Treating Substance Abuse and Trauma Symptoms in Incarcerated Women

Abstract

Incarcerated women report high rates of trauma exposure and substance use. The present study evaluated an integrated treatment program, Helping Women Recover/Beyond Trauma (HWR/BT), supplemented with additional modules on domestic violence, relapse prevention, and a 12-step program. The HWR/BT combined treatment program was compared with a matched comparison sample that did not receive the target treatment. Self-report measures were collected from 95 incarcerated women, with 56 women in the completer sample. Women in the treatment condition attended a 4-month group treatment. Results indicated statistically significant between-group differences, favoring the treatment condition, for negative posttraumatic cognitions. Pre–post, but not between-group, differences were also observed for posttraumatic stress disorder (PTSD) symptoms and substance-related self-efficacy, whereas no differences were observed for depression, dissociation, tension reduction, or anxious arousal. The present study indicates some promise for specific aspects of the treatment, although results question the overall benefit of the program over standard prison services.

Keywords

alcohol and drugs PTSD treatment women offenders

Incarcerated women evidence particularly high rates of trauma exposure compared with community samples (e.g., Acierno et al., 2001; Barth, Bermetz, Heim, Trelle, & Tonia, 2013; Browne, Miller, & Maguin, 1999; Wolff et al., 2011). Reactions to trauma exposure can be significant and long lasting and may include the development of posttraumatic stress disorder (PTSD) and substance use disorders (SUD; e.g., Pietrzak, Goldstein, Southwick, & Grant, 2011). A needs assessment by Wolff et al. (2011) found that 88% of the incarcerated women in their sample reported full or subthreshold PTSD, and 87% were classified as having an SUD. Furthermore, in incarcerated women with SUD, prevalence rates of PTSD range from 38.6% (Harrington & Newman, 2007) to 79% (Wolff et al., 2011).

Overall, poorer treatment prognosis has been associated with comorbid PTSD and SUD versus either disorder alone (Kubiak, 2004; Najavits et al., 2007). Because of the relationship between SUD and PTSD, researchers and clinicians have suggested that treatments for substance abuse should incorporate a trauma-informed or trauma-specific component, as treatment is unlikely to be effective if the issue driving the substance use is not addressed (Najavits, 2005). Several such integrated treatments have been created and evaluated in the past decade. Overall, research has shown positive pre–post outcomes for integrated treatments in reducing PTSD and substance use, although less consistent support has been found for unique benefits compared with control groups (see Najavits & Hien, 2013, for a full review). When between-group differences have been found, they have more often been for PTSD, indicating additional research may be needed to identify integrated treatments that effectively treat both PTSD and SUD relative to other forms of treatment.

The focus of the present study is to evaluate one emerging integrated treatment approach, Helping Women Recover: A Program for Treating Addiction (HWR; Covington, 1999/2008) and Beyond Trauma: A Healing Journey for Women (BT; Covington, 2003), delivered sequentially in a group format as a combined treatment package (HWR/BT). HWR/BT is manualized, targets both substance use and symptoms following trauma exposure, and is based broadly on relational theory and gender responsiveness (Covington, 2007). Gender-responsive approaches consist of “comprehensive services that take into account the content and context of women’s lives” (Covington & Bloom, 2006, p. 30) and address “gender-specific adversities” (p. 10). HWR consists of 17 sessions spanning four modules, including Self, Relationships, Sexuality, and Spirituality. BT consists of 11 additional sessions organized into three modules, including Violence, Abuse, and Trauma; the Impact of Trauma; and Healing From Trauma. Both HWR (Covington, 1999/2008) and BT (Covington, 2003) include psychoeducation and cognitive–behavioral techniques, as well as expressive arts (e.g., drawing, music), and are based on increasing coping skills and emotional expressiveness following trauma. Both HWR (Covington, 1999/2008) and BT (Covington, 2003) are designed to be delivered over a 3- to 4-month period in either outpatient, residential, or corrections settings.

Messina and colleagues have conducted a series of studies evaluating HWR/BT and other gender-responsive approaches. Women attending HWR/BT were compared with women attending either a prison-based therapeutic community (Messina, Grella, Cartier, & Torres, 2008, 2010) or a mixed-gender outpatient substance abuse treatment program (Messina, Calhoun, & Warda, 2012). Regardless of setting, both HWR/BT and the comparison groups showed significant reduction in drug use and psychological symptoms. No significant between-group differences were observed in either study. Of note, in a larger secondary analysis, Messina and Calhoun (2014) combined data from several previous studies and found significant between-group differences for PTSD, favoring gender-responsive programs (i.e., Seeking Safety or HWR/BT). In summary, research has shown HWR/BT to decrease both trauma symptoms and substance use. However, similar to many other integrated treatments, its superiority to other interventions has not yet been well-established. Given the complexity of integrated treatments and the required clinical resources, additional evaluation is warranted to determine whether HWR/BT provides unique advantage over standard therapeutic services.

The purpose of the present study was to extend previous research through a pre–post treatment evaluation of an HWR/BT-based program in a sample of incarcerated women. In the present study, HWR/BT was supplemented with additional modules on relapse prevention (Black, 2000, for Cohort 1; Gorski & Grinstead, 2010, for Cohorts 2-3), domestic violence (Kubany, McCaig, & Laconsay, 2003), and 12-step addiction recovery as part of a comprehensive treatment program (Covington, 2000). These modules were presented simultaneously with the main treatment (i.e., presented as a portion of each treatment session). It should be noted that the integration of additional modules into the treatment program is specific to the site of the present study and does not reflect the original HWR/BT treatment program. It was hypothesized that incarcerated women with SUDs and a history of trauma enrolled in the integrated treatment program (HWR/BT) would report a statistically significant improvement from pre- to post-treatment on PTSD symptoms, self-efficacy related to substance use, negative posttraumatic cognitions, and other mental health symptoms (anxiety, dissociation, tension reduction, depression) versus those in a non-treatment comparison condition receiving standard prison services.

