Stigma has been implicated in poor outcomes of human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) care in clinical and community settings (Mukolo, Villegas, Aliyu & Wallston 2013). Stigma reduction is regarded as key to the long-term success of HIV/AIDS prevention and treatment efforts (Pulerwitz, Michaelis, Lippman, Chinaglia & Diaz 2008; Roura, Urassa, Busza, Mbata, Wringe & Zaba 2009; Sengupta, Banks, Jonas, Miles & Smith 2011). Stigma is generally sustained by a complex set of factors that are not easy to address (Alonzo & Reynolds 1995; Corrigan, Watson & Barr 2006; Goffman 1963; Green, Davis, Karshmer, Marsh & Straight 2005; Logie & Gadalla 2009; Mahajan, Sayles, Patel, Remien, Sawires, Ortiz, et al. 2008; Major & O'Brien 2005). While it is known that some people are more vulnerable to stigma than others, it is not clear what accounts for variance in vulnerability to stigma in general as well as in specific settings (Mukolo, Heflinger & Wallston 2010). Stigma reduction strategies that work in some contexts (e.g. Western nations) might not work in other contexts, e.g. resource limited, linguistically and culturally diverse regions in sub-Saharan Africa (Mutalemwa, Kisoka, Nyigo, Barongo, Malecela & Kisinza 2008; Nyblade, Stangl, Weiss & Ashburn 2009; Pulerwitz, Michaelis, Weiss, Brown & Mahendra 2010).

Literature describing HIV/AIDS stigma is extensive, but accounts of stigma reduction are few (Brown, Macintyre & Trujillo 2003; Heijnders & Van Der Meij 2006; Sengupta et al. 2011). A recent review points to lack of dedicated stigma reduction interventions and good quality efficacy assessment studies (Sengupta et al. 2011). Therefore, more descriptions are needed to highlight the diversity and efficacy of stigma reduction interventions that are being tried and tested in relation to HIV/AIDS. While progress has been made to develop theoretic models to assist in the identification and classification of anti-stigma strategies (Heijnders & Van Der Meij 2006; Holzemer, Uys, Makoae, Stewart, Phetlhu, Dlamini, et al. 2007; Mahajan et al. 2008; Nyblade et al. 2009), there is a need to develop and document strategies informed by comprehensive models of stigma, covering dimensions of stigma that tend to be overlooked, such as internalized and institutional stigma (Sengupta et al. 2011). As noted by Sengupta et al. (2011) there is need for studies and/or interventions whose primary goal is to reduce stigma and for such studies to address issues peculiar to target populations and the context in which those populations experience stigma.

In Mozambique, the need to address the negative consequences of stigma is widely acknowledged and supported among associations of people living with HIV/AIDS (PLWHA), government agencies and nongovernmental organizations (NGOs) involved in HIV/AIDS care. However, to the best of our knowledge, there are no reported systematic (rigorously evaluated and published) studies of HIV/AIDS stigma reduction in Mozambique. Therefore, the domains of stigma that characterize the problem of HIV stigma in Mozambique are not widely reported in the literature and little is known about variance in the manifestation of stigma across socio-geographic contexts, more so between rural and urban settings. Furthermore, interventions to reduce HIV stigma in Mozambique appear limited in scope, most notable are mass media campaigns (TV and bill board advertisements and the use of drama and theatre), the enactment of anti-discrimination legislation by the national government in 2002 and 2009 (UNAIDS 2013), and indirectly, through the scale-up of antiretroviral treatment programs for HIV infected patients (Pearson, Micek, Pfeiffer, Montoya, Matediane, Jonasse, et al. 2009). There is also need for a comprehensive theoretic framework to guide the development and critique of context-specific anti-stigma strategies in Mozambique.

We describe one attempt at generating consensus on critical characteristics of HIV stigma and anti-stigma interventions suitable for Zambézia Province, Mozambique, a region that has been impacted by the HIV epidemic and has been targeted for the scale-up of anti-retroviral treatment (ART) since 2006. For example, HIV prevalence in Zambézia Province is estimated at 12.6% among adults 15–49-year-old and 15.3% among women vs. 8.9% among men (INSIDA 2009). This represents a decline in prevalence over time, since in 2004 the adult prevalence for the central region of Mozambique in which Zambézia Province is located was estimated at 20.4% (INSIDA 2009). We describe key dimensions of stigma that a diverse group of stakeholders identified and some of the stigma reduction interventions presumed most suitable for Zambézia Province.

