Content uploaded by Gregory J Mchugo
Author content
All content in this area was uploaded by Gregory J Mchugo on Apr 19, 2016
Content may be subject to copyright.
PLEASE SCROLL DOWN FOR ARTICLE
This article was downloaded by:
[Davis, Rebecca]
On:
18 May 2011
Access details:
Access Details: [subscription number 916358921]
Publisher
Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-
41 Mortimer Street, London W1T 3JH, UK
Journal of Dual Diagnosis
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t792306890
The Trauma Recovery and Empowerment Model: A Quasi-Experimental
Effectiveness Study
Roger D. Fallota; Gregory J. McHugob; Maxine Harrisa; Haiyi Xieb
a Community Connections, Washington, DC, USA b Dartmouth Psychiatric Research Center,
Dartmouth Medical School, Lebanon, New Hampshire, USA
Online publication date: 11 May 2011
To cite this Article Fallot, Roger D. , McHugo, Gregory J. , Harris, Maxine and Xie, Haiyi(2011) 'The Trauma Recovery and
Empowerment Model: A Quasi-Experimental Effectiveness Study', Journal of Dual Diagnosis, 7: 1, 74 — 89
To link to this Article: DOI: 10.1080/15504263.2011.566056
URL: http://dx.doi.org/10.1080/15504263.2011.566056
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
JOURNAL OF DUAL DIAGNOSIS, 7(1–2), 74–89, 2011
Copyright C
Taylor & Francis Group, LLC
ISSN: 1550-4263 print / 1550-4271 online
DOI: 10.1080/15504263.2011.566056
SERVICES & POLICY
The Trauma Recovery and Empowerment Model:
A Quasi-Experimental Effectiveness Study
Roger D. Fallot, PhD,1Gregory J. McHugo, PhD,2
Maxine Harris, PhD,1and Haiyi Xie, PhD2
Objective: A quasi-experimental study tested the effectiveness of the Trauma Recovery and Em-
powerment Model (TREM), a group intervention for women trauma survivors, in comparison to
services as usual. Methods: Two hundred fifty-one women with histories of physical and/or sexual
abuse and co-occurring serious mental illnesses and substance use disorders completed comprehen-
sive study assessments at baseline and at 6 and 12 months. TREM groups were added to standard
services at two community mental health agencies in Washington, DC (n=153). Comparison group
participants received usual services at two agencies in Baltimore, MD (n=98). Results: TREM
participants showed greater reductions in alcohol and drug abuse severity, anxiety symptoms, and
current stressful events, and they showed greater increases in perceived personal safety. There were
no group differences in change for posttraumatic stress disorder and global mental health symptoms,
physical and mental health–related quality of life, and exposure to interpersonal abuse. Changes in
trauma recovery skills were associated positively with gains in study outcomes for TREM group par-
ticipants. Conclusions: Despite design limitations, this study provides preliminary evidence for the
effectiveness of the TREM intervention for a heterogeneous population of women trauma survivors
with co-occurring disorders when added to usual services. (Journal of Dual Diagnosis, 7:74–89,
2011)
Keywords trauma, childhood abuse, co-occurring disorders, integrated treatment, gender-specific
group interventions
Surveys have documented the high prevalence rates of trauma exposure, especially of physical
and sexual abuse, among women with diagnosed serious mental health disorders (Goodman,
Dutton, & Harris, 1995; Mueser et al., 1998) and those with substance use disorders (Najavits,
Weiss, & Shaw, 1997). The consequences of such trauma are both profound and wide-ranging
1Community Connections, Washington, DC, USA
2Dartmouth Psychiatric Research Center, Dartmouth Medical School, Lebanon, New Hampshire, USA
Address correspondence to Roger D. Fallot, Community Connections, 801 Pennsylvania Ave. SE, Washington, DC
20003, USA. E-mail: rfallot@ccdc1.org
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 75
(van der Kolk, Roth, Pelcovitz, Sunday, & Spinazzola, 2005), affecting many life domains and
increasing the risk of social and physical difficulties as well as psychological ones. Trauma is
especially central to the understanding and effective treatment of co-occurring substance use and
mental disorders. Histories of exposure to interpersonal violence not only increase the risk of
developing a substance use disorder (Hedtke et al., 2008) but complicate treatment for substance
abuse, leading to poorer outcomes (Rosen, Ouimette, Sheikh, Gregg, & Moos, 2002). Further,
posttraumatic stress disorder (PTSD) has been tied to poorer general mental health functioning
among individuals in substance use treatment (Ouimette, Goodwin, & Brown, 2006). In mental
health settings, the sequelae of interpersonal violence extend beyond PTSD to other psychiatric
difficulties, primarily depression (Hedtke et al., 2008) and generalized anxiety (Grant, Beck,
Marques, Palyo, & Clapp, 2008).
Violence against girls and women has been well-documented in both general populations
(Moracco, Runyan, Bowling, & Earp, 2007) and among women with diagnosed severe mental
disorders (Mueser et al., 1998). Recent reviews have confirmed that PTSD is more likely to develop
in women than men (Olff, Langeland, Draijer, & Gersons, 2007), along with the complications
of related comorbidity. In establishing a priority focus on the impact of violence in the lives
of women, the Substance Abuse and Mental Health Services Administration funded a multisite
project, the Women, Co-Occurring Disorders, and Violence Study (WCDVS; 1998–2003), to
evaluate the effectiveness of integrated, comprehensive, trauma-informed, consumer-involved
services for women abuse survivors with co-occurring mental health and substance use disorders.
Each of the nine sites used a quasi-experimental design to compare this integrated approach,
including a trauma-specific group intervention, to a “services as usual” comparison condition. A
meta-analysis of study results demonstrated overall advantages for the experimental conditions in
reducing participants’ self-reported problems: mental health symptoms and drug and alcohol use
severity at 6 months after baseline (Cocozza et al., 2005) and mental health and trauma symptoms
at 12 months after baseline (Morrissey et al., 2005). In addition to these overall differences, one of
the study’s key findings was the specific advantage of “integrated counseling,” that is, counseling
that addressed trauma, mental health, and substance abuse in the same time period, in achieving
better outcomes.
