Incarcerated Women's Relationship-based Strategies to Avoid Drug Use After Community Re-Entry

Abstract

While recent research has stressed the supportive role that family and friends play for incarcerated persons as they reenter the community, drug-using incarcerated women reentering the community often have to rely on family, community, and intimate relationships that have played a role in their substance abuse and criminalization. In this study the authors conducted qualitative analysis of clinical sessions with rural, drug-using women (N = 20) in a larger prison-based HIV risk reduction intervention in Kentucky during 2012–2014 to examine incarcerated women's perceptions of the role of their family, community, and intimate relationships in their plans to decrease their substance abuse upon community re-entry. Women stressed the obstacles to receiving support in many of their family and drug-using relationships after community re-entry. Nonetheless, they asserted that changes in their relationships could support their desires to end their substance abuse by setting limits on and using their positive relationships, particularly with their children, to motivate them to change. Interventions to promote incarcerated women's health behavior changes—including substance abuse—must acknowledge the complex social environments in which they live.

Introduction

In the last two decades, women have been incarcerated at a higher rate than men (Heimer and Kruttschnitt 2006). Many of these women have been incarcerated for drug offenses, particularly as sentencing has changed (Mumola and Karberg 2006). Though initially public health advocates drew attention to the adverse health effects of incarceration itself (Massoglia 2008), more recent research has stressed the health risks that they face beyond prison walls—in their social relationships (Cook et al. 2005; DeHart 2008; Fogel et al. 2014; Fuentes 2014).

Cross culturally women seek out relationships—positive and negative—because relationships are important for their self-identity (Jack and Ali 2010). Because women understand and communicate their life experiences through their relationships with others, it is important to examine how women maintain healthy relationships and grow through their relationships (Tuchman 2010). Yet, despite the fact that relationships may be crucial for how women understand themselves, incarcerated women often suffer from inter-related, comorbid health problems that emerge from their relationships—mental health challenges, substance abuse, and risky health behaviors which put them at high risk for transmission of STIs (Fogel et al. 2014; Kuo et al. 2013). These inter-related health problems frequently stem from women's extensive histories of trauma and physical and sexual victimization (Fuentes 2014; Messina and Grella 2006). Many experience multiple traumas over their lifetimes, including abuse and neglect during childhood and sexual violence and interpersonal violence as adults (DeHart 2008).

Nonetheless, many substance-using women remain in “risky relationships”: those that endanger their health (Knudsen et al. 2008; Staton-Tindall et al. 2011) and lead to their criminalization. Women's desire for connection may lead them to stay in abusive relationships (Fogel et al. 2014; Fuentes 2014) and to normalize their feelings of self-worth within these relationships (Staton-Tindall et al. 2007). To cope with the mental health outcomes of their traumatic, stressful relationships, many women use drugs and alcohol (DeHart 2008; Fuentes 2014). Women may also exchange sex for drugs, making it difficult to demand sexual protection (DeHart 2008; Fogel et al. 2014) or remain cognizant of the need for protection.

Increasingly, interventions to tackle the complex health problems of incarcerated women have targeted their risky relationships. Correctional settings can offer a venue where incarcerated women can contemplate behavior change while they are isolated from the relationships that might put them at risk (Abad et al. 2013; Leukefeld et al. 2012). As challenging as the conditions of incarceration are, many women report that, in such settings, they find themselves insulated from the grind of their stressful relationships (Moe and Ferraro 2006). While programming may vary significantly across jail and prison institutions, some examples of successful programs have included content on empowerment, relationship power, and interpersonal violence to help women recognize that their relationship dynamics may contribute to their health risks (Fasula et al. 2013).

