HIV-related knowledge in Nigeria: a 2003-2013 trend analysis

As of 2015, 3.5 million people were estimated to be living with HIV in Nigeria, and the disease is estimated to have resulted in 180,000 deaths in that year [1]. Given Nigeria’s status as the country with the second highest number of people living with HIV globally, and 9% of the total global burden of HIV being attributable to Nigeria alone in 2013 [2], improving our understanding of the nature of the HIV epidemic in Nigeria is crucial.

Previous studies have emphasised that increasing HIV-related knowledge is a critical aspect of HIV prevention [3‐5], underlining its influence on the likelihood of engagement in preventive behaviours [6], but few studies have examined trends in HIV-related knowledge in Sub-Saharan Africa, with one study doing so in Uganda [7], and others in Ethiopia [8] and South Africa [9]. Importantly, despite the fact that low HIV-related knowledge, and consequently lower engagement in preventive behaviours, is considered a relevant factor in the transmission of HIV in Nigeria, HIV-related knowledge levels in the country have not been studied in detail [10]. Although a prior study in Nigeria examines the association of HIV transmission and prevention knowledge indicators with HIV-related stigmatization [11], trends in these knowledge indicators themselves were not investigated, and these remain to be analysed, particularly in the context of national efforts to improve awareness and knowledge of HIV.

In working towards the millennium development goal of halting and beginning to reverse the spread of HIV/AIDS by 2015, Nigeria launched the National Strategic Plan (NSP) to combat HIV/AIDS. This program ran from 2010 to 2015, and focused on prevention, aiming to reduce the transmission of the disease through the modification of behavioural practices and improving public HIV-related knowledge [10]. Importantly however, despite including the support of research activities and the reduction of gender inequities in its mandate, Nigeria’s National Agency for the Control of AIDS (NACA) reports that evidence-based programing and gender-based approaches in the strategy remain to be improved [10]. As such, this study will investigate the trends in HIV-related knowledge between 2003 and 2013, and stratify the investigation of these trends by socio-demographic characteristics such as gender and income, in order to determine whether HIV-related knowledge has increased since the implementation of the NSP, and whether this increase differs among socio-demographic groups. The selection of these socio-demographic factors is based on a recent study by the authors, in which logistic regression analyses indicated that factors including poverty, low literacy, and being female, among other factors, are associated with a higher likelihood of having low HIV-related knowledge [12].

As prior studies in Sub-Saharan Africa have reported that wealth inequality, rather than absolute wealth or poverty, is a stronger driver of HIV transmission [13‐16], and a prior study using Nigerian data has found that, particularly among females, high wealth inequality is associated with lower HIV-related knowledge [12], the observed trends in HIV-related knowledge over time will also be described with respect to state-level wealth inequality rather than solely absolute wealth. Investigating changes in the levels of HIV-related knowledge prior to and following the implementation of the NSP, and observing differences in these changes across various socio-demographic strata will shed light on whether or not the NSP was effective in increasing general population-level knowledge and understanding of HIV transmission, and in which socio-demographic groups this was most or least successful, in order to inform future national HIV education efforts and more specific targeting of such efforts among priority groups.

When considering the implications of an analysis of trends in HIV-related knowledge, it is relevant to note at the outset that several Sub-Saharan African studies have reported the effectiveness of behaviour change interventions regarding HIV risk reduction and prevention measures [20‐22]. This underlines firstly that interventions improving HIV-related knowledge, when delivered along with behaviour change elements, are associated with a greater likelihood of adoption of appropriate preventive behaviours, and in turn therefore a reduced risk of HIV transmission [22], and secondly that HIV-related knowledge is a modifiable factor that can and should be targeted through intervention, as part of efforts to prevent further HIV transmission. Regarding the socio-demographic determinants of HIV-related knowledge in Nigeria, a recent paper has shown that factors including poverty, unemployment, low literacy, rural residence, sex, and wealth inequality are significant predictors of HIV-related knowledge in the country [12], With this in mind, the implications of the current trend analysis in the Nigerian context are discussed.

The low level of HIV-related knowledge seen in 2008 in the current analysis may be explained in part by the fact that HIV-related knowledge in Nigeria may have been low prior to the launching of the National Strategic Plan to combat HIV, which was initiated in 2010. The subsequent rise in 2013 suggests that the plan may have to some extent been effective in increasing overall HIV-related knowledge at the national level, however, due to the limited amount of data, the influence of other societal dynamics and events on this trend remains to be better understood.

