The Evolution of Gender-and Trauma-Responsive Criminal-Justice Interventions for Women

For over five decades, the development of risk classification assessments, corrections-based treatment, and the associated outcome research have been focused on men. Thus, it is no surprise that existing treatment frameworks and correctional policies have been established from a male perspective. Women have also been incarcerated for over five decades, without suitable recognition of the body of literature to guide policy and procedures specifically for their needs. Compared with their male counterparts, justice-involved women have different pathways into, and out of, crime and substance use; they respond to supervision and custody differently, they have a higher prevalence of co-occurring mental health issues and lifelong trauma and abuse, and higher rates of other complex interpersonal and financial disadvantages [1-9].

hierarchies [9,12,13,33]. Such hierarchies have created differences across gender and gender roles (e.g., patriarchy and sexism) that speak to the lived realities of women [34]. These complex disadvantages, intersectional inequalities, and differences in social capital continue for women during incarceration [7].
Based on the numerous research results showing that women's early childhood adversity is correlated with subsequent harmful behaviors, studies also began to explore distinctive factors associated with treatment and criminal justice outcomes for women relative to men. To begin to untangle treatment outcome data, Pelissier et al., [50] assessed commonly analyzed predictors of post-release recidivism among 1,842 men and 473 women who participated in gender-neutral treatment. Among the 32 variables included in the model, only one variable was significantly unique to women (i.e., a history of mental health treatment increased the likelihood of recidivism for women). Thirteen variables were uniquely associated with recidivism for men, but only four were significant for both men and women and in opposite predictive directions. Variables that increased recidivism for men but decreased recidivism for women included disciplinary infractions during incarceration, counseling during supervision, number of monthly collateral contacts, and previous criminality. Interestingly, prior criminality has been a consistent predictor of return to criminal behavior in samples of men and is a risk factor often generalized to women.
Another study compared recidivism risk factors among a large sample of gender-neutral treatment participants (4,386 incarcerated women and 4,164 incarcerated men) and also found that there was a notable lack of predictive factors for women [45]. Of the 11 variables in the models, the strongest predictor of return to prison for both men and women was co-occurring disorders. The single unique predictor for women was previous education, with higher education reducing the likelihood of return to prison. In contrast, previous employment significantly decreased return to prison for men, but not for women. Notably, a much smaller proportion of women reported any employment in the year prior to incarceration compared to men.
Hamilton et al., [51] included women-centered variables in their analytical model and found that the predictive factors of recidivism for 8,815 women were primarily related to social support (e.g., minor children, no child support, legal contact restrictions) and victim/ offender characteristics prevalent among women (e.g., IPV and prostitution). Brennan et al.,[52] identified eight reliable yet complex pathways to women's recidivism, linking multiple women-centered factors to previous literature, including sexual/physical abuse, lower social capital, poor relational functioning, and extreme mental health issues. Other studies contend that risk factors that are more prevalent among women are trauma-related factors associated with co-occurring disorders, IPV, involvement with child protective services, homelessness, and dependency on others for financial support [2,9,31,[53][54][55][56][57][58][59].
Thus, the literature reveals pattens that indicate that justice-involved women may be at a differential risk for recidivism than their male counterparts given their life realities. At the very least, treatment outcome and recidivism data should be analyzed separately for men and women with examination of women-centered variables included in the analyses.

Risk and need assessments
It follows that the predictive validity of gender-neutral risk assessments are also not as robust for women as for men [32,52,53,60,61]. 2 There is evidence showing the increased predictive validity for women when assessments are inclusive of women-centered factors. Van Voorhis et al., [31,32] created the Women's Risk Needs Assessment (WRNA) as a stand-alone needs assessment or as a supplement for gender-neutral tools, such as the Level of Service Inventory-Revised [62] and the Northpointe COMPAS [63,64]. The WRNA and the WRNA Trailor (WRNA-T) account for factors that are empirically more persistent in the lives of justice-involved women and included measures of trauma and abuse, unhealthy relationships, depression, parental stress, safety, financial considerations, anger, housing safety, family support and personal strengths such as self-efficacy [32, [65][66][67].
In their application of WRNA, Salisbury et al., [59] assessed whether the inclusion of measures of women's needs (as risk factors) contributed to poor prison adjustment and recidivism among 156 women admitted to the department of corrections in a western state. Although different patterns were found across prisons, child abuse and relationships were associated with prison adjustment and victimization, while limited self-efficacy and parental stress were identified as risk factors for women upon release. Patterns were replicated across eight separate prison samples, seven pre-release samples, and six probation samples and resulted in recommendations for women-centered needs assessments for each type of setting [32,67]. Women's gender-related needs are the pivotal factors to address in guiding assessment, treatment development, and gender-responsive policies to aid in women's recovery.

