/living together
.174
Divorced/separated
.325
Primary drug problem [Opiates] (a)
.107
Cocaine
.281
Alcohol/other
.297
Methamphetamme
.429
Abused as child
.743
Aftercare participation
.000 **
Constant
.000 **
Variable
Wen (N = 1,744)
Time in prison treatment
Cooccurring disorders
Age
Education (years)
Recent employment history
Number of years incarcerated
Race [Hispanic]
White
Other
Black
Marital status [Never married] (a)
.036
.113
.135
.164
-.236
-.166
.222
-.235
-.157
.049
-.007
1.667
Exp(B)
1.000
1.402
.964
.970
.713
1.077
1.548
1.331
1.627
Effect
size
(%)
40
4
29
8
55
63
Married/living together
Divorced/separated
Primary drug problem [Opiates] (a)
Cocaine
Alcohol/other
Methamphetamine
Aftercare participation
Constant
Women (N = 1,192)
Time in prison treatment
Cooccurring disorders
Age
Education (years)
Recent employment history
Number of years incarcerated
Race [Hispanic] (a)
White
Other
Black
Marital status [Never married] (a)
Married/living together
Divorced/separated
Primary drug problem [Opiates] (a)
Cocaine
Alcohol/other
Methamphetamme
Abused as child
Aftercare participation
Constant
.925
1.012
.922
.982
.936
.993
2.872
.999
1.576
.963
.936
.803
1.036
1
187
0.1
58
4
6
4
1.120
1.145
1.178
.789
.847
1.248
.790
.854
1.050
.993
5.297
1
430
(a) Brackets indicate reference category.
* p < .05.
** p < .01.
Table 4. Summary of significant predictors of outcomes in
men
and women
Significant predictors for
men
likelihood
Outcomes
Increased likelihood
Decreased
of outcome
of outcome
Aftercare
employment
participation
in prison
Whites > hispanics
Prior
Blacks > hispanics
Total years
Age (older)
Alcohol/other
< opiate
Total days in prison
treatment
Motivation
Return-to-custody
Whites > hispanics
Blacks > hispanics
Age (older)
Prior
Cooccurring disorders
Total days in
employment
aftercare
Total years in prison
Significant predictors for
women
Increased likelihood
Decreased
of outcome
of outcome
Aftercare
opiate
participation
Total days in prison
Meth. <
Return-to-custody
Cooccurring disorders
Total years in prison
likelihood
Outcomes
treatment
Motivation
Age (older)
Prior
education
Total days in
prison
treatment
Total days in
aftercare
Source Citation:Messina, Nena, William Burdon, Garo
Hagopian, and Michael Prendergast. "Predictors of prison-based
treatment outcomes: a comparison of men and women
participants." American Journal of Drug and Alcohol Abuse 32.1 (Feb
2006): 7(22). General OneFile. Gale. LIRN. 7 July 2009
<http://find.galegroup.com.prx-01.lirn.net/ips/start.do?prodId=IPS>.