Assessing Domestic Violence Shelter Workers Views and Practices Pertaining to HIV Prevention Services for Women Residing in Domestic Violence Shelters

Abstract

There is a need for studies to assess domestic violence (DV) shelter workers views about brief HIV prevention interventions for shelter residents to improve these workers’ provision of HIV prevention interventions to shelter residents. This mixed methods study assessed DV shelter workers’ views about the following: (a) the need for and appropriateness of HIV prevention services within DV shelters, (b) the utility (i.e., acceptability, systems support, understanding, and feasibility) of an HIV Risk Assessment and Safety Plan (HIV RASP) for women in DV shelters, and (c) suggested changes to or concerns about using the HIV RASP. Workers from DV shelters located in the 10 states in the United States with the highest rates of HIV reviewed the HIV RASP and answered survey questions about it including the Usage Rating Profile–Intervention (URP-I) Questionnaire and two open-ended questions. Although workers felt it was appropriate to provide HIV prevention interventions within DV shelters, only 23% reported that HIV prevention interventions had ever been implemented at their shelter and only 42% had provided residents with educational brochures about HIV prevention. Workers generally agreed that the HIV RASP was acceptable, understandable, and feasible. They somewhat disagreed about their ability to implement the tool independently. Findings suggest that little progress has been made in engaging DV shelter workers in HIV prevention efforts for residents during the past decade and reveal ways to improve the HIV RASP and overcome barriers to implementing it. The study findings may be used to help reduce gaps between the science and practice of HIV prevention for abused women.

There is a need for domestic violence (DV) shelter workers to integrate brief HIV prevention interventions into routine services provided to women residing in DV shelters. Women who experience intimate partner violence (IPV) are at high risk for contracting HIV (Kouyoumdjian, Findlay, Schwandt, & Calzavara, 2013; Li et al., 2014; Siemieniuk, Krentz, & Gill, 2013). Yet evidence-based interventions for preventing HIV among this high-risk population of women are lacking (Prowse, Logue, Fantasia, & Sutherland, 2014). DV shelters are ideally suited to provide HIV prevention services for abused women (Cavanaugh, Campbell, Braxton, Harvey, & Wingood, 2016) because they are community-based and provide services for hundreds of thousands of abused women across the United States each year with 40,000 women and children sheltered on a typical day in 2015 (National Network to End Domestic Violence, 2017). Yet we are aware of only three HIV prevention interventions designed specifically for women residing in DV shelters (Cavanaugh et al., 2016; Johnson, Johnson, Beckwith, Palmieri, & Zlotnick, 2017; Rountree, Bagwell, Theall, McElhaney, & Brown, 2014). Residents in the treatment conditions of these studies had significant improvements in HIV knowledge (Rountree et al., 2014) and condom use self-efficacy (Cavanaugh, Campbell, Whitt, & Wingood, under review) as well as condom use (Johnson et al., 2017). However, there were no statistically significant differences in outcomes between the treatment and control groups in one study (Rountree et al., 2014), one study lacked a control group (Cavanaugh et al., under review), and none of these studies involved a randomized, control design to determine the efficacy of the interventions (Cavanaugh et al., under review; Johnson et al., 2017; Rountree et al., 2014). Only one of these interventions was delivered entirely by shelter staff to shelter residents (Cavanaugh et al., under review). Brief HIV prevention interventions could be integrated into routine services DV shelter workers provide for shelter residents.

A number of factors may affect DV shelter workers’ implementation of HIV prevention interventions for abused women including these workers’ views about the need for and appropriateness of HIV prevention interventions within DV shelters (Durlak & DuPre, 2008; Proctor et al., 2011) as well as the acceptability and feasibility of brief HIV prevention interventions for shelter residents (Proctor et al., 2011). However, we are aware of only two studies that have assessed DV shelter workers views about the need for HIV prevention interventions in DV shelters and we are unaware of any published studies that assessed shelter workers’ views on the appropriateness, acceptability, and feasibility of such interventions. Previous reports have found shelter workers did not disseminate HIV/AIDS information to residents, speak with residents about HIV risk-reduction strategies (Rountree, Goldbach, Bent-Goodley, & Bagwell, 2011), nor did they implement HIV prevention interventions in their shelter (Cavanaugh et al., 2016). These studies assessed DV shelter workers in only two states, and one study was based upon data collected 11 years ago (Rountree et al., 2011). Thus, it is imperative to assess the current need for HIV prevention in more DV shelters across the United States, especially among those that are located in areas with high rates of HIV. The aims of the current study were to assess the following views of workers from DV shelters located within the 10 states in the United States with the highest rates of HIV (Centers for Disease Control and Prevention [CDC], 2017): (a) the need for and appropriateness of HIV prevention services within shelters, (b) the utility of an HIV Risk Assessment and Safety Plan (HIV RASP) that was developed and integrated into a 6-hr, adapted evidence-based intervention for women residing in DV shelters (Cavanaugh et al., 2016; Cavanaugh et al., under review), and (c) suggested changes to or concerns about using the HIV RASP with residents in DV shelters.

