To ensure that program development was relevant and consistent with the needs and values of our priority population, we established a Steering Committee. This committee consisted of barbers and barbershop owners; an HIV outreach worker; a physician with over 40 years of experience providing care in the community; a pastor; and Black men who live in the community, frequent area barbershops, and represent our priority population. At the time of the study, all Steering Committee members lived in and/or provided services in Brooklyn and had a deep understanding of and connection with the community for whom the program was being developed. The Steering Committee provided suggestions for approaches to recruitment and retention, offered valuable feedback on evaluation measures and intervention components, and worked with us on naming the project. When key objectives for the program were developed, the committee was vocal in guiding the investigators to minimize the burden on the barbers in their delivery of the intervention, on whether intervention delivery should be in multiple sessions or a single session, and in terms of what type of graphics and health education materials would be appropriate for barbers and in barbershops. We also received input from colleagues at the Centers for Disease Control and Prevention.

Our program team developed a set of guiding principles that helped to facilitate intervention development and establish a sense of cohesiveness within our program team. These principles included: a commitment to avoiding reinforcement of stereotypes of Black men; the importance of a strengths-based perspective that acknowledges and capitalizes on men’s roles as providers and leaders in the community; a view of health that incorporates the social determinants of HIV risk along with behavioral risk factors; and development of an approach that leverages the natural roles of barbers as community supporters and advisors. Our goal therefore was not just to change individual-level risk, but also to build community empowerment in support of better overall health.

The study was conducted in partnership with barbershops situated in the neighborhoods of East Flatbush, Flatbush, and Bedford-Stuyvesant. These neighborhoods have some of the highest rates of HIV incidence among Black men in New York City (New York City Department of Health [NYCDOH], 2010). For the formative phase of intervention development, men were recruited from four barbershops and completed study activities both in the barbershop (a brief quantitative risk assessment) and off-site at either SUNY Downstate or at AAIUH (focus groups and individual interviews); in-depth analyses of these qualitative data are reported elsewhere (Taylor et al., in press). For the pilot test of the program, we recruited an additional seven barbershops (eleven in all) and completed program activities in a small group format either at participating barbershops or in offices at SUNY Downstate or AAIUH.

To inform the content of the program, we utilized a sequential, mixed methods approach in which eligible men first completed a brief quantitative risk assessment delivered in barbershops. Eligible participants included men ages 18–45, who identify as Black or African American, and who reported at least one female partner in the past 3 months. Among 122 men screened, 86% (N = 105) were eligible and agreed to participate. Participants completed audio computer-assisted self-interviews (ACASI) in barbershops, using privacy screens and headphones to protect participant privacy. From the men who completed the ACASI, we then purposively selected a subgroup of men for participation in focus groups and individual interviews. Men were ineligible to participate in the focus groups and interviews if they reported injection drug use in the past 3 years, oral or anal sex with a man in the past 5 years, or involvement in an HIV or substance use prevention study in the past 6 months. Sixty of the 105 were eligible for focus group and individual interviews based on these criteria. We then used data from the ACASI to purposefully select a subset of 22 of these 60 men to represent diversity in age, levels of sexual risk, and country of origin. Of the 22 participants selected, 8 chose to participate in focus groups only, 1 in an interview only, and 13 in both focus group and individual interviews. Participants completing the ACASI were provided a $20 gift certificate that could be used toward services at the participating barbershop. Focus group participants were provided $60 plus a $10 barbershop gift certificate; participants in the in-depth interviews were also provided $60 plus the $10 certificate.

Topics included masculinity, partner selection and networks, sexual risk behaviors, attitudes, beliefs and intentions regarding HIV-testing and condom use, myths and stigma surrounding HIV, and suggestions for program development. The focus group questions targeted a general understanding of these topics among Black heterosexual men, while the individual interviews focused more specifically on the participant’s individual experiences.