Despite this decrease observed in most knowledge domains in 2008, an overall increase is seen from 2003 to 2013 in all knowledge domains except knowledge of mother-to-child transmission, which was significantly lower in 2013 than in 2003. This is particularly alarming given the substantial contribution of MTCT to the continuation of the Nigerian HIV epidemic, with a MTCT prevalence rate of 27.3% in 2014 [10], This suggests that future national programs with an HIV prevention and education mandate should place an increased focus on the prevention of MTCT. Specific recommendations regarding MTCT prevention based on the findings of the current study include that educational interventions emphasizing safe infant feeding practices and encouraging antenatal and postnatal care seeking should be targeted at expectant mothers and females of childbearing age, in particular those who are HIV positive or whose serostatus is unknown. In addition, as mother-to-child transmission knowledge is the only knowledge area in which males scored lower than women, it may be of interest to also include mother-to-child transmission prevention content in HIV education efforts targeted at males. Given that males may often be the primary household decision-makers, they may be more likely to support their partners in seeking maternal care or in making alternative infant feeding choices if adequately informed of the risks of mother-to-child transmission.

The overall increase in the other HIV-related knowledge domains over time seen in our analysis however aligns with the findings of studies that have examined trends in HIV-related knowledge in other sub-Saharan African countries. The aforementioned study in Ethiopia for example reports an increase in HIV-related knowledge between 2005 and 2011 [8]. As knowledge levels in the Ethiopian study were based on only 3 knowledge indicators, one being having heard of HIV, and two relating to risk reduction measures, these results are comparable to the upward trend in the general knowledge and HIV risk reduction domains observed in the current study.

Importantly, although the trends in HIV-related knowledge are similar across socio-demographic strata, there are marked disparities in the levels of knowledge between strata. For example, females generally have lower HIV-related knowledge across most knowledge domains in comparison to males, and those in higher wealth quintiles have higher mean HIV-related knowledge levels than those in lower wealth quintiles. This, as well as the disparities seen in levels of HIV-related knowledge among males compared to females at each level of wealth, and in particular the fact that mean risk reduction knowledge is similar among males and females at low wealth inequality, but is much lower in females than in males in states with higher wealth inequality ratios, suggests that future HIV awareness and education campaigns should be targeted at the most marginalized, particularly those experiencing the confluence of gender and wealth inequalities.

Moreover, the fact that disparities in HIV-related knowledge between literate and illiterate respondents persist from 2003 to 2013 across all knowledge domains suggests an urgent need for the tailoring of future national HIV education programs to the needs of those to whom complex or text-based information is less accessible. This should include the use of non-written media (for example diagrams or pictograms) for the dissemination of HIV-related information, including transmission mechanisms and prevention measures.

Furthermore, the results of this analysis regarding the observation of growing disparities in certain domains of HIV-related knowledge between states suggest that increased focus should be devoted to improving HIV-related knowledge in these specific states in which it is currently low, such as in Zamfara, Kebbi, and Bauchi. In particular, efforts to improve HIV-related knowledge in these states should focus on the specific knowledge subdomains that are currently most poorly understood.

Regarding the trend analysis using ARIMA modelling, the fact that the proportion of literate respondents as a predictor provided the best fit model for total HIV-related knowledge, risk reduction knowledge, and knowledge of mother-to-child transmission suggests that at the national level, improvements in literacy over time may in part explain and facilitate improvements in HIV-related knowledge. Consequently, this indicates that not only should efforts be made to ensure that HIV education campaigns are more accessible to individuals with low literacy levels, but also that investments into national education and literacy in general will equip individuals with a greater capacity to acquire, understand, and use HIV-related information.

