Witchcraft is frequently attributed to be the cause of HIV/AIDS, with many believing that it can be cured by traditional medicine alone, or accompanied by spiritual healing [10]. The reasons for this are complex, and partially reflect the pervasive stigma surrounding HIV infection, which creates an environment that is conducive to transferring blame away from an individual's own behaviour, towards an external power, so that the infected individual becomes a victim of a malevolent third party, rather than being responsible for his/her own actions [11]. Furthermore preferences for traditional healers over health facilities are common due to familiarity, trust, accessibility and their cheaper cost [12].

Another common misconception in some rural areas of Tanzania is that the lubricant oil present on condoms contains the HIV virus, such that their use propagates, rather than prevents, the spread of the epidemic [13]. Furthermore, these rumours are sometimes accompanied by claims that the HIV virus has been deliberately placed there by white people to infect and kill Africans; or that they knowingly distribute old and inferior products [13]. In addition it has been reported by some individuals that both the man and the woman should use a condom during sexual intercourse, suggesting that even when they are considered acceptable, their effectiveness may be compromised by a lack of correct knowledge about their use [13].

The existence of myths and misconceptions around HIV occurs in the context of treatment, as well as in relation to the aetiology of infection and prevention interventions. ART is provided free of charge in Tanzania and has been available in Mwanza city since 2005 and Kisesa Health Centre since 2008 [14]. Following the introduction of free ART, some HIV-related stigmas have declined, but others have emerged in their place, including the perception among some community leaders that improvements in ART patients' health has led to increased sexual activity, further spreading HIV [3]. Another misconception recorded in a separate study was that engaging in sex whilst on ART accelerated death, and that the use of the drugs requires sexual abstinence [10]. The mistrust and confusion associated with ART may discourage the initiation of treatment for infected individuals and may lead to de-motivation among those already enrolled in care and treatment programmes [10]. Furthermore, since decision-making with regards to ART commonly reflects the views of sexual partners, family members and friends as well as influential community members, misconceptions around ART within these social networks can either generate support or discouragement for sustained adherence to therapy [10].

One potential vehicle for providing communities with HIV education, and for addressing the common misconceptions that can undermine prevention and treatment programmes are village AIDS committees (VACs). VACs were formally established in 2003 in Tanzania, in order to give local governments responsibility for reducing HIV prevalence [15,16]. The aim of the VACs is for its members to disseminate information to the community on various aspects of HIV transmission, prevention and treatment, in order to increase understanding about HIV amongst the local community [15]. Each village in Tanzania is required to have its own committee consisting of approximately 15 members. Recruitment of VAC members is based upon Tanzanian policy, with national guidelines stating that the VAC should include both men and women, and should consist of a diverse range of individuals including prominent citizens from the village council, youth representatives, religious representatives and residents living with HIV/AIDS [15]. Any monitoring and evaluation work that may have been conducted previously with VACs in Tanzania is undocumented.

However, VACs are unlikely to be able to meet their responsibilities unless they are equipped with comprehensive HIV education and an explanation of their expected roles. HIV curricula have previously been piloted in a variety of settings, often targeting young people prior to sexual debut. Mema Kwa Vijana (MKV) was a curriculum implemented and evaluated in 62 primary schools in rural Mwanza between 1999 and 2001 [17]. This teacher-led and peer-assisted programme incorporated various methods including scripted dramas, role-plays and stories to educate students on aspects of sexual health. School-based HIV curricula have since been expanded to other areas within Tanzania and elsewhere in sub-Saharan Africa [18-20]

Despite the success of school-based interventions, they are unlikely to reach many sexually active members of the community with poor levels of knowledge about HIV and at risk of HIV infection. Interventions that are delivered in a community setting have the potential to target higher risk youths who have dropped out of school, as well as reaching the rest of the village population [21]. The development of curricula focusing on sexual health, HIV and STIs targeting people in schools, clinics or in community settings have been shown to be promising interventions towards reducing risky behaviours [22-24]

Nevertheless, this pilot study clearly suggests that providing VACs with HIV education and training can increase HIV knowledge levels. The post-training surveys provided positive results of the curriculum's effectiveness, illustrating that fairly short educational sessions can result in an immediate increase in knowledge. Furthermore, the baseline surveys highlighted that VAC members who had received previous HIV education had better HIV knowledge levels, indicating that knowledge can be retained, but does need reinforcing. These findings encourage the development of a system of continuing education for the VAC in order to improve and maintain good knowledge of HIV and ART. Education needs to be appropriate to the cultural context, and may vary across the country where different beliefs, behaviours and cultural practices exist. The curriculum incorporated participatory learning action techniques to encourage learning and engage VAC members with the trainer.