Method

Participants and Procedures

The CDC identified the following 10 states in the United States as having the highest number of HIV diagnoses: Florida, California, Texas, New York, Georgia, Illinois, Maryland, North Carolina, New Jersey, and Pennsylvania (CDC, 2017). DV shelters within those 10 states in the United States were identified through an online directory of DV services across the United States. The shelter listings compiled were then compared with state government or DV coalition websites for verification.

After receiving approval from the appropriate institutional review committee on the protection of human participants, trained research assistants contacted 532 DV shelters by phone. Assistants either spoke with the directors or left a voicemail for them explaining the study protocol and requesting assent to send study materials by email. Assenting directors were asked to complete the study materials and forward the survey link to case managers and counselors who would be in a reasonable position to discuss HIV prevention with shelter residents. The survey took approximately 30 min to complete and consisted of workers providing informed consent, reviewing the HIV RASP tool, and answering survey questions.

The HIV RASP was developed to assess abused women’s risks for contracting HIV and facilitate related safety planning; It takes 30 to 60 min to implement and was included in a 6-hr, adapted evidence-based intervention for preventing HIV for women in DV shelters (Cavanaugh et al., 2016; Cavanaugh et al., under review). The HIV RASP consists of four parts. The first part includes psychoeducation about abused women’s consensual and nonconsensual (e.g., sexual assault) risks for contracting HIV as well as mental health and substance use problems that are both common among abused women and associated with risky behavior and sexual assault. The second part involves an assessment of abused women’s consensual and nonconsensual risks for contracting HIV. The third part consists of a personalized safety plan that asks women to identify harm reduction strategies to reduce their risks for contracting HIV. The fourth part consists of information about local service providers to assist women with implementing harm reduction strategies, for example, mental health treatment, pre-exposure prophylaxis (PrEP), HIV/sexually transmitted infection (STI) testing. For the purposes of this study, shelters workers were asked only to review the HIV RASP and specifically told not to implement it with residents given that the tool is still under development.

Each time workers participated in the survey, their shelter was entered into a drawing and one shelter was randomly selected to receive US$1,000. Data were collected between January and May of 2016. See Table 1 that indicates the number of shelters contacted per state and the number of workers who completed the survey by state for each of the three study aims. As shown, the final analytic sample for study Aims 1, 2, and 3 consisted of 95, 76, and 50 DV shelter workers respectively. For study Aims 1, 2, and 3, the number of responses from each state varied from 2 to 16, 0 to 12, and 0 to 11, respectively. Due to the small number of workers who answered these questions by state, study findings are reported for the study sample as a whole and not by state.

Table 1.

Flowchart of Shelters Contacted and Shelter Workers Who Completed Survey Questions by State.

	CA	FL	GA	IL	MD	NC	NJ	NY	PA	TX	Any
Number of shelters identified	117	41	35	22	22	70	16	97	42	70	532
Number of shelter directors who agreed  to participate in study	14	13	15	6	8	19	5	15	15	12	122
aNumber of shelter workers (shelters) in  the final analytic sample for Study Aim 1	20(2)	5(3)	8(7)	12(4)	10(3)	7(7)	10(2)	9(7)	16(9)	16(5)	95(49)
bNumber of shelter workers (shelters) in  the final analytic sample for study Aim 2	0	4(2)	6(5)	9(4)	10(3)	6(6)	8(2)	9(7)	12(6)	12(5)	76(40)
Number of shelter workers (shelters) in  the final analytic sample for study Aim 3	0	4(2)	3(3)	4(3)	6(2)	3(3)	5(2)	7(7)	11(4)	7(5)	50(31)

Note. RASP = Risk Assessment and Safety Plan.