Outputs have been synthesized into a logical framework and are being turned into an implementation plan. The following eight consensus positions were captured.

Consensus one: Stigma has declined over time in tandem with and corresponding to the history of the HIV/AIDs epidemic and of the scale-up of public health interventions in Zambézia Province, Mozambique.

The key changes have been in the toning down of negative public attitudes toward the HIV infected – more precisely consensus is that the uninfected are less inclined to enact their stigmatizing attitudes and behaviors and that tolerance for PLWHA has increased in the general population. Key drivers of reductions in the levels of stigma have been (a) acquisition of knowledge that most known modes of HIV transmission can be managed/controlled to prevent HIV infection, boosting confidence that risk of contagion can be minimized and enhancing prevention self-efficacy and, (b) knowledge that efficacious therapies exist and thus the HIV infected (even those with advanced HIV disease) need not die from infection or experience poor quality of life. The stigma of HIV infection/disease is mitigated when the dominant (socially held) image and prognosis of HIV disease changes from life-threatening to chronic disease and through decrement in fear and anxiety about HIV infection among both the infected and uninfected. For detailed description of the link between the HIV stigma and the natural history of HIV disease, see Alonzo and Reynolds (1995) and Brown et al. (2003). The change is attributed to national level interventions to reduce stigma, notably legal reforms and informational stigma reduction strategies (i.e. mass media-delivered public education campaigns). As has been noted in relation to other anti-stigma initiatives (Brown et al. 2003), we do not know how deep or superficial these perceived changes have been and which domains of stigma have been impacted the most. Therefore, we need properly designed studies that are dedicated to tracking measurable changes in public stigma over time in Zambézia Province.

The scale-up of efficacious anti-retroviral therapies, to the extent that it changes the image of HIV disease from a life threatening condition (extremely stigmatizing due to existential threat) to a chronic illness (less stigmatizing and induces hope due to the probability of clinical remission and enhanced quality of life) interferes with the stigma trajectory. Hence, we considered the scale-up of combination ART (cART) as a clinical or biomedical stigma reduction strategy, particularly in high HIV prevalence contexts like Zambézia Province. The strategy might be particularly effective where there are highly visible treatment experienced patients who have experienced dramatic clinical remission following cART commencement.

Nonetheless, we were cognizant that stigma reduction might not occur in a linear fashion. For example, it has been noted in some contexts that cART efficacy can have both abating and incremental effects on HIV stigma (Roura et al. 2009). Roura et al. (2009) observed that improvements in the health and functioning of persons on long-term cART engenders positive attitudes toward HIV disease among the non-infected, but does not change dominant (lay) attributions about HIV transmission and the belief that the HIV infected are fundamentally flawed and incapable of changing their risky and socially deviant behaviors. cART efficacy likely makes HIV infection invisible (i.e. purges externalized bodily characteristics by which to tag the infected), making it challenging to apply lay criteria for screening the HIV infected and exerting social control over them. As noted by Roura et al. (2009), cART efficacy might create new bases for social anxiety about HIV infection in some settings. In Fig. 2, a similar outcome is captured in the cognitive map of one of the key interviewees.

Consensus two: Levels of perceived stigma are likely to be greater than enacted stigma because of the intensity of public stigma enacted in earlier phases of the epidemic and the gravity of the consequences of stigma experienced in these earlier times.