A primary mode for implementing such integrated counseling is through a single intervention,
delivered either individually or in group settings, that explicitly and simultaneously addresses
recovery in at least two of the three domains: trauma, mental health, and substance use. For
example, a number of interventions focus on PTSD and substance abuse: Substance Dependence
PTSD Therapy (Triffleman, Carroll, & Kellogg, 1999); Concurrent Treatment of PTSD and
Cocaine Dependence (Back, Dansky, Carroll, Foa, & Brady, 2001); and Seeking Safety (Najavits,
2002). Other approaches attend primarily to the connections between PTSD and severe psychiatric
disorders, for example, cognitive-behavioral treatment for PTSD in severe mental illness (Mueser
et al., 2008). Still others emphasize a broad-based model of overlapping recovery from trauma,
addictions, and/or mental health disorders, such as Addictions and Trauma Recovery (Miller &
Guidry, 2001) and Trauma Affect Regulation: Guide for Education and Therapy (Ford & Russo,
2006).
This study was designed to assess the effectiveness of one of these integrated models, the
Trauma Recovery and Empowerment Model (TREM) group (Fallot & Harris, 2002; Harris,
1998). TREM groups are designed specifically to be responsive to the complex needs of women
with histories of interpersonal violence who have received diagnoses of severe mental disorders
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
76 R. D. Fallot et al.
and, very frequently, alcohol or other drug use disorders. TREM is not an adaptation of a
preexisting approach; it was designed for women members with women group leaders and was
initially developed by women professionals with iterative feedback from group members. Its
gendered approach draws significantly on feminist/relational theories of development as well as
the unique responses of women survivors to sexual and physical violence.
Previous studies of TREM have been reported from primary substance abuse treatment settings
within WCDVS (Amaro et al., 2007; Toussaint, VanDeMark, Bornemann, & Graeber, 2007). In
both the Boston (Amaro et al., 2007) and the Colorado (Toussaint et al., 2007) sites, modifications
(e.g., shorter overall length, more frequent sessions, and open-ended structure) were made to the
usual implementation of TREM. With these modifications, outcomes for women in TREM were
significantly better than those for women receiving usual services in trauma-related and general
mental health symptom domains, although not in drug or alcohol use severity.
The current study reports the results of the District of Columbia Trauma Collaboration Study
(DCTCS), the Washington, DC, site of the WCDVS. This study was designed to test the effec-
tiveness of TREM in comparison to services as usual, and it is the first to implement TREM
according to its developers’ model (full 33-session length, weekly sessions, closed groups after
a specific period) and to report formal outcomes of the TREM intervention in mental health
settings. In line with WCDVS, the primary hypothesis was that TREM would be more effective
than services as usual in reducing PTSD symptoms, general mental health symptoms, and drug
and alcohol use severity among women abuse survivors with co-occurring mental health and sub-
stance use disorders. Group differences were also tested on the following secondary outcomes:
depressive, anxiety, and hostility symptoms; health-related quality of life; personal safety; and
current exposure to interpersonal abuse and other stressful life events.
METHODS
The DCTCS shared the quasi-experimental design of the WCDVS. Women at two mental health
agencies in Washington, DC, received usual community support services plus the TREM groups;
women at two mental health agencies in Baltimore, MD, received usual community support
services but no trauma-specific services. Standardized interviews were conducted at baseline, 6
months, and 12 months, thereby enabling assessment before, during, and after participation in the
8-month TREM group. A full description of the WCDVS design is available elsewhere (McHugo,
Kammerer, et al., 2005).
Study Participants
Women were recruited from two community mental health agencies in Washington, DC, (TREM
condition) and two in Baltimore, MD (Comparison condition). Community support specialists and
clinical supervisors referred women to the research team for eligibility determination. Inclusion
criteria for WCDVS and thus for DCTCS were as follows: (a) women 18 years or older, (b)
a history of sexual and/or physical abuse, (c) co-occurring mental health and substance use
disorders based on chart review (one current; both current within past 5 years), and (d) at least
two service episodes within the formal mental health or substance abuse service system. Due to the
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 77
effectiveness nature of this study, there were no formal exclusion criteria, although clinicians were
unlikely to refer women who were unable to provide informed consent. Prospective participants
were informed fully about the purpose and procedures of the study and about their responsibilities
prior to providing their written consent. The Dartmouth Committee for the Protection of Human
Subjects approved the DCTCS, and it was conducted in accordance with the Declaration of
Helsinki.
Women enrolled over a 14-month period from February 2001 through March 2002. Two
hundred eighty-seven referred women were eligible, and 255 of them provided informed consent
and completed the baseline assessments. Four participants were subsequently dropped from the
study for administrative reasons, leaving 251 in the intention-to-treat study group. One hundred
fifty-three women were enrolled at the intervention sites in Washington, DC (131 at site A and
22 at site B), and 98 women were enrolled at the comparison sites in Baltimore (85 at site A
and 13 at site B). Study participants were paid $30 for the baseline interview and $20 for the 6-
and 12-month follow-up interviews.
Overall, the mean age of the study group was 42 (±8.6) years. The majority (82.1%) was
African American, 14.7% were White, 76.8% completed high school, 33.9% had never been
married, 25.5% were currently married or living as married, 59.7% lived in their own or someone
else’s house or apartment, 17.5% were employed, and 78.9% had children. Most participants
had either a mood disorder (70.1%) or a schizophrenia spectrum disorder (23.9%). The majority
(75.3%) reported at least one psychiatric hospitalization among these; the average number of
psychiatric hospitalizations was 7.9 (±16.2). The most common substance use disorders were
alcohol (34.3%), crack/cocaine (22.7%), opioids (10.4%), and polysubstance use (23.1%). At
baseline, 51% reported abstinence from alcohol, 44.6% reported abstinence from drugs of abuse,
and 29.6% reported abstinence from both alcohol and drugs. These women reported exposure to
an average of 16.5 (±4.8) out of 31 types of stressful life events (as assessed by the Life Stressor
Checklist–Revised [LSC-R]), and the rates of interpersonal abuse were uniformly high: childhood
sexual abuse (72.1%), adulthood sexual abuse (57.4%), childhood physical abuse (59.4%), and
adulthood physical abuse (55.8%).
Sites and Conditions
The TREM sites were two community agencies within the public mental health system in
Washington, DC. TREM groups were offered in the context of integrated trauma services teams
at both agencies. Integrated trauma services teams at both agencies had six to eight community
support specialists who were cross-trained in trauma, mental health, and substance abuse. Each
clinician worked with 15 to 20 women to ensure access to a range of support services. The
clinicians also supported the TREM groups in three ways: (a) by addressing obstacles to TREM
group attendance; (b) by offering to review that week’s material with women who missed a
TREM session, using a self-help workbook that parallels the TREM manual (Copeland & Harris,
2000); and (c) by developing recovery goals that supported the participant’s acquisition of trauma
recovery skills as assessed by the Trauma Recovery and Empowerment Profile (TREP; Harris &
Fallot, 2001).
