In most parts of the world, unprotected sex continues to be the leading cause of HIV infections [1]. In the United States, approximately 56,000 new HIV infections were diagnosed in 2006, and over 80% were attributed to unprotected sex [2, 3]. High-risk heterosexual contact accounted for 31% of the new cases, whereas unprotected sex among men who have sex with men accounted for 53% of the new cases [2, 3]. In the absence of an effective vaccine, protective behavior such as consistent and correct condom use continues to be the most effective strategy to reduce HIV transmission risk among sexually active heterosexual and MSM populations.

There are various existing approaches that promote condom use among people at high risk for sexual transmission of HIV. Individual-level interventions (ILI) and group-level interventions (GLI) directly address knowledge, attitudes, skills, and behaviors related to condom use in persons participating in intervention activities in one-on-one settings or in existing (e.g., couple, family) or newly formed groups. Community-level interventions (CLI) also directly and indirectly address knowledge, attitudes and behaviors with the focus on the entire community and often with a strong emphasis on changing social norms [4]. Evidence suggests that ILIs, GLIs, and CLIs demonstrate moderate to high success in promoting condom use [5]. Disseminating these types of interventions, however, may not be sufficient in combating the HIV epidemic, since they do not address the social and structural system or environments in which individuals reside. These contextual factors have been found to be a critical influence on individuals’ behaviors [6].

Structural-level interventions (SLIs) are particularly attractive in HIV prevention efforts since they are designed to address contextual factors that impact personal risk for HIV. Blankenship et al. proposed three structural-level intervention components that address behaviors of interest in public health: availability, accessibility, and acceptability [7]. When applied to HIV prevention tools such as condoms (see Table 1), SLIs target availability by making establishments responsible for the behaviors of their clients or for providing settings or materials that would reduce client risk [7]. Examples include placing condom bowls in clinics or 100% condom use policies in brothels. SLIs that target accessibility typically do so by making condoms and prevention messages widely accessible (e.g., massive distribution of free condoms at multiple venues simultaneously) without focusing on a specific establishment or client type. Increasing accessibility to condoms is a particularly attractive option for reaching marginalized populations in an effort to compensate for social inequities in access to prevention tools. SLIs target acceptability by changing social norms towards condoms (e.g., public service announcements about benefits of using condoms). These three SLI components can be directed to individuals, organizations, or the environment (Table 1). Combinations of these SLI components are likely to positively influence condom use behaviors since they are implemented in settings where individuals reside, an aspect frequently overlooked by existing ILIs, GLIs, and CLIs [7].Table 1

Examples of condom distribution as structural interventions for HIV prevention

Previous reviews have focused mainly on evaluating ILIs, GLIs, and CLIs that address condom promotion (e.g., teaching individuals how to use a condom, promoting a positive attitude toward condoms, giving participants condoms to use after intervention) rather than larger-scale distribution [8, 9]. Johnson et al.’s meta-analysis of HIV risk reduction interventions among adolescents found that educational, psychosocial, or behavioral interventions advocating sexual risk reduction for HIV prevention were significantly more efficacious among intervention participants compared to participants in the control arm [8]. Although 18% of these interventions provided condoms as part of the intervention, nearly all also included provision of condoms for control participants as well. Findings from Smoak et al.’s meta-analysis showed that safer sex programs that provide condoms do not inadvertently increase sexual frequencies [9]. While those reviews are informative, the interventions evaluated in the reviews were limited to the small scale delivery (individual or group) and did not directly focus on or test structural-level approaches (i.e., improving availability, acceptability, and accessibility) in condom distribution programs.

The goals of this systematic review are to summarize the available research literature evaluating structural-level condom distribution interventions (SLCDIs), to meta-analyze quantitative data to assess the overall efficacy of SLCDIs on HIV-risk sex behaviors and sexually transmitted infections (STIs), and to assess whether structural-level intervention components (availability, accessibility, and acceptability) are associated with intervention efficacy. We also examine whether the level of implementation of SLI components (i.e., individual, organizational, or environmental) and the presence of any combination of additional ILI, GLI, or CLI components has a differential impact on the efficacy of SLCDIs. Additionally, we test whether intervention effects are applicable to various risk groups and to studies conducted in different countries or with different study designs.