This is partially indicated by a near pervasive fear of sero-status disclosure and considerable investments in the management of information about one's sero-status and use of HIV-related services (Alonzo & Reynolds 1995). The discrepancy might also reflect the presence of internalized stigma. As noted in the methods section, internalized can lead to profoundly negative cognitive and behavioral consequences such as self-esteem decrement, psychological distress and maladaptive coping. Therefore, residual fear of public stigma is likely to mask real positive changes in public attitudes and behavior toward PLWHA in Zambézia Province. Such deep-seated fear (Brown et al. 2003) might also indicate the prevalence of maladaptive coping with public stigma among some PLWHA or those who worry about HIV infection risks. There is need, therefore, to investigate this assertion because, if true, one might expect generational disparities in the discrepancy between perceived and enacted stigma – with lesser discrepancy and greater willingness to utilize HIV services among younger than older persons in Zambézia Province. The underlying psychological impact of public stigma might require that counseling and support strategies for reducing stigma be implemented in combination with informational strategies (Heijnders & Van Der Meij 2006).

Consensus three: As a direct product of observations in consensus one and two, social withdrawal and secrecy (theoretically consistent consequences of public/community stigma) might be maladaptive responses to public stigma, since public stigma has declined over time in Zambézia Province.

The persistence of such responses indicates potential internalization of public stigma by PLWHA. According to the self-stigma model (Corrigan et al. 2006), victims of public stigma might differ in their responses to stigma by the extent to which they consider public stigma to be legitimate. Variance in perceived legitimacy primarily reflects differences in the tendency to concur with negative stereotypes widely held in one's social milieu. According to this theory, those who consider public stigma to be illegitimate tend to be represented among persons who reject negative social stereotypes and respond to the public stigma with righteous anger – via protests, litigation and other forms of communicating disapproval and seeking redress for damages (Corrigan et al. 2006). Therefore, maladaptive coping with stigma in this context is likely among PLWHA who concur with negative public stereotypes. Internalized stigma might be indicated by the elevated levels of psychological distress often observed in samples of HIV infected persons, either pre-and post-HIV testing or among the treatment experienced (Arendt 2006; Atkinson, Higgins, Vigil, Dubrow, Remien, Steward, et al. 2009; Berger, Schad, von Wyl, Ehlert, Zellweger, Furrer, et al. 2008; Burack, Barrett, Stall, Chesney, Ekstrand & Coates 1993; Eller, Corless, Bunch, Kemppainen, Holzemer, Nokes, et al. 2005; Gore-Felton & Koopman 2008; Ickovics, Milan, Boland, Schoenbaum, Schuman & Vlahov 2006; Kaharuza, Bunnell, Moss, Purcell, Bikaako-Kajura, Wamai, et al. 2006; Leserman 2008; Owe-Larsson, Sall, Salamon & Allgulander 2009). Therefore, screening for psychosocial distress might help identify PLWHA who are at risk of internalizing public stigma in Zambézia Province. Psycho-educational anti-stigma strategies (Brown et al. 2003; Sengupta et al. 2011) should be focused on those at risk of psychosocial distress, helping them acquire knowledge of alternative (externalized) responses to public stigma (such as advocacy/protest, strategic sero-status disclosure and help seeking). In Mozambique, PLWHA have the opportunity to belong to voluntary associations of PLWHA and healthcare-provider-initiated peer-support groups. Some of these social networks have a strong anti-stigma agenda while others are focused on non-stigma aspects of HIV/AIDS care such as ensuring fidelity to care and ART through improvements in logistics and treatment self-efficacy. Some local support groups are coupled to complex social networks that have far-reaching social and legal reform agendas pursued at multiple geo-political levels. These social networks provide potential avenues for developing theory-driven, targeted and context-specific interventions that help build up positive psychological resources (including coping skills), mitigating self-stigmatization and its negative consequences. Obtaining the support of popular opinion leaders in local communities would be critical to the success of such interventions.

Consensus four: Male participation in maternal and child health programs, coupled with stigma focused educational interventions, are likely to reduce the disparity in knowledge of sero-status and women's vulnerability to stigma.

Despite consensus one, two and three above, enacted stigma still exists in some segments of Zambézia Province, as evidenced by participants’ recall of recent (publicized) instances of physical and verbal abuse and neglect of persons with advanced disease that culminated in acute hospitalization and, in some cases, suicide. We have not been able to confirm the suicide claims. However, the consensus is that the burden of enacted stigma is disproportionately skewed toward women of reproductive age as they are more likely than men to be tested for HIV infection, e.g. via maternal and child health interventions to prevent mother-to-child transmission (PMTCT) of HIV that are being scaled-up by the Ministry of Health in the province. As advocated elsewhere in the Southern African region (de Paoli, Manongi & Klepp 2004), increasing male involvement in PMTCT programs is likely to moderate the degree of stigma enacted toward women in the province.