Research on incarcerated men's health has stressed the potentially supportive role that community relationships can play in ending drug use and reducing the risk of recidivism (Grieb et al. 2014; Naser and Visher 2006). In contrast, as incarcerated women re-enter their communities, they are likely to return to relationships that threaten whatever hopes they may have for ending their drug use. Often, women's economic reliance on these relationships only increases as they return to the community without housing arrangements or employment (Nelson, Deess, and Allen 2011). Women often are poorly prepared to be re-immersed in the range of risky relationships that they face in the community when they are released (Fogel et al. 2014) and resume risky behaviors because of their return to the same communities of drug use (Abad et al. 2013). In some cases, women rely on familial and community networks that are challenged to provide the type of support they desire, and may include the same patterns of violence and addiction that are at the root of their drug use (DeHart 2008).

Therefore, incarcerated women re-entering the community must navigate both the relationships that put them at risk for poor health and the relationships that they will rely on for support—complicated by the fact that the same relationships may be relied upon to supply both. Overwhelmingly, research has focused on the specific health risks that women face due to their risky relationships (Fasula et al. 2013; Fogel et al. 2014) and the challenges they face as mothers (Berry and Smith-Mahdi 2006; Moe and Ferraro 2006). Scholars have focused less on the perceptions of incarcerated women, for whom these relationships are meaningful for their identity. While we are aware of the power inequities and potential for victimization within these relationships, little understanding exists of the methods by which women assert themselves within their broad range of relationships—including the choices, limits, and desires they delineate. In this study the authors examine the risks women perceived in their relationships and the changes they proposed making in them to reduce drug use and related risk behaviors following incarceration.

Methods

Research context

This research took place in rural Appalachian Kentucky, a region with some of the nation's highest rates of prescription drug abuse in the country (Zhang et al. 2008). Within Appalachia, Appalachian Kentucky has the highest percentage of residents in poverty, with a rate of 25.1% in 2008–2012 in comparison to 15.9% nationally (Pollard and Jacobsen 2014). Community members have argued that the poor economy and widespread unemployment contribute to high rates of substance abuse (Zhang et al. 2008).

Study background, sampling procedures

As part of a larger study (blinded for review), rural drug-using women were recruited from three rural jail facilities located Appalachian Kentucky counties. The jails are similar in terms of size, female populations, and availability of programming and resources. Eligibility criteria for the larger study included: (1) National Institute on Drug Abuse (NIDA) -modified ASSIST score of 4+ for any drug indicative of moderate risk substance abuse (NIDA, 2009); (2) engagement in at least one sex risk behavior in the past 3 months; and (3) incarceration period of at least 2 weeks but no longer than 3 months. All participants were provided with informed consent and signed consent forms prior to study screening. Following screening and baseline interviews, participants were randomized into one of two intervention conditions: (1) HIV/HCV Education, which included the NIDA standard pre- and post-test counseling, HIV and HCV rapid testing, and an information packet on existing community drug abuse and HIV/HCV resources; or (2) Motivational Interviewing-based HIV Risk Reduction (MI-HIV) which, in addition to what was received in the HIV-Ed group, also included an evidence-based brief intervention for high-risk women focused on an individualized plan for enhancing motivation to reduce risk behaviors and to use community-based health services. Of the total sample, 100 participants were randomized to the intervention. With permission from participants, MI-HIV sessions were audio recorded, and the transcripts from those sessions were used for coding in this study. At the time of submission, 222 participants screened eligible for the study (56% of the 397 screened for eligibility), and 203 completed baseline interviews (91% participation rate, 17 people were released early between screening and the time of the interview). From the total intervention sample when qualitative data analysis began (N = 100), we generated a purposive sample of those participants who completed 4+ sessions and agreed to have all the sessions audiotaped (N = 20) to allow analysis of themes that developed over the course of multiple sessions. The Institutional Review Board (University of Kentucky) approved this study's research protocol.