TREM groups included 33 weekly sessions for 75 minutes with two or three co-leaders.
Based on a fully manualized curriculum, each session has a designated topic, specific goals, and
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
78 R. D. Fallot et al.
guiding questions, as well as at least one skills-oriented exercise. TREM draws on cognitive
restructuring, psychoeducation, and skills-building exercises as well as group support to facilitate
the development of members’ recovery capacities. TREM explicitly limits the use of exposure
techniques, focusing on the current impact of trauma and alternative coping skills.
A TREM group was started when 12 to 15 women had enrolled in the project. If a participant
did not attend her assigned group, she was offered another group at a later time. Fourteen TREM
groups were conducted. The number of participants per group ranged from 6 to 12; the average
was 9.5 participants. Of the 153 women in the TREM condition, 133 (86.9%) attended at least one
TREM group session. Only eight women used one-on-one sessions to make up missed TREM
group sessions; the total number of make-up sessions was 23. Attendance, defined as exposure
to TREM sessions, was computed by adding the number of one-on-one make-up sessions to the
number of TREM group sessions attended. Given this definition, the average number of sessions
attended was 18.7 (SD =11.5); of those who attended at least one session, the average number
attended was 21.4 (SD =9.46).
The comparison sites were two community agencies within the public mental health system
in Baltimore, MD. Clinicians offered an array of usual services, including psychiatric evaluation
and follow-up, psychotherapy, and linkage to other needed supports via case management. These
sites did not offer trauma-specific individual or group therapy. Substance abuse and dual diagnosis
counseling and groups were available.
The TREM group fidelity scale assessed the consistency and quality of implementation.
Independent fidelity assessors rated key elements of the treatment context and leader behaviors
based on documentation and audio recordings. One to three sessions from each TREM group
were chosen randomly for audio recording; an overall fidelity score was computed for each rated
session. Based on 31 group sessions, the overall fidelity scores ranged from 4.04 to 4.93 (on a
1-to-5 scale). The mean rating was 4.73, indicating strong fidelity to the manual’s content and
leadership process.
Measures
An in-person interview that combined cross-site measures from WCDVS and site-specific mea-
sures from DCTCS was administered at baseline and at 6 and 12 months. The average duration of
the interview was 86.4 minutes (SD =25.0) at baseline, 67.8 minutes (SD =21.0) at 6 months,
and 72.0 minutes (SD =21.1) at 12 months. The cross-site interview from WCDVS assessed per-
sonal history, behavioral health, service utilization, and consumer satisfaction. The four primary
outcome variables for WCDVS, and thus for DCTCS, were described in McHugo, Kammerer,
et al. (2005) and are described briefly here.
Posttraumatic symptom severity was assessed by the Posttraumatic Symptom Scale (Foa,
Cashman, Jaycox, & Perry, 1997). Respondents rate how often each of 17 symptoms of PTSD
has bothered them in the past month. Ratings range from 0 (not at all or only one time)to3(5or
more times per week/almost always), and the sum of the ratings indicates posttraumatic symptom
severity. One-week test-retest reliability (N=186) for the Posttraumatic Symptom Scale in the
WCDVS was .79, as determined by the intraclass correlation coefficient.
Alcohol and drug problem severity were assessed by the corresponding sections from the
Addiction Severity Index (McLellan et al., 1992). The Alcohol and Drug Composite Scores
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 79
are based on reported use and perceived problem severity during the past 30 days and range
between 0 and 1, with higher scores indicating greater problem severity. One-week test-retest
reliabilities for the alcohol and drug composite scores in the WCDVS were, respectively, .82
and .86.
Mental health symptom severity was assessed by the global severity index (GSI) from the
Brief Symptom Inventory (BSI; Derogatis, 1993). The BSI has 53 items that cover nine symptom
domains. Respondents rate how much each symptom has bothered them in the past week (ranging
from 0 =not at all to 4 =extremely), and the GSI is the mean of the 53 items. One-week test-retest
reliability for the GSI in the WCDVS was .87. Subscales from the BSI were used in DCTCS to
assess the severity of depression, anxiety, and hostility symptoms.
Trauma history and personal safety measures from the WCDVS were also used in DCTCS.
Lifetime and current trauma history were assessed by the LSC-R (Wolfe & Kimerling, 1997),
which was modified for the WCDVS (McHugo, Caspi et al., 2005). Primary modifications
involved removing follow-up probes from items that were not about interpersonal abuse (e.g.,
age at the time, subjective response, and current effect on life), adding items pertinent to women
with co-occurring disorders (e.g., institutional abuse, discrimination, hate crimes), and limiting
probes for the interpersonal abuse items to age and frequency. The modified LSC-R contained
30 stressful life events and a final open-ended item. Each item asked about lifetime exposure
(yes/no) and, if yes, about current exposure (yes/no in past 6 months). For the 14 items that dealt
with interpersonal abuse and neglect, there were standard probes to determine frequency and age
at onset.
Four summary variables were computed from the LSC-R: (a) lifetime exposure to stressful
events is the count of the 31 items that were positively endorsed, (b) lifetime frequency of inter-
personal abuse uses probes for 9 items to quantify the lifetime frequency of sexual and physical
abuse (range: 0–36; higher scores indicate more interpersonal abuse), (c) current exposure to
interpersonal abuse is the count of positive endorsement of eight items pertaining to recent expe-
rience of sexual or physical abuse, and (d) current exposure to other stressors counts how many
of 14 other stressful events were experienced in the past 6 months. One-week test-retest relia-
bility from WCDVS was acceptable for each of these scales (intraclass correlation coefficient =
.77 to .88). The LSC-R was also used to create four categorical variables that encoded (yes vs.
no) whether each woman had experienced childhood sexual abuse, childhood physical abuse,
adulthood sexual abuse, and adulthood physical abuse.
A six-item scale was developed for WCDVS to measure perceived personal safety in the past
6 months. The items asked about feeling unsafe in general and in the home specifically. The
frequency rating scale ranged from 1 =not at all (or only one time)to4=almost always (or five
or more times per week), and the total score was the mean of the six items; lower scores indicate
greater feelings of safety. Internal consistency (Cronbach’s alpha =.81) and test-retest reliability
(intraclass correlation coefficient =.78) were acceptable in WCDVS.