Consensus five: Notwithstanding consented national-level legal and social norm reforms, institutional stigma, as reflected by negative attitudes and behaviors of healthcare workers and discriminatory policies and practices of healthcare institutions, persists. Institutional stigma is likely the most significant form of HIV stigma in Zambézia Province and is probably the single most significant psychosocial barrier to the utilization of HIV services.

Institutional stigma is one of four types of stigma highlighted, based on deliberations about who is being stigmatized and the source (or context) of stigma – the other three being public stigma, courtesy/associative stigma and self-stigma (Andrewin & Chien 2008; Mukolo et al. 2010). Institutional stigma was further subdivided into professionals’ attitudes and behaviors and organization policies and operational procedures. The consensus was that, at the level of healthcare professionals, institutional stigma cannot be explained solely by lack of factual knowledge about HIV transmission, prevention and treatment, but more likely by unintended biases in healthcare professionals’ beliefs and behaviors toward HIV infected patients and/or inadvertent consequences of implementing current care delivery practices. The assertion is partially supported by the fact that no targeted systematic stigma reduction interventions have been developed for healthcare settings in Zambézia Province. Elsewhere it has been noted that, while there is an increasing willingness to treat HIV infected patients, factual knowledge of HIV transmission routes and poor infection control practices might inadvertently increase the fear of contagion among health workers (Brown et al. 2003). Hence, interventions are needed among healthcare professionals to improve stigma awareness and enhance the quality of provider-patient interaction.

Consensus six: Stigmatizing and discriminatory standard operating procedures (SOPs) in healthcare institutions are a barrier to the uptake of HIV testing, care and treatment services.

Many participants, particularly those enrolled in care narrated specific instances of enacted stigma or stigma triggers they experience in the process of accessing and utilizing existing care. These had to do with policies and standard operational procedures implemented at healthcare facilities. Most representatives of the healthcare providers that participated in the consensus-building exercise seemed unaware of these institutional aspects of stigma. In addition, systems to facilitate patient input in healthcare quality improvement were poorly developed and most patients ascribe to a culture of deference to authorities, particularly in rural settings. Similar observations have been made in other countries (Mahendra, Gilborn, Bharat, Mudoi, Gupta, George, et al. 2007; Nyblade et al. 2009). Nyblade et al. (2009) reviewed studies aimed at mitigating stigma in healthcare facilities and noted provider unawareness of what stigma looks like as well as lack of effective ethics/norm enforcement mechanisms needed to deviantize (Schur 1983)¹ stigma enacted toward patients.

While acknowledging that treatment efficacy has a positive impact on public attitudes and behaviors, study participants considered current guidelines on cART commencement and HIV care as favoring persons with advanced HIV disease, which inadvertently perpetuates the negative image of HIV disease, i.e. most people in Zambézia Province only know about advanced forms of HIV disease. Policy shifts toward earlier cART commencement and pre-ART care that emphasize health maintenance at early stages of HIV disease, would reinforce the positive image of HIV disease that emanates from treatment efficacy. The public health significance of such policy changes might be an increase in the number of HIV infected persons who present early for HIV diagnosis.

The consensus was that reducing institutional stigma requires a partnership between healthcare providers and HIV infected patients because, on one hand, patients are better placed to note unintentional bias in SOPs and personal attitudes and behaviors that are oblivious to providers, while providers can facilitate patients to take a more active role in ensuring the success of clinical/bio-medical approaches to reducing HIV stigma. Hence, the consensus was to adopt a partners-in-health approach to address institutional stigma in healthcare settings.

Consensus seven: Lack of adequate information about HIV, stigma and the legal rights and entitlements of PLWHA are still the main drivers HIV/AIDS stigma in all settings (public/community, self and institutional) in Zambézia Province.