^aNinety-seven workers answered questions about the need for HIV services, but two respondents had missing data and were removed from the final analytic sample for study Aims 1 and 2.

^bSeventy-eight workers answered questions about the usability of the HIV RASP, but two respondents were missing data on more than two of the related subscale items and were removed from the final analytic sample.

Measures

Need for HIV services at the shelter.

The following three questions were used to assess the need for HIV prevention interventions within the DV shelters: (a) Have you ever implemented an HIV prevention intervention at your shelter? (b) Do you currently have HIV prevention services at your shelter? and (c) Do you provide shelter residents with educational brochures about HIV prevention? Response options were yes/no (α = .76).

Appropriateness of HIV prevention interventions in DV shelters.

Workers were asked to indicate the extent to which they agreed or disagreed with the following four statements: (a) HIV/AIDS prevention is an important issue for residents at my shelter, (b) shelters are an appropriate place to address HIV/ AIDS, (c) shelter involvement in HIV/AIDS prevention could have positive impacts on residents’ health, and (d) shelters should not address HIV/AIDS risk with residents. Responses were rated on a 6 point Likert-type scale ranging from 1 (strongly disagree) to 6 (strongly agree). Responses to the fourth question were reverse scored so that higher responses for each of these four questions reflected stronger views about the appropriateness of HIV prevention in DV shelters. The four items were combined to yield a total score for the appropriateness of HIV prevention in DV shelters (α = .73).

Resident referrals.

Workers were also asked to indicate whether they had local contacts with and/or regularly referred clients to any of the following six services or organizations: (a) mental health screening and treatment, (b) substance abuse treatment, (c) methadone or suboxone dispensaries, (d) needle exchange programs, (e) STI/HIV testing and/or treatment, and (f) free condom distribution. They also had the option to select none of the above.

Usability of the brief HIV RASP.

Shelter workers were asked to review the brief HIV RASP that was developed for women residing in DV shelters and answer 33 questions about the tool usability that were adapted from the URP-I scale (Chafouleas, Briesch, Riley-Tillman, & McCoach, 2009). The language used throughout the original measure includes references to children, parents, and classroom settings. Adaptations of this instrument included replacing references to “child” with “resident” and the statement “parental collaboration is required in order to use this intervention” with “collaboration with professionals outside of this shelter is required in order to use this intervention.” Two items were also removed from the scale because they were not relevant to the current study. Responses were rated on 6-point Likert-type scale ranging from 1 (strongly disagree) to 6 (strongly agree). Some items were reverse scored. Mean scores were calculated for the following four URPI subscales: Acceptability (α = .95), Understanding (α= .90), Feasibility (α = .88), and Systems Support (α = .75). Items for the System Support subscale were scored so that high scores on all subscales indicated more favorable responses. An overall score was also obtained by summing the mean scores from the four factors (α = .84) with higher scores indicating greater usability.

Suggested changes to the instrument and concerns/reservations about using it.

Participants were also asked the following two open-ended questions about the HIV RASP: (a) What changes would you suggest for the HIV RASP? and (b) What concerns or reservations would you have about using the HIV RASP with residents?

Data Analysis

Scale reliabilities were evaluated, and mean scores for the URPI subscales were computed for respondents who were missing no more than two items of the subscales. Descriptive statistics were used to describe the data. The subscale means were interpreted in relation to the scale response anchors.Thematic analysis was used to code responses to the open-ended questions about the HIV RASP (Braun & Clarke, 2006). The first and fourth authors independently reviewed the open-ended responses to the first question and identified brief discourse units to represent the content of participants’ statements. Themes and subcategories representing the discourse units were identified along with the number of participants reporting the same theme/ subcategories. The two coders then discussed their coding schemes together. They identified 12 themes, eight of which were identified by both coders, yielding an interrater agreement of 67%. The raters discussed the themes/subcategories and agreed upon seven themes and nine subcategories. The same process was followed for coding responses to the second question. However, the raters each identified eight categories, seven of which both raters identified as the same themes found for Question 1 (78% reliability) and both raters agreed to two new themes yielding nine themes and 13 subcategories. Both raters then independently reanalyzed all participant responses to both questions using the nine themes and 17 subcategories identified and later discussed findings together. Discrepancies were resolved after discussion and agreement about the most accurate way to capture the content. Excerpts from participant responses to the open-ended questions were used to illustrate the themes and subcategories that emerged. Each of the direct excerpts from participants who are reported below are followed by the participants identification number.