Qualitative data analysis

In a first round of open coding, the study team read transcripts of six participants, noting a preliminary list of codes that were concept-driven (emerging from the research literature) and data-driven (emerging from the data), an approach often used within code construction (Gibbs 2008). In subsequent readings, the team used constant comparison to compare the existing codes with data in additional transcripts to refine and re-categorize the codes, their definitions, and sub-codes (Saldaña 2012). Through discussion, team members identified whether their disciplinary backgrounds in social work, behavioral science, and anthropology shaped their varied understandings of the codes (Creswell 2007). After reaching consensus, the team defined the codes, using typical and atypical exemplars while identifying exclusions. Next, two team members tested inter-coder reliability, using Cohen's Kappa to calculate the degree of agreement for each code until it reached at least 80% (Bernard 2006). Throughout, coders explained their coding choices and made necessary revisions to the codebook to clarify meaning. Using MAXQDA qualitative data analysis software, the team coded the transcripts of the sample, checking for confirmatory data or conflicting statements. After reading through sixty-seven transcripts from twenty participants, concept saturation was reached in this relatively homogenous population (Bernard 2006). Finally, the lead author looked for patterns and variations within each code that appeared most predominantly within the transcripts, using memos to reflect on the interrelationships between codes (Saldaña 2012).

Results

As shown in Table 1, sample demographics indicated that participants were White (100%) with a mean age of 32.2 years. Most had completed a high school education (mean 11.6 grades completed). The majority were currently unemployed or disabled, with roughly one third reporting their primary income source as family and friends, and 20% reporting illegal activities. Overall this was a low-income sample accruing on average $7,363 in the past 6 months. A majority (55%) were married and parents (90%), characteristics similar to the larger Appalachian Kentucky population (Pollard and Jacobsen 2014). The mean time spent in jail during the previous year was 120 days. In comparison to urban incarcerated populations (Fogel et al. 2014; Messina and Grella 2006), this sample had a similar mean age and education. All women in the sample used at least one drug, and the vast majority were polydrug users (Table 2). Most women had comorbid mental health disorders (Table 3), as measured by Global Appraisal of Individual Needs (GAIN) criteria (Dennis 1998).

Table 1.

Sample demographics (N = 20).

Characteristic	Mean or %
Age	32.2
Race	100% White
Housing status	100% housed in rented or owned place
Marital status	55% married
Highest grade completed	11.6
Employment in past 6 months	65% unemployed or disabled
Have children	90%
Average number of children	2.5
Income sources	35% from family/friends
	30% wages from job
	20% illegal activities
	15% other
Income past 6 months	$7,363
Average number of days incarcerated	120.2

Table 2.

Drug use in sample (N = 20).

Drug used	% Ever used
Marijuana	100
Lortab/Hydrocodone	95
OxyContin	90
Cocaine	85
Methamphetamines	65
Heroin	45
Average number of days using multiple substances in past 6 months	112 days
Average number of days being high on any substance in past 6 months	115 days

Table 3.

Mental health in sample (N = 20).

GAIN criteria*	% Endorsed
Major depression	75
Generalized anxiety	50
Posttraumatic stress	65

Note.

^*GAIN criteria (Dennis 1998) was assessed by research staff, not trained mental health professionals. The percentages here reflect endorsement of symptoms consistent with these disorders, but diagnoses should not be inferred.

Three themes predominated among women's comments about their family, intimate, and community relationships: (1) the need to prepare for unsupportive relationships upon their return; (2) the critical assessment of how relationships were involved in their drug use; and (3) the use of positive relationships to motivate their desire to reduce drug use.

I. Many social demands and little social support

In their clinical sessions, all women were ambivalent about their upcoming return to the community because of their re-immersion in family and community relationships that they relied on for social support, but were laden with risks that could cause them to relapse. Therapy techniques encouraged women to identify relationships that would support their basic survival needs (e.g., housing, transportation, social support) when they reentered the community as well as reinforce their efforts to end or at least limit the health risks involved with their drug use. Most women realized how crucial support would be to their behavior change, but frankly acknowledged that it was in short supply. One woman reflected, “I just have to find the right support. And I really don't have much.” Many women had to choose between staying at family households where drug use was still taking place or hoping for space at a homeless shelter. Others contemplated how they would begin a new life—which would include seeing their children, seeking a job, and engaging in recovery services—when they had no one to count on for transportation in this mountainous, rural area.