Method

Participants

Participants were women recruited from a minimum-security correctional facility in the Midwest. The study was approved by the Institutional Review Boards for both the sponsoring university and the Department of Corrections. Women were enrolled in an integrated treatment program for trauma-related symptoms and substance use or were members of a non-treatment matched comparison condition. Data were collected from three cohorts over a 1-year period (March 2011 through March 2012). An a priori power analysis indicated 48 women were needed to reach adequate power. A total of 95 women completed the baseline assessment, including 34 women in Cohort 1 (18 treatment, 16 comparison), 34 women in Cohort 2 (17 treatment, 17 comparison), and 27 in Cohort 3 (20 treatment, 7 comparison), with 56 women completing both pre- and post-assessments.

Complete demographic information can be found in Table 1. Of those with complete demographic information at baseline, 95.7% (n = 90) reported experiencing a potentially traumatic event (PTE) in their lifetime. The most frequently reported PTE in the present sample was a history of being attacked with a weapon, reported by 51% of responders. This was followed closely by unwanted sexual contact before the age of 13 (48%), witnessing someone being killed or seriously injured (43%), being attacked without a weapon (42%), and being forced to have sexual contact before the age of 18 (42%). In the baseline sample, about 48% of the women met the recommended cutoff for a likely diagnosis of PTSD, whereas approximately 64% met the cutoff for depression. The most common primary substance of choice in the present study was methamphetamine, reported by 36 women (38%), followed by cocaine (n = 18, 19%) and marijuana (n = 12, 13%). Criminal history was assessed for all crimes ever committed (not restricted to those resulting in arrest). The most commonly reported crime was possession or distribution of drugs, reported by 88% of the sample, followed by shoplifting (68%) and parole or probation violations (54%).

Table 1.

Demographic Characteristics (N = 94 Unless Noted).

Characteristic	n	%
Racial background
Caucasian	54	57.4
African American/Black	18	19.1
American Indian/Alaskan Native	16	17.0
Mixed/multiracial	5	5.3
Other	1	1.1
Current sexual orientation
Heterosexual	72	76.6
Bisexual	16	17.0
Gay/lesbian	3	3.2
Unsure	3	3.2
Current relationship status
Single	35	37.2
Divorced/separated	34	36.2
Married	23	24.5
Widowed	2	2.1
Highest education
Less than 8th grade	3	3.2
9th grade	7	7.4
10th grade	7	7.4
11th grade	6	6.4
Graduated high school/GED	36	38.3
Some college	32	34.0
Graduated college	1	1.1
Professional school	2	2.1
Pre-incarceration employment (n = 95)
Unemployed	55	57.9
Full-time	30	31.6
Part-time	8	8.4
Student	2	2.1
Pre-incarceration income
US$0-US$12,000	51	54.2
US$13,000-US$20,000	24	25.5
US$21,000-US$30,000	12	12.8
US$31,000-US$40,000	3	3.2
US$41,000-US$50,000	2	2.1
More than US$50,000	2	2.1

Note. Data available for 94 of the 95 women on most items. GED = general education development.

Procedure

Data were collected at a minimum-security facility housing about 800 women. Women in the treatment condition attended the HWR/BT comprehensive treatment program consisting of 3 weekly sessions, each held for 3 hr per session for 4 consecutive months. Women in the matched comparison condition did not attend the target treatment. Women in both groups continued to attend standard prison-based treatments, which may have included individual counseling or other elective treatment programs (e.g., Genesis One, Alcoholics Anonymous).

Inclusion/exclusion criteria

Inclusion criteria were pre-determined by the prison clinical staff for selection into the treatment and were adopted in the present study for compiling the matched comparison sample. The Level of Service Inventory–Revised (LSI-R; Andrews & Bonta, 1995) was originally used to identify women for the treatment and comparison conditions, using a cutoff score of 30 or higher to qualify for the study (as determined by prison clinical staff). However, post-data analyses indicated that some women in the treatment condition had lower LSI scores, so risk scores were ultimately entered as a covariate. Women were required to be 18 years of age or older, to have 1.5 years or less remaining in prison, a case plan or staff member identifying substance abuse treatment need, and ability to read and comprehend as evidenced by at least a sixth-grade reading level on the Test of Adult Basic Education (TABE; Tests of Adult Basic Education: Forms 9 & 10 Technical Report, 2004). Once inmates were screened and selected for HWR/BT participation by prison staff, women in each cohort were assigned to smaller treatment groups based on their LSI-R risk scores (low LSI-R scores in one group and high LSI-R scores in another). The groups were separated by risk level as part of standard prison procedures, based on research findings that treatment response may vary according to offender risk (Lowenkamp & Latessa, 2005). All of the women in the HWR/BT program received the same treatment regardless of risk level.

A statistical analyst compiled a matched comparison sample electronically, based on the inclusion criteria described above. The only exclusionary criterion for the comparison condition was prior HWR/BT treatment completion or being scheduled for upcoming treatment. The list of names was sent directly to the treatment director and was combined with the treatment group list before being sent to the primary investigator, to eliminate any potential observer bias. Once data collection was complete for each cohort, the treatment status of each participant was given to the primary investigator to aid in analyses.

All of the women in both conditions were given the option to participate in the present study by completing two assessment packets (one pre- and one post-treatment). In addition, the women were informed that their decision to participate would not affect treatment. The women in the comparison condition were assessed at each time point that the treatment condition was assessed, but they did not receive the HWR/BT treatment or ancillary components between testing sessions. However, they remained eligible to receive the treatment at a later date. Those who decided to participate completed an informed consent form and the self-assessment packet. No names were used on the measures; all participants were assigned a confidentiality code that was used to link their pre- and post-assessments. The post-treatment assessment was conducted in the same manner. Following both assessments, the inmates were allowed to select an item from a grab-box. The grab-box included items worth approximately US$2 or less, such as candy, snack cakes, and hair clips, purchased from the correctional center canteen, as approved by prison policy.