The DCTCS supplemented the cross-site interview with site-specific measures of both process
and outcome. The additional outcome measure was the SF-12 Health Survey, which is a 12-item
self-report measure of health-related quality of life that yields physical and mental component
summary scales (Ware, Kosinski, & Keller, 1996). The SF-12 is reliable and valid for people
with serious mental illness (Salyers, Bosworth, Swanson, Lamb-Pagone, & Osher, 2000). The
component scales are standard scores (M=50, SD =10), with higher scores indicating higher
levels of functioning.
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
80 R. D. Fallot et al.
Site-specific process measures were collected on TREM condition participants. TREM group
leaders provided attendance data for the TREM groups. TREM-condition clinicians used the
Substance Use Scale (SUS) to rate participants’ substance use disorders on a five-point scale that
combined the Alcohol Use Scale and the Drug Use Scale into a single rating (1 =abstinence,
2=non-problematic use,3=abuse,4=dependence,5=severe dependence). The Alcohol
Use Scale and Drug Use Scale are reliable and valid ratings that are based on DSM criteria for
substance use disorder (Drake et al., 1990). TREM-condition clinicians rated the status of the
alcohol/drug use disorder during the past 6 months at baseline and 12 months.
The TREP (Harris & Fallot, 2001) gathers clinician ratings on 11 trauma recovery skills (self-
awareness, self-protection, self-soothing, emotional modulation, relational mutuality, accurate
labeling of self and others, sense of agency and initiative taking, consistent problem-solving,
reliable parenting, possessing a sense of purpose and meaning, and judgment and decision
making). Each skill is rated on a five-point scale, which has anchors specific to each skill domain.
The total score is the average of the 11 ratings, and higher scores indicate more skills. TREP
ratings were obtained at baseline and 12 months.
Independent interviewers were trained in standardized interviewing techniques, and they were
monitored throughout the study by submitting audio recordings of both eligibility and outcome
interviews. Regular conference calls were held with the interviewers to discuss assessment
issues and to develop consensual solutions. Although separate sets of interviewers covered the
Washington, DC, and Baltimore sites, they crossed over for about one-quarter of the interviews
in order to reduce confounding. TREM-condition clinicians were trained at the outset in the use
of the rating scales (SUS and TREP), and refresher training was provided throughout the study.
Data Analysis
Because this study used a nonequivalent control group design, the first step was to characterize
the TREM (n=153) and comparison (n=98) groups in depth and to test for differences between
them. Baseline group differences across a range of background and outcome variables (see
Table 1) were examined by independent-groups t-tests for continuous measures and chi-square
tests for categorical measures. The second step was to evaluate the effect of attrition by contrasting
participants who completed neither a 6- nor 12-month follow-up interview (n=44) against
those who completed one or both follow-up interviews (n=207) on the same set of variables.
Independent-groups t-tests and chi-square analyses were conducted to test overall differences
between research dropouts and follow-up participants. Two-way ANOVAs (study condition by
study status) and multiway chi-square tests were conducted to determine whether there was
differential dropout between the TREM and comparison conditions.
The third step was to determine the effect of treatment. Because analyses of baseline variables
revealed differences between groups (see Results), we computed propensity scores and included
them as inverse probability of treatment weights in outcome analyses (Harder, Stuart, & Anthony,
2010; Stuart et al., 2009). In essence, each treatment group is weighted up to the “population,”
which is actually the study group itself, thereby reducing baseline differences between the TREM
and comparison groups. The analog of repeated-measures analysis of variance in the mixed-effects
regression framework (SAS PROC MIXED) was used to analyze differences between the TREM
and comparison groups over time. We used covariance pattern models to contrast mean changes
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 81
TABLE 1
Demographic and Clinical Characteristics of the Study Groups at Baseline
TREM Group Comparison Group
Variable Values (n=153) (n=98) Statistic p
Age M(SD) 42.03 (8.66) 42.04 (8.50) t=0.01 .99
Race/ethnicity n(%) Black 133 (86.9) 73 (74.5) χ2=7.61 .02
n(%) White 15 (9.8) 22 (22.4)
n(%) Other 5 (3.3) 3 (3.1)
Education level M(SD) years 11.30 (2.38) 11.42 (2.20) t=0.41 .68
Relationship status n(%) Married/partnered 31 (20.3) 33 (33.7) χ2=5.87 .05
n(%) Widowed/
separated/divorced
65 (42.5) 37 (37.8)
n(%) Never married 57 (37.3) 28 (28.6)
Currently employed n(%) Yes 26 (17.1) 18 (18.4) χ2=0.07 .8
Ever had children n(%) Yes 123 (80.1) 75 (76.5) χ2=0.53 .46
Primary residence (past 30 n(%) Homeless 29 (19.2) 9 (9.4) χ2=23.03 <.001
days) n(%) Institution 4 (2.6) 0 (0)
n(%) Independent 78 (51.7) 78 (81.3)
n(%) Supervised 40 (26.5) 9 (9.4)
Primary psychiatric
diagnosis
n(%) Schizophrenia
spectrum disorders
42 (27.5) 18 (18.4) χ2=3.34 .34
n(%) Mood disorders 101 (66.0) 75 (76.5)
n(%) Anxiety disorders 5 (3.3) 3 (3.1)
n(%) Other 5 (3.3) 2 (2.0)
Ever treated in psychiatric
hospital
n(%) Yes 116 (75.8) 73 (74.5) χ2=0.06 .81
Psychiatric
hospitalization–lifetime
M(SD) Admissions 4.86 (7.61) 5.56 (7.59) t=0.72 .47
PTSD symptom severity
(PSS)
M(SD) 24.74 (11.77) 27.03 (11.79) t=1.48 .14
Global severity index (BSI) M(SD)1.39 (0.77) 1.66 (0.82) t=2.69 .008
BSI depression subscale M(SD)1.49 (1.03) 1.84 (1.10) t=2.61 .01
BSI anxiety subscale M(SD)1.42 (0.95) 1.64 (1.04) t=1.70 .09
BSI hostility subscale M(SD)1.18 (0.94) 1.42 (1.02) t=1.86 .06
Primary substance use n(%) Alcohol 41 (26.8) 45 (45.9) χ2=23.74 <.001
disorder n(%) Cocaine 43 (28.1) 14 (14.3)
n(%) Opioid 10 (6.5) 16 (16.3)
n(%) Polysubstance 45 (29.4) 13 (13.3)
n(%) Other 14 (9.20) 10 (10.2)
Alcohol problem severity
(ASI)
M(SD)0.15 (0.26) 0.11 (0.18) t=1.17 .24
Alcohol use at baseline n(%) Yes 72(47.1) 51 (52.0) χ2=0.59 .44
Drug problem severity
(ASI)
M(SD)0.10 (0.14) 0.09 (0.10) t=0.15 .88
Drug use at baseline n(%) Yes 74 (48.4) 64 (66.0) χ2=7.45 .006
Personal safety scale M(SD)1.90 (0.77) 1.87 (0.80) t=0.10 .92
SF-12: Physical component
score
M(SD) 40.83 (7.10) 39.55 (8.05) t=1.31 .19
(Continued)
