|Home | About | Journals | Submit | Contact Us | Français|
Little is known about the transition to adulthood for adolescent females and young women who are impoverished and homeless. Co-occurrence of drug use and abuse, HIV risk, and victimization is notable among homeless women, highlighting the need for comprehensive interventions. Unfortunately, evidence-based prevention approaches addressing these inter-related problems among impoverished women transitioning into adulthood are lacking. To address this gap, we designed an innovative prevention program by utilizing open and closed ended interview data from impoverished women (n = 20), focus groups with community experts and providers (2 groups; n = 9), and a theoretical framework to direct the research. Information provided by our focus groups and interviews with women supported our theoretical framework and highlighted the importance of addressing normative information, providing skills training, and utilizing a non confrontational approach when discussing these sensitive issues.
Little is known about the transition to adulthood for young women who are impoverished and homeless. The period of emerging adulthood, typically defined as ages 18 to 25 (Arnett, 2000a), deserves attention for several reasons. During this period, many developmental changes take place, including increased independence as adolescents leave home, full-time employment, and developing or strengthening relationships with sex partners (Bachman et al., 2002). These changes may lead to increased participation in and exposure to risks. Alcohol and drug use typically peak during this period (Windle, Mun, & Windle, 2005). In addition, many young women may not use condoms consistently with new sex partners, placing them at risk of HIV/AIDS and other STDs (Leigh, 2002). Finally, surveillance data show that women between the ages of 16 and 24 are at greater risk of domestic violence than any other age and gender group (Rosewater, 2003).
All of these problems are more pronounced among homeless young women in shelters. Specifically, 44% of women ages 18 to 25 living in shelters in Los Angeles County have used marijuana during the past year (Wenzel et al., 2006), compared to 24% of adolescents and young women ages 18 to 25 in the National Survey on Drug Use and Health (SAMHSA, 2004). Sexual risk-taking and experience of victimization by violence are also more common among homeless than low-income housed women (Wenzel et al., 2004).
Substance use, HIV risk, and victimization are inter-related (Maman, Campbell, Sweat, & Gielen, 2000; Wenzel et al., 2004), but current interventions for homeless young women are limited in that many focus on one specific risk behavior such as HIV (e.g., Rotheram-Borus et al., 2001; Woods et al., 1998). Programs focusing on substance use tend to focus exclusively on tertiary prevention (i.e., treatment) (e.g., El-Bassel, Witte, Wada, Gilbert, & Wallace, 2001), and there are few programs to prevent physical or sexual victimization (Gunderson, 2002; Wathen & MacMillan, 2003). Thus, all three problems are rarely addressed together, and if a problem is addressed, programs are typically more treatment focused versus prevention focused.
To address this gap in services for homeless young women, we developed a prevention program that targeted alcohol and other drug (AOD) use, HIV risk behaviors, and victimization by intimate partner violence. Because risks tend to co-occur (Gerrard, Gibbons, Benthin, & Hessling, 1996), particularly among homeless and other impoverished persons (Amaro et al., 2001; Wenzel et al., 2004), it is important to simultaneously address all three behaviors to decrease overall risk. We utilized a theoretical framework, interviews with women and focus groups with community experts and providers to develop content for these three topics. Our theoretical framework was based on Social Learning Theory (SLT) and Decision Making Theory (DMT). Both theories have been used extensively to develop interventions with youth to address how beliefs may impact behavior and to educate youth to examine the pros and cons of behavior before making choices (Borsari & Carey, 2000; D'Amico et al., 2005).
Specifically, SLT suggests that a person's learned beliefs may significantly impact subsequent behavior (Bandura, 1986; Maisto, Carey, & Bradizza, 1999). Beliefs may be inaccurate and can increase potential risk taking behavior. For example, research has shown that youth who overestimate the AOD use of their peers are more likely to use these substances (Ames, Sussman, & Dent, 1999; D'Amico et al., 2001). In addition, women may underestimate the use of condoms among their peers (Latkin, Forman, Knowlton, & Sherman, 2003), which may lead to increased sexual risk-taking. Perceptions of the prevalence of dating violence, as well as personal experience with this kind of violence, may influence beliefs that dating violence is normative (Wingood, DiClemente, McCree, Harrington, & Davies, 2001). Stereotyping is also relevant to violence and victimization and is evident among homeless women's expressions of self blame for being victimized due to their drug use (Tucker et al., 2005). Thus, it is important to discuss normative information in the context of all three of these behaviors.