We identified 21 SLCDIs (Fig. 1; Table 2) [18–35]. Two-thirds of the studies were conducted in international settings (k = 14) [24–35], including 5 in Africa (2 in Tanzania, 2 in Cameroon, 1 in Ghana), 4 in East & Southeast Asia (2 in China, 1 in Indonesia, 1 in Thailand), 2 each in the Caribbean and Mexico, and 1 in Central America. Among the 7 studies conducted in the U.S. [18–23], two either specifically targeted (100%) or consisted of majority of African Americans (71.5%) and one specifically targeted Latinos; 2 studies consisted of majority of whites (75.4 and 51%) and the remaining 2 studies included mixed ethnic/racial groups.Fig. 1

Systematic search and study selection

This meta-analysis of 21 U.S. and international studies indicates that structural-level condom distribution interventions (SLCDIs), evaluated as a whole, are efficacious in reducing various HIV sex-risk behaviors and incident STIs. The magnitude of intervention effects observed here are comparable, and in most cases, stronger, than those reported in other meta-analyses that examined intervention effects on HIV risk among various populations [36–40]. Risk reduction of this magnitude is also well within the range considered to be cost-effective when translated into final health outcomes [41].

Several characteristics of SLCDIs contribute to intervention efficacy based on our analysis of condom use behaviors. For example, interventions consisting solely of a structural component (SLI only) are efficacious, while interventions that combine SLIs with additional individual, small group or community-level activities show even greater efficacy. One possible reason for the increased efficacy of combining multi-level intervention components is that these different modalities address various aspects of influencing factors (e.g., norms, knowledge, skills, motivation and access) as well as prevention needs of individuals in affected communities. Additionally, interventions that increased the availability of condoms, or increased accessibility to condoms, as a distribution strategy were efficacious in increasing condom use behaviors. Interestingly, no intervention used acceptability as the only strategy as acceptability was addressed in combination with availability and/or accessibility. There were also no SLCDIs that were implemented at the environmental level only. SLCDIs implemented at the individual level were found to be efficacious as well as SLCDIs implemented at all three levels (individual, organizational, and environmental).

With regard to applicability, it is very encouraging that SLCDIs promote condom use in various populations, such youth, adults, commercial sex workers, high STD populations and males. Our review further shows that SLCDIs targeting youth generally combined the SLI component with ILI/GLI components and focused on increasing the availability of condoms. Making condoms available at venues frequented by youth combined with ensuring that important prevention messages are relayed through individual and small group sessions is likely to result in greater condom use behaviors. Most importantly, our overall findings also show that SLCDIs have protective effects on sexual initiation among youth.

Commercial sex workers also benefitted from condom distribution programs. These interventions, from the international literature, targeted sex establishments (e.g., sex brothels) by trying to change the social and structural environment, in addition to making condoms more available or accessible. This finding is not too surprising because programs that aim to increase accessibility are designed to compensate for social inequities by making products more available for marginalized populations who may not have equal access to those resources (e.g., massive distribution of free condoms). In addition, creating an enabling environment through supportive policies and social norms around condom use may empower sex workers to refuse unprotected sex.

Several limitations of this meta-analysis warrant comment. Some of the more notable structural-level condom distribution programs in international settings were included in this review (e.g. 100% condom campaign in China [35]), however, several similar programs were not. A few 100% Condom Use Programs that have been conducted across many countries in Asia were excluded for various methodological reasons [42–46]. These reasons included study design issues (e.g., not reporting both pre-intervention and post-intervention data from independent cross-sectional samples or using a separate comparison group) or not providing sufficient data for meta-analyses (e.g., not providing sample sizes or variance estimates or citing unpublished data). These 100% Condom Use Programs have been successfully implemented throughout Asia over the years, and, in general, they have resulted in increases in condom use by commercial sex workers (CSWs) or their clients over time and declines in HIV/STD prevalence in the CSW community or the population as a whole [42–46]. The findings in our review are consistent with these results and we believe not including these did not bias our overall conclusions. Based on the framework proposed in our review (Table 1), these programs usually have addressed all three components (availability, acceptability, and accessibility) at all three levels (individual, organizational, and environmental) which also supports the notion that multi-level or multi-component condom distribution programs may have the strongest efficacy. Some common elements of these programs that may play a role in their success include providing political support, focusing on owners and managers of sex work establishments (e.g., brothels, bars), conducting public campaigns or social marketing campaigns to normalize condom use, making condoms more available or accessible, and obtaining organizational support [43, 47].