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
82 R. D. Fallot et al.
TABLE 1
Demographic and Clinical Characteristics of the Study Groups at Baseline
(Continued)
TREM Group Comparison Group
Variable Values (n=153) (n=98) Statistic p
SF-12: Mental component
score
M(SD) 41.43 (7.66) 40.48 (8.12) t=0.93 .35
Lifetime exposure to
stressful events
M(SD) 16.15 (4.88) 17.02 (4.59) t=1.41 .16
Lifetime frequency of
interpersonal abuse
M(SD) 16.32 (7.59) 16.26 (7.36) t=0.06 .95
Childhood sexual abuse n(%) Yes 111 (71.4) 70 (71.43) χ2=0.04 .85
Adulthood sexual abuse n(%) Yes 88 (57.5) 56 (57.1) χ2=0.03 .95
Childhood physical abuse n(%) Yes 88 (57.5) 61 (62.2) χ2=0.55 .46
Adulthood physical abuse n(%) Yes 82 (53.6) 58 (59.2) χ2=0.76 .38
Current exposure of
interpersonal abuse
n(%) Yes 64 (43.5) 40 (41.2) χ2=0.13 .72
Current exposure to other
stressors
M(SD)2.87 (2.33) 2.97 (2.14) t=0.32 .75
over time between groups (Hedeker & Gibbons, 2006). The between-subjects factor was group
(TREM vs. comparison), and the repeated measure was time (baseline, 6 months, and 12 months).
Time was treated as a classification variable, and the covariance structure was left unspecified
(i.e., fully estimated), due to having only three time points. The group by time interaction tested
the treatment effect. Because mixed-effects regression models can accommodate missing data and
analyses of attrition did not indicate differences between dropouts and completers (see Results),
all enrolled participants were included in the outcome analyses (intention-to-treat analyses).
The fourth step was to examine differences among participants within the TREM condition.
Changes over time (baseline vs. 12 months) on TREM condition-specific measures were assessed
by paired t-tests. The final step was to examine the association (Pearson product-moment corre-
lations) between two process measures (TREM group attendance and TREP change) and study
outcomes. A .05 level of significance was used in baseline and outcome analyses, and .01 was
used for the TREM condition analyses.
RESULTS
Group Differences at Baseline
Table 1 presents the characteristics of the TREM and comparison groups at baseline. The groups
were similar in age, education level, current employment, and having children. The TREM group
had proportionally more African American, unmarried, and non–independently housed women
than the comparison group. In the clinical domain, the two groups were similar in primary
psychiatric diagnoses and lifetime hospitalization. The groups were similar in PTSD symptom
severity, but the comparison group had higher scores on the GSI and the BSI depression subscale.
The comparison group had more alcohol and fewer drug use disorders as well as more participants
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 83
who reported drug use in the past month, but the two groups did not differ in either alcohol or
drug problem severity. The groups did not differ in health-related quality of life. No differences
were found between groups in trauma history, current exposure to traumatic events, or personal
safety. We note that these analyses were powered to detect small effects, and no correction for
multiple tests was applied.
Based on these analyses, six baseline variables on which the two groups differed significantly
(race/ethnicity, relationship status, primary residence, GSI, substance use disorder, and drug use
at baseline) were used in a logistic regression analysis to compute a propensity score for each
participant (i.e., the likelihood of being in the TREM group). Several participants were missing
at least one of the variables used to compute the propensity scores, thereby reducing the TREM
group size from 153 to 151 and the comparison group size from 98 to 95 for the outcome analyses.
Attrition Analysis
Although longitudinal analyses retained all participants regardless of missing outcome data, we
examined differences between participants who completed one or both follow-up assessments and
those who completed neither. One hundred thirty-one (85.6%) participants in the TREM group and
76 (77.6%) participants in the comparison group completed one or both follow-up assessments,
which was a non-significant difference in retention rate. A range of baseline variables was
examined for the effect of attrition (see variables in Table 1), and only one difference was found.
Participants who completed neither follow-up interview had less severe alcohol problems (n=
44, M=.05, SD =.13) than those who participated in follow-up (n=207, M=.15, SD =.25,
t(249) =2.76, p=.006). Likewise, 31.8% (14/44) of the dropouts reported recent alcohol use
at baseline, whereas the rate was 52.7% (109/207) among those who participated in follow-up,
χ2(1, N=251) =6.31, p=.01). No evidence for differential attrition by condition was found
when examining study condition by study status (completer vs. dropout) interactions.
Primary Outcomes
The results for the four primary outcome measures are shown in the top part of Table 2. The
group by time interaction was significant for two of the four outcome measures—alcohol problem
severity and drug problem severity—and in each case the TREM group had better outcomes. The
results for these two outcomes in Table 2 are for those participants who reported using alcohol or
drugs at baseline. The results were similar when the entire study group was analyzed. The groups
did not differ in change in PTSD or global mental health symptom severity. Although the focus
was on differential change over time (i.e., the group by time interaction), it should be noted that
all outcomes, except the two SF-12 scales and the two current trauma exposure variables, showed
a significant main effect of time, indicating that both groups improved.