Decision Making Theory suggests that decisions about using substances are often emotional (Kahneman, Slovic, & Tversky, 1992; Kahneman & Tversky, 2000) and therefore problem focused coping skills are needed. In addition, most youth do not see long-term consequences as relevant (e.g., Arnett, 2000b). This is particularly true of homeless youth (Rew, Chambers, & Kulkarni, 2002). Thus, it is important to focus on short-term probabilistic consequences of behavior that may be more meaningful to young people rather than long-term consequences that may not be as relevant. Decision making also involves competing motivations because there are pros and cons associated with making changes and staying with the status quo (Miller & Rollnick, 2002). For example, many youth may perceive benefits from substance use and risky sexual behavior (Fromme, Katz, & Rivet, 1997). Thus, according to our framework, it is crucial to provide a discussion of the pros and cons of AOD use, HIV risk behavior, and victimization. Providing skills training can also emphasize how to weigh choices in high risk situations and help young women make healthy, informed decisions.
Program presentation is also important to consider in program development. A motivational approach (Miller & Rollnick, 2002) focused on non-judgmental and nonconfrontational communication can be valuable. Motivational approaches are based on expressing empathy through acceptance that facilitates change, developing discrepancy by letting the client present arguments for change, rolling with resistance by avoiding arguing, and supporting self-efficacy by encouraging a person's belief in the possibility of change (Miller & Rollnick, 2002). Such techniques have been used effectively to help adolescents (D'Amico & Edelen, in press) and young adults (Borsari & Carey, 2000) reduce harmful substance use. This approach has also shown promise in reducing sexual risk-taking among HIV positive young people (Naar-King et al., 2006.). Because motivational interviewing emphasizes an interactive process in which people are active participants, this approach can help ensure that program content is both culturally appropriate and acceptable.
Over six months, we completed 20 semi-structured interviews with young women staying in shelters and conducted two focus groups with community experts and shelter providers. To maximize variation across shelters, we sampled sites by size (i.e., total shelter beds/number of persons served), geographic location, and number of female residents, ages of 18 through 25. Because our intervention was prevention focused, we excluded shelters for domestic violence, and residential AOD treatment sites. Of the 39 sites screened, 19 met our eligibility criteria in that they served young women ages 18-25, did not serve an exclusively Spanish-speaking population, and served a majority homeless population. Participating shelters received a $100 honorarium.
We recruited female residents through a scripted site visit presentation. After each presentation, our staff screened residents interested in participating in the study. We oversampled African American women because they are disproportionately represented in shelters. We interviewed 2-3 women at 9 different sites for a total of 20 women. Women were African American (n = 9), Hispanic/Latina (n = 6), white (n = 4), or mixed ethnicity (n = 1). Women were 18-19 (n = 7), 20-22 (n = 7), and 23-25 (n = 6). We also recruited community experts and providers (N = 9) to participate in focus groups (N = 2) based on their work in shelter settings or other work for the benefit of homeless women.
We developed a protocol to elicit feedback on developing a prevention program for impoverished women. Table 1 provides a list of key questions. We discussed participants’ beliefs about the transition to adulthood, the types of programs that might be available for women in the shelter setting that target the three risk behaviors, cultural and contextual considerations in developing programs, and barriers and facilitators that might impact implementation of programs in the shelter setting. We concluded by asking participants to address other issues they felt were important. Because these persons were participating in their professional capacity, an oral consent procedure was used. Each focus group was audio taped and detailed notes were generated.