This might not be unique to Zambézia Province or Mozambique as similar observations have been made in some regions of neighboring South Africa (Campbell, Nair, Maimane & Nicholson 2007). Under consensus one, reductions in externalized forms of public stigma at the national level are attributed to better knowledge about HIV transmission routes and preventions. However, emerging findings from a province-wide survey of health-related attitudes and behaviors indicates that most adults in Zambézia Province have limited knowledge about HIV transmission and prevention (Vergara, Blevins, Vaz, Manders, Calvo, Arregui, et al. 2010). As many as 70% of those surveyed in the province could not provide one correct mode of HIV transmission or stated they did not know how HIV is transmitted (Vergara et al. 2010). A link was also reported between willingness to negatively label and socially exclude PLWHA and low levels of HIV transmission knowledge in data from the same survey (Mukolo, Blevins, Victor, Vaz, Sidat & Vergara 2013). Although further investigations of these observations and assertions are needed, public health education interventions to improve knowledge of facts about HIV transmission and prevention might still be the way to reduce HIV stigma in this community. Mukolo et al. (2013) also reported low levels of willingness to stigmatize PLWHA among female heads of households who were confident that the legal system would adequately protect them if they needed protection. There is a need to increase awareness of the anti-discrimination legislation that was enacted by the national government in 2002 and 2009, particularly in rural provinces and districts. Implementation of these legal provisions is likely to increase public confidence in the legal protections that are now available to PLWHA. Follow-up operational research should track the impact of improvements in the knowledge of HIV transmission/prevention and the legal rights of PLWHA in the four domains of stigma highlighted in this consensus-building process – public, courtesy, institutional and self-stigma.

Consensus eight: Stigma reduction approaches likely to work include (1) public education (informational strategies) to address under and misinformation about HIV transmission, prevention and treatment as well as lack of knowledge about the nature of stigma and how it occurs, (2) skills-building to provide positive coping skills among those susceptible to negative consequences of public stigma, (3) dissemination of information about and/or enforcement of laws, policies and procedures that protect the stigmatized and punish stigmatizers particularly among service providers, (4) role modeling by community leaders (traditional, political & religious) to promote social norm reform and enforcement (breaking silence, denial, fear and institutionalized violence) and (5) protest by PLWHA and their associates against stigma enacted in community settings and healthcare institutions. These strategies have been reported in Southern Africa and elsewhere (Pulerwitz et al. 2010; Sengupta et al. 2011). Follow-up operational research studies should evaluate the comparative efficacy of these stigma reduction strategies, particularly in rural settings.

Despite these key observations, the consensus is not representative of all NGO/CBOs and associations in the Maputo and Zambézia Province, nor is it representative of all PLWHA in the Maputo and Zambézia Province. We consulted with those PLWHA who have joined established associations/organizations and were willing to share their views and opinion at public forums. There was also less than optimal female participation in the group discussions we facilitated and most of the key informants we consulted were male. Even in the group meetings that were dominated by women, we had to ensure that women's voices were heard because the few males in attendance tended to dominate proceedings. However, when directly asked to contribute, most of the women in attendance had no problems articulating their views and opinions.

We have also relied on qualitative and exploratory methodology for building consensus. The approach was deemed appropriate because our primary goal was to identify meta-themes in a rapid appraisal fashion. We did not prioritize individuals’ opinions but institutional and community level experiences with stigma as perceived by informants who were knowledgeable about the institutional and community dimensions of HIV stigma. The themes and hypothesized relationships and patterns outlined, therefore, need to be tested (and hopefully confirmed) via more rigorous and systematic (quantitative) studies.

However, the main strength is that this is one of few in-depth inquiries into HIV/AIDS stigma in Zambézia Province. We also assembled and evaluated views from a broad cross-section of the community of interest, including service-users and PLWHA, clinicians and others involved in the provision of services, policy-makers, advocates, governmental and nongovernmental institutions and civil society organizations, traditional healers and religious leaders and men and women from urban and rural locales. All the national organizations and association of PLWHA we consulted acknowledged the need for a nationwide systematic study of stigma and its consequences as well as documentation of effective stigma reduction strategies and approaches. We hope that the themes and stigma-reduction strategies we highlight provide the backdrop and rationale for nationwide assessment of HIV stigma in Mozambique.

