Discussion

To the best of our knowledge, this is the first quantitative study to analyse the association between HIV-related misconceptions and HIV status in Congo. We detected a high prevalence of HIV-related misconceptions in Kinshasa and a positive association between these and HIV infection.

In agreement with previous Congolese studies, most participants had heard of AIDS. However, 64% of our respondents acknowledged having little/no HIV knowledge, which was more prevalent in cases.

Around 15% of the participants were unaware of the viral basis of AIDS, which was associated with an HIV-positive test. This has never been evaluated in Congo and has only been addressed in a handful of African studies. Given that infectious origin of AIDS was first described three decades ago, providing this basic information should be a priority. Also, erroneous beliefs involving supernatural forces were particularly prevalent among both cases and controls. The 2007 DRC-Demographic Health Survey reported that half of the respondents held this belief, and the recently published 2014 survey keeps showing that near the same percentage continue to believe in supernatural transmission.1 These misconceptions are similarly prevalent in Ghana and South Africa, and even more in other African countries.3 ,10 Scepticism about the role of infectious agents leads people to believe that HIV can only be cured by traditional healers and not by antiretrovirals. We need to be aware of these beliefs in order to develop culturally adapted HIV information. Moreover, as religion is an important part of daily life in Kinshasa,11 close collaboration between religious leaders and HIV prevention programmes is necessary to ensure success.

Another highly prevalent misconception was believing that an HIV-positive individual is visibly unhealthy. Nearly half of the cases held this, a significantly larger proportion than in the control group. The 2014 DRC Health Survey reveals that around 30% of 15–24 year olds do not know that a healthy person can be seropositive.1 If people believe that only those visibly unhealthy can transmit HIV, the infection risk can increase. Around 10% of our participants believed that HIV could be transmitted by mosquitoes (more cases than controls). Comparing with previous data from Kinshasa,4 ,5 the prevalence of misconceptions has decreased considerably.

Regarding the misconceptions we found about HIV prevention, it is important to consider them in condom promotion campaigns in Africa, where the prevalence of multiple partnerships is on the rise in some countries and condom use is inconsistent.12 Indeed, only around 30% of our HIV-positive cases said they would use a condom if they/their partner were infected.

Basic socio-demographic determinants also need to be considered. First, education is crucial to change misconceptions. Cases were more likely to have no education or primary education only, and reported more misconceptions than controls. It is thus crucial that accurate information reaches people from all educational levels.

In addition, the economic context must also be considered: some African studies have shown that a high education level is sometimes linked to HIV-related risk behaviours or even a higher rate of infection,1 as those with the highest economic status may have a higher frequency of premarital sex or access to commercial sex.

Another factor that must be considered is sex: women in sub-Saharan Africa are particularly vulnerable to HIV due to low socioeconomic status/educational level, few job opportunities, limited access to health services or forced sex, frequently associated with an increased HIV risk.2 As cases in our study were mostly women, special attention should be paid to this group when developing prevention campaigns.

All these findings, and the fact that around 40% of participants consider that little information on HIV prevention is available in Kinshasa (more cases than controls), underscore the need for culturally adapted and gender-orientated basic information in Congolese HIV prevention programmes. Current HIV campaigns in Congo do not include any messages to clear up common misconceptions. This information should be provided in Lingala, the language people use to speak about intimate/taboo issues. The most effective information channels must also be identified. Most of our respondents, particularly cases, had regular access to a radio/TV, which could both be used to send preventive messages. Visits to health services for VCT/blood donation provide useful opportunities to inform about HIV/AIDS, to offer testing and provide antiretrovirals for those who test positive.

The present study has some limitations. The study population is probably not representative of the general population, especially VCT attendees that are likely to be at higher risk of HIV and potentially have a better HIV knowledge. However, even in our participants we observed a low level of knowledge, which would likely have been poorer still if we had recruited from a general, non-HIV-specific health clinic. Nonetheless, what made them think they were at risk should be further analysed. On the other hand, family replacement blood donors are more similar to the general population because the only reason for going to the health centre is that they are requested to do so to give back the blood used by their relatives. Second, our study is based on self-reporting. Nonetheless, under-reporting was probably low as same-sex interviewers and private rooms were available. Also, retrospective responses could have led to recall bias; however, most questions were specific enough to reduce errors. Lastly, we did not use a validated scale to measure HIV knowledge. The HIV-Knowledge-27-Scale, specifically suited to sub-Saharan populations, was not available when we implemented our questionnaire.

Despite these limitations, our study has several strengths. This is the first analytic study in the DRC to evaluate knowledge, attitudes and sexual behaviour of individuals newly diagnosed with HIV. Also, it has assessed many HIV-related aspects not considered in other Congolese studies. Finally, we had a high response rate and recruited 1614 Congolese, a sufficiently large sample with which to estimate relevant adjusted associations.

This study provides evidence of the high prevalence of HIV-related misconceptions and their association with HIV. There remains a great need to challenge these misconceptions, tackling them one by one, through media/HIV campaigns and hospital visits. People that attend health services are useful participants in which to evaluate HIV-related knowledge and behaviours as they can transmit preventive messages to people within their community without access to health facilities, influencing knowledge and behaviour.