Using the above methods, a total of 141 women were approached to complete the study. Of those, 95 women entered the study and completed a baseline assessment, and 56 were available for final data analyses (34 treatment, 22 comparison). Figure 1 shows a flowchart documenting how participants progressed through the study, by treatment status.

Figure 1.

Consolidated Standards Of Reporting Trials (CONSORT) 2010 flow diagram.

Treatment fidelity

All treatment sessions were conducted by master’s-level licensed professional counselors. Therapists were trained in the HWR/BT curriculum by the treatment director. It was originally hoped to monitor audio recordings of treatment sessions to ensure therapist fidelity. A section discussing this audio recording of sessions was included in the informed consent and required a separate initial from participants. However, at least one participant in each cohort refused to provide consent for recording the sessions. As such, recording of sessions was not possible in the present study. Instead, treatment fidelity ratings were completed by each therapist and included a list of required activities for each component of the treatment program. Therapists initialed and dated next to each activity to indicate completion of the activity. At completion of the study, fidelity checklists indicated that 100% of the required activities were completed across all cohorts and treatment groups for the HWR/BT components. The only additional module that was not 100% compliant was Relapse Prevention, which averaged 88% compliance with required activities. In addition, an interview with the treatment director indicated that no violations of procedure were observed, and no abnormal incidents or negative reactions were noted during the study period.

Measures

Each participant completed a packet consisting of the following self-report measures:

Demographics and History Questionnaire

The demographics and history questionnaire was created for the present study. Descriptive information was gathered from this questionnaire regarding general demographic information in addition to frequencies of substance use, criminal behavior, and temporal patterns of behavior (e.g., age of first substance use or age of first criminal act).

Trauma Assessment for Adults–Modified (TAA)

Parts of the following description were taken from Avant, Swopes, Davis, and Elhai (2011). The TAA (Resnick, Best, Kilpatrick, Freddy, & Falsetti, 1993) was administered to assess for lifetime history of 15 specific types of traumatic events, including interpersonal traumas, natural disasters, witnessing a violent crime, and being diagnosed with a serious illness. A modified version of the original assessment was used, with reformatted questions to aid in understanding, three added questions about “other” traumatic events that were not included in the standard assessment, and additional follow-up questions to assess diagnostic criteria. The psychometric properties of the modified version have not been established, but adequate reliability has been reported on the original version (Gray, Elhai, Owen, & Monroe, 2009). In the present study, the types of PTEs endorsed on the TAA were summed to represent the frequency of different traumatic events experienced.

PTSD Checklist–Civilian Version (PCL-C)

The PCL-C (Weathers, Litz, Huska, & Keane, 1994) is a widely used 17-item self-report checklist that corresponds to the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM-IV-TR; American Psychiatric Association [APA], 2000) criteria for PTSD, using a 5-point Likert-type scale ranging from 1 (not at all) to 5 (extremely). Acceptable psychometric properties have been identified (Ruggiero, Del Ben, Scotti, & Rabalais, 2003). In the present study, a total score was computed to represent PTSD symptom severity. A cutoff of 38 was used to represent a likely PTSD diagnosis, as recommended by Harrington and Newman (2007). Cronbach’s alpha for the total score was .94 at the pre-assessment and .93 at the post-assessment.

Posttraumatic Cognitions Inventory (PTCI)

The PTCI (Foa, Ehlers, Clark, Tolin, & Orsillo, 1999) is a 36-item self-report measure of cognitions related to traumatic experiences, rated on a 1 (totally disagree) to 7 (totally agree) Likert-type scale. A total score and three subscale scores are assessed by the PTCI, including Negative Cognitions About Self, Negative Cognitions About the World, and Self-Blame. Adequate reliability and validity have been established (Foa et al., 1999). Foa and colleagues identified statistically significant relationships between the overall PTCI total score and PTSD symptom severity, depression, and anxiety. As such, the current study used only the total score as a primary outcome variable to minimize the Type I error rate, with Cronbach’s alphas ranging from .94 to .95 for the total score across both time points.

Drug-Taking Confidence Questionnaire–Drug Scale (DTCQ)

The DTCQ (Sklar, Annis, & Turner, 1997) is a 50-item self-report assessment of confidence and self-efficacy to refuse drugs or alcohol across situations of varying risk levels. Participants are asked what percentage of confidence they have to cope with a particular situation (e.g., “If I had a headache or was in physical pain,” “If I were anxious or tense about something,” p. 658) without using their drug of choice. Items are rated on a 6-point Likert-type scale (0 = not at all confident to 100 = very confident). In the current study, participants were instructed to respond to the DTCQ for the substance that has caused them the most problems. Instead of asking the participant to respond to the appropriate drug or alcohol scale relevant to their worst substance (as specified in the manual), only the drug scale was used in the present study (due to participant confusion found during pilot testing). Participants were instructed to respond based on their drug of choice, including alcohol. Good psychometric properties have been identified for the DTCQ (Sklar et al., 1997), although no studies have examined the use of drug scales as used in the present study (including alcohol on the drug scale). In the current study, an overall mean percentage of confidence was used as an outcome measure of substance use. Cronbach’s alpha for the total score was .99 at both time points.

Trauma Symptom Inventory (TSI)

The TSI (Briere, 1995) is a 100-item self-report measure that assesses 10 clinical domains following trauma, including Anxious Arousal, Depression, Anger/Irritability, Intrusive Experiences, Defensive Avoidance, Dissociation, Sexual Concerns, Dysfunctional Sexual Behavior, Impaired Self-Reference, and Tension Reduction Behavior. Three additional validity scales include Atypical Response, Response Level, and Inconsistent Response. Items are assessed on a Likert-type scale ranging from 0 (never) to 3 (often). Acceptable reliability and validity have been demonstrated for the TSI (Briere, 1995; Briere, Elliott, Harris, & Cotman, 1995). The current study found alpha coefficients ranging from .73 to .94 at the pre-assessment (M = .85), and from .80 to .91 at the post-assessment (M = .87).