Moreover, the finding that knowledge of HIV transmission in Nigeria was best approximated in the ARIMA model using the mean state-level wealth inequality ratio as the single predictor is of particular interest in light of recent studies on the social determinants of HIV transmission indicating that, in Sub-Saharan Africa, wealth inequality may be a more significant predictor of HIV transmission than absolute poverty or wealth [13‐16]. Additionally, in Nigeria specifically, a recent study investigating wealth inequality as a predictor of HIV-related knowledge [12] indicated that under circumstances of inequality, females in particular are at higher risk of low HIV-related knowledge. The observation that the trend in knowledge of HIV transmission to some extent follows the trend in state-level wealth inequality therefore underlines that under circumstances of inequality, individuals experience both greater barriers to accessing HIV-related health information, as well as greater barriers to the actual application of this information through the adoption of preventive or care-seeking behaviours. Apart from indicating that, particularly in a country as socioeconomically heterogeneous as Nigeria, HIV prevention education should be especially targeted at areas of high wealth inequality, the observed HIV transmission knowledge and wealth inequality trends also suggest that efforts towards reducing wealth disparities in Nigeria could address an important driver of HIV transmission in the country, and consequently substantially reduce future transmission. More specifically, as the current study shows a more pronounced decrease in risk reduction knowledge among females than among males at higher levels of wealth inequality, females living in areas of high wealth inequality should be particularly prioritized for HIV risk reduction interventions. This is particularly relevant given that circumstances of wealth inequality have been shown to be associated with an increase in high HIV risk activities, such as engagement in informal transactional sex [23, 24].

The limitations of this study include, firstly, the limited number of time points in this analysis, which limited the time series analysis as it precluded the production of trend forecasts from the ARIMA models, therefore limiting conclusions regarding future HIV-related knowledge levels in the country. In addition, the small number of time points between 2003 and 2013 limits the level of detail in our understanding of the trends in HIV-related knowledge over time in the Nigerian context, making it difficult to interpret what these knowledge levels suggest regarding the effectiveness of HIV awareness and education campaigns implemented over the years, or to determine which other events or dynamics may be contributing to the observed trends in knowledge levels.

In addition, the fact that NDHS data is not longitudinal – i.e. not collected from the same individuals over the multiple time points – means that longitudinal data analysis methods are not applicable, and conclusions from this analysis are thus unable to take into account how individual-level changes in absolute wealth, wealth inequality, educational attainment or other socio-economic indicators influence changes in individual-level HIV-related knowledge over time.

It should also be noted that as the DHS sampling procedure includes women in all sampled households and the corresponding men in only a subset of the households from the original female sample, there is a greater representation of women than men in all survey years, however, the effects of this are taken into account through the application of individual sample weights (as provided in the DHS) to the male and female datasets.

Lastly, although health-related knowledge has been shown to lead to favourable health behaviours and engagement in preventive measures and can therefore be considered a relevant factor influencing potential HIV infection risk, the lack of individual HIV serostatus information in the NDHS limits our ability to corroborate the contribution of low HIV-related knowledge to HIV transmission risk. Consequently, the conclusions drawn from an analysis of HIV-related knowledge are of limited value in terms of their direct translation into the evidence-based targeting of HIV preventive interventions among high-risk groups. Future research in Nigeria could therefore focus on the collection of individual-level HIV serostatus data for the determination of whether HIV-related knowledge is a valuable predictor of HIV transmission risk. Moreover, more detailed evaluations of national HIV education and prevention programs in terms of their effectiveness in disseminating HIV preventive information to vulnerable groups, improving HIV-related knowledge, and ultimately leading to preventive behaviours are needed.

Overall, through the identification of states, population subgroups and knowledge domains (such as knowledge of MTCT) in which HIV-related knowledge levels have been low or decreasing over time, the understanding of recent trends in HIV-related knowledge as investigated in this paper provides insight for the evidence-based design and targeting of further HIV education efforts. In light of the current study’s findings, we suggest several specific strategies for HIV-related knowledge improvement in Nigeria, relating to both the content and method of dissemination of HIV education interventions. Taking continued steps towards improving HIV-related knowledge in Nigeria is particularly relevant in light of the fact that the aforementioned study on attitudes towards HIV in Nigeria demonstrated that individuals with greater understanding of HIV transmission were also less likely to have negative or stigmatizing attitudes towards people living with HIV [11]. This indicates that appropriate HIV education and subsequent improvements in HIV-related knowledge are crucial to reducing the social stigmatization of the disease, which consequently will contribute to increasing at-risk individual’s propensity and ability to seek preventive services, as well as facilitating easier access to treatment and strengthening social support for HIV-positive individuals. In addition however, as a prior Sub-Saharan African study has noted discrepancies between HIV awareness and the adoption of protective behaviours [25], it will be crucial that future HIV education efforts also incorporate elements of behaviour change interventions, in order to maximize their likelihood of resulting in actual increases in preventive behaviours.