Qualitative data were transcribed, subjected to thematic analysis, and coded by one member of the SUNY team and three members of our AAIUH team. AAIUH team members were trained in qualitative research methods as part of this project. Themes linked to risk included low HIV transmission knowledge, low perceived HIV risk, higher emotional attachment/trust with a partner, heat of the moment/impulsive decision making, and concerns regarding discussing safer sex, particularly with long-standing partners when involved in concurrent sexual relationships. Participants also spoke of their perceptions of masculinity and their roles as protectors and gatekeepers in the community, the need for HIV prevention activities that originate from trusted sources in the community, and the expressed willingness to advocate for change. Details on the methods and key findings from the formative phase have been previously described (Taylor et al., in press).

We also conducted a focus group with five barbers and barbershop owners to gain insights on what could be feasibly implemented and housed within barbershops. Although we had originally thought that barbers would serve as interventionists for the program, both the barbershop focus group findings and feedback from our steering committee identified the need not to burden the barbers, but to instead utilize their trusted roles in the community as a conduit to the program through referrals, to provide pared-down messages surrounding HIV prevention and communicative strategies to discuss HIV naturally within the context of a client/barber interaction, and to reinforce messages provided in the BTWB program. We also gained insight into the fact that barbers wanted and would need considerable support in terms of health education to be able to speak accurately and with confidence to clients if they were to provide a supporting role in program activities.

As part of our collaborative process, the formative program team utilized an Intervention Mapping process (Bartholomew, Parcel, & Kok, 1998) to develop a matrix of objectives for the program, select strategies and approaches to meeting objectives, and develop the intervention components. At each of these steps, we engaged core Intervention Mapping processes of reviewing previous research, incorporating theory, and applying the formative findings. We applied Social Cognitive Theory (Bandura, 2001, 2004) as a theoretical basis for program development, given that there are clear intrapersonal factors (e.g., condom use attitudes, condom use self-efficacy) and features of the social environment (e.g., community empowerment, HIV stigma) that impact HIV risk among Black heterosexual men, and these factors influence one another continuously and dynamically over time (Poundstone, Strathdee, & Celentano, 2004). The concept of a reciprocal relationship between individual characteristics and the environment in which risk occurs is a central tenet of Social Cognitive Theory. We therefore developed program components that not only sought to improve attitudes and build skills around safer sexual behavior but that also aimed to reduce HIV stigma and improve collective efficacy for community changes related to HIV risk. HIV stigma toward persons with HIV/AIDS has been shown to be a factor that can mitigate safer sex and HIV testing outcomes in prevention trials (Mahajan et al., 2008). The entire program was iteratively pre-tested and modified prior to pilot testing in focus groups reflecting our priority population.

The BTWB program takes place in a single group session, divided into three modules—Module 1: Wake Up, Module 2: Gear Up, and Module 3: Build Up (Table 1). The program takes less than two hours to administer. Each module provides opportunities for skills building and feedback and includes written support materials. We used the concept of a “fire” as a metaphor to underpin the modules, moving from the example of community efforts to put out real fires to introducing the concept of public health problems as less visible fires that both impact the community and that can be addressed by the community. HIV is then discussed as one particular fire that engulfs many Black communities and that can be extinguished through social action and personal risk reduction.

The Wake Up module promotes community responsibility and motivation to engage in health promotion with men and women in the participant’s social network. In this module, participants are oriented to the program objectives and ground rules for participation. Participants discuss the theme of the module, which is to wake up to the fire of local community health problems and of HIV specifically; the fire metaphor seeks to demonstrate to men how their natural inclination to take care of others (i.e., how they would react if they witnessed their next-door neighbor’s house on fire) can be applied in the service of other, less visible problems. During this module, participants engage in discussions of how they view themselves as men in our society, and explore the idea of how different conceptualizations of masculinity may influence men’s health risk or protective behaviors (e.g., whether it is seen as a weakness or vulnerability to go to a doctor, how being a protector of others necessitates being healthy). Discussions of health also focus on social networks, and how to comfortably leverage these networks to improve community health. During this module we discuss men’s ability and motivation to address health, and discuss a movement in Brooklyn toward control of the community’s own health.