Gender-and trauma-responsive treatment outcomes among justice-involved women
In 2003, the National Institute of Corrections published a groundbreaking report, Gender-Responsive Strategies: Research, Practice and Guiding Principles for Women Offenders [54]. This report documented the need for a new vision that recognized the need to focus and integrate trauma services into the justice system. Since this time, supporters have been proposing to move corrections forward by adopting the Guiding Principles and other published "Blue-Prints" outlining gender-and trauma-responsive policies and practices.
There is now a growing evidence base and multiple Randomized Controlled Trials (RCT) documenting the effectiveness of genderand trauma-responsive interventions for justice-involved women, at With funding from NIDA, Messina et al., [77] conducted an experimental study comparing post-release outcomes of 115 prison-based treatment participants. Women were randomized to a 20-session gender-and trauma-responsive treatment program (i.e., Helping Women Recover, Covington [27], and 16-session Beyond Trauma, Covington, 2013) or a prison-based therapeutic community model. Helping Women Recover and Beyond Trauma are manualized curricula with a facilitator guide and participant workbook. The gender-responsive treatment group had significantly greater reductions in post-release substance use, remained in voluntary residential aftercare longer (2.6 months vs. 1.8 months, p < .05), and were less likely to have been re-incarcerated within 12 months after parole (31% vs. 45%, p < .05; a 67% reduction in recidivism). While both groups improved on mental health outcomes, the findings show the beneficial effects of treatment components responsive to women's needs.
The second experimental study, also funded by NIDA compared women in mixed-gender drug court programs with those in genderand trauma-trauma responsive drug court programs [76]. The gender-and trauma-responsive intervention groups across four outpatient drug courts showed the experimental intervention group had less disciplinary sanctions during the second and most intensive phase of drug court treatment (Gender-responsive group = 0.65 sanctions; Mixed-gender group = 1.2; p < .03) and were had less sanctions resulting in remand to jail, compared with the mixed-gender control group (Gender-responsive group = 1.9 jail remands; Mixed-gender group = 2.4 jail remands; p < .05).
A series of recent research studies (data collected from 2014-2019) conducted with 1,118 women convicted of serious or violent offenses who participated in brief or intensive interventions designed for women also showed consistent and positive results. The first study included a sample of 39 women in a Security Housing Unit (SHU: used to house residents at the highest risk of committing violent offenses against staff, other residents and the public). The pilot study assessed the efficacy of a six-session manualized intervention designed for women who have experienced trauma associated with ACEs (i.e., Healing Trauma: A Brief Intervention for Women, Covington & Russo, [28]. Results demonstrated preliminary support for the effectiveness and feasibility of the brief intervention for women in the highest risk classification. The SHU women exhibited significant improvement across measures of depression, anxiety, Post-Traumatic Stress Disorder (PTSD), aggression, anger and social connectedness from the intervention [78]. Effect sizes were moderate to large, with the largest impact on physical aggression (Cohen's d . 82).
The Healing Trauma SHU pilot study was replicated with 682 high-need incarcerated women (i.e., those with co-occurring disorders, frequent disciplinary infractions, or conflict with staff/others). Using a peer-facilitated model, the women exhibited improvement on over 90% of the outcomes measured [79]. Significant reductions were found for anxiety, depression, PTSD, psychological distress, aggression, and anger. Significant increases were found in empathy, social connectedness, and emotional regulation. Effect sizes were small to moderate, with the largest impact on depression, PTSD and angry feelings (Cohen's d ranged from 0.51, 0.41, 0.42 respectively). Anger expression measures approached significance (p = 0.061; p = 0.051).
Moreover, Messina and Schepps [36] found that a greater number of ACEs increased the likelihood of program gains for all mental health and aggression outcomes.
The findings of the pilot studies showed that the Healing Trauma six-session brief intervention had a positive impact on trauma-related outcomes for high-risk/high-need women, and those with the highest incidence of childhood trauma and abuse derived the most benefit. However, these pilot studies were limited to measures of pre-and post-change, without the benefit of a comparison group. Building upon the pilot studies with funding from the National Institute of Justice, Messina and Calhoun [40] conducted an experimental study assessing an intensive 20-session manualized violence intervention (i.e., Beyond Violence, Covington, 2014) among 123 women primarily incarcerated for violent crimes (e.g., murder, attempted murder, manslaughter, assault). Results from the participants randomized to the Beyond Violence (BV) program had significantly lower mean scores than the control participants on depression (F=4.97), anxiety (F=9.12) and PTSD (F=4.68). Findings also showed that the BV participants had significantly lower mean scores than the control participants on physical aggression (F=6.11), hostility F=4.23), indirect aggression (F= 9.42), and expressive anger (i.e., anger used to manipulate or threaten) (F=7.15). Due to nature of the crimes and the lengthy sentences, post-release outcomes could not be explored.
A previous experimental study comparing BV with a 44-session Assaultive Offender Program in a women's prison in Michigan, Kubiak et al., [74] found similar positive changes in anger and aggression for the BV participants. While both groups experienced improvement in anger and mental health, women randomized to the BV intervention had stronger declines in anxiety (F=5.32) and state anger (i.e., outward expression or control of others) (F=8.84) than women in gender-neutral anger program. Furthermore, a longitudinal follow-up study showed that the women who participated in the BV program were significantly less likely to recidivate (i.e., arrest or time in jail) than women in the gender-neutral anger program during the first 12 months following their release from prison [74].
In summary, women with complex problems, histories of ACEs, and serving sentences for property, drug, or violent offenses benefited from various gender-and trauma-responsive interventions when compared to treatment as usual. These curricula evaluated were designed specifically for the primary needs of justice-involved women, addressing the gaps in programs focused on trauma, substance use and violence prevention. The content of the interventions, the method of delivery, and the applicability to the needs of the population are the essential components for enhancing women's recovery.

