Barriers to Psychosocial Services among Homeless Women Veterans

Abstract

Veterans comprise a disproportionate fraction of the nation's homeless population, with women veterans up to four times more likely to be homeless than non-veteran women. This paper provides a grounded description of barriers to psychosocial services among homeless women veterans. Three focus groups were held in Los Angeles, CA, with a total of 29 homeless women veterans. These women described three primary, proximal (current) barriers: lack of information about services, limited access to services, and lack of coordination across services. Compared to non-veteran homeless women, women veterans potentially face additional challenges of trauma exposure during military service, post-military readjustment issues, and few services specific to women veterans. Understanding their service needs and experiences is critical to the development of relevant and appropriate services that move homeless women veterans away from vulnerability, into safety.

Women veterans are up to four times more likely to be homeless than non-veteran women (Gamache, Rosenheck, Tessler, 2003). Current limited evidence indicates that homelessness among women veterans is on the rise, especially as there are more women veterans than ever and they are continually returning from recent conflicts (e.g., in Iraq and Afghanistan). A recent report released by the US Department of Housing and Urban Development (2011) indicates that women veterans are more likely to be homeless than female non-veterans in the U.S. and female non-veterans in the U.S. poverty population.

Although services for homeless women are growing in response to federal efforts to end homelessness, there are still numerous barriers to care and, consequently, homeless women may not receive the services that they need. Given that most women veterans utilize services outside VA (Washington, Yano, Simon, Sun, 2006), it is likely that homeless women veterans—and women veterans at risk for homelessness—are being seen in community-based organizations that may not be equipped to address the potentially unique needs and histories of women veterans. Understanding homeless women veterans’ perceptions of and experiences with psychosocial services is critical to the development of more comprehensive care and programs that could ameliorate risk for homelessness.

Little is known about perceptions of care among homeless women veterans, but guidance can be gleaned from representative studies of women veterans. In a nationally representative sample of women veterans, perceptions of VA care were least positive regarding the availability of needed services and the logistics of receiving VA care, with problems related to ease of use as the most significant barrier (Vogt et al., 2006). However, it is important to note that women veterans are using VA services at unprecedented rates (Duggal et al., 2010; Haskell et al., 2011), and their level of satisfaction with VA care is high (Kimerling et al., 2011), particularly when it is tailored to the needs of women (Washington et al., 2011). Perceptions of VA care differ for women with histories of military sexual trauma (Kelly et al., 2008; Kimerling et al., 2011), which may be important when examining the particular vulnerabilities of homeless women veterans.

To place into context homeless women veterans’ perceived treatment needs and treatment utilization, it is necessary to draw upon non-veteran-specific studies of homeless women. For example, Tucker and colleagues (2011) found that perceived treatment need was more likely among homeless women with drug-using sex partners and an arrest history, but less likely for those with a minor child and a longer history of homelessness. Receiving treatment was more likely among women who received informational support from their sex partners and who had an arrest history, but less likely among those who had a more street-based social network, had a minor child, considered themselves homeless, and recently needed mental health treatment. Additionally, Tam and colleagues (2008) found that homeless women with mental illnesses were more likely to be linked to services than those without mental illness, who may lack a defined reason for entry into services.

The purpose of this study was to examine perceived proximal barriers to psychosocial services among homeless women veterans. We consider proximal barriers to be those that women faced in their current homelessness, in contrast to distal factors that may have occurred earlier in women's lives and may have contributed to their risk for homelessness (e.g., childhood trauma; see Zlotnick et al., 2010), which we describe in a separate paper (Hamilton, Poza, Washington, 2011). We chose focus group methodology to augment survey findings (Washington et al., 2010) with more nuanced information about homeless women veterans’ experiences. Focus groups capitalize on group interaction to produce data and insights that might be less accessible without interaction among individuals with common experiences (Morgan 1996). Furthermore, qualitative methods such as focus groups are appropriate when the topic under study is complex and/or when the target population is hidden or hard-to-reach (Sofaer 1999), as is the case with homeless women veterans.

METHODS

Design and Sample

We conducted 3 focus groups among homeless women veterans in Los Angeles, California. Enrollment criteria were being a woman veteran and spending at least one night of the prior 30 in a shelter or transitional residential facility, a hotel paid for with a voucher, a car, an abandoned building, a nonresidential building, or another non-dwelling, or on the street. Exclusion criteria included military dishonorable or other than honorable discharge, current service on active duty in the armed services, and inability to complete the screening questions. All procedures were approved by the University of California Los Angeles and the VA Greater Los Angeles Healthcare System Institutional Review Boards.

