Adapting A Family-Based HIV Prevention Program for Homeless Youth and Their Families: The HOPE (HIV prevention Outreach for Parents and Early adolescents) Family Program

Abstract

As rates of HIV infection increase in adolescents, it is important to provide prevention programs targeting this population. Homeless adolescents living with their families in shelters are at greater risk of participating in risky sexual behavior and incurring negative health outcomes. A family based HIV-prevention pilot study was conducted with eight homeless families in a New York City shelter to explore: 1) the perceived impact of family communication, parental monitoring, family understanding of puberty, STD's and HIV on preventing risky behavior for the participating youth, and 2) the feasibility of conducting such a program within the shelter system. Qualitative and quantitative results indicate increased family communication, parental monitoring and decreased parental depressive symptoms.

Adolescents are the fastest growing population at risk for HIV/AIDS in the United States (Centers for Disease Control, 2003; ^{2000; 1998}), with females and minority youth disproportionately affected by STD's and the AIDS epidemic (Centers for Disease Control, 2003; ^2001; DiLorenzo & Hein, 1993^; Jemmott, Jemmott & Fong, 1998). One in four of all new cases of HIV infection occurs among people under 21 years of age (Center for Disease Control, 2001), with most adolescents (80%) reporting their first sexual intercourse between 14 and 19 years of age (Goldman & Goldman, 1988; The Alan Gutmacher Institute, 1994; Moore & Rosenthal, 1993). Over the last decade, the incidence of HIV and AIDS infection has risen dramatically in low income, minority neighborhoods. African American and Latino youth, who are overrepresented in such poor neighborhoods, are at increased risk of exposure to HIV due to overall rates of neighborhood prevalence, along with poorer access to preventive healthcare, early detection and treatment services (Rotheram-Borus, Mahler & Rosario, 1995).

The overrepresentation of African American and Latino youth living in poverty is magnified in New York City, where youth account for almost half of all residents living in the city's homeless shelters (Partnership for the Homeless, 2003), with the number of homeless families having increased by almost 78% in the last 5 years (Citizen's Committee for Children, 2003). Moreover, while African Americans and Latinos represent 53% of the city's total population, they represent the majority (90%) of the city's homeless shelter population.

The detrimental effects of homelessness on youth development have been previously examined, with findings that suggest disruption in housing is associated with youth mental health difficulties (Bassuk & Rosenberg, 1990; Masten, Miliotis, Graham-Bermann, Ramirez & Neemann, 1993; Rescola, Parker & Stolley, 1991; Zima, Wells, & Freeman, 1994; Bassuk & Buckner 1997), educational underachievement (National Coalition for the Homeless, 1997; Rafferty, 1998^; Shinn & Weitzman, 1996; U.S. Department of Education, 1997), substance abuse (Athey, 1991; Huba et al., 2000; Rotheram-Borus, Koopman & Ehrhhardt, 1991) and numerous other health-related problems (Riemer, Cleve & Galbraith, 1995). Further, experiences of homelessness have been found to disrupt parenting practices (Hausman & Hammen, 1993; Rossi, 1994) and are often compounded by adult caregivers who attempt to cope with high levels of stress (Burt & Cohen, 1989), their own mental health issues (Banyard & Graham-Bermann, 1995; Lipton, Nutt & Sabatini, 1988; Drake, Yovetich, Bebout, Harris, & McHugo, 1997) and drug abuse (Weinreb & Bassuk, 1990; Selzer, Vinokur, & van Rooijen, 1975). For youth within homeless families, the aforementioned factors converge to place them at high risk for a range of overlapping difficulties, which can include engaging in HIV risk behavior, drug abuse, and mental health difficulties, particularly as youth enter the critical developmental period of adolescence.

This article describes the development and preliminary examination of a family based HIV prevention program designed to meet the needs of homeless youth and their families living within New York City's Tier II shelter system. Tier II shelters are designated specifically for families (caregivers and children) as temporary housing to facilitate their return to permanent housing. These families are likely to live in the shelter for an average of one year and are provided with case management services by shelter staff which includes, child care, employment services, independent living planning and assistance in searching for housing. Issues specific to this population, including the stigma of homelessness, barriers to resources and impact of shelter life on family dynamics, are discussed. The preliminary impact of the program and related feasibility and implementation challenges are presented and next steps for research and program development are highlighted.