Conclusion and Recommendations
Acknowledging the existing literature on the needs and recovery processes of justice-involved women is vital to the implementation of appropriate assessments, treatment services, supervision, policy recommendations and continued research for further advancement. One must only recognize the plethora of available research, RCTs, and meta-analyses [22,24,40,76,77,85]. Although, movement is gradual, California has been responsive to this process of change as their female population grew. Beginning in 2020, ten years after the published findings from the RCT on Helping Women Recover [77] Overtime the conclusions regarding corrections-based treatment has shifted from "what works" later interpreted as "nothing works" to "some things work, for some people, some of the time" [87][88][89][90]. Covington and Bloom [91] suggested an important shift of the field's central question of "what works" to "what is the work?" The authors state that the work requires a theoretically based model recognizing the psychological development of women and a treatment model that supports gender-responsive programs and policy development. A gender-responsive and trauma-informed approach considers the social issues of gender inequalities and individual factors that impact justice-involved women. 3 An interpersonal approach to programming would address substance use, trauma, economic marginality, relationships, and mental health issues through comprehensive, integrated, and culturally relevant services and supervision. Service providers need to be cross-trained in areas of gender-responsiveness and trauma-informed principles, and resources must be allocated for women's programs and continuous rigorous evaluation.
Although men continue to be the majority of the imprisoned population in the United States, there are still over 230,000 women in prisons and jails across the country [17]. Funding restraints often require service provisions be focused on the larger population of men and those at the highest risk of recidivism. Prison administrators and government officials may feel that rehabilitation programs are not a proper investment for women who often have short-term sentences. Yet, brief gender-and trauma-responsive interventions have been shown to be feasible and could be effective re-entry services. Ignoring the critical needs of women has long-term consequences and high costs to society given the involvement of social services and the intergenerational cycle of trauma, substance use, and criminal involvement.
In addition, focusing on recidivism as the sole determinant of a predictive model of rehabilitation is antiquated and based on research on men and goals for public safety. Measures of recovery should go beyond criminal activity or abstinence to include reductions in IPV, increased psychological wellbeing, education/employment, financial independence, housing, family reunification, etc. Assessing multiple outcome measures, during confinement and post-release, are necessary to fully determine program effectiveness. Recidivism does not capture the full picture of post-release challenges or successes. Ward and Stewart [92] question whether rehabilitation ends with risk management (i.e., reduced crime for public safety) or if it should incorporate services toward personal enhancement (i.e., improved quality of life/well-being). Rehabilitation and sustainable recovery after release go far beyond involvement with the criminal justice system. It is time other measures of change are expected and required in peer-reviewed journals seeking to increase the knowledgebase on what works for women and men.
Women's gender-related needs are the pivotal factors to address in guiding assessment, treatment development, and gender-responsive policies to aid in women's recovery. The recommendation of the 3 Becoming gender-and trauma-responsive are terms which are inclusive of men, women, and gender-diverse populations. Gender identity and histories of trauma are important factors that should be included in treatment opportunities for all justice-involved populations. Men can also benefit from trauma-specific programming, as histories of trauma are not unique to women. The prevalence, type, and impact of lifelong trauma may vary by gender, but that is not an argument against incorporating treatment components that address trauma for both men and women [36].
Gender-Responsive Theoretical Framework and Guiding Principles for Corrections as a paradigm of care for justice-involved women was essential in 2003 and remains so as we begin 2022.

Conflict of Interest Statement
Dr. Messina is the president of Envisioning Justice Solutions, Inc. and a research criminologist at the UCLA Integrated Substance Abuse Programs. The author declares that there is no conflict of interest and there was no funding provided for this commentary from any person or organization.

Acknowledgment
I am grateful to those who contributed decades of scholarly works, dedication to theoretical and program development, research, dissemination of information, committee involvement, and the creation of policy guidelines for criminal-justice-involved women. These criminologists, psychologists, sociologists, LCSWs, nurses, substance use treatment professionals, corrections professionals and relentless thinkers are cited throughout this article. I am especially grateful for the guidance of my mentors, friends, and colleagues Dr. Barbara Bloom, Dr. Stephanie Covington, Velda Dobson-Davis, Dr. Christine Grella, Rochelle Leonard, Rita Marmolejo and Dr. Barbara Owen. I am also forever grateful to the peer facilitators who provided years of program facilitation in prison, as well as care and attention to the program participants. Finally, I wish to acknowledge the bravery of the women with lived experience who volunteered to participate in the programs and the research and for sharing their stories over the years.