Results

Participant Demographics

The average length of time participants had worked in the field of DV services was 9.44 years, and the average time in their current shelter was 7.02 years.

AIM 1: Need for and Appropriateness of HIV Prevention Interventions Within DV Shelters

Table 2 shows that fewer than one quarter of workers (23.2%) reported an HIV prevention intervention had ever been implemented at their shelter. Slightly fewer (21.1%) reported their shelter was currently providing HIV prevention services. Less than half (42.1%) reported they provide residents with educational brochures about HIV prevention. The most common service contacts/referrals were for mental health treatment (94.7%), and the least common service contacts/referrals were for needle exchange programs (4.2%; Table 2). As shown in Table 3, the mean item scores indicate that shelter workers somewhat agreed to agreed that HIV/AIDS prevention is an important issue for residents at their shelter and that shelters are an appropriate place to address HIV/AIDS. Average item responses indicated that shelter workers agreed to strongly agreed that shelter involvement in HIV/AIDs prevention could have a positive impact on residents and that shelters should address HIV/AIDS risk with residents.

Table 2.

Domestic Violence Shelter Workers Reports of HIV Prevention Services and Referrals Provided to Domestic Violence Shelter Residents by State (n = 95).

	n	%
HIV prevention services
Ever implemented an HIV prevention intervention at your shelter?	22	23.2
Currently have HIV prevention services	20	21.1
Do you provide shelter residents with educational brochures about HIV  prevention?	40	42.1
Resident referrals
Mental health treatment	90	94.7
Substance abuse treatment	82	84.2
Methadone/suboxone	19	20.0
Needle exchanges	4	4.2
STI/HIV testing	40	42.1
Free condoms	30	31.6
None of the above	1	1.0

Note. STI = sexually transmitted infections.

Table 3.

Domestic Violence Shelters Views About the Appropriateness of HIV Prevention Services for Shelter Residents (n = 97) and the Utility of the HIV RASP for Shelter Residents (n = 76).

	M	SD	Scale Response Anchors
Appropriateness (n = 97)
1. HIV/AIDS prevention is an important issue for residents at my shelter	4.47	1.29	Somewhat agree (4) to agree (5)
2. Shelters are an appropriate place to address HIV/AIDS	4.94	1.05	Somewhat agree (4) to agree (5)
3. Shelter involvement in HIV/AIDS prevention could have positive impacts on resident’s health	5.10	1.04	Agree (5) to strongly agree (6)
4. Shelters should address HIV/AIDS risk with residents	5.25	0.80	Agree (5) to strongly agree (6)
5. Total appropriateness	4.94	0.78	Somewhat agree (4) to agree (5)
Utility of HIV RASP (n = 76)
Acceptability	4.49	0.94	Somewhat agree (4) to agree (5)
Understandability	4.59	0.74	Somewhat agree (4) to agree (5)
Feasibility	4.13	0.86	Somewhat agree (4) to agree (5)
Systems Support	3.03	0.77	Somewhat disagree (3) to somewhat agree (4)
Total Score	4.05	0.58	Somewhat agree (4) to agree (5)

Note. RASP = Risk Assessment and Safety Plan

AIM 2: Usability of HIV RASP

The subscale means shown in Table 3 indicate that shelter workers somewhat agreed to agreed that the HIV RASP was acceptable, understandable, and feasible. The mean rating for systems support was close to the scale anchor of “3,” which indicates that workers somewhat disagreed that they would be able to implement the tool independently. The total URPI score indicates that workers somewhat agreed to agreed that the HIV RASP was usable.

AIM 3: HIV RASP Suggestions, Concerns/Reservations

The following nine themes emerged from the participants’ responses to the two open-ended questions about suggested changes to the instrument and concerns or reservations about using it: (a) none, (b) duration, (c) content, (d) resources, (e) inclusivity/adaptability, (f) practices and procedures, (g) client concerns, (h) facilitator concerns, and (i) strengths and 17 corresponding subcategories (see Table 4).

Table 4.

Domestic Violence Shelter Workers’ Suggested Changes to the HIV RASP and Concerns About the Tool (n = 50).