Face-to-face semi-structured interviews were conducted by trained female interviewers. Open-ended questions covered topics such as women's goals, AOD use and abuse, potential sexual risks, and victimization (see Table 2). Closed-ended questions assessed sexual relationships, partner violence, and AOD use (see Measures). Interviews lasted approximately 1 hour, were audio taped and transcribed. Transcripts were entered into Atlas.ti, a qualitative text-management program. The interview was anonymous and did not require the women to provide any identifying information. Women were told to use only their first names when talking with the interviewer and were also told not to provide identifying information about any other person during the interview (i.e., use first names only). The research protocol was approved by the RAND Institutional Review Board. Women were paid $30 for their participation.
Upon completion of the interview, women completed a set of closed ended questions. Measures on substance use, sexual partners, and condom use were adapted from our previous work (Wenzel et al., 2006; Wenzel et al., 2004).
Participants were asked about use of alcohol, marijuana, crack or cocaine (including freebase), and “any other drugs” in the last six months. Lifetime information was gathered on outpatient or residential treatment, and on self-change efforts for AOD use.
Women reported whether they had any steady (e.g., boyfriend or husband), casual (e.g., friend you might be dating but not going steady with), and need-based partners. They reported the total number of partners for the last six months.
Women reported condom use, comfort in using condoms, and how often they refused sex with their partner in the last six months if a) they did not want to have sex and their partner did and b) their partner refused to wear a condom. Consistent condom use was defined as using a condom every time that they had sex.
Victimization was assessed with items from the Conflict Tactics Scale (Straus, Hamby, Boney-McCoy, & Sugarman, 1996), the National Women's Study (Kilpatrick, Edmunds, & Seymour, 1992), and our previous work (Wenzel, Leake, & Gelberg, 2000). Women were asked about physical and sexual violence and verbal abuse. The terms “violence,” “abuse,” “rape,” and “assault” were not used in the interviews because of their potential biasing effect on women's responses.
To analyze the qualitative data from women's interviews, we first marked text related to each of the three key behaviors, and then examined responses to each of the key questions (Table 2) and sorted them into different categories based on their thematic similarity (Ryan & Bernard, 2003). We built a formal codebook and applied it to the entire text. To assess the saliency of each thematic response category, we counted the number of times each was mentioned by different individuals. A similar analytic procedure identified thematic categories from transcripts and notes from focus groups with community experts and providers.
Table 1 highlights the most salient thematic responses and quotes for each of the key questions asked in the focus groups sessions. Several thematic responses in the focus groups with community experts and providers reoccurred. These included: (1) Many young women see these problems as normative (e.g., “AOD use, violence, sexual risk-taking seem normal; so talking about how this is NOT normal would be helpful”); (2) You need to empower the young women so they will learn (e.g., “What would help them is allow them to set their own goals”); (3) Use a harm reduction approach (e.g., “We can't expect them to stop on the spot, but can start the process”); and (4) The facilitator needs to be non judgmental-- the issues transcend cultures (e.g., “You have got to do something else, something non confrontational; use a non value laden approach; don't point a finger at them”). Thus, participants emphasized topics that fit within our framework, such as providing normative information, teaching skills, and using a non confrontational approach when presenting information.
Barriers to women's successful transition to adulthood were noted, including lack of housing and health care, poor decision making skills, and attention to personal safety (see Table 1). A successful transition should therefore be reflected in “benchmarks” different than for non-homeless women, such as helping homeless women make better decisions and care for themselves. Participants emphasized that the larger context must also be addressed in this population. Low self-worth due to homelessness, and working to survive day-to-day may make it difficult to mature emotionally and negotiate service systems.
Eight women reported using alcohol to the point of being drunk, 5 reported marijuana use, and 1 reported using crack cocaine in the past six months. The majority (n = 12) reported that they had made reductions in their AOD use on their own. Almost half (n = 9) reported experiencing verbal abuse, 6 reported physical and 2 reported sexual victimization.