The Los Angeles-based Homeless Women Veterans Coordinator assisted with recruitment, which primarily took place during an annual open house for women veterans who are homeless or who have accessed homeless services. This open house draws from a number of outreach sites, including four shelters, one drop-in center for homeless people, six transitional housing programs and residential substance abuse rehabilitation programs, the county jail, several soup lines, and direct outreach to people on the streets in an area in downtown Los Angeles with a high concentration of homeless individuals. The Homeless Women Veterans Coordinator was in contact with approximately 150 women at the time of this study. Participants were recruited between December 2005 and January 2006. Twelve women were recruited for each focus group, with the goal of seating at least 9. A total of 29 women veterans participated in the 3 focus groups.

Procedures

Focus groups were held at a large, urban VA site where a homeless women veterans program is based. All focus groups were moderated and co-moderated by authors IP and DW, respectively. The study used a semi-structured moderator guide, which allowed the facilitators to follow certain topics and open new lines of inquiry when appropriate. The moderator guide elicited general information about participants’ personal contexts (e.g., homelessness history, military experience, perceived needs and priorities), use of VA and non-VA healthcare and homeless services, and perceptions and experiences with VA use. Each focus group participant was each reimbursed $25 and given a gift bag of hygiene supplies valued at $25.

Analysis

All focus groups were recorded and professionally transcribed. We analyzed the interview transcripts using the constant comparative approach (Glaser, 1965). This approach involves four iterative stages: 1) comparing incidents (i.e., discrete narratives or dialogues) within categories (i.e., themes), 2) integrating categories, 3) delimiting a theory for how the categories relate to each other, and 4) writing the theory. In this study, the categories (i.e., themes) of “experiences with social services” and “preferred/desired services” were derived a priori from the interview guide and used as primary (top-level) codes in the data set. Within the “experiences” category, barriers to (e.g., “lack of information”) and facilitators of (e.g., “women-only programs”) service utilization were identified and subcoded. In addition, the theme of “substance abuse” (including pre-, during, and post-military substance abuse) was incorporated into the analysis via axial coding (Boeije, 2002) in order to examine how substance abuse was perceived to be related to homelessness and service utilization. Data coded to these themes was reviewed independently by AH and DW and then discussed and synthesized.

RESULTS

Characteristics of the participants are presented in Table 1. Focus group participants represented a diversity of current age and age at post-military entry into homelessness. No women were currently married or employed, almost 2/3 were disabled, and over half had completed an Associate's Degree or higher. Over women's lifetime, the most commonly used substance (other than alcohol) was cocaine, followed by marijuana. One quarter of participants reported using alcohol two or more times per week in the past year, but no women reported having alcohol or drug problems in the past 30 days. Most women had made use of some type of homeless assistance services within the past 30 days, with highest reported utilization of medical services (76%) and mental health services (59%).

Table 1.

Characteristics of Focus Group Participants (n=29)¹

Age, mean (range), years [n=25]	48 (32-68)
Age at discharge from military, mean (range), years [n=25]	26 (17-45)
Age at first homelessness, mean (range), years [n=17]	36 (17-62)
Race/ethnicity (%) [n=24]
African American	46
White	33
Mixed	13
Hispanic	4
American Indian Alaskan Native	4
Marital status (%) [n=17]
Single	41
Divorced/Separated	47
Married	0
Widowed	12
Education level (%) [n=17]
High school/GED	18
Vocational School	12
Associate's Degree	59
Bachelor's Degree	12
Work status (%) [n=17]
Employed	0
Unemployed	29
Disabled	65
Retired	6
% on probation or parole [n=17]	47
Service connection (%) [n=24]
Non-service connected	58
Branch of military (%) [n=29]
Army	69
Navy	14
Air Force	10
Marines	3
Coast Guard	3
Period of military service (%) [n=29]
Vietnam Era	24
Post-Vietnam Era	45
Period around Persian Gulf War	21
Between Persian Gulf War and 9/11	10
Use of drugs more than 5 times in lifetime (%) [n=17]
Cocaine/crack/free base	71
Marijuana/hashish	53
Self-administered sedatives	35
Self-administered pain-killers	24
LSD or other hallucinogens	24
Frequency of alcohol consumption in past year (%) [n=17]
Never	35
Monthly or less	35
2-4 times per month	0
2+ times per week	24
Use of homeless assistance services in past 30 days (%) [n=17]
Medical services	76
Mental health care program	59
A shelter, mission, or transitional housing program	47
Alcohol or drug treatment program	47
Daytime drop-in center	24
Job or employment services	24
Vocational rehabilitation services	12