Themes and Subcategories	Suggested Changes n (%)	Concerns/ Reservations n (%)
1. None	22 (44)	13 (26)
2. Duration	10 (20)	7 (14)
a. Too long	9 (18)	6 (12)
b. Other (i.e., will take longer when implemented)	1 (2)	1 (2)
3. Content	13 (26)	7 (14)
a. Language	8 (16)	2 (4)
b. Clarity needed	3 (6)	1 (2)
c. Trauma informed	3 (6)	1 (2)
d. Other (e.g., talk about all STI’s to make more   palatable to clients)	7 (14)	4 (8)
4. Resources	5 (10)	10 (20)
a. Available	1 (2)	--
b. Needed	4 (8)	10 (20)
5. Inclusivity/adaptability	10(20)	2 (4)
a. LGBTQ	5 (10)	1 (2)
b. Gender	2 (4)	--
c. Adaptable	3 (6)	1 (2)
6. Practices and procedures	3 (6)	14 (28)
a. Current	1 (2)	4 (8)
b. Needed	3 (6)	10 (20)
7. Client concerns	12 (24)	18 (36)
a. Victim blaming	6 (12)	2 (4)
b. Causing client to experience adverse feelings	4 (8)	12 (24)
c. Client safety	2 (4)	3 (6)
d. Other	4 (8)	4 (8)
8. Facilitator concerns	1 (2)	13 (26)
9. Strengths	11 (22)	9 (18)

Note. RASP = Risk Assessment and Safety Plan; STI = sexually transmitted infections; LGBTQ = lesbian, gay, bisexual, transgender, and queer/questioning.

No suggestions or reservations/concerns and duration.

Forty-four percent of the 50 participants had no suggested changes to the instrument, and 26% reported having no concerns or reservations about using the instrument. Duration was mentioned by 20% for the question about suggested changes. The duration theme was comprised of two subcategories—too long and other. All of the respondents noted that the instrument was too long, except two who mentioned that it would take more time to implement the tool than the study suggested.

Content and resources.

One in four respondents suggested a change to the instrument’s content, and 14% mentioned a concern or reservation about the instrument’s content. The content theme was comprised of the following four subcategories that captured respondents’ input about the language, clarity, making the content more “trauma informed,” and other comments. Responses coded for the content subcategories language, clarity needed, and trauma informed included statements such as “I am not sure many of the women would use the word risky to describe long-time partners (#7),” “Not clear on the differences of pressured and coerced sex? (#42),” and “Make the questions more trauma informed (#49)” respectively. A response coded under the “other” content category suggested expanding the instrument content to refer to all STIs to make the instrument more “palatable to clients (#24).” A resource theme also emerged that was comprised of two subcategories—available and needed. Several respondents mentioned resources that they would need to implement the tool including onsite HIV testing, crisis counseling for clients, local resources to help newly diagnosed HIV cases, training, and participant incentives. Resources available to support the implementation of this tool included a health specialist on staff and urgent care clinic in a shelter.

Inclusivity/adaptability and practices and procedures.

A theme emerged about inclusivity and adaptability that was comprised of three subcategories— LGBTQ (lesbian, gay, bisexual, transgender, and queer/questioning), gender, and adaptability. Twenty percent of respondents suggested that the instrument should be more inclusive of LGBTQ residents and men, and adapted to individual client personalities and needs. Another theme about practices and procedures emerged that was comprised of two subcategories—current and needed. For example, one participant coded under the needed practices and procedures subcategory noted,

While the safety plan is great for HIV prevention, for some of the ladies it will be difficult and dangerous for them to implement a good portion of it if they are still dealing with their abuser. For this reason I think that additional safety planning should also be done at the same time to try and minimize the risks of using the HIV safety plan (#1).

Other participants coded in this subcategory mentioned needing to establish rapport with a client prior to administration and maintain client confidentiality, finding the right way to approach the topic and engage participants, and administering the instrument after a client has settled into the shelter. An example of a response coded under the current subcategory for practices and procedures was “Unless they [clients] brought up concerns about STI/HIV or sexual practices, we do not ask about these things specifically (#13).”

Client and facilitator concerns.

Client concerns was one of the most common themes for both questions and the theme consisted of the following four subcategories: victim blaming, causing clients to experience adverse feelings, client safety, and other. Several respondents mentioned concerns about the instrument being victim blaming and/or causing clients to experience adverse feelings including discomfort, guilt/shame, and self-blame. For example, one worker coded for the subcategory victim blaming said, “I also do not like the use of the phrase the following things I have chosen to do because it implies they had a choice in it which with abuse victims this is often not the case (#13).” Concerns about client’s safety were also mentioned. For example, one participant coded in this subcategory said, “Using a female condom without telling the partner is very dangerous for our clients (#49).” An example of a response coded as “other” questioned whether clients could “absorb the content of this information (#8).” Another theme focused on facilitator concerns. For example, one participant coded in this theme stated, “The assessment seems like more paperwork and less interaction (ID #2)” and another mentioned concerns or reservations about “the overall success rate.”