Secondary Outcomes
Significant group differences were found for the anxiety subscale of the BSI, for the personal
safety scale, and for current exposure to other stressors (Table 2). In each case, the TREM group
had better outcomes than the comparison group. There were no significant group differences for
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
84 R. D. Fallot et al.
TABLE 2
Outcome Means (
SD
) at Baseline, 6, and 12 Months for TREM and Comparison Groups
TREM Comparison Group by Time
Variable Month MSDMSDF p
PTSD symptom severity (PSS)a024.72 11.83 27.42 11.58 1.72 .18
620.77 11.86 25.68 12.31
12 18.41 10.73 24.32 12.08
Alcohol problem severity (ASI)b00.31 0.31 0.21 0.20 7.58 <.001
(baseline alcohol users only) 6 0.10 0.18 0.18 0.25
12 0.15 0.25 0.15 0.21
Drug problem severity (ASI) 0 0.20 0.14 0.14 0.09 10.16 <.001
(baseline drug users only) 6 0.07 0.09 0.12 0.12
12 0.12 0.13 0.11 0.11
Global symptom severity (BSI) 0 1.39 0.78 1.69 0.81 1.20 .30
61.18 0.81 1.60 0.81
12 1.14 0.75 1.55 0.81
Depression subscale (BSI) 0 1.49 1.03 1.86 1.08 1.23 .29
61.23 0.99 1.80 1.13
12 1.19 0.97 1.70 1.09
Anxiety subscale (BSI) 0 1.42 0.96 1.67 1.03 3.33 .04
61.14 0.94 1.70 0.99
12 1.11 0.94 1.60 1.00
Hostility subscale (BSI) 0 1.18 0.94 1.45 1.02 1.44 .24
60.93 0.86 1.27 0.91
12 0.90 0.89 1.33 0.94
Personal safety scale 0 2.02 0.78 2.04 0.79 3.55 .03
61.66 0.63 1.93 0.77
12 1.69 0.69 1.78 0.65
Physical component score (SF-12) 0 40.81 7.14 39.55 8.11 0.09 .91
641.06 7.67 40.54 6.81
12 41.34 7.12 41.74 7.68
Mental component score (SF-12) 0 41.44 7.71 40.48 8.05 2.17 .12
641.97 7.89 41.43 6.83
12 42.38 6.61 39.61 7.91
Current exposure to interpersonal 0 1.10 1.63 0.92 1.51 1.85 .16
abuse 6 0.93 1.43 0.99 1.58
12 0.68 1.46 1.00 1.57
Current exposure to other stressors 0 2.88 2.34 2.99 2.15 3.83 .02
62.24 1.85 3.26 2.55
12 2.42 2.20 2.88 2.09
Note. Statistical results are from repeated measures analysis of variance with inverse probability weights based on
propensity scores.
aAcross all variables, except substance abuse, the n’s for the TREM group vary from 151 to 145 at baseline, from 126
to 122 at 6 months, and from 107 to 103 at 12 months. The n’s for the comparison group vary from 95 to 91 at baseline,
from68to65at6months,andfrom60to56at12months.
bFor alcohol problem severity, the n’s were 72, 60, and 54 for the TREM group and 51, 42, and 38 for the comparison
group at baseline, 6 months, and 12 months, respectively. For drug problem severity, the n’s were 74, 60, and 49 for the
TREM group and 63, 48, and 40 for the comparison group at baseline, 6 months, and 12 months, respectively.
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 85
TABLE 3
Correlations of TREM Group Attendance and TREP Change With Change on Study Outcomes
(Baseline to 12 Months)
TREM Group Change
Variable Attendanceain TREPb
PTSD symptom scale −.19 −.29∗
Global Severity Index −.10 −.32∗
ASI alcohol .01 −.28∗
ASI drug −.19 −.36∗
BSI depression .01 −.26
BSI anxiety −.10 −.24
BSI hostility −.12 −.32∗
Safety scale −.08 −.29∗
SF-12 physical −.02 .08
SF-12 mental −.03 .05
Current exposure to interpersonal abuse .01 −.21
Current exposure to other stressors .05 −.11
aThe n’s for the correlations with TREM group attendance range from 107 to 104.
bThe n’s for the correlations with TREP change range from 92 to 89.
∗p<.01.
depression and hostility symptoms, the physical or mental component scores from the SF-12, or
current exposure to interpersonal abuse.
TREM Condition
Changes over time within the TREM condition were significant for the clinician ratings on
the TREP (M0=1.86, SD =0.66; M12 =2.85, SD =0.80, t(105) =12.83, p<.001) and
the SUS (M0=3.05, SD =1.61; M12 =1.80, SD =1.10, t(109) =8.00, p<.001). TREM
participants improved in trauma recovery skills and decreased their substance use disorder status
on average from abuse to use without impairment. Next we examined the relationship between
attendance at TREM groups and changes in trauma recovery skills. TREM group attendance was
significantly correlated with changes in TREP scores (12 months–baseline; r=.39, n=105,
p=.001). Women who attended more TREM group sessions showed greater gains in TREP
scores. Finally, Table 3 shows the correlations between changes from baseline to 12-month
follow-up on study outcomes and both TREM group attendance and change in TREP score.
TREM group attendance was not associated significantly with outcomes at p<.01, but changes
in TREP scores were associated significantly with 6 of the 12 outcomes.
DISCUSSION
The results of this study provide partial confirmation of the hypothesis that TREM condition
participants would have better outcomes than those receiving services as usual. Significant
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
86 R. D. Fallot et al.
differences were found for two of the four primary outcomes and for three of eight secondary
outcomes. The results are consistent with findings from the multisite WCDVS by showing that
TREM has advantages over usual services in reducing alcohol and drug use severity. This result
was corroborated among TREM participants by community support specialist ratings of substance
use disorders. TREM did not show the expected advantage in reducing posttraumatic symptoms
or overall mental health symptoms, although there was a significant difference in reduction of
anxiety symptoms. There was also a group difference in change in perceived safety, which is a
core component of TREM, and in the experience of recent stressful life events. Expected dif-
ferences were not found in depressive and hostility symptoms, health-related quality of life, or
current experience of interpersonal abuse.
The drug and alcohol findings are in contrast to those reported by the Boston and Colorado
sites of the WCDVS, which found no differences for TREM group participants on drug or alcohol
use severity but did find differences in mental health and posttraumatic symptoms. Two factors
may account for this different outcome pattern. First, these sites offered a modified version
of TREM in which duration and dosage of TREM was lower than that in Washington, DC.
Second, whereas the Washington, DC, study served women recruited mostly through the public
mental health system, the Boston and Colorado projects were based in primary substance abuse
treatment settings, which may have enhanced usual service outcomes in drug and alcohol use. The
mental health settings involved in the current study may have provided comparatively stronger
mental health (as contrasted with substance abuse) services, attenuating the differences somewhat
between TREM and services as usual. Addressing trauma recovery along with substance abuse
in the TREM groups in DCTCS, though, may have added a significant element to the TREM
condition, strengthening the outcomes in drug and alcohol use, especially.