Numerous women acknowledged that the web of relationships they depended on constantly shifted. During their time away, the fragile conditions of life outside and the frequency of drug use in the community may have altered their social worlds. After identifying one support person, one woman lamented: “He doesn't use, but I've heard he has been high since I have been in here, so I really don't know.” One woman discovered that, while she was in jail, her boyfriend had been driving her car under the influence, and her sister had sold many of her belongings. “I figured my boyfriend would be the one that would stab me in the back,” she explained, “instead my sister is doing the one thing I figured she'd never do to supply her current boyfriend with drugs.” Having purposefully signed over her assets to her sister to protect them during her incarceration, she now felt cynical about her calculations about which relationships to rely on upon release.

Several women responded to the vacuum of support by imagining their lives without a need for any at all. Some women reflected that relationships they had sought to support their drug use in the past had forced them to compromise themselves through prostitution or criminal activity. One woman resisted the idea of any relationships, stating, “I hate depending on people, I hate staying with people. I hate that.” Rather than seek out supportive relationships, she stressed the need to build her self-sufficiency: “I need to find me a job when I get out of here, have my own money, buy my own f—ing cigarettes.” Other women proclaimed their need for independence because they felt that the people they counted on had failed them. One woman vowed that, if her family continued to restrict control of her children's custody, she had no other option than to leave those relationships behind: “If it comes down to it I will leave. I will pack up and I will leave.”

II. Re-assessing relationships

In addition to their efforts to secure social support for their basic needs, women re-assessed how they would engage in relationships that had been central to their identity. Most women were now critical of the community relationships in which they had used drugs. One woman sighed as she planned how she would explain her new limits: “I want to be clean. I'll just tell them, and if they don't like it, then oh well. They are not really my friends anyway. They are just people I associate with because of drugs.” In revising her feelings about the relationships, she hoped to release herself from them. Yet, others were not as confident in their critiques of friends, admitting the hole left by a lack of support: “You know for one thing you don't got a friend. I ain't got nobody to talk to, ain't got nobody to get this off your chest.” In other cases, women feared the temptations revived by old relationships, rather than discounting the people themselves.

Nearly all women struggled to redefine the relationships with the partners with whom they used drugs. They weighed the support provided by these partners and their skepticism that shared drug use drove the relationship. One woman noted: “I don't really know if I love him or not. I think it was more of attachment, controllment [sic] … He was all I knew. He was there for everything. We got high together.” From these reflections, women assessed these relationships by examining whether they could change to support their desires for quitting. “I know that if I'm with him and he's not going to try to stay clean then I can't,” explained another woman. “And I can't struggle with that anymore, I'm done. I just don't want to struggle with it, as much as I love him, I just can't.” Though they clearly stated their desires for connection, these women asserted that they needed to avoid the risks they could now see in these relationships.

Many women concentrated on the risks that their relationships posed to their mental health, triggering substance abuse. A number confronted difficult family situations in which they endured other family members' drug use histories. One woman described the stress of interacting with her mom, whose drug use and related jail time prevented them from having a relationship. “So if we're arguing about something or, um … if she's throwing up my past to me,” she explained, “you know [I] think well, I just go get high … and forget about it all.” These women acknowledged that unless their family members could engage in relationships, the stress of being involved with them posed more risk than benefit. However, sometimes, the emotional distress of relationships resulted from women's tendency to care for family over themselves. Describing how hard it would be to resist urgent text messages for help that she had encountered in the past, one woman commented, “I feel guilty because at some point I have to quit trying to fix everybody else.” In these cases, women had to reconcile the social expectation that they support others with their choices to protect themselves.

As much as women identified how their relationships put them at risk for relapse, many emphasized that ultimately they needed to account for their agency within relationships. Responding to conversations among inmates, one woman posed that “[some people] want to say, ‘Well, if it wasn't for him I wouldn't of done this’… No, you did it because you wanted to do it.” Going even further, a few women identified the role that they played in increasing others' risk for substance abuse.