¹Data was incomplete for some demographic items, so the sample size for each item is provided.

Perceptions and Utilization of Social and Psychosocial Services

Participants described three main barriers to social and psychosocial services: (1) lack of information about services available to them, (2) limited access to services, and (3) lack of coordination across services. A sense of isolation and abandonment permeated women's descriptions of their experiences of seeking and receiving services that were inappropriate or uncomfortable. In addition, the issue of substance abuse was often interwoven with women's homelessness experiences and help-seeking. Several participants who received appropriate and effective services were eager to share their experiences and concrete information about which services were worth utilizing.

Lack of Information About Services

Participants in all three focus groups expressed a lack of awareness of services available to them as veterans, and in particular, as women veterans. They expressed that this lack of information contributed to their homelessness:

I do not believe [that] I would be homeless and going through the tremendous, tumultuous time I've been going though in the last four years had I heard a long time ago that women's services were available for women Vets, like through the newspaper or the TV. I would have gone to get help a lot earlier.

I wouldn't have even been [homeless]...I needed this help a long time ago. I didn't even know [VA care] was free. I was clueless.

Women were unsure about many aspects of services, e.g., eligibility requirements, location, and terms of use (e.g., duration of stay). As one woman said, “You really have to ask a lot of questions.” Women in the second group discussed their experiences of calling around about services, and having to clarify that they were indeed women, and veterans, and that they needed services “specifically for WOMEN veterans.” Several women in the third group remarked that they did not become aware of any women veterans’ services for years after discharge.

Perhaps not surprisingly given the pervasive lack of information, participants used the focus groups as venues for sharing their experiences and providing one another with details about programs and services that were particularly helpful (or not helpful), such as VA employment services. Four women in the first focus group had been in the same 12-week sexual trauma program, about which the other women were eager to hear. The problem with the program, however, was that it was only 12 weeks, and there was no housing available for women after they graduated.

Nobody gets housing after they finish all this stuff. You have to find it yourself...Many times when you go into recovery, you got to be recovered [in a month], or six months, nine months or a year. And they are not supplying housing.

Participants expressed in all focus groups that their peers—other women veterans—would be the most appropriate people to help homeless women obtain information and negotiate the system of care. As one woman stated, “They need to get women veterans to work in these different positions in order to help other women veterans.”

Limited Access to Services

Access to services was limited by a number of factors: lack of gender-appropriate care, geographic barriers, lack of long-term housing options, and restrictive entry criteria of many programs.

Across all three focus groups, women stated that there were more homeless services for male than female veterans, and, as noted above, that there was little housing available for women and women only.

We all think that they need to make a place for homeless Vet women like us to live in. Be responsible to pay our own rent. To stand up on our own two feet to be independent. We can't find nothing out here...for women and only women.

Women who attempted to access mixed-gender (or what they called “co-ed”) programs often had problems with “fraternization” rules; they had experiences of being viewed as “problematic” because staff felt that they might distract or “entice” men in the program. In the second focus group, five of seven women said that they felt “overlooked” or “ignored” when they tried to get into mixed-gender programs.

I was told that there was a fraternization policy...and I was asked if I was going to [have] a problem with that...I didn't know what they were talking about...There was a nice shelter [available], but they let me fall through the cracks. That made me feel terrible. I already felt like I was a problem, OK? [A problem because you are...?] Because I'm a female!

With regard to lack of gender-appropriate care, women described both psychological and physical safety concerns in mixed-gender programs that they had utilized.

I was raped in the military...Whether you've been in a war zone, whether it's rape, whatever, you suffer from PTSD... and after-care for women is atrocious. Even if you've had drug-related problems, they want you to get [into programs] with 55 men and one female...and lock the doors on you...