In addition, the relationships between trauma skill development, as assessed by the TREP, and
study outcomes point to the potential value of the skills training component of TREM. Because
TREP scores represent the clinician’s perspective and the study outcomes represent participants’
interview responses, the confluence of these findings warrants further investigation. In contrast
to interventions that focus more intensively on symptom reduction, TREM adopts an overtly
skills- and strengths-oriented approach. A formal mediation analysis was not possible in this
study, because the TREP and other process measures were not collected from the comparison
group. Additional research is needed to examine more directly the mechanisms of action for such
interventions as TREM, including the links between intervention dose (e.g., attendance), skill
development, and clinical outcomes.
There are limitations in the current study, the primary of which is inherent in the quasi-
experimental (non-randomized) design. Although the two groups were similar at baseline across
a range of measures, there were several significant differences, including a complicating difference
in mental health symptoms, although it did not pertain to PTSD symptoms. There was also a
difference in residential setting at baseline; a higher proportion of participants lived in supervised
housing or were homeless in Washington, DC, whereas the proportion in independent housing
was higher in Baltimore. These pre-enrollment residential differences declined during the 1-year
study period, but they may have contributed to the group difference in change in substance
abuse severity over time. To the extent possible, outcome analyses controlled for these baseline
differences by using propensity scores as inverse probability of treatment weights.
The nonequivalence of the TREM and services as usual groups at baseline is heightened
by possible differences between the two conditions’ sites. The Washington, DC, mental health
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 87
agencies may have differed in other aspects of service delivery—that is, other than the TREM
intervention—from the Baltimore sites. Although all sites were private, not-for-profit agencies
funded primarily through the public mental health system, the intensity, organization, and quality
of usual services likely varied among the sites, and the impact of this variation is unknown. That
both groups changed over time, non-differentially in many cases, weakens the case for overall
differences in service effectiveness between conditions. It is also customary to acknowledge the
questionable validity of self-report data, especially of substance use, but the focus on differences
in change over time between conditions minimizes this limitation.
There are also noteworthy strengths in this study. The pattern of findings is consistent across
substance abuse domains and, to a lesser degree, in areas especially relevant for trauma survivors
(anxiety, personal safety, and exposure to current stressors). Although not all outcomes showed
a significant treatment effect, this consistency makes more credible TREM’s effectiveness in
working with women whose presenting difficulties are multiple and complex. In addition, the
findings relating trauma skill development to outcomes are consistent with each other. Because
the clinicians had no knowledge of the women’s own self-reported outcomes in any domain, the
coherent story that emerges between clinician ratings of trauma recovery skills on one side and
participant self-reports on the other is worth noting.
Given the prevalence and powerful impact of interpersonal violence in the lives of women
and the special challenges faced by women with co-occurring mental health and substance use
disorders, there is a need for effective interventions that are engaging, that can be implemented
with fidelity in usual mental health settings, and that facilitate recovery for women in this
population. The results of this study suggest that TREM, a gender-specific group developed with
and for women, is one such promising intervention.
ACKNOWLEDGMENTS
Funding was provided to Community Connections (Grant No. 6 UD1 TI11400-04) by the Sub-
stance Abuse and Mental Health Services Administration under the Women, Co-Occurring Dis-
orders, and Violence Study. The authors would like to express their gratitude for the contributions
of staff members at Community Connections and Lutheran Social Services in Washington, DC,
and at the North Baltimore Center (especially Suzanne Bates-Crandall, the site coordinator) and
People Encouraging People in Baltimore, MD.
The views expressed in this article are those of the authors and do not necessarily represent
those of the Substance Abuse and Mental Health Services Administration.
DISCLOSURES
The authors report no financial conflicts of interest regarding the subject of this manuscript.
Dr. Fallot consults with and provides training through the Connecticut Women’s Consortium
(Hamden, CT) and has been compensated as a member of the Substance Abuse and Mental Health
Services Administration’s Advisory Committee for Women’s Services.
Dr. McHugo consults with Psychological Applications, a small company that develops and
tests computer-based assessments in several areas of physical health care.
Drs. Harris and Xie have no compensation for professional services to report.
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
88 R. D. Fallot et al.
REFERENCES
Amaro, H., Dai, J., Ar´
evalo, S., Acevedo, A., Matsumoto, A., Nieves, R., & Prado, G. (2007). Effects of integrated trauma
treatment on outcomes in a racially/ethnically diverse sample of women in urban community-based substance abuse
treatment. Journal of Urban Health: Bulletin of the New York Academy of Medicine,84, 508–522.
Back, S. E., Dansky, B. S., Carroll, K. M., Foa, E. B., & Brady, K. T. (2001). Exposure therapy in the treatment of PTSD
among cocaine-dependent individuals: Description of procedures. Journal of Substance Abuse Treatment,21(1),
35–45.
Cocozza, J. J., Jackson, E. W., Hennigan, K., Morrissey, J. P., Reed, B. G., Fallot, R., & Banks, S. (2005). Outcomes for
women with co-occurring disorders and trauma: Program-level effects. Journal of Substance Abuse Treatment,28(2),
109–119.
Copeland, M. E., & Harris, M. (2000). Healing the trauma of abuse: A woman’s workbook. Oakland, CA: New Harbinger
Publications.
Derogatis, L. R. (1993). Brief Symptom Inventory (BSI): Administration, scoring and procedures manual (4th ed.).
Minneapolis, MN: National Computer Systems.
Drake, R. E., Osher, F. C., Noordsy, D. L., Hurlbut, S. C., Teague, G. B., & Beaudett, M. S. (1990). Diagnosis of alcohol
use disorders in schizophrenia. Schizophrenia Bulletin,16(1), 57–67.
Fallot, R. D., & Harris, M. (2002). The Trauma Recovery and Empowerment Model (TREM): Conceptual and practical
issues in a group intervention for women. Community Mental Health Journal,38(6), 475–485.
Foa, E. B., Cashman, L., Jaycox, L. H., & Perry, K. (1997). The validation of a self-report measure of posttraumatic stress
disorder: The Posttraumatic Diagnostic Scale. Psychological Assessment,9, 445–451.
Ford, J. D., & Russo, E. (2006). Trauma-focused, present-centered, emotional self-regulation approach to integrated
treatment for posttraumatic stress and addiction: Trauma adaptive recovery group education and therapy (TARGET).
American Journal of Psychotherapy,60(4), 335–355.