Adolescence and risk behavior

Adolescence is marked by the onset of physical and emotional maturation, which is accompanied by the challenges of adapting to social, emotional, and cognitive changes. As children move through adolescence towards adulthood they have basic developmental and psychosocial needs, including being valued as a member of a group, receiving family support, participating in caring relationships, acquiring skills to cope adaptively with everyday life, and believing in a future with real opportunities (Hamburg, 1990; Hauser, 1999). Youth under adverse circumstances are often ill prepared to effectively cope with these normative challenges, making this period particularly problematic and potentially associated with high rates of school dropout, early and risky sexual behavior, pregnancy, drug abuse, and suicide (Hamburg, 1990). Moreover, adolescents with elevated mental health needs, or levels of emotional distress, are known to be at increased risk for engaging in unsafe sexual behavior and drug use due to impaired judgment, poor problem-solving ability, low self-esteem, self-destructive behaviors, and poor interpersonal relationships (Brown et al., 1997; Jemmott & Jemmott, 1994; Walter, Vaughan & Cohall, 1991).

Adolescent risk behavior and homelessness

Homelessness during a critical developmental transition, such as entering adolescence, might prove particularly devastating within a national context that finds teenagers dramatically affected by the shifts in the AIDS epidemic over the last decade. Opportunities for HIV exposure increase at this young age, as adolescents have limited life experience and may not have the skills to negotiate sexual risk situations. In addition, adolescents may be more impacted by homelessness than their younger counterparts as a result of increasing cognitive capacities that allow for heightened awareness of family circumstances, experiences of stigma related to homelessness, and the consequences of disruption on peer relations and school placement (Buckner, et al. 1999; Tower, 1992). Thus, this particular group of youth are at substantial risk for engaging in early, high risk sexual activity and drug abuse when faced with compounding issues such as, lack of stability, stress, school disruption, the physical environment of shelters, and the frequent struggles of adult caregivers with stress, mental health and substance abuse issues. The unique needs of homeless adolescents living in shelters with their families, compounded by their increased likelihood for negative outcomes, specifically in relation to HIV exposure and infection, create a distinct context for prevention efforts.

Families in the context of homelessness

Families, primarily mothers with their children, constitute a significant proportion of the nation's homeless population (Hausmann & Hammen, 1993). In 1995, the U.S. Department of Education estimated that 744,000 school age children and adolescents were homeless during the course of one school year; this estimate increased to one million youth in 2002 (U.S. Department of Education, 2002). Approximately 87 percent of homeless youth are enrolled in school, but only 77 percent attend school regularly (U.S. Department of Education, 2002). In New York City, families with children are the fastest growing sector of the homeless population. In September 2002, an estimated 8,900 families with 16,000 children resided each night in the city's shelter system.

Previous studies tapping into the strengths of homeless families show that prevention programming might capitalize on parent's motivation in relation to their children. In a study conducted by Banyard and Graham-Bermann (1995), homeless mothers described high levels of determination related to changing their life situation “for [their] children.” In addition, the mothers sampled described specific parenting behaviors that were meant to protect their children from experiences within the shelter, particularly in relation to potential victimization and contact with “negative influences.” Further, there is a small body of evidence indicating that homeless parents maintain regular contact with family members, and most have sizable overall social networks (Farr, Koegel, & Burnam, 1986; Toro & Wall, 1991), which may help mitigate the experiences of homelessness for their children (Jarret, 1995). Therefore, family-focused HIV prevention programs that encourage social support and social service, mental health, and substance abuse resource utilization might enhance the possibility for effectiveness and sustainability of change in youth and parent behavior over time.

The absence of evidence based prevention programs that target homeless pre and early adolescents and their families, led to the collaborative development of The HOPE (HIV prevention Outreach for Parents and Early adolescents) Family Program by university researchers and community members. HOPE is based on, and adapted from a nationally and internationally tested model of family based prevention programming.

