HIV-related knowledge and antiretroviral therapy outcomes (ART) in HIV infected women initiating ART during pregnancy

The characteristics of South Africa’s HIV epidemic mean that approximately 28% of women presenting for antenatal care, are HIV-infected. Maternal HIV-infection can lead to