In the present study, age-based normative T-scores were calculated, with scores of 65 or greater considered clinically significant (Briere, 1995). Scores from the three validity scales were first used to determine validity of the clinical scales, and participants with invalid protocols at the pre- or post-assessment were excluded from analyses. To minimize the likelihood of Type I errors, only 3 of the 10 clinical scales (Dissociation, Anxious Arousal, and Tension Reduction) were evaluated in the present study, chosen based on relevance to the present study as well as previous research findings (e.g., Covington, Burke, Keaton, & Norcott, 2008; Najavits & Walsh, 2012).

Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977)

This is a 20-item self-report scale designed to measure depressive symptoms. Respondents rate the frequency over the past week of 20 symptoms ranging from 0 (rarely or none of the time) to 3 (most or all of the time). A total score is calculated by summing all items, and ranges from 0 to 60, with higher scores indicating increased depression, and a cutoff of 16 indicating depression in the diagnostic range. Acceptable reliability has been found across a wide variety of demographic characteristics, including age, education, geographic area, and racial, ethnic, and language groups (Radloff, 1977). The current study found Cronbach’s alphas ranging from .89 to .93 across the two time points.

The Level of Service Inventory–Revised (LSI-R)

The LSI-R (Andrews & Bonta, 1995) is used to identify both risks and needs in offenders. It consists of a 54-item assessment completed from semi-structured interview and records review. The assessment covers 10 subscales, including Criminal History, Education and Employment, Financial Circumstances, Family and Marital Situation, Accommodations or Housing, Leisure and Recreation, Companions, Drug and Alcohol Abuse, Emotional and Personal Characteristics, and Attitudes and Orientations. A total score is calculated to estimate offender risk levels. In the present study, the LSI-R was administered by prison staff as part of their standard procedure, and the total risk score from this screening instrument was provided to the principal investigator following completion of the study to aid in data analysis. Adequate predictive validity of the LSI-R has been reported for various treatment outcome variables (e.g., rearrest, recidivism) across a variety of populations (e.g., Kelly & Welsh, 2008; Simourd, 2004).

Data Analysis

In the current study, total scores for each of the outcome measures were calculated if participants had completed at least 75% of the items for that scale. For the TSI, age-based normative T-scores were calculated. Completers were defined as women from either condition (treatment or control) who had a total score for at least one of the primary outcomes at both the pre- and post-assessments. For women in the treatment condition, they also must have graduated treatment according to the agency’s pre-determined standards. Preliminary analyses were conducted between completers versus non-completers, and between treatment versus control conditions. In addition, comparisons were made to assess for differences between study cohorts (Cohort 1 vs. 2 vs. 3). Finally, for those who received treatment, differences were assessed by therapist (Therapist 1 vs. 2), individual treatment groups within each cohort (high vs. low risk), and the version of Relapse Prevention received (RP 1 vs. 2).

ANOVA and chi-square analyses were used to examine baseline differences in demographic and dependent variables, whereas mixed repeated-measures ANCOVA and ANOVA analyses were used to assess for significant differences between the treatment and control conditions on the dependent variables. Bonferroni adjustments were used for all multiple comparisons to minimize the Type 1 error rate. Partial eta-squared was used as a measure of effect size, and Cohen’s (1988) guidelines were used for interpreting eta-squared (.01 = small, .06 = medium, .14 = large).

Results

Of the 95 women who completed baseline assessments, no significant differences were found between completers and non-completers on any of the baseline demographic or dependent variables. As such, only the 56 completers were included in subsequent analyses. Baseline demographic information for the total sample is found in Table 1, whereas baseline differences for the completer sample and the two conditions (treatment vs. control) can be found in Table 2. Significant differences were identified at baseline for treatment versus control based on age, F(1, 55) = 13.43, p = .001, and LSI-R total score, F(1, 54) = 5.48, p = .023. Statistically significant differences were also found at baseline for the smaller treatment groups within each cohort (high vs. low risk) for pre-incarceration employment status, Yates’s corrected χ²(1, N = 34) = 9.06, p = .003. As would be expected, baseline LSI-R total scores also statistically significantly differed by risk group, F(1, 32) = 18.29, p < .001. All of these variables were considered covariates. However, neither age nor employment was significantly related to any of the dependent variables, and thus, both were dropped from the model. LSI-R total score was statistically significantly related to pre-treatment PTCI (r = .34, p = .012) and post-treatment depression (r = .30, p = .025). As the LSI-R is an indicator of offender risks and needs relevant to treatment, and because it has established predictive validity in various domains (e.g., Kelly & Welsh, 2008; Simourd, 2004), it was decided to retain the LSI-R as a covariate for PTCI and depression in the main analyses.

Table 2.

Baseline Differences by Treatment Condition, Valid Trauma-Exposed Completers (N = 56).