To our knowledge, this was a first attempt to develop consensus about the nature of HIV stigma and ways to reduce it in Zambézia Province, Mozambique. Even though some of these theoretic frameworks and concepts, e.g. self-stigma model (Corrigan et al. 2006) or the identity threat construct described by Major and O'Brien (2005) in respect to stigma, have not been directly applied to HIV or studies of stigma in Mozambique, they have been found to be applicable across diverse cultures and countries. For example, Cheng-Fang, Cheng-Chun, Yu, Tze-Chun, Ju-Yu and Chih-Hung (2005) applied the self-stigma model among depressed people in Taiwan, and Angermeyer et al. (Angermeyer & Dietrich 2006; Angermeyer & Matschinger 2003) have tested the model in Germany. Yang et al. (2007) have adapted the model for testing among samples in China or Americans of Chinese cultural affiliation. We also borrowed Brown et al.‘s (2003) classification of stigma reduction strategies – informational, skill-building, counseling/support and contact with affected groups (Brown et al. 2003; Sengupta et al. 2011). Since the latter was not useful for describing strategies to reduce institutional stigma (which involves changes in national legislation and organization policies and procedures), we also utilized strategy classification taxonomies used in organization behavior/analysis literature (Chaffee 1985; Chrisman, Hofer & Boulton 1988).

The overriding viewpoint from a broad spectrum of reviewers is that responses to the HIV/AIDS epidemic have had a positive effect on enacted forms of stigma, particularly at community levels. However, there is likely excessive residual fear of HIV disease and community attitudes that some individuals (i.e. the already infected and worried well) struggle with. The lingering concern about public stigma reflects the strength of stigma enacted in earlier phases of the epidemic. This internalization of public stigma (and subsequent psychological distress) likely restricts access and utilization of available health care, particularly among the less targeted subsets of the rural population (i.e. those of higher socio-economic status and men). Given strong evidence indicating the negative impacts of psychological distress on HIV disease progression among the treatment experienced (Chida & Vedhara 2009; Mukolo & Wallston 2012), addressing the psychological implications of internalized stigma should be prioritized among HIV care and treatment programs.

Pregnant women testing positive through maternal and child health services (like PMTCT) are vulnerable to enacted stigma and deleterious consequences such as divorce/separation and withdrawal of essential social support. The positive externalities of current pre- and post-test counseling seem less protective to those subgroups of men and women who struggle with stigma. Given residual fear of public stigma and the likely enactment of stigma post-voluntary counseling and testing (VCT), it might be beneficial to have culturally sensitive interventions dedicated to addressing psychosocial distress post-VCT. Interventions to increase men's support and participation in maternal and child health services would also help reduce women's vulnerability to stigma.

In addition, our consensus population, which included persons who use public HIV/AIDS care services, singled out unintentional biases among healthcare providers (individuals and institutions) as critical psychosocial barriers to current healthcare access and utilization. However, there are few (if any) interventions that are specifically designed to address stigma in healthcare settings in Zambézia Province and elsewhere (Nyblade et al. 2009; Sengupta et al. 2011). The consensus is that stigma reduction strategies in Zambézia Province should be led by associations of people living with HIV working in partnership with healthcare providers. Associations of PLWHA need to be resourced to acquire requisite capacity to partner with providers to improve the quality of HIV care. Providers need to be supported to effectively address the unintended consequences of standard operational procedures and deep seated biases that likely compromise patients’ access and use of services. This likely entails a paradigm shift in the patient-provider relationship that is likely, also, to generate positive externalities for the broader system of care. The culture change requires support by policy-makers and healthcare providers and the change process needs to be strategically managed.

It would be advantageous if similar stigma audits were to be conducted in other provinces, to tease out the extent to which our findings are mirrored nationwide. At the time of this study, a number of institutions, supported by UNAIDS-Mozambique, were preparing to undertake a national study of HIV stigma based on the Stigma Index compilation methodology (IPPF & UNAIDS/WHO 2010).