Strengths.

Finally, approximately one fifth of the responses to both questions mentioned perceived strengths of the instrument such as educational information, comprehensiveness, and ability to facilitate conversations with clients. One respondent said, “The safety plan could be easily implemented and I believe would be a useful tool for some clients and would probably open up conversations with clients that may otherwise not disclose some of their experiences (#12).” Others wrote, “I love the tool and we would easily be able to use it and other intervention materials (#50)” and “There are no concerns about using the HIV RASP with clients. In this line of work we empower the client, nothing is more powerful than knowledge (#48).”

Discussion

This study assessed DV shelter workers’ views about the needs for and the appropriateness of HIV prevention interventions in DV shelters, and their views about a brief HIV RASP assessment tool for women in DV shelters. Findings indicated that although DV shelter workers feel that addressing HIV is appropriate in DV shelters, very few shelter workers reported having implemented HIV prevention, providing HIV education to residents, or referring clients regularly to HIV prevention services. These findings are similar to those involving data that were collected 11 years ago (Rountree et al., 2011) and suggest that despite the substantial scientific literature about the intersection of HIV and IPV (Kouyoumdjian et al., 2013; Li et al., 2014; Siemieniuk et al., 2013), little progress has been made in engaging DV shelter workers in HIV prevention efforts for residents during the past decade. However, some workers expressed a great deal of support for HIV RASP and felt as if it could and should be implemented, and 22 respondents had no suggested changes to the tool.

Shelter workers identified barriers to implementing the HIV RASP including the need for support and resources to help implement the tool (e.g., onsite HIV testing, training, and crisis counseling), and both client and facilitator concerns. Respondents expressed concerns about negative effects the tool may have on residents including the danger residents may face when implementing some suggested HIV prevention strategies. One in four workers felt the tool could cause residents to experience negative feelings and was victim blaming. These findings suggest that there are more barriers to implementing HIV/AIDS prevention services by DV shelter staff (e.g., causing the resident to experience distress or harm), than those reported in another study (e.g., lack of training, need for additional staff and funding; Rountree, Pomeroy, & Marsiglia, 2008). As addressing HIV among shelter residents may result in residents revealing they are HIV positive, a number of other related ethical issues warrant consideration. Such considerations include keeping residents’ HIV status confidential, not limiting shelter services among residents who are HIV positive, and helping HIV-positive residents access HIV care and overcome obstacles to engaging in HIV care that women who experience IPV face (Hatchera, Smoutc, Turand, Christofidesb, & Stocklc, 2015). In addition, disclosure of women’s HIV status to current and former sexual partners may be needed to inform those individuals about their risks for acquiring HIV. However, this is complicated among abused women who may fear disclosure will lead to IPV (Hatchera, 2015).

Practice Implications

DV shelters should implement policies requiring that 100% of their workers who work directly with residents (i.e., case managers and therapists) and 100% of shelter residents receive educational information about (a) the intersection of IPV and HIV, (b) local HIV prevention service information, and (c) local HIV treatment services. For example, each case manager and therapist working with residents could be given the brochure about the intersection of IPV and HIV from the CDC that was created for social workers and therapists (CDC, 2014). Using this information, shelter staff could ask residents about their HIV risks, encourage residents to seek HIV testing (Cavanaugh et al., 2016), and integrate educational discussions about PrEP and HIV testing into health and safety plans. Recent studies demonstrate that women experiencing IPV find PrEP an acceptable HIV prevention strategy (Willie, Kershaw, Campbell, & Alexander, 2017). Shelters could also consider developing strong collaborations with organizations to deliver in-house HIV testing and counseling for their residents because it is conveniently accessed by women in a supportive environment during a time when many women are making life changes (Draucker et al., 2015). Shelter directors could also obtain information from local health departments about local HIV prevention services and HIV treatment services and that information could be provided to all residents at the shelter. All shelters should consider implementing other evidence-based behavioral interventions for preventing HIV among shelter residents, particularly DV shelters located in high-risk areas of the 10 states in the United States with the highest rates of HIV (i.e., California, Florida, Texas, New York, Georgia, North Carolina, Illinois, Louisiana, Pennsylvania, and New Jersey), although such efforts would require financial support to implement. We are aware of three HIV prevention interventions designed specifically for women in DV shelters (Cavanaugh et al., 2016; Cavanaugh et al., under review; Johnson et al., 2017; Rountree et al., 2014) that were described above and warrant consideration, but none of these studies have been rigorously evaluated to determine their efficacy in preventing HIV among residents.