The Gear Up module introduces a specific fire of HIV that is damaging the community’s health. The module objectives are to promote understanding of HIV prevention and increase perceptions of the threat of HIV, while also increasing outcome expectancies and self-efficacy related to condom use to reduce that threat. This module addresses the local epidemiology of HIV/AIDS, highlighting the burden of heterosexually transmitted infection in the area, and includes discussions about the role of community stigma as a barrier to effectively battle HIV/AIDS. The module also includes a section on truths and misconceptions about HIV transmission. The discussion of HIV transmission addresses modes of HIV transmission and correction of misconceptions that arose during the formative work (e.g., HIV is not transmitted during anal sex between men and women; all HIV transmissions are the result of men’s drug use or same sex behavior; a “quickie” without a condom does not put you at risk for HIV). Different types of condoms and lubricants are reviewed and instructions and demonstrations are provided on the proper use of condoms. This module incorporates methodologies that have been shown to be standard components of effective HIV interventions (Albarracin et al., 2005), but also integrates messages that emerged from our formative research that speak specifically to local concerns and beliefs surrounding condom use, HIV testing, and HIV transmission.

The third module, Build Up, prepares participants to take what was learned in the prior modules and share it with sexual partners and others in their social network. The objective of this module is to provide men with the motivation and skills to share and communicate effectively the information that they learned about HIV with their sexual partners as well as throughout their community through exercises and feedback (Patterson, Grenny, McMillan, & Switzler, 2002). This module includes role play with scenarios that were developed based on stories of risk that were presented during our formative work. These involve varied physical and emotional contexts of risk, including sexual risk as a function of emotional motivations (e.g., anger, loneliness), high-risk interactions (e.g., sexual pressures from female partners to not use condoms, a theme which came up frequently in our formative work), and high-risk settings (e.g., when there is limited perceived time or opportunity to have sex, in clubs, etc.). As part of this module, men are asked to develop and commit to a step that they could take to reduce their risk of HIV or that could help reduce risk for someone they know. Participants are encouraged to maintain contact with the program team after completion of the module in order to have additional questions answered and to receive support and encouragement.

To successfully implement the BTWB program, we needed to ensure that barbers were prepared to deliver HIV messages and refer clients to the BTWB program (e.g., it is important to know your HIV status, it is important to protect yourself and partner by using condoms). On-site trainings with participating barbers and other barbershop personnel were conducted within each participating barbershop. The trainings consisted of an overview of the project and methods, participation in the BTWB intervention, and training on methods to encourage their clients’ participation in the program. In addition, each barber was provided with a workbook, summarizing key information from the BTWB program. Barbers were also provided with T-shirts, buttons, and barber capes branded with the BTWB logo, and condoms to distribute at the barbershops. Throughout a barbershop’s participation in the BTWB program, barbers received ongoing support from field staff.

Of the 11 shops whose owners originally agreed to participate, we completed training at 10 (one shop owner dropped out after determining that he did not have the capacity to house program activities). At those 10 shops, we trained 30 of the 39 employees. Barbers were compensated $40 for their participation in the training.