III. Relationships that motivate behavior change

While all women stressed that their relationships made them vulnerable to drug use and might inhibit their desire for behavior change, they also underscored how transforming their relationships could be vital to their recovery. Despite the fact that many families in this study had non-traditional kinship patterns due to drug use within the family, most all women prized cultural values about the family that are typical to Appalachia, in which family identity is crucial. Women described their dedication to family, with nearly all highlighting their commitment to motherhood. As one woman commented, being a mother is “the only thing in my life. Only thing.” Some women prioritized the care of their children above all the other tasks that faced them upon release. This could mean letting go of other goals as well as the desire for a better life. One woman explained: “I ain't aiming high, I'm going to shoot low because I need to not get bored and I need to have money … I don't need money for myself; I don't care. It's for my kids.” Another woman stressed that she would forgo all of her other relationships in order to prioritize her child. She argued, “the rest can kiss my a—, they ain't in here with me, they didn't do nothing for me.” As women envisioned what behavior change could bring for them, most pictured themselves not just immersed in family, but caring for them. If she could quit, one woman imagined, in five years, “I take care of my kids; I take care of my family. I always have a home for them and I never let them down.”

Yet most women admitted that they were responsible for being separated from their children. One sighed, “I'm tired of not seeing my kids more, and I'm tired of me being responsible for that.” Re-establishing relationships, they acknowledged, could mean a painful process of building trust again. Feeling that her children often forget her while she is in jail, one woman resolved, “I'm going to try to show them that I'm not going to leave them anymore. I'm not going to be took away from them. I'm going to try really hard to show them that.” Many women looked critically at how their past drug use had prevented them from being the kind of caregiver they wanted to be. A few contrasted their “real” familial self with a foreign, drug-using version of their self: “When I clean up I can get a decent job, I can be a good mommy … When I'm strung out, I go to hell.”

Many women proudly reported that their family members supported their plans for behavior change. Yet some were frustrated with the discordance between their desire to be a family caregiver and the fact that others were caring for them by waiting for them to stop using drugs. Being a good mom, explained one woman, meant, “being able to take care of my kids with a sound mind and not have them ask me if I'm okay today.” A few realized that they might be facing their last chance to prove themselves to family members when they re-entered the community. The prospect of disappointing family members who had been there for them was unbearable: “Mamaw and papaw would be devastated if I went back to doing what I do … did!… I don't want them carrying to their grave knowing that's what I do.” Most women felt strongly that they could recover the identity of being a reliable, caregiving family member despite their histories of drug use.

Yet a few women were hurt when their dual efforts to reclaim familial responsibility and quit substance abuse were rebuffed by family members. Some felt that their families were skeptical of their efforts at behavior change because it would threaten their custody of children they currently cared for. One woman's family treated her as a threat to the family form they had established in her absence, rather than embracing her potential contribution to care for her children again. Before, she had been close to her family, but “I feel like once I got on drugs they abandoned me.” She perceived this withdrawal of support as a failure to fulfill kinship obligations and demonstrated their shame: “I'm an embarrassment; that's the way it feels. To me, if you aren't going to be there when I need you the most [then] don't be there whenever I'm doing good either.” For the several women who experienced this tension with their families, they struggled to imagine how they could re-establish their relational selves while trying to negotiate the hurt of their families' rescinded relationships. By focusing on their family members' rejections, these women emphasized that the obligation to care was mutual, rather than their individual prerogative as a former substance user.

Discussion

In this study we explored incarcerated women's assessments of the risks embedded within the relationships to which they were returning after incarceration and the relational strategies they identified to reduce their substance use. Etic perspectives on incarcerated women's risky relationships—from outside researchers and policy-makers—have stressed that these ties are frequently abusive, unequal, and lead to criminalization (Fogel et al. 2014; Kuo et al. 2013). Women captured within these relationships through economic dependence and poor mental health abuse substances to cope (DeHart 2008; Fuentes 2014). By contrast, incarcerated women's emic perspectives affirmed their capacity to assess and act within their relationships even as they condemned their abuse, bad influence, and neglect. Though women's accounts demonstrated the inherent stress of their family and community relationships, most all affirmed their obligation to care for others and acknowledged their responsibility within their relationships.