The current review is limited to English-language publications. A post hoc supplemental search using the same search protocol as described in the “Method” section indicated that only a handful of foreign language citations out of almost 300 identified were potentially relevant to this review. Upon further examination, only two were identified as new studies that would potentially meet eligibility criteria [48, 49]. Since we did not have the capacity to translate the full text of these reports to confirm eligibility, we believe our search missed at most two studies due to the language restriction during our review period. Based on the abstracts, both studies support condom distribution for HIV prevention.

We reviewed the HIV prevention literature through September 2007. Since completing this review, we have identified several newly published evaluations of condom distribution programs in other international settings and with different populations (e.g., men who have sex with men) that may be relevant to our review [50–58]. The recent emphasis by UNAIDS on national level comprehensive HIV prevention approaches aimed towards universal access to prevention, treatment, care and support for all those in need [59] has led to evaluations of these types of programs in various parts of the world [53–58]. Many of these comprehensive HIV prevention approaches have included wide-scale condom promotion and distribution in combination with other HIV/STD prevention efforts, such as STD care or sexual health care. Our findings do support the 2007 UNAIDS Guidelines Toward Universal Care in the UNAIDS, which recommends universal and uninterrupted condom availability and integrated condom promotion into other health services as part of a comprehensive HIV prevention approach [59]. These recently published studies could build upon the evidence summarized in this review to better understand what works, how it works, in what settings, and with what populations.

We observed moderate to high levels of heterogeneity among studies, even after conducting stratified analyses. We initially attempted to conduct a multivariate random-effects metaregression analysis including covariates for the type of structural component, level of implementation, and additional intervention components in an effort to identify independent predictors of intervention effects. Given the small number of studies in our review, and the concern for multicollinearity across variables, the findings were difficult to interpret and, thus, not presented. The studies included in this review are heterogeneous in terms of target group, intervention components, level of implementation, and study design. Although the heterogeneity was not substantially reduced with the univariate stratified analyses, our findings do support overall efficacy of SLCDIs and point to important directions for future consideration.

First, the remaining heterogeneity does suggest that additional factors related to efficacy, beyond those considered in this review, may exist. Our findings suggest that addressing availability, accessibility, or acceptability, as key barriers to condom use, does increase condom use. However, these programs do employ a variety of delivery methods, implementation strategies, or operational techniques in order to promote and distribute condoms. Questions related to implementation, such as how condoms were distributed, who distributed condoms, where they were distributed, duration of the program, types of social marketing, who delivered messages, additional services provided, etc., were beyond the scope of this review. Exploring these types of questions in future reviews can help to improve program implementation.

Second, condom distribution programs can significantly impact condom use behaviors among at-risk populations (e.g., youth, adults), as well as high-risk populations (e.g., commercial sex workers). Given the efficacy of these programs on condom use as well as STD incidence, future research should explore how other high-risk populations that are disproportionately affected by HIV and other STDs, such as African Americans, MSM and those in correctional facilities, may also benefit from such programs. It would also be important to include cost analyses to determine how the programs can be most effectively and efficiently delivered.

Finally, more than half of the studies included in our review were conducted in international settings, and those interventions were found to be significantly more efficacious than those conducted in the U.S. This finding suggests that much can be gained by understanding what makes those condom distribution programs more efficacious. Several evaluations of structural-level condom distribution programs, particularly in the international setting, have been published since conducting this review that could build upon the evidence summarized here to better understand the differences between the U.S. and international programs. Identifying effective strategies in one setting could also inform how to adapt and implement programs for different target populations or in new settings.

This systematic review supports the structural-level condom distribution intervention as an efficacious approach to increasing condom use and reducing HIV/STD risk. Given the urgency of the HIV epidemic, making condoms more universally available, accessible, and acceptable, particularly in communities or venues reaching high-risk individuals, should be considered in any comprehensive HIV/STD prevention program. Further exploration around how best to implement condom distribution programs to maximize their reach and impact should be considered.