Development of the HOPE Family Program

A Collaborative model of program development

The Collaborative HIV-Prevention and Adolescent Mental Health Family Program (CHAMP) served as a template for the HOPE Family Program. The original CHAMP Family Program was developed in response to the increasing need to reduce urban adolescent HIV exposure in inner-city populations. CHAMP's original goal was to increase understanding of sexual development and HIV risk within the urban context, while applying that understanding of development to an intervention program (Paikoff, 1997; McKay et al., 2000). The content and structure of the CHAMP Family Program was influenced by research findings and a collaborative partnership. The primary goals of this evidence-based intervention were to: 1) foster discussion of sexual possibility situations; 2) make links between family processes and children's participation in sexual possibility situations (in particular, stressing family communication, rule setting, monitoring, support, and provision of clear values); and 3) increase caregiver and youth understanding of puberty and HIV/AIDS in order to prepare families for the coming changes of adolescence.

Although CHAMP was reaching families through their children's schools, school officials and recruitment efforts indicated that a large segment of the targeted population was being missed. Families experiencing homelessness were less likely to participate in the program with domiciled families where information about their housing status may have come to light, potentially exposing their children to teasing from classmates and stigmatizing judgments from other parents. Development of the HOPE Family Program was a direct result of this situation in an attempt to reach the large numbers of homeless families in the initial targeted communities.

Bridging the gap with the HOPE Family Program

Given the goals of promoting communication and managing high risk behavior, the CHAMP model offered an opportunity in collaboration with Tier II shelters, to address the need for family based prevention with homeless families. The resulting HOPE curriculum maintained the delivery methods of the CHAMP Program using multi-level group modalities, which include both multiple family sessions and parent and child specific group sessions.

As in the CHAMP program, the HOPE program sessions are very interactive and rely on an open dialogue between participants and facilitators. Each session focuses on different topics, such as communication, establishing rules, puberty, understanding HIV/AIDS and preparing kids for adolescence, to achieve the goals and objectives of the program. Consequently, it is the hope that these intra-family discussions, grounded in the development of family communication skills, bring participants to a deeper understanding of the impact of the virus and its risk factors, with the aim of preventing and/or decreasing behaviors that place children at risk of contracting HIV.

A major component that was maintained and especially important for the HOPE curriculum was the “family support” section. This session helps families recognize who supports them in times of need and in helping raise and monitor their children. For families residing in a shelter, the identification and recognition of their internal and external support system was considered extremely useful in highlighting their available resources for managing and buffering stressful situations inside and outside of the shelter.

Significant changes made include: shortening the length of each session in order to retain participation and maximize time and scheduling constraints placed on the project by the hosting shelter. A trained social worker was also present at each session to provide clinical support to the group if needed, due to the perceived vulnerability of participating children and parents. The pilot project staff, which included university researchers and community based group facilitators, collaborated closely with shelter staff in addressing potential obstacles and setting ground rules to ensure meaningful dialogue to benefit the future development of the HOPE Family curriculum and program.

Methods

Setting

The HOPE Family Program was pilot-tested in 12 sessions over a 6 week time span at the HELP Morris Tier II Family Shelter in Bronx, New York. As one of the largest shelters in NYC, HELP Morris houses up to 241 families, 80% of which are headed by single mothers (www.helpusa.org). Open since 1992, the shelter's objective is to provide assessment and transitional housing to dislocated families. The shelter provides case management services for families towards obtaining permanent housing, and youth programs through the KIPS Bay Boys & Girls Club on site for the children.

Recruitment

Families were recruited based on the study's eligibility requirements. In order to participate in the HOPE Family Program, families were required to have children in the 4^th and 5^th grade within the household and at least one available parent to participate with the child. To facilitate recruitment, a flyer advertising the program was placed in the mailbox of each family with eligible children and asked to contact the Program Director at KIPS Bay Boys & Girls Club for details. Due to limited funding and in order to keep the pilot manageable, while obtaining pertinent information for potential impact and feasibility, researchers opted to work with a limited number of participants. Eight families (n=8 parent/child dyads) totaling 16 participants, were recruited for the HOPE Family Program pilot study.

Participants

The eight families initially recruited to participate consisted of seven mothers, one father, five daughters and three sons. Three of the families identified as African American and the other five as Latino. Caucasian families were not excluded, although their lack of presence within the study sample is representative of the shelter's population. The age of the child participants ranged from 9 to 11 years of age, in line with the age group for 4^th and 5^th grade students. The participating caregivers were all biological parents of the children and ranged in age from 27 to 34 years of age. All of the families had resided at HELP Morris anywhere from 2 months to 2 years.