Through the results of this study we contribute to the existing knowledge on women, relationships, and risk behavior. First, by expanding the range of relationships that incarcerated women identify as contributing to their drug use. Second, results draw attention to women's efforts at agency, even within unequal or unsupportive relationships. Third, questions were raised about the specific challenges that rural incarcerated women—an understudied population—face upon re-entry. The summary that follows concentrates on this unique data set, however, and may not be generalizable to all populations of incarcerated women who use drugs.

While the literature on incarcerated women's health has focused predominantly on their relationships of shared drug use (DeHart 2008; Fogel et al. 2014; Staton-Tindall et al. 2007) and the traumatic and violent relationships that harm their behavioral health (Green et al. 2005; Knudsen et al. 2008; Messina and Grella 2006), in this study, women identified a broader constellation of relationships that they see affecting their substance abuse. Our results confirmed others' findings that positive relationships, such as those with children, motivated many women's recovery (Moe and Ferraro 2006; Shamai and Kochal 2008). Women in our study envisioned their transition from being a recipient of care by worried family members and children to being a reliable provider of care once they changed their drug use behavior. While the stress of women's relationships has been repeatedly noted by scholars (Abad et al. 2013; Fuentes 2014), in these data, women explicitly identified the stressful relationships that caused poor mental health and related substance abuse. Women's descriptions in our study of the shifting landscape of support revealed the difficult calculations that they made about which relationships to invest in to support their behavior change. Following researchers who have advocated for the role of including family members in interventions to re-integrate inmates (Grieb et al. 2014; Naser and Visher 2006), these findings suggest that interventions to address incarcerated women's complex health problems could benefit from reaching out to families as well as focusing on the intimate relationships that have been successfully targeted to date. To address the challenges of accurately assessing relationships once back in a community setting, community-based therapy sessions could reiterate themes of prison-based programs.

Much research focusing on incarcerated women's relationships has asserted that their relationships have failed them, endangering their health through trauma, stress, and unsafe sex and drug use (DeHart 2008; Fogel et al. 2014; Staton-Tindall et al. 2011) and denying them support (DeHart 2008; Shollenberger 2009). Though interventionists and abstinence programs like Alcoholics Anonymous (AA) and Narcotics Anonymous (NA) might advocate for women to do away with all risky relationships—a suggestion echoed by some participants who expressed the desire to escape—women's economic dependence and family ties generally eliminated the behavior change strategy of “clean breaks.” Findings from this study supported women's accounts of their agency within relationships complicated by violence, questions about support, and doubts about the future. In evaluations of their relationships of drug use, women critically assessed their emotional valences while considering the possibility of ending stressful or dangerous relationships. Women in our study resolved to avoid relationships they deemed usury or unreliable. Yet many narratives revealed women's varied abilities to assess their relationships accurately: they admitted previous naiveté and vowed future vigilance. Frank criticisms about the betrayals within their relationships may have been due to their current sobriety and social distance from jail, begging the question of whether, back in the community, with accompanying economic difficulty, the desire for connection and substance abuse might cloud their critical assessments in the present. Nearly all women acknowledged that their success for achieving their relationship goals was tenuous: living up to their ideals of themselves as a family member and establishing boundaries required consistent control over their addiction and mental health self-care. In some cases, women resisted the claim that their relationships played any role in their drug use, arguing for the need to account for their own actions. This individualistic orientation mirrored the philosophy of abstinence-based support groups like AA and NA (Humphreys and Kaskutas 1995) and contradicted the challenges of the social environment that women described.

