Using DHS data to describe scale and pattern of HIV epidemic

Clinic-based studies routinely report a higher prevalence of domestic violence among HIV-positive women compared to HIV-negative women. This same association is not consistently found in population-based cross-sectional studies. The usual interpretation, if an association is detected, is that domestic violence leads to an increased risk of HIV. Yet a direct effect is unlikely. If domestic violence increases the odds of a woman having HIV, it must do so indirectly through her own risk behaviors, those of her partner, or her partner’s HIV status. Population-based studies seldom explicitly articulate and model the intervening paths through which domestic violence may influence HIV status; Variation in their findings may well depend on variations in the conceptualization of violence, analytic methodologies, and included covariates. Additionally, most studies are based on women’s data with limited information on partners’ HIV status and risk behaviors. This paper clarifies the ways in which domestic violence contribute to women’s HIV status, using recent DHS surveys from five Sub-Saharan countries with data on domestic violence, HIV test results and risk factors for both partners of married/in-union matched couples and guided by a conceptual framework depicting possible pathways by which violence could indirectly effect HIV status.

It is important to clarify whether gender inequalities in human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) prevalence in sub-Saharan Africa are explained by differences in the distributions of HIV risk factors, differences in the effects these risk factors, or some combination of both.
Data from Demographic and Health Survey and the AIDS Indicator Survey coming from Kenya, Lesotho and Tanzania were used. We used extension of the Blinder-Oaxaca method to decompose the gender inequality in HIV/AIDS prevalence.
After adjusting, female gender was associated with a higher prevalence of HIV/AIDS in Kenya and Lesotho, but not in Tanzania. Decomposition analyses demonstrated that in Kenya and Lesotho, gender inequality in HIV/AIDS was attributable to the difference in the effects of characteristics on HIV/AIDS. In Tanzania this inequality was explained by the differences in distributions of HIV risk factors between men and women.
These different patterns have important implications for policies to reduce gender inequalities in HIV/AIDS

Extensive efforts to improve reproductive health generally, and to increase awareness of HIV and AIDS specifically, have taken place across Central Asia over the past two decades. Such programs have coincided with significant increases in HIV/AIDS awareness. However, among women comprehensive HIV/AIDS knowledge remains modest across the region, and misconceptions concerning HIV transmission remain. Differentials by age, education, and rural/urban residence in awareness and knowledge regarding HIV and AIDs persist among women in the countries of Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan. Using Demographic and Health Surveys and Multi-Cluster Indicator Surveys from 1999 through 2011 and reports, records, and interviews from health programs in the region, I examine increases in HIV and AIDS knowledge, tracing patterns of improvement across demographic, social and cultural sub groups within each country. Findings indicate a concerning trend in poor knowledge relating to needle transmission (the major transmission route for HIV in the region) and a persistent gulf in reported knowledge (at all levels) by marital status, age, education and urban residence. These results raise questions concerning the relevance of programmatic content (emphasizing ABCs and sexual transmission) in the region and program access and reception.

The link between HIV infection and poverty has attracted considerable research attention in recent years, but the relationship is rather complex and findings from existing studies remain inconclusive. While some argue that poverty increases vulnerability, existing empirical evidence from sub-Saharan Africa largely support the view that wealthier men and women, especially rural residents, have higher prevalence of HIV. In this paper, we focus on urban/rural differentials in the association between HIV infection and poverty in Kenya. The study is based on secondary analysis of data from the Kenya Demographic and Health Surveys conducted in 2003 and 2008. We apply multilevel logistic regression models, allowing the poverty risk factor to vary across rural and urban communities (i.e. clusters) in Kenya. The results suggest that the urban poor in Kenya have significantly higher odds of HIV infection than their urban non-poor counterparts, despite poverty being associated with a significantly lower risk among rural residents. Interesting urban/rural differentials are observed in poverty risk factor by key demographic and socio-cultural characteristics, including gender disparities and ethnicity. For example, the rich-poor gap among urban residents is wider for women, with poor urban women being particularly vulnerable.

Though global HIV/ AIDS incidence is declining, HIV/AIDS remained the leading cause of death in women of reproductive age in low-and middle-income countries, particularly in sub-Sahara Africa. Given the high costs of HIV treatment, evidence-based targeted interventions are advocated as cost-effective strategy to fighting HIV/AIDS.
This study aims to assess the socioeconomic predictors of HIV infection and identify the most-a-risk groups among women in Malawi using a Chi-square Automatic Interaction Detector. The analysis was performed on a sample of 8,596 women aged 15-49 years from the 2004 and 2010 Malawi Health and Demographic Surveys.
The results revealed that marital status is the most significant predictor of HIV infection among women in Malawi. Women who are no longer in union and living in the households within the highest wealth quintiles were the most-at-risk group of HIV infection. The less-at-risk group included nulliparous never married women living in the rural areas. In the light of these findings, we recommend targeted interventions taking into account HIV prevalence and the demographic size of different groups at risk groups; and reinforcement of integration of family planning and HIV/AIDS services through community health workers, households and school based campaign and reproduction health services.

Consent rates for HIV testing in population surveys are often low, which may cause a bias in prevalence estimates if refusal to test is correlated with HIV status. Interviewer identity represents a plausible variable that affects testing, but not HIV status, and can be used in a Heckman-type selection model that provides consistent prevalence estimates. We innovate by adopting an interviewer random effects estimator which improves on the existing interviewer fixed effect approach in three respects. Firstly, using our model allows the effects of interviewer identity to be estimated even for those whose interviewers conducted a small number of interviews. Secondly, this methodology facilitates the use of bootstrapped standard errors which are necessary to correct for regression parameter uncertainty in the correlation between consent and HIV status. Thirdly, we propose a Bayesian model averaging approach that gives estimates that are consistent and unbiased. We report results for Zambia and Ghana. For Zambia we estimate a prevalence rate of 32% among males who refuse consent compared with 12 % among those who agree to test, substantially increasing the estimated population prevalence rate.