Research Implications

More research is needed to understand barriers such as those reported by DV workers in this study and other factors affecting shelter workers views about the appropriateness, acceptability, and feasibility of tools such as the HIV RASP. The present study did not collect data about workers demographics including their racial/ethnic backgrounds, age, or the degree to which they are personally affected by HIV (such as knowing individuals outside the shelter who are HIV positive), which may affect their views about the HIV RASP. Research is also needed that assesses DV shelter residents’ views toward HIV prevention interventions including their views about a version of the brief HIV RASP that incorporates some of the improvements suggested by shelter workers (e.g., making the tool more inclusive for diverse residents, for example, LGBTQ and male residents). Such studies should also examine whether residents’ age, race/ethnicity, and other cultural variables affect their views about the HIV RASP to better understand how this instrument is viewed among diverse DV shelter residents.

Although workers identified a number of ways the HIV prevention tool evaluated in this study may be improved, 22 of the 50 workers (44%) who responded to the open-ended questions about the HIV RASP had no suggested changes for the instrument, and workers generally agreed that the HIV RASP was acceptable, understandable, and feasible. One in five workers also identified a number of the HIV RASP’s strengths including the educational information provided and the tool’s ability to help facilitate conversations between workers and residents. These findings suggest that the HIV RASP was acceptable although it may be improved based upon the feedback received.

Study limitations need to be noted. The response rates were low and therefore the findings may not be generalized to shelter workers in the states assessed. Many factors may have influenced the poor response rate, including requirements for statewide coalition approvals prior to shelters participation in the study as well as limited time or incentive for staff to complete the survey. Study strengths also warrant attention. To our knowledge, this is the largest study of DV shelter workers’ views toward HIV prevention and data were collected from workers in the 10 states in the United States with the highest rates of HIV.

Conclusion

We addressed gaps between science and practice of HIV prevention for abused women by soliciting DV shelter workers’ views about a brief HIV RASP that was developed for women residing in DV shelters : (Cavanaugh et al., 2016; Cavanaugh et al., under review). The findings underscore the need to utilize DV shelter workers in HIV prevention for shelter residents, suggest the HIV RASP is acceptable with some modifications, and provide suggestions for improving the instrument.

Author Biographies

Courtenay E. Cavanaugh is an associate professor in the Department of Psychology at Rutgers University in Camden, New Jersey. She received her PhD in clinical psychology from the California School of Professional Psychology and completed two National Institute on Drug Abuse–funded postdoctoral fellowships at Yale University and Johns Hopkins University. Her research, which has been supported with internal and external grant funding, is focused on improving the health and well-being of diverse and marginalized women and children affected by interpersonal violence.

Jenna Harvey completed both her bachelor’s and master’s degrees in the Department of Psychology at Rutgers University, Camden, New Jersey. She is interested in dissemination and implementation science and improving the health of high-risk individuals in the community.

Kamila A. Alexander is an assistant professor in the School of Nursing at Johns Hopkins University. Her research focuses on prevention of sexual health outcome disparities and the complex roles that structural determinants such as intimate partner violence, societal gender expectations, and limited economic opportunities play in the experience of intimate human relationships. She has received funding to conduct research investigations through the National Institutes of Health, American Nurses Foundation, Sigma Theta Tau International, and the Society for the Scientific Study of Sexuality.

Samantha Saraczewski recently completed her bachelor’s degree in psychology at Rutgers University and is now attending a master’s program in counseling psychology at the University of Pennsylvania. Her interests include clinical and research activities with underserved populations.

Jacquelyn C. Campbell is a professor in the School of Nursing at Johns Hopkins University. She has been principal investigator of 12 National Institute of Justice, National Institutes of Health, Department of Defense, and Centers for Disease Control and Prevention funded research studies of IPV and homicide and more than 220 articles and seven books on the subject.