	Total (N = 56)		Treatment (n = 34)		Comparison (n = 22)
Variable	M	SD	M	SD	M	SD	F(df)
Age	35.48	8.56	38.53	8.88	30.77	5.47	13.43 (1, 54)**
PTEs	5.39	2.66	5.21	2.75	5.68	2.55	0.42 (1, 54)
LSI-R total score	29.36	5.26	28.06	4.98	31.32	5.17	5.48 (1, 53)*
Baseline dependent variables
PTSD	37.46	14.25	37.44	13.96	37.50	15.03	0.00 (1, 54)
PTCI	98.37	33.82	96.86	32.40	100.64	36.51	0.16 (1, 53)
Depression	20.00	13.12	20.18	12.43	19.73	14.42	0.02 (1, 54)
TSI anxiety	52.17	10.21	52.16	9.80	52.18	11.05	0.00 (1, 54)
TSI dissociation	53.84	9.02	53.98	8.17	53.63	10.39	0.02 (1, 54)
TSI tens. reduc.	52.38	11.25	52.03	10.04	52.91	13.14	0.08 (1, 54)
DTCQ	67.41	28.94	67.66	27.57	67.04	31.60	0.01 (1, 54)
	n	%	n	%	n	%	χ²
Racial background							1.63^a
Caucasian	31	55.4	16	47.1	15	68.2
Not White/Other	25	44.6	18	52.9	7	31.8
Current relationship status							1.31^a
Never married	24	42.9	12	35.3	12	54.5
Prev. married	32	57.1	22	64.7	10	45.5
Highest education							3.54
Less than HS	12	21.4	10	29.4	2	9.1
HS/GED	18	32.1	9	26.5	9	40.9
College	26	46.4	15	44.1	11	50.0
Pre-incarceration employment							0.00^a
Part- or full-time	21	37.5	13	38.2	8	36.4
Not employed	35	62.5	21	31.8	14	63.6
Pre-incarceration income							0.14^a
US$12,000 or below	29	53.7	16	50.0	13	59.1
US$13,000 or higher	25	46.3	16	50.0	9	40.9

Note. Percentages may not add up to exactly 100 due to rounding. PTEs = potentially traumatic events; LSI-R = Level of Service Inventory–Revised; PTSD = posttraumatic stress disorder; PTCI = posttraumatic cognitions; tens. reduc. = tension reduction; DTCQ = drug-taking confidence questionnaire; prev. = previous; HS = high school; GED = general education development.

Yates’s continuity correction.

p < .05. **p < .01.

To test the main hypotheses, a mixed model repeated-measures ANCOVA was conducted for PTCI and depression, and ANOVA was used for the remaining dependent variables. Assumptions were tested prior to data analyses. All were met with the exception of normality of data, requiring transformation of two variables (log to the base 10, or LG10, for TSI tension reduction, and square transformation for DTCQ). Relevant to the models including covariates (PTCI and depression), additional assumptions for ANCOVA were examined and met.

Results are summarized in Table 3. Results indicated a statistically significant main effect of Time, but not for the Time × Treatment interaction, for PTSD symptoms and substance use self-efficacy, suggesting overall pre–post improvements, but with no specific advantage for the treatment condition. For posttraumatic cognitions, results indicate a statistically significant Time × Treatment interaction. Specifically, although no significant differences were found at baseline, the simple effect of Treatment was significant at post-treatment, such that the treatment condition reported significantly fewer negative posttraumatic cognitions than the control condition following treatment. Examination of simple effects of Time by treatment condition showed no statistically significant pre–post differences for the control condition, but a statistically significant pre–post reduction for the treatment condition. No differences were found by Time or Treatment condition for the remaining mental health symptoms (depression, anxious arousal, dissociation, and tension reduction).

Table 3.

Analysis of Variance/Covariance Results for Treatment Versus Comparison Conditions (N = 56).

Variable (df)	F	MSE	$η_{p}^{2}$
PTSD (1, 54)
Time	10.22**	82.36	.16
Treatment	0.67	256.09	.01
G × T	1.99	82.36	.04
PTCI (1, 51)
Time	1.87	301.21	.04
Treatment	0.81	1,721.29	.02
G × T	8.83**	301.21	.15
CES-D (1, 52)
Time	0.17	49.46	.003
Covariate	4.27*	185.68	.08
Treatment	0.02	185.68	.00
G × T	2.95	49.46	.05
TSI anxious arousal (1, 54)
Time	0.33	39.24	.01
Treatment	0.29	162.16	.01
G × T	1.17	39.25	.02
TSI dissociation (1, 54)
Time	2.45	24.91	.04
Treatment	0.25	141.05	.005
G × T	2.45	24.91	.04
TSI tension reducing behaviors (log to the base 10 transformed; 1, 54)
Time	0.30	0.004	.01
Treatment	0.57	0.01	.01
G × T	0.84	0.004	.02
DTCQ (square transformed; 1, 53)
Time	6.70*	6,208,004.30	.11
Treatment	0.93	15,116,581.28	.02
G × T	2.13	6,208,004.30	.04

Note. Time = pre-treatment versus post-treatment. G × T = group (condition) by time interaction. PTSD = posttraumatic stress disorder; PTCI = posttraumatic cognitions; CES-D = Center for Epidemiologic Studies Depression Scale; TSI = Trauma Symptom Inventory; DTCQ = drug-taking confidence questionnaire.

p < .05. **p < .01.

Discussion

The present study evaluated the effectiveness of a multifaceted HWR/BT-based treatment program in reducing symptoms of PTSD and other mental health conditions, decreasing negative posttraumatic cognitions, and increasing self-efficacy regarding substance use. The hypotheses were largely unsupported, in that of the seven dependent variables, only posttraumatic cognitions showed significant between-group decreases. PTSD symptoms and substance-related self-efficacy demonstrated pre–post, but not between-group differences. The present findings are consistent with many previous studies of integrated trauma-focused treatments for PTSD-SUD that have failed to find a unique advantage over comparison groups (e.g., Hien et al., 2009; Messina et al., 2012). However, other studies have found success for integrated treatments in reducing PTSD and SUD symptomology over and beyond treatment as usual (e.g., Hien, Cohen, Miele, Litt, & Capstick, 2004). With such inconsistency remaining in the literature, additional exploration is needed.

The fact that the treatment condition demonstrated statistically significant reduction with a large effect size in negative cognitions following treatment, above and beyond the non-treatment comparison condition, is curious given that between-group differences were not observed for PTSD symptom severity, especially considering the strong relationship between these variables (Foa et al., 1999). Because there were multiple elements included in the treatment as delivered in the present study, it is difficult to assess which components were associated with the observed change in cognitions. Future research should examine potential mechanisms through which the HWR/BT-based treatment program instigates this change.