Other scholars have stressed that incarcerated women may report isolation upon returning to the community and frustration that family members are unreceptive to their plans for behavior change (Abad et al. 2013), having exhausted social support networks due to their past drug use and criminal activity (Staton-Tindall, Royse, and Leukefeld 2007). Family members may fear the potential for their substance abuse relapse as they return home (Naser and Visher 2006) and remain skeptical about the impact of incarcerated female relatives' return on her children (Shollenberger 2009). In a rural setting in which women are morally judged by how well they care for their family members (Seiling, Manoogian, and Son 2011), few scholars have identified how rural women morally redefine their relationship obligations even though, as we find, it may be crucial for their health outcomes. In our findings, women's strategies to reduce their substance use directly confronted the ambivalence of their support networks. Though they embraced their responsibility to care for family, they defined limits in their financial and social support and articulated their family members' mutual obligations to them.

The vast majority of the research on incarcerated women and men's community re-entry has focused on poor urban communities (Grieb et al. 2014; Kuo et al. 2013; Naser and Visher 2006), raising the question of what separate issues may be faced by rural populations. In a remote region with no public transportation and limited public services, this predominantly low-income group of women faced difficult choices about support upon re-entry to support their proposed health behavior changes. For some, the only alternative to home environments with drug use was a homeless shelter—if even that was available. Others, who wished to avoid relying on risky relationships forfeited work or rehabilitation opportunities. In a region with a limited population, women struggled to identify alternative social networks beyond the complicated intimate, family, and community relationships that were involved in their past drug use. Women in this sample typically had higher rates of marriage than urban populations, yet still many rural women struggled to support their behavior changes in households with revolving membership. Though Appalachia has a long tradition of kin- and community-based support (Coyne, Demian-Popescu, and Friend 2006), female-headed households strained by high rates of poverty are increasingly common (Lichter and Cimbaluk 2012). Further research could illuminate how moral concepts of family support in rural areas (Sherman 2006) may distinguish rural incarcerated women's strategies for behavior change.

This study has noteworthy limitations. First, the qualitative analysis focused on clinical session transcripts from a small sample of incarcerated female substance abusers from rural jails in one Appalachian state. While the sample was randomly selected and randomly assigned to the intervention group, the sampling frame may still limit generalizability to other samples of female offenders and substance users. In particular, this study's sample is differentiated from other populations of incarcerated females because of its racial/ethnic make-up (100% White) and its setting in jails, rather than prisons. In addition, while no jail staff or personnel were present during clinical sessions, the potential existed for response bias in the jail and audio-taping. One-on-one clinical sessions took place in a private office in the jail and no identifying information was collected for the audio files to increase confidentiality. It is possible that these findings about women's agency were a limitation of the data, as we analyzed motivational interviewing sessions in which the goal was for incarcerated women to identify their capacity for changing their drug use and related health behaviors. Because the open-ended nature of the interviews was qualitatively different from baseline data on women's health and relationships, it is impossible to discern whether participant statements represented intervention outcomes or women's “authentic” views. Nonetheless, these data demonstrate women's dynamic involvement in their relationships and their aspirations to reshape these relationships in spite of the challenges they face in doing so.

Conclusion

Despite these limitations, in this study we addressed critical issues for rural women as they re-enter the community after incarceration. Specifically, drug-using rural women face unique challenges in navigating their intimate, family, and community relationships. Though they rely on these relationships for support in re-establishing themselves, these relationships may threaten their mental health, involve histories of shared drug use or ties to criminalization—all of which can inhibit women's efforts to end their drug use. This study demonstrated women's strategies to sustain abstinence from drug use which were often relationship-based, using critical assessment, limit-setting, and positive motivation to realistically appraise the community context to which they return. In highlighting incarcerated women's perceptions of their strategies for behavioral change, we draw attention to a range of relationships beyond those targeted by interventionists as most “risky” while demonstrating the agency that women carve out for themselves despite their unequal positions. In so doing, we draw attention to the fact that interventions to promote incarcerated women's health behavior changes must acknowledge the complex social environments in which they live.