The main themes that emerged from the interviews with women differed somewhat depending upon the topic discussed. Table 2 highlights the most salient thematic responses and quotes for the key questions asked in the semi-structured interviews with the women. For example, when women were asked how AOD use could stand in the way of young women meeting their goals, two themes emerged: loss of self and functional impairment. When women were asked this same question for HIV risk behavior, two different themes emerged: consumed self and risks/taking chances. We used these themes to help us develop content specific examples throughout the intervention protocol, and to ensure that these issues were addressed in the different sessions. For example, when women were asked to describe what may lead impoverished women to use AOD, they indicated both internal and external events. This interview data led us to include a discussion of both internal and external triggers that might contribute to use when we created the AOD session. Similarly, a theme that emerged in the interviews on victimization was “know your partner”; thus, we included a discussion on warning signs for this session.
Table 3 describes the draft program components we developed based on our theoretical framework, provider focus groups, and interviews with women. We organized this information into three sessions that can be conducted in any order. The information in each one of these sessions complements the other sessions. Women may choose to participate in one session, such as AOD use, or more than one. We also developed a 12-page color brochure to complement the information provided in all three sessions. Some components are included in all three sessions, such as providing normative feedback and discussing resources, whereas other components are session specific, such as discussing the difference between a caring versus harmful relationship and discussing internal and external triggers for drinking. In all sessions, we incorporated role plays so women could view an example of a high risk situation and how it could be navigated.
The current study contributes significantly to the literature as it provides important information on the development of an innovative tripartite prevention program for impoverished young women transitioning to adulthood. The current intervention is unique in that it focuses on AOD use, HIV risk behavior, and victimization by intimate partner violence and adopts a prevention based approach for this vulnerable population.
In developing this program, we drew from the theoretical literature, the opinions of the women for whom this intervention is designed, and the expertise of providers who work for the benefit of impoverished women on a daily basis. Focus groups with providers and interviews with women supported our theoretical and prevention program framework. Specifically, experts, providers, and women all highlighted the importance of utilizing a non-confrontational and nonjudgmental method when presenting the information. The value of motivational interviewing was therefore supported in discussing these sensitive issues. In addition, women and providers emphasized that many young women may view risky sex and other harmful behaviors as normative, thus making it difficult for women to make changes because the behavior may not seem problematic. Feedback on the frequency of these behaviors in this population is thus a valuable educational tool. Focus group and interview data indicated the value of empowering women to make healthier choices. Providing women with knowledge and conducting skills training are key components; for example, many women may not know how to make a safety plan. Preparing women for high risk situations can increase self-confidence in handling them.
The program discussed here was designed to empower homeless young women in the transition to adulthood, and the findings from this study indicate that such a program is needed for this population. Any such program for homeless persons will be limited, however, if prevailing contextual factors such as lack of stable housing are not also addressed by providers and government agencies.
In sum, characteristics of the women we interviewed suggest a high level of need for prevention services designed to address the overlapping problems of AOD use, HIV risk behaviors, and victimization. Our framework and findings promote confidence that the content and presentation of the material are appropriate for this population. In addition, a prevention based approach, which incorporates principles of harm reduction and motivational interviewing techniques, corresponds well to the women's needs. Our next step is to pilot test the session content and presentation method with impoverished young women so that we can make any necessary modifications and finalize our prevention manual. Efficacy will then be examined in a larger clinical trial. If findings are positive, this program can then be offered more broadly to a population of young women for whom few such services are available.
Elizabeth J. D'Amico has her Ph.D. in Clinical Psychology, Dionne Barnes has her Masters in Social Work, Mary Lou Gilbert has her Masters in Latin American Studies and her J.D., Gery Ryan has his Ph.D. in Anthropology, and Suzanne Wenzel has her Ph.D. in Community Psychology. Work on this article was supported by a grant from the National Institute of Drug Abuse (R21DA019183). The authors wish to thank the women, providers, community experts, and shelters who participated in this research. We would like to acknowledge Dr. Marcia Ellison for her contributions on an earlier part of the study. We would also like to thank Michael Woodward and Shirley Cromb for their assistance and creativity in developing project materials.