With regard to substance use, the most common drug of choice was methamphetamines, consistent with other studies of HWR/BT (e.g., Covington et al., 2008; Messina et al., 2010). Statistically significant pre–post increases in drug refusal self-efficacy with medium effect size were observed in the present study for both conditions. Messina and colleagues (2008, 2010) found that self-efficacy significantly improved in the control condition, but not in the condition that received HWR/BT, whereas other research (Messina et al., 2012) has indicated support for reductions in general self-efficacy for both treatment and comparison conditions. The current study supports Messina and colleagues’ (2012) research for improved self-efficacy following HWR/BT, with the present study adding a substance-specific component. Self-efficacy for ability to refuse substances has been related to post-treatment abstinence for up to 1 year following treatment (Ilgen, McKellar, & Tiet, 2005). As such, substance-related self-efficacy may be an important variable to consider in future studies of integrated treatments.

In the present study, no pre–post or between-group differences were found for depression, anxious arousal, dissociation, or tension reduction. These results were surprising and concerning, given that other studies have found pre–post reductions in depression following integrated treatment (Najavits & Hien, 2013). Specific to HWR/BT, previous studies have also found significant pre–post reductions in depression, anxiety, and dissociation (Covington et al., 2008), as well as improvements in general psychological well-being (Messina et al., 2012). The HWR/BT program as implemented in the present study was heavily focused on reducing trauma- and substance-related symptoms. Thus, depression and other mental health symptoms may not have been specific targets of the treatment, although improvements in these areas would have been expected as other symptoms improved. Alternatively, it is possible that the observed reductions in PTSD symptoms were related more to reductions in general negative emotions or distress than to depression or anxious arousal (e.g., Simms, Watson, & Doebbeling, 2002; Watson, 2009). Such reductions may not be reflected in the total scores of mental health measures. However, future investigations can examine mental health variables at the subscale or even symptom level to assess for between-group differences.

Overall, the lack of present findings presents questions regarding the effectiveness of HWR/BT, especially given the number of additional components that were added in the present study. More may not necessarily be better (Foa et al., 2005), and if stand-alone treatments can accomplish the same results, then they may be a better option for some patients, as they are often more acceptable to the clinicians who implement them (Marks, Lovell, Noshirvani, Livanou, & Thrasher, 1998). Alternatively, if an integrated treatment can be adapted to better address substance abuse in addition to trauma symptoms, this may ultimately be the most efficient solution. Research has already begun to isolate subgroups of women who are most likely to benefit from integrated treatments (Hien, Campbell, Ruglass, Hu, & Killeen, 2010). However, more research is needed in this domain.

Limitations

The present study was a quasi-experimental effectiveness study conducted in a prison setting. Although statistical control was utilized where possible, without random assignment and adequately controlled conditions, accurate statements about cause and effect cannot be made. Several factors limit the generalizability of the present findings. The final sample represented significant loss from baseline recruitment. It is possible that the women who completed the study were characteristically different from those who initially entered the study. However, no significant differences were found between completers and non-completers on any of the demographic or dependent variables at baseline. The final sample also consisted of unequal sample sizes between conditions. Measures relied solely on self-report, and some measures were modified to maximize utility in the present sample. In addition, multiple treatment components were utilized, limiting interpretation of which component was responsible for change. Due to these limitations, the findings should be interpreted only in the context of the treatment package as implemented in this study. Future studies should utilize more thorough assessment of PTSD and substance use symptoms, using clinical interviews and biological measures (e.g., urinalysis) where possible. A more controlled randomized study would also greatly improve the interpretation of findings.

Due to the scarcity of significant findings in this study, future studies of HWR/BT should follow Hien and colleagues’ (2010) research and examine whether specific subsets of women are most likely to benefit from the treatment program. As a whole, the HWR/BT comprehensive program utilized in the present study was not much more effective than standard prison services at reducing psychological symptoms, with the exception of posttraumatic cognitions, calling into question the overall cost-effectiveness of the treatment. Additional research with a more rigorous design may help determine which components of the treatment are most effective, and for which specific subsets of the incarcerated women population. Although many questions remain, the present study has contributed to the growing literature of integrated treatments, and particularly of HWR/BT-based programs. More research is needed regarding its unique advantage over standard prison-based services, and whether the cost-effectiveness of the program is consistent with its potential benefits.

Footnotes

Acknowledgements

The authors wish to thank members of the Trauma Research: Assessment, Prevention, and Treatment (TRAPT) Center, as well as The University of Tulsa Institute of Trauma, Adversity and Injustice (TITAN), for commentary and support throughout the research process. The authors also thank the women, clinician, and staff of the Oklahoma Department of Corrections.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was based on the first author’s doctoral dissertation and was partially supported by the Bellwether Fellowship, awarded by the University of Tulsa Graduate School and a Student Research Grant from the University of Tulsa Office of Research.

Author Biographies

Rachael M. Swopes, PhD, received her doctorate in clinical psychology from the University of Tulsa, Oklahoma, completed an internship at Duke University Medical Center, and completed her postdoctoral residency at Behavioral Healthcare of Rural Missouri in Stockton, Missouri. As a graduate student, she conducted research in the Trauma Research: Assessment, Prevention, and Treatment (TRAPT) Center and The University of Tulsa Institute of Trauma, Adversity and Injustice (TITAN). She is now a licensed psychologist in full-time clinical practice at Citizens Memorial Hospital in Bolivar, Missouri. Her clinical specializations include treatment of trauma symptoms and behavior disorders in children and families. She continues to stay involved in research on identification of evidence-based treatments for trauma. She is also interested in the dissemination of evidence-based practice and integration of mental health services in rural communities.