Procedure

During the first session of the intervention, families signed informed consent forms informing them of the research study, the possible benefits and risks to participating, and their right to confidentiality. Families also gave permission for the groups to be videotaped to facilitate a process evaluation, which was further aided by the process notes taken during each session by a trained observer. In addition, every parent and child completed pre and post- test quantitative intervention measures in order to gather preliminary data on the potential effectiveness of the intervention on significant individual and family processes.

Parent mental health was measured using the anxiety and depression subscales of the Symptom Checklist-90-Revised (SCL-90-R; Derogatis, Lipman, & Covi, 1973; Derogatis, Rickels, & Rock, 1976). Respondents rated the extent to which each symptom (e.g., feeling blue) had been bothersome in the past week on a 5-point Likert scale (1 = not at all, 5 = extremely). The subscales chosen for this study were selected on the basis that lower SES individuals, represented in the current sample, are more likely to evidence symptoms of depression and anxiety (Dohrenwend, Levav, Shrout, et al., 1992; Johnson, Cohen, Dohrenwend, et al, 1999; Ritsher, Warner, Johnson, & Dohrenwend, 2001). Higher scores represent higher mental health difficulty. The degree of parental involvement & supervision and discipline effectiveness were measured using an adaptation of Tolan, Gorman & Smith's Parenting Skills Questionnaire (1991).

Due to small sample size, quantitative data was analyzed using paired sample comparison means at baseline and post-intervention. Although quantitative measures were utilized, the authors felt that due to the expected small sample size, confidential in-depth exit interviews would provide solid pilot data that could be beneficial when evaluating the program for effectiveness and feasibility. In-depth interviews were conducted by non-group facilitators with every participant at the last program session. This interview consisted of questions that assessed participant satisfaction, perceptions of family change, impact of homelessness on family, social support and program feasibility. Interview data was assigned themes from frequency of participant responses. Program feasibility was assessed based on program outcomes.

RESULTS

The multi-level quantitative and qualitative evaluation of the HOPE Family Program pilot study suggests that, in general, participants were satisfied with the program and viewed their involvement as beneficial for themselves and their families. Study findings also indicate increased parent/child communication, increased parental supervision and monitoring of potential high risk possibility situations and a decrease in parental depressive symptoms. There is also significant evidence of program feasibility within the targeted setting with the intended population. Specific findings are presented below.

Participant Satisfaction

The qualitative data shows that parents and children voiced satisfaction and pleasure with the program overall, with some wanting the program to have been longer than it was. One parent stated during an interview that “HOPE was the only positive thing to come out of being at HELP Morris.” All of the parents agreed in the interviews that the information being shared during the program was new to them and applicable to their lives. One parent stated, “My mom has been positive [for HIV] since 1990, [my] sister since '86, I didn't know how HIV physically effects people.” All of the children reported at the interviews that the information was also new to them, specifically the information on puberty. The consistent level of attendance throughout the program supported participant satisfaction with an average of 12 participants per session. (There was a one week low attendance of 8 participants and a high of 14 participants in another week). 10 participants (5 parent/child dyads) completed the program by attending the last session. The program only lost 6 participants, and 2 of those can be attributed to obtaining permanent housing. Initially group members who obtained housing were not able to attend the program due to the hosting shelter's no visitor policy; this policy was later adjusted to allow post-housing access to the group. Two other families later received housing and were allowed to visit the shelter to participate.

Parent/Child Communication

Overall, there was a strong sentiment that family communication had increased and parents and children were talking more often. Both parents and children reported in interviews and on the self-reporting measures that since participating in the program, they were able to communicate about “tough topics,” such as puberty, sex, drugs, STD's and HIV in a more effective manner (parents: M=23.6, SD=4.3 to M=24.6, SD=3.1 and children: M=15.2, SD=9.4 to M=20.4, SD=8.1). During the interviews, parents and children expressed that increased communication was one of the best outcomes with one parent stating, “I can talk to my son better. He doesn't ask me why I'm asking about his life.” Another parent explained it by stating, “I'm communicating with my child in a better way by being comfortable and not getting mad.” Many of the children clearly stated that they talk to their parents more, with one child explaining that “me and my mom get along much more”. Another child stated that HOPE had helped them “talk stuff out with [my] mom and others instead of fighting. It's good to talk about views instead of holding it in.”

Parental Monitoring

As parents became more conscious of possible high risk situations, there was also a slight increase in parent monitoring. Parents reported an increase in how often they check for adult supervision when their child is going to be outside of the shelter (M=4.6, SD=.5 to M=5.4, SD=.9). This was supported by the children who reported an increase in after school supervision by their parents (M=2.6, SD=2.2 to M=1.0, SD=.0); indicating that they were now being supervised more than at baseline. One parent reported that, “having to obey [shelter] rules all the time” made it difficult for her to monitor her child the way she saw fit. Another parent reported that their concerns surrounding the difficulty of monitoring children in the shelter decreased after attending the program because “my daughter is now aware of what is out there and dealing with boys and sexual possibility situations.”

Parental Depression

It is important to note that on the Symptom Checklist-90-Revised (SCL-90-R) which measured parent depression and anxiety, the pre and post test means decreased with a pre-test mean of M=42.8, SD=15.1 and a post test mean of M=39.8, SD=10.3. This decrease was felt by parents who expressed, during interviews, that they felt supported in HOPE and consistently stated that their home life with their children had improved. All of the children reported during the interviews that they felt HOPE had been beneficial for their parents. When asked how he felt the program had helped his mother, one child stated “it calmed her down.”

Family Stressors

The qualitative interviews also presented findings regarding significant family stressors in the context of shelter life. One particular theme was the impact of homelessness on childhood experiences. Children reported that they had lost friendships and felt stigmatized at school by teachers and classmates due to living in a shelter, which intensified the difficulty of trying to deal with the stigma and stress of living in a shelter. One child when describing how living in the shelter had affected them stated during an interview that “certain kids at school made fun of me…[my] teacher announced to class that I was in a shelter…my dad is an adult he can handle it better than I do.”

Aside from the daily issues of shelter life (lack of space, no visitors allowed, imposed curfews), participants also consistently discussed the high levels of violence that they were exposed to from other shelter residents. Participants reported witnessing fighting inside and outside of the shelter. This was further supported by HOPE Family Program facilitators who were on site to witness an altercation between shelter residents (not involving program participants).

Social Support

When discussing support systems, parents indicated that they felt supported by the HOPE Family Program which helped them make connections with other families in the shelter that they were previously isolated from. Participating in the program and increasing their understanding of HIV and AIDS also prompted some parents to contact HIV positive relatives that they had previously avoided. Parents also initiated discussions with their children about family members who were infected, thereby increasing their family's support network. One participant stated, “[My] uncle had HIV, [I ]never went to see [him], because [I] thought I could get it from sharing soap or hugs. Now [we] see him a lot more because we understand these things.” However, when exploring other internal shelter supports, families clearly stated a need for supportive and relevant programs in the shelter. Many of the parents stated displeasure with their caseworkers and felt unsupported in their struggles to find permanent housing while raising their children within a shelter.

Identification of Barriers

Evaluation of the HOPE Family Program pilot study also revealed barriers to participation in the intervention. Obstacles included both personal and programmatic barriers which are detailed below.

Personal Barriers

Participant personal level barriers included lack of time, competing priorities, and some initial discomfort in discussing sensitive topics. Participating in the program required families to commit to three hours every Saturday morning for 6 weeks, with two sessions being covered weekly. This is a big commitment for any family to give, much less for families overly stressed by their present living situation and their efforts to resolve it. This was also related to competing priorities, as families were involved in trying to obtain permanent housing, visiting their families outside of the shelter (families were not allowed visitors in the shelter), and taking care of other personal needs.

Programmatic Barriers

During the preparation for and implementation of the HOPE Family Program pilot, some facility level barriers were identified, including: 1) space constraints; 2) shelter staff intrusions; 3) confidentiality issues; 4) coordination of childcare and 5) the “no visitors” policy. Shelters in general appear to have limited non-residential space; consequently space for groups was limited and often less private than the average intervention program. This led to concerns regarding confidentiality as shelter staff was often intruding on group time. Each of these issues was addressed in conjunction with the shelter staff to improve feasibility and success. An important concession from the shelter, which was implemented after the programs onset, was to allow families that obtained housing during the course of the program to visit the shelter to complete the program, allowing for improved retention.

Curriculum Development

Although pilot data indicates that the program was successful and beneficial for families, the process evaluation revealed several areas that must be addressed through curriculum development. Based on participant responses, it is clear that stress, stigma, and coping with the exposure to violence are components that should be adapted into a prevention curriculum for these families.

Families reported a need for programs like HOPE in the shelters, with one participant offering suggestions on what other shelters could benefit from the intervention. Parents and children expressed that the information shared was relevant to their lives, was presently helpful, and would continue to be useful in the future, with one child getting to the goal of the HOPE Family Program by stating “you don't need to know some things when you are at a certain age, but its good to know before you face it at a later age.”

Discussion

Homeless families with early adolescent children are less likely to participate in HIV and drug use prevention programs due to multiple life stressors, but have a higher risk than domiciled families for early onset of sexual activity and drug use in children. In the pilot study described in this article, an evidence based HIV prevention program provided an important theoretical framework and foundation to create a program geared towards homeless families. The findings derived from this pilot study suggest that prevention programs with homeless families can be successfully implemented using shared recruitment responsibilities and genuine collaboration with shelters, specifically relevant program information and well-trained facilitators. Families who participate in such programs may find them valuable in improving family communication, and parental monitoring and supervision of youth inside and outside of the shelter towards reducing potential risk situations, especially in a living situation which at times may be disempowering for parents.

Beyond acknowledging the successes of the pilot program, it is clear that there are still some themes that must be included to further meet the needs of these families. Issues related to violence exposure, and the related stigma and stress of homelessness must be addressed, as these are unique and distinct concerns for these families in crisis. It can be reasoned that not focusing on these issues may impact participant attendance and retention, which can already be problematic for such a transitional population. Moreover, failure to address these issues may undermine the significance and success of any program attempted.

There were also notable limitations in this study. Although researchers chose to keep the pilot at a modest size, the small sample resulted in quantitative results that weren't statistically significant and made the rate of attrition seem high; slightly detracting from the fact that sixty percent of families completed the program. Obtaining permanent housing, while necessary and highly desirable for participating families, also contributed to the attrition rate as these families could not be contacted by researchers because initial consent forms only provided contact information at the shelter. In addition to gaining permission from the shelter for former residents to participate, retention could have been aided by obtaining some form of permanent contact information prior to a family's departure from the shelter. This would also be beneficial in a larger study to facilitate post-intervention follow-up to measure impact of the program on risky behaviors.

Conclusion

Although the concrete needs of homeless families in shelters are being addressed, it is clear that these families are still in need of supportive interventions that target family communication, coping, stress and exposure to risky behaviors to prevent negative outcomes for youth and their caregivers.

The supportive framing of a multi-family program, such as the HOPE Family Program, can be beneficial for decreasing mental health issues for these families as well. This can increase family communication and improve parental monitoring, reducing possible risk situations and delaying onset of risky behavior. Homeless families are not very different in their needs and concerns from domiciled families, though their issues are magnified due to their circumstances and seemingly higher levels of past high risk behavior which may influence future behaviors.

Although this was an exploratory pilot study with a small sample, the results are encouraging. It is clear that with flexible and purposeful collaboration with shelters, these families can be reached for successful intervention programs. Fortunately, researchers were able to partner with a shelter that was receptive to the program and did not present many obstacles towards conducting the pilot. A program model like HOPE that uses an evidence based curriculum designed by community stakeholders, community members and researchers can be relevant in addressing the needs of homeless families who may be more likely to participate in a shelter-based project. Participant's satisfaction with HOPE and the positive changes in family processes noted in the results also speak to the potential impact of this type of program. All of this information can be useful in developing a more comprehensive prevention program to be implemented with a larger study sample. The lack of such programs targeting homeless youth and their families is at odds with the growing numbers of HIV infection and substance abuse among adolescents of color, who comprise a large percentage of the homeless population, and are at a greater risk because of the factors associated with homelessness. A family-based model that targets parental, youth and family processes may not only support families through their disruption in housing, but provide a buffer that could reduce the potential negative behavioral and health outcomes associated with homelessness.