A key success of this pilot curriculum was the observed change in belief surrounding condoms. Prior to HIV training, 57% of participants believed that unused condoms contain the HIV virus compared to none in the follow-up survey. Kisesa is not isolated in these beliefs surrounding condoms, with similar reports documented in other regions of Tanzania [13]. For condom interventions to be successful, negative beliefs and attitudes need to be addressed with accurate information.

Univariate analysis of the baseline survey revealed initial knowledge levels to be lower amongst female members of the VAC (p = 0.041). This concurs with results from other studies in rural Tanzania, which have illustrated that girls in school have significantly poorer HIV/AIDS knowledge than boys [26]. Boys and men tend to have greater exposure to HIV information than girls, which is likely to explain their more accurate knowledge on average, about the virus. Exposure to such information may come from within and outside the classroom, including through entertainment education or public awareness campaigns. Adolescent women in poor rural communities have few opportunities to access television or radio, and often have less access to schooling and lower levels of educational attainment [26]. It is therefore particularly important that female VAC members are well trained about HIV, and are encouraged to provide their contemporaries in the community with general health education and HIV prevention information, as well as ways to negotiate safe sexual behaviour. It may be appropriate on occasion for the VAC to provide female-only educational sessions, which may help to reduce or eliminate the HIV knowledge disparities between the sexes that are evident across the continent [26].

This study suggested that knowledge regarding the free availability of ART in Tanzania was higher amongst VAC members compared with the general community. However the discrepancy could be attributed to the fact that the data from the general community were collected several years earlier in 2007, when there was less knowledge about ART across the whole community. As free treatment only became available in Mwanza city in 2005 and within Kisesa health centre in 2008, it is likely that awareness about this treatment has increased over time.

VAC guidelines promote recruitment of a diverse array of individuals from across the community [15,16]. However, the composition of the VACs in Kisesa indicated a predominantly male committee, averaging 40 years of age with the absence of youths, elders, teachers and religious representatives. This imbalance may reduce opportunities for dissemination of HIV education, as opportunities to interact with youths and women within these communities may not be available. In addition, the absence of teachers among the committees in Kisesa may lead to missed opportunities for HIV education within schools, as primary school teachers can be trained to deliver HIV education messages [27]. The MKV programme rolled out in rural Mwanza implemented peer-assisted methods of education in addition to teacher-led education. In MKV peer educators successfully created a drama serial to aid in promoting desirable behaviours [17]. Written curricula, which can be implemented outside the school environment in a community setting can potentially reach higher risk youths who have dropped out of school [21]. Recruitment of youths onto the VAC may provide opportunity to educate and inform similar age groups about HIV; especially those high-risk individuals.

Tanzanian guidelines also advise that people living with HIV/AIDS (PLHA) are recruited onto VACs [15], although it was not possible to determine from this study whether this has been adopted within the VACs of Kisesa. As religion is highly influential in Tanzanian society, incorporating faith leaders into VACs is desirable, particularly since religious leaders often influence people's beliefs and behaviours as well as offering support and care for PLHA [7,28]. In particular, previous studies have shown that some PLHA gain support from their religion, helping them to live positively with HIV and providing them with courage to face their condition [7]. Providing religious leaders with education, through membership on the VAC could allow them to encourage their followers to access services for testing and care, although it is important that religious leaders emphasise that prayer and medical care can be combined for PLHA [7].

The turn-over of VAC recruits was most apparent in the trade-centre and roadside villages. Accordingly, to accommodate this faster turnover of recruits, HIV education may need to be delivered more frequently to VAC members in villages that are closer to urban areas, compared to the more remote villages, in order to ensure that each new member is sufficiently informed.

Several steps were taken to address the potential for bias in the results. In order to limit selection bias, invitations were distributed to every member of the VAC several days prior to training. However participation in the trade centre was lowest, which may be explained by members of this VAC being more likely to have other work commitments compared to those in the more rural areas. Social desirability bias may have resulted in an over-estimate of the success of the intervention, if participants reported what they felt the interviewer would like to hear, particularly in the post-training survey. To reduce this bias as far as possible, a research assistant who had previously established a rapport with these individuals, conducted both the baseline and post-training surveys after briefing the participants on the purpose of the study and explaining that their responses would be anonymous.

The HIV curriculum was translated from English to Kiswahili by a native Kiswahili research assistant fluent in English, however some words could not be directly translated. To aid in this process as much as possible, both the KiSwahili and English versions of MKV were used as a reference tool.