Joanne L. Davis, PhD, received her doctorate in clinical psychology from the University of Arkansas, completed an internship at the Medical University of South Carolina, and completed a postdoctoral fellowship at the National Crime Victims Research and Treatment Center in Charleston, South Carolina. She is currently an associate professor of clinical psychology at the University of Tulsa, Oklahoma, where she oversees the Trauma Research: Assessment, Prevention, and Treatment Center, and serves as the co-director of TITAN. As a scientist-practitioner, she focuses on identifying cognitive, emotional, and behavioral aspects of traumatic experience and applying evidence-based treatments to difficulties faced by victims of trauma. She recently wrote a book on a treatment she developed for trauma-related nightmares, Treating Post Trauma Nightmares: A Cognitive Behavioral Approach. She continues her research in the area of nightmares and sleep problems in trauma-exposed persons and is currently evaluating a video intervention for sexual assault survivors.

James A. Scholl is a clinical psychology doctoral student at the University of Tulsa. He has served as an undergraduate assistant and graduate student member of the Trauma Research: Assessment, Prevention, and Treatment Center, as well as TITAN. He has spent time studying sexual assault, trauma-related nightmares, and trauma and substance use in incarcerated women. Other interests include studying the sequela of trauma and infectious disease in minority populations and the relationship between culture and cognitive and emotional responses to trauma.

References

Acierno

Gray

Best

Resnick

Kilpatrick

Saunders

Brady

(2001). Rape and physical violence: Comparison of assault characteristics in older and younger adults in the national women’s study. Journal of Traumatic Stress, 14, 685-695. doi:10.1023/A:1013033920267

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. doi:10.1176/appi.books.9780890423349

Andrews

D. A.

Bonta

J. L.

(1995). LSI-R: The Level of Service Inventory-Revised. Toronto, Ontario, Canada: Multi-Health Systems.

Avant

E. M.

Swopes

R. M.

Davis

J. L.

Elhai

J. D.

(2011). Psychological abuse and posttraumatic stress symptoms in college students. Journal of Interpersonal Violence, 26, 3080-3097. doi:10.1177/0886260510390954

Barth

Bermetz

Heim

Trelle

Tonia

(2013). The current prevalence of child sexual abuse worldwide: A systematic review and meta-analysis. International Journal of Public Health, 58, 469-483.

Black

(2000). A hole in the sidewalk: The recovering person’s guide to relapse prevention. Bainbridge Island, WA: MAC Publishing.

Briere

(1995). Trauma Symptom Inventory professional manual. Odessa, FL: Psychological Assessment Resources. doi:10.1001/archpsyc.1997.01830130087016

Briere

Elliott

D. M.

Harris

Cotman

(1995). Trauma Symptom Inventory: Psychometrics and association with childhood and adult victimization in clinical samples. Journal of Interpersonal Violence, 10, 387-401. doi:10.1177/088626095010004001

Browne

Miller

Maguin

(1999). Prevalence and severity of lifetime physical and sexual victimization among incarcerated women. International Journal of Law and Psychiatry, 22, 301-322. doi:10.1016/s0160-2527(99)00011-4

10.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

11.

Covington

S. S.

(2000). A woman’s way through the twelve steps workbook. Center City, MN: Hazelden.

12.

Covington

S. S.

(2003). Beyond trauma: A healing journey for women. Center City, MN: Hazelden.

13.

Covington

S. S.

(2007). The relational theory of women’s psychological development: Implications for the criminal justice system. In Zaplin

(Ed.), Female offenders: Critical perspectives and effective interventions (2nd ed., pp. 135-164). Sudbury, MA: Jones & Bartlett.

14.

Covington

S. S.

(2008). Helping women recover: A program for treating addiction (with a special edition for the criminal justice system). San Francisco, CA: Jossey-Bass. (Original work published 1999). doi:10.1080/02791072.2008.10400665

15.

Covington

S. S.

Bloom

B. E.

(2006). Gender responsive treatment and services in correctional settings. Women & Therapy, 29, 9-33. doi:101300/J015v29n0302

16.

Covington

S. S.

Burke

Keaton

Norcott

(2008). Evaluation of a trauma-informed and gender-responsive intervention for women in drug treatment. Journal of Psychoactive Drugs, 387-398. doi:10.1080/02791072.2008.10400666

17.

Foa

E. B.

Ehlers

Clark

D. M.

Tolin

D. F.

Orsillo

S. M.

(1999). The posttraumatic cognitions inventory (PTCI): Development and validation. Psychological Assessment, 11, 303-314.

18.

Foa

E. B.

Hembree

E. A.

Cahill

S. P.

Rauch

S. A. M.

Riggs

D. S.

Feeny

N. C.

Yadin

(2005). Randomized trial of prolonged exposure for posttraumatic stress disorder with and without cognitive restructuring: Outcome at academic and community clinics. Journal of Consulting and Clinical Psychology, 73, 953-964.

19.

Gorski

T. T.

Grinstead

S. F.

(2010). Relapse prevention workbook: Identifying early warning signs related to personality and lifestyle problems (Rev. ed.). Independence, MO: Herald Publishing House.

20.

Gray

M. J.

Elhai

J. D.

Owen

J. R.

Monroe

(2009). Psychometric properties of the trauma assessment for adults. Depression and Anxiety, 26, 190-195. doi:10.1002/da.20535

21.

Harrington

Newman

(2007). The psychometric utility of two self-report measures of PTSD among women substance users. Addictive Behaviors, 32, 2788-2798. doi:10.1016/j.addbeh.2007.04.016

22.

Hien

D. A.

Campbell

A. N. C.

Ruglass

L. M.

M. C.

Killeen

(2010). The role of alcohol misuse in PTSD outcomes for women in community treatment: A secondary analysis of NIDA’s women and trauma study. Drug and Alcohol Dependence, 111, 114-119. doi:10.1016/j.drugalcdep.2010.04.011

23.

Hien

D. A.

Cohen

L. R.

Miele

G. M.

Litt

L. C.

Capstick

(2004). Promising treatments for women with comorbid PTSD and substance use disorders. American Journal of Psychiatry, 161, 1426-1432. doi:10.1176/appi.ajp.161.8.1426

24.