Following training at each barbershop, we began pilot recruitment. To increase the likelihood that our recruited sample represented the population of barbershop clients, we varied recruitment times at sites to include weekday and weekend shifts, each with a daytime and evening recruitment period. Customers were introduced to study staff by their barber and asked whether they would be willing to meet with our research assistant to screen for the study. The research assistant described the study and assessed eligibility via an ACASI screening form. All eligible participants who were interested in participating provided individual, written informed consent. Men were eligible for participation in the pilot study if they reported on the ACASI screening form that they were 18–45 years old, self-identified as an African American or Black man, and reported unprotected vaginal or anal intercourse with two or more women in the past 3 months. Men were excluded if they could not complete study activities in English, were HIV-positive, reported any injection drug use in the past 3 years, reported any oral or anal sex with a man in the past 5 years, or participated in any HIV or substance use research in the past 6 months (including participation in the formative phase of our study). We excluded HIV-positive men due to the different types of messages that are offered in HIV primary prevention interventions, versus those that prioritize HIV-positive participants (e.g., medication adherence, etc.). Participants were provided $20 for participation in the BTWB program; they also received $20 for completing the baseline assessment and $30 for the follow-up assessment. In addition, participants were reimbursed up to $20 for transportation costs to and from the study site. All procedures, including those of the formative phase, were approved by the Institutional Review Boards at SUNY Downstate

Following the consent procedure, participants completed an ACA-SI-administered baseline assessment, participated in the BTWB program, and then completed a 3-month follow-up ACASI assessment; the window for the 3-month follow-up was 11 to 14 weeks from baseline. The baseline assessment was completed either at the barbershop on the day of screening or on the day of the BTWB program. The BTWB program was administered either at the barbershop during a time that it was closed to customers or in a private room within the Arthur Ashe Institute of Urban Health. To reduce bias, project staff responsible for recruiting and administering evaluation assessments did not administer program activities. Our main outcome measure was unprotected vaginal or anal sex with a female partner over the past 3 months. To create this variable, we asked men to report whether they had any main partners in the past 3 months, and if they did to report whether there was at least one time when they did not use condoms during vaginal or anal sex. These questions were repeated for casual partners (Lansky, Thomas, & Earp, 1998). Those who reported having at least one unprotected episode with either main or casual partners were coded as having unprotected vaginal or anal sex. We also asked men about whether they had engaged in unprotected sex with two or more women in the past 3 months. The 3-month time frame for recall of behaviors is short enough to minimize memory distortions, but long enough to allow for an accurate picture of sexual behavior over time (Jaccard & Wan, 1995).

We also assessed a set of mediating variables, as defined by Social Cognitive Theory. To assess attitudes toward condom use, we asked men to report how favorable they felt about always using condoms with main female sexual partners in the next 3 months; 7-point response options ranged from extremely unfavorable to extremely favorable; this question was repeated for casual sexual partners. The responses to these two questions were then combined such that those who reported being extremely or moderately favorable to always using condoms in the next 3 months with both main and casual partners were categorized as having favorable attitudes toward condom use, as compared to those who reported being only slightly favorable, unsure, or responding in the unfavorable range with either main or casual partners. To assess self-efficacy for condom use, we asked men to report how confident they felt about always using condoms with their female main sexual partners in the next 3 months; 7-point response options ranged from extremely unconfident to extremely confident; this question was repeated for casual partners. Those who reported that they were extremely or moderately confident in their ability to use condoms with both main and casual partners were categorized as having adequate self-efficacy for condom use, as compared with those who reported being slightly confident, unsure, or reporting a lack of confidence with either main or casual partners. We assessed variables related to community empowerment using a 5-item Likert scale developed and validated by Barbara Israel and colleagues (Israel, Checkoway, Schulz, & Zimmerman, 1994); higher scores on the scale indicate higher empowerment (range = 6–25; Cronbach’s alpha among respondents at baseline = 0.73). In addition, we assessed HIV-related stigma using another widely implemented and validated 11-item Likert scale (Darrow, Montanea, & Gladwin, 2009; Herek, 2002; Herek, Capitanio, & Widaman, 2002), with higher scores indicating higher stigma (range = 11–55; Cronbach’s alpha at baseline = 0.89). We conducted a paired-samples t-test for the comparisons of stigma and empowerment in the pre- and post-test period, and McNemar’s test for assessments of changes in behavioral measures and for the assessments of condom use attitudes and self-efficacy.