Goodman, L. A., Dutton, M. A., & Harris, M. (1995). Episodically homeless women with serious mental illness:
Prevalence of physical and sexual assault. American Journal of Orthopsychiatry,65(4), 468–478.
Grant, D. M., Beck, J. G., Marques, L., Palyo, S. A., & Clapp, J. D. (2008). The structure of distress following trauma:
Posttraumatic stress disorder, major depressive disorder, and generalized anxiety disorder. Journal of Abnormal
Psychology,117(3), 662–672.
Harder, V. S., Stuart, E. A., & Anthony, J. C. (2010). Propensity score techniques and the assessment of measured
covariate balance to test causal associations in psychological research. Psychological Methods,15, 234–249.
Harris, M. (1998). Trauma recovery and empowerment: A clinician’s guide for working with women in groups.New
York, NY: The Free Press.
Harris, M., & Fallot, R. D. (2001). The Trauma Recovery and Empowerment Profile (TREP). Unpublished manuscript,
Community Connections, Washington, DC.
Hedeker, D., & Gibbons, R. D. (2006). Longitudinal data analysis. Hoboken, NJ: Wiley.
Hedtke, K. A., Ruggiero, K. J., Fitzgerald, M. M., Zinzow, H. M., Saunders, B. E., Resnick, H. S., & Kilpatrick, D. G.
(2008). A longitudinal investigation of interpersonal violence in relation to mental health and substance use. Journal
of Consulting and Clinical Psychology,76(4), 633–647.
McHugo, G. J., Caspi, Y., Kammerer, N., Mazelis, R., Jackson, E. W., Russell, L., ... Kimerling, R. (2005). The
assessment of trauma history in women with co-occurring substance abuse and mental disorders and a history of
interpersonal violence. Journal of Behavioral Health Services Research,32(2), 113–127.
McHugo, G. J., Kammerer, N., Jackson, E. W., Markoff, L. S., Gatz, M., Larson, M. J., ...Hennigan, K. (2005). Women,
Co-occurring Disorders, and Violence Study: Evaluation design and study population. Journal of Substance Abuse
Treatment,28(2), 91–107.
McLellan, A. T., Kushner, H., Metzger, D., Peters, R., Smith, I., Grissom, G., ...Argeriou, M. (1992). The fifth edition
of the Addiction Severity Index. Journal of Substance Abuse Treatment,9(3), 199–213.
Miller, D., & Guidry, L. (2001). Addictions and trauma recovery: Healing the body, mind, and spirit. New York, NY:
Norton.
Moracco, K. E., Runyan, C. W., Bowling, J. M., & Earp, J. A. (2007). Women’s experiences with violence: A national
study. Women’s Health Issues,17(1), 3–12.
Morrissey, J. P., Jackson, E. W., Ellis, A. R., Amaro, H., Brown, V. B., & Najavits, L. M. (2005). Twelve-month outcomes
of trauma-informed interventions for women with co-occurring disorders. Psychiatric Services,56(10), 1213–1222.
Journal of Dual Diagnosis
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011
Trauma Recovery and Empowerment Model 89
Mueser, K. T., Goodman, L. B., Trumbetta, S. L., Rosenberg, S. D., Osher, C., Vidaver, R., ...Foy, D. W. (1998). Trauma
and posttraumatic stress disorder in severe mental illness. Journal of Consulting and Clinical Psychology,66(3),
493–499.
Mueser, K. T., Rosenberg, S. D., Xie, H., Jankowski, M. K., Bolton, E. E., Lu, W., ...Wolfe, R. (2008). A randomized
controlled trial of cognitive-behavioral treatment for posttraumatic stress disorder in severe mental illness. Journal of
Consulting and Clinical Psychology,76(2), 259–271.
Najavits, L. M. (2002). Seeking safety: A treatment manual for PTSD and substance abuse. New York, NY: Guilford.
Najavits, L. M., Weiss, R. D., & Shaw, S. R. (1997). The link between substance abuse and posttraumatic stress disorder
in women. A research review. American Journal of Addictions,6(4), 273–283.
Olff, M., Langeland, W., Draijer, N., & Gersons, B. P. (2007). Gender differences in posttraumatic stress disorder.
Psychological Bulletin,133(2), 183–204.
Ouimette, P., Goodwin, E., & Brown, P. J. (2006). Health and well-being of substance use disorder patients with and
without posttraumatic stress disorder. Addictive Behavior,31(8), 1415–1423.
Rosen, C. S., Ouimette, P. C., Sheikh, J. I., Gregg, J. A., & Moos, R. H. (2002). Physical and sexual abuse history and
addiction treatment outcomes. Journal of Studies on Alcohol,63(6), 683–687.
Salyers, M. P., Bosworth, H. B., Swanson, J. W., Lamb-Pagone, J., & Osher, F. C. (2000). Reliability and validity of the
SF-12 health survey among people with severe mental illness. Medical Care,38(11), 1141–1150.
Stuart, E. A., Marcus, S. M., Horvitz-Lennon, M. V., Gibbons, R. D., Normand, S. T., & Brown, C. H. (2009). Using
non-experimental data to estimate treatment effects. Psychiatric Annals,39, 719–728.
Toussaint, D., VanDeMark, N., Bornemann, A., & Graeber, C. (2007). Modifications to the Trauma Recovery and
Empowerment Model (TREM) for substance-abusing women with histories of violence: Outcomes and lessons
learned at a Colorado substance abuse treatment center. Journal of Community Psychology,35(7), 879–894.
Triffleman, E., Carroll, K., & Kellogg, S. (1999). Substance dependence posttraumatic stress disorder therapy. An
integrated cognitive-behavioral approach. Journal of Substance Abuse Treatment,17(1–2), 3–14.
Van der Kolk, B. A., Roth, S., Pelcovitz, D., Sunday, S., & Spinazzola, J. (2005). Disorders of extreme stress: The
empirical foundation of a complex adaptation to trauma. Journal of Traumatic Stress,18(5), 389–399.
Ware, J., Jr., Kosinski, M., & Keller, S. D. (1996). A 12-Item Short-Form Health Survey: Construction of scales and
preliminary tests of reliability and validity. Medical Care,34(3), 220–233.
Wolfe, J., & Kimerling, R. (1997). Gender issues in the assessment of posttraumatic stress disorder. In J. P. Wilson & T.
M. Keane (Eds.), Assessing psychological trauma and PTSD (pp. 192–238). New York, NY: Guilford.
2011, Volume 7, Numbers 1–2
Downloaded By: [Davis, Rebecca] At: 14:51 18 May 2011