Hien

D. A.

Wells

E. A.

Jiang

Suarez-Morales

Aimee

N. C. C.

Cohen

L. R.

. . . Nunes

E. V.

(2009). Multisite randomized trial of behavioral interventions for women with co-occurring PTSD and substance use disorders. Journal of Consulting and Clinical Psychology, 77, 607-619. doi:10.1037/a0016227

25.

Ilgen

McKellar

Tiet

(2005). Abstinence self-efficacy and abstinence 1 year after substance use disorder treatment. Journal of Consulting and Clinical Psychology, 73, 1175-1180. doi:10.1037/0022-006X.73.6.1175

26.

Kelly

C. E.

Welsh

W. N.

(2008). The predictive validity of the Level of Service Inventory-Revised for drug-involved offenders. Criminal Justice and Behavior, 35, 819-831. doi:10.1177/0093854808316642

27.

Kubany

E. S.

McCaig

M. A.

Laconsay

J. R.

(2003). Healing the trauma of domestic violence: A workbook for women. Oakland, CA: New Harbinger.

28.

Kubiak

S. P.

(2004). The effects of PTSD on treatment adherence, drug relapse, and criminal recidivism in a sample of incarcerated men and women. Research on Social Work Practice, 14, 424-433. doi:10.1177/1049731504265837

29.

Lowenkamp

C. T.

Latessa

E. J.

(2005). Increasing the effectiveness of correctional programming through the risk principle: Identifying offenders for residential placement. Criminology & Public Policy, 4, 263-289.

30.

Marks

Lovell

Noshirvani

Livanou

Thrasher

(1998). Treatment of posttraumatic stress disorder by exposure and/or cognitive restructuring: A controlled study. Archives of General Psychiatry, 55, 317-325.

31.

Messina

Calhoun

(2014). Trauma-informed treatment decreases posttraumatic stress disorder among women offenders. Journal of Trauma & Dissociation, 15, 6-23. doi:10.1080/15299732.2013.818609

32.

Messina

Calhoun

Warda

(2012). Gender-responsive drug court treatment: A randomized controlled trial. Criminal Justice and Behavior, 39, 1539-1558.

33.

Messina

Grella

C. E.

Cartier

Torres

(2008). Enhancing prison treatment for women offenders: An in-depth follow-up study. Los Angeles: University of California, Los Angeles.

34.

Messina

Grella

C. E.

Cartier

Torres

(2010). A randomized experimental study of gender-responsive substance abuse treatment for women in prison. Journal of Substance Abuse Treatment, 38, 97-107. doi:10.1016/j.jsat.2009.09.004

35.

Najavits

L. M.

(2005). Theoretical perspective on posttraumatic stress disorder and substance use disorder. Australian Psychologist, 40, 118-126. doi:10.1080/00050060500094621

36.

Najavits

L. M.

Harned

M. S.

Gallop

R. J.

Butler

S. F.

Barber

J. P.

Thase

M. E.

Crits-Christoph

(2007). Six-month treatment outcomes of cocaine-dependent patients with and without PTSD in a multisite national trial. Journal of Studies on Alcohol and Drugs, 68, 353-361.

37.

Najavits

L. M.

Hien

(2013). Helping vulnerable populations: A comprehensive review of the treatment outcome literature on substance use disorder and PTSD. Journal of Clinical Psychology: In Session, 69, 433-479. doi:10.1002/jclp.21980

38.

Najavits

L. M.

Walsh

(2012). Dissociation, PTSD, and substance abuse: An empirical study. Journal of Trauma & Dissociation, 13, 115-126. doi:10.1080/15299732.2011.608781

39.

Pietrzak

R. H.

Goldstein

R. B.

Southwick

S. M.

Grant

B. F.

(2011). Prevalence and Axis I comorbidity of full and partial posttraumatic stress disorder in the United States: Results from Wave 2 of the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Anxiety Disorders, 25, 456-465. doi:10.1016/j.janxdis.2010.11.010

40.

Radloff

L. S.

(1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401. doi:10.1177/014662167700100306

41.

Resnick

H. S.

Best

C. L.

Kilpatrick

D. G.

Freddy

J. R.

Falsetti

S. A.

(1993). Trauma assessment for adults (TAA). Charleston: National Crime Victim Center for Research and Treatment, Medical University of South Carolina.

42.

Ruggiero

K. J.

Del Ben

Scotti

J. R.

Rabalais

A. E.

(2003). Psychometric properties of the PTSD Checklist–Civilian Version. Journal of Traumatic Stress, 16(5), 495-502.

43.

Simms

L. J.

Watson

Doebbeling

B. N.

(2002). Confirmatory factor analysis of posttraumatic stress symptoms in deployed and nondeployed veterans of gulf war. Journal of Abnormal Psychology, 111, 637-647.

44.

Simourd

D. J.

(2004). Use of dynamic risk/need assessment instruments among long-term incarcerated offenders. Criminal Justice and Behavior, 31, 306-323. doi:10.1177/0093854803262507

45.

Sklar

S. M.

Annis

H. M.

Turner

N. E.

(1997). Development and validation of the drug-taking confidence questionnaire: A measure of coping self-efficacy. Addictive Behaviors, 22, 655-670. doi:10.1016/s0306-4603(97)00006-3

46.

Tests of Adult Basic Education: Forms 9 & 10 technical report. (2004). Monterey, CA: CTB/McGraw-Hill.

47.

Watson

(2009). Differentiating the mood and anxiety disorders: A quadripartite model. Annual Reviews of Clinical Psychology, 5, 221-247.

48.

Weathers

F. W.

Litz

B. T.

Huska

J. A.

Keane

T. M.

(1994). PTSD Checklist–Civilian version. Boston, MA: National Center for PTSD, Behavioral Sciences Division.

49.

Wolff

Frueh

Shi

Gerardi

Fabrikant

Schumann

(2011). Trauma exposure and mental health characteristics of incarcerated females self-referred to specialty PTSD treatment. Psychiatric Services, 62, 954-958. doi:10.1176/appi.ps.62.8.954