A minority of women acknowledged, however, that it was sometimes helpful to be in programs with men, and that above all, women should be able to choose: “Coed for some is good, and not for others. There should be various facilities. Women-specific, male-specific and coed, and there should be more of a choice into where the placement is in regards to that.”

Many participants had to wait to get into specialized programs, such as women veteran's sexual trauma programs. One woman said she had been waiting 10 years for a bed in one of the few well-known sexual trauma programs. Another said she got into what was known to be a good program for homeless veterans, only to discover that she was one of four women being housed with 80 men.

Many programs required clients to have current or past substance abuse, or mental health symptoms. As noted above, no participants reported having alcohol or drug problems in the past 30 days, but several described their experience of feeling compelled to say that they did have such problems, in order to access services.

“In order for homeless female Vets to get help they either got to say they're alcoholic or dope fiends.” “Right! You got to go in there as an alcoholic or you been using drugs. [If] you go in there sober, you're going to definitely get turned back around.” “But you would think they would help a sober veteran first!”

I called all around and the only place they can send me to is sober living homes. [They ask], ‘Do you have a problem with drugs or alcohol?’ [If I say yes], then they can tell me a place to go to.

In the first focus group, participants discussed the lack of services for women without substance abuse or mental illness. The moderator asked, “There's nothing if you're just homeless?” Participants agreed: “They only have winter shelters.” And one woman added,

I was told that I could have seven days of transitional housing unless I went in for my depression. And so that's how I have shelter right now. Now I have to go through and deal with all these issues that may or may not [be a problem for me]. I had to go and break down again what I just tried to build up over the past two years, just to have shelter.

Similarly, women in the second group discussed pretending to have a drug or alcohol problem in order to access services, even though certain program requirements went along with utilizing the services (e.g., attending a mandatory number of 12-step meetings). One woman said, “You got to go to the meetings, you got to go to get your card signed, you've got to do all this. You have no choice if you want housing.” And another responded, “Yeah, but that's okay. It's better than [living] under the bridge, OK? I would have gone anywhere and said anything to get help.” And in the third group, women had the same perception that services mainly focused on individuals with substance abuse problems: “You are either a drug addict or you're an alcoholic. There is no one else. They will do everything for a drug addict or they'll do all this stuff for an alcoholic...but if you're a person that just has a physical problem, forget it.”

In addition, some programs were only available to women with children, in the case of non-VA women's programs; or to women with no children, in the case of VA women's programs. In the second group the moderator asked if it had been difficult to find housing that would allow women to keep their children with them. Participants laughed and one said, “No, but finding housing without children was!” Others went on to say, “No, if you've got children they're going to find you a place to live,” and, “Oh, they're going to find housing for you quick—fast if you got kids.” Some women also described barriers related to being able to stay with their non-veteran lesbian partners: “I've been together with [my partner] for five years and we just came out here [to Los Angeles] together and became homeless the same time, and we have to be in separate places. She's not a Vet so she can't stay where I am, so she's in a place [that's] much, much worse.”

Several women described having to move away from home in order to obtain needed services. These geographic barriers affected women's social support networks, and often prolonged homelessness, unemployment, and vulnerability.

Lack of Coordination Across Services

Participants experienced little coordination between services, for example between screening and receipt of services identified as necessary during screening. Participants noted that they had already had the experience during military service of reporting assault and being ignored, or even worse, being stigmatized and further harassed, so reporting alone was not viewed as sufficient for achieving the desired goals. Indeed women had been encouraged not to report experiences of abuse and violence, to suffer in silence and “get stronger” as a result. This attitude of persevering through adversity continued into women's post-military lives and contributed to substance abuse and a reluctance to report problems and seek help. As one participant stated, “My stint in the military made me even more independent. You don't seek out that help that other women would seek out because they're lost. We don't feel like we're lost. We can do it.”

In addition, women described having to leave temporary shelters without being screened for potential service needs:

In the temporary shelters...they don't do any exit interviews with us. [If they did interviews], they would know if I'm ready or not to go.

Women who had been incarcerated described not being connected to services upon their release into the community.

When women veterans are released from county jail or state prison...they should be referred to a veteran representative...When I was last in a jail, I put in a request after request after request to meet with the veteran representative so that I could set up myself for [housing] after my release. But nothing came.

In addition, women faced challenges with money management and with negotiating across VAs and between VA and county-funded services.

You can't please all those entities at once. If you go to the VA, you go to the VA. If you go to the County, you go to the County. You can't mix the VA with housing assistance. They all don't run together. And no one is on the same page. Now, don't you think those agencies should communicate? Everyone should be on the same page. Each government agency should have the same information, the same availability of services. They should communicate.

Women's sense of isolation and abandonment

Participants consistently expressed a sense of isolation related to their homelessness and to the lack of appropriate and sufficient services.

After the military I felt so lost. I had no self-esteem. I didn't know what to do. I thought everyone hated me. I couldn't go back to my family. I felt I had to just take off somewhere and just isolate myself. I felt so detached from society.

If there was something in [my home state]—somewhere that would be wonderful for me and I wouldn't feel so isolated—I could start working on more of what I need, like finding a place to live, a job...

DISCUSSION

The homeless women veterans we interviewed faced numerous barriers to social services, including lack of information about services, lack of access to services, and lack of coordination across services. Women in all of our focus groups expressed a desire for an integrated source of information on all of the services that are available to them as women, as veterans, and as homeless individuals. They felt that other women veterans would be ideal for outreaching to homeless women veterans. Women who learned of services faced numerous obstacles in accessing many of the services: some services were inappropriate, insufficient, inconvenient, or limited in their entry criteria. Many women had to leave their home towns in an attempt to obtain services, and thereby left sources of social support, employment, and housing. Participants consistently expressed a preference for women-only programs, but found limited availability of these programs, with the few existing programs having long waiting lists. Women agreed that they should at least have a choice as to accessing “coed” or women-only programs. Substance abuse was often a factor contributing not only to prolonged homelessness but also to utilization of services: many women found that they could only access services if they had—or claimed they had—substance abuse problems. For women in recovery from substance abuse, and women with no history of substance abuse, this limitation meant that they often had to “go backwards” in their trajectories toward self-sustainability, just to obtain shelter and needed social services. Throughout these homeless women veterans’ experiences of homelessness was a pervasive sense of isolation, abandonment, and confusion.

Johnson (as reported in Johnson & Cnaan, 1995) suggested well over a decade ago that the service needs of homeless individuals are additive across several types of needs: basic needs, stabilization needs (e.g., needs for professional services such as mental health care), change-oriented needs (e.g., employment, education), and emergency needs. Our findings support this notion of additive needs among homeless women veterans, with added complexity among the stabilization needs in that services available to women veterans are often not prepared for their unique needs, for a variety of reasons (e.g., the VA is not thoroughly equipped to accommodate homeless women veterans with children, and community-based providers may not be equipped to treat homeless women veterans with histories of military sexual trauma and/or combat exposure). Our findings also support the work of Wenzel and colleagues (2001) and others who have found that homeless individuals with a history of substance abuse are more likely to receive treatment than those without this history. Specifically, our data points to the perception among homeless women veterans that they will be more likely to obtain the services they need if they indicate that they have current substance abuse problems. To the best of our knowledge, no other study has revealed that homeless women without current substance abuse problems may have less access to care than those with substance abuse problems.

Gelberg and colleagues (2004) studied homeless women's access and barriers to family planning and women's health care, and found that women faced significant barriers in their efforts to prevent pregnancies. Our study did not reveal this particular set of barriers, but we should acknowledge that women were not directly asked about barriers to family planning; this would be an important direction for future research, especially as increasing attention is being paid to the reproductive health of younger women veterans (Mattocks et al., 2011).

Study Limitations

This study does have notable limitations. First, the sample of focus group participants is not representative of all homeless women veterans. In order to actively and coherently participate, women in the focus groups could not be actively psychotic or under the influence of drugs or alcohol. Thus our findings may be more heavily informed by women without current substance abuse problems. More research needs to be conducted on the ways in which substance abuse (or lack thereof) impacts access to and utilization of homelessness-related services (Nyamathi et al., 1999). Second, our sample included women from a wide range of military service eras, thereby generating a wide spectrum of predisposing characteristics (e.g., beliefs about care in the VA system; see Washington, et al., 2007). Studies of narrower bands of homeless women veterans may reveal more depth as to issues related to particular service eras (e.g., combat exposure in recent conflicts; Street, Vogt, Dutra, 2009). Third, the focus groups took place approximately five years ago, and significant improvements have been made in VA services for women veterans even since the time of data collection. More services are now available for homeless veterans (see below), though we hasten to note that homelessness does not currently appear to be declining among women veterans (US Department of Housing and Urban Development, 2011).

IMPLICATIONS FOR SOCIAL WORK PRACTICE

Ending homelessness has become a top federal priority, with President Obama and the Secretary of Veterans Affairs (VA) pledging to end homelessness among veterans by 2015 by carefully assessing the problem and developing a number of federal programs (see http://www.va.gov/homeless/). For example, there is the National Call Center for Homeless Veterans (877-4AID VET) and the National Coalition for Homeless Veterans (nchv.org); and there are Homeless Veteran Program Coordinators and Women Veteran Program Managers at VA Medical Centers. As of July 2011, US Housing and Urban Development (HUD) and the Department of Veterans Affairs Supportive Housing Program (HUD-VASH) have collaborated with local housing agencies to provide permanent housing for homeless veterans (see http://www.va.gov/HOMELESS/HUD-VASH.asp). In addition, in August 2011, the Department of Labor awarded grants to agencies in 15 states that will provide homeless veterans with job training (see http://www.dol.gov/opa/media/press/vets/VETS20111184.htm), and the VA launched homeless prevention grants that will serve approximately 22,000 homeless and at-risk veteran families as part of the new Supportive Services for Veteran Families (SSVF) program (see http://www.va.gov/homeless/ssvf.asp).

Social work has much to offer in the development of services specifically for homeless women veterans. At an individual level, social workers have expertise in strengths-based care, i.e., care that builds on women's problem solving-skills (Thrasher 1995). Considering that women veterans have likely honed their problem-solving skills through their military service, such an approach could be particularly effective, particularly in enhancing women's self-esteem, which may be compromised after military discharge, and which typically plummets in the context of homelessness (Diblasio & Belcher, 1993). Compatible with the strengths-based approach is the empowerment-oriented approach, another promising avenue for homeless women veterans. Within this approach, practice strategies can be used to encourage homeless women to identify their needs, determine their goals, and set the terms of the helping process (Cohen, 1989). This emphasis on self-determination and autonomy, again, could be particularly appealing for homeless women veterans who pride themselves on their independence and self-reliance. Johnson and Cnaan (1995) have noted that social work care for homeless individuals will often take on more of a technical rather than therapeutic orientation, in that social workers can assist women with meeting their basic needs and getting linked to needed services. These services may differ depending on whether women have children in their custody (Johnson and Kreuger, 1989). Furthermore, at a community level, social workers can train shelter and community-based programs—and VA programs—to teach clients advocacy and self-help skills (Johnson & Cnaan, 1995).

CONCLUSION

As Wenzel, Koegel, and Gelberg (2000) pointed out over a decade ago, homeless women deserve consideration as a distinct group within the homeless population, a group that perhaps exhibits unique ways of becoming homeless, responses to being homeless, and needs to prevent cycling in and out of homelessness. We contend further that homeless women veterans should be considered as a distinct subgroup of homeless women, especially considering their pathways into homelessness (Hamilton, Poza, Washington, 2011), including experiences specific to their military service. Our findings suggest the need for: (1) safe and stable housing for women veterans; (2) greater geographic availability of women-only treatment programs; (3) tailoring of mixed-gender programs to address safety concerns of women veterans and to improve coordination of care; (4) greater attention to the needs of women veterans without substance abuse and/or other mental health problems, and women veterans without children; and (5) interventions that incorporate a peer support and strengths-based or empowerment-oriented approach, as discussed below. There is growing evidence that integrated care (i.e., co-located homeless, primary care, and mental health services) for homeless individuals can improve access to primary care and reduce emergency department utilization (e.g., McGuire et al., 2009), with some indication that gender-specific substance abuse treatment interventions should be included in programs for homeless women with mental illness (Cheng & Kelly, 2008).

Understanding homeless women veterans’ service needs and experiences, and the phenomenological underpinnings of those experiences, is critical to the development of relevant and appropriate services that move women away from vulnerability, into safety. Findings from this study improve our understanding of these issues, and inform actions to begin to address these needs. Future research should be directed toward identifying best practices for implementing these actions in VA clinical settings as well as settings outside the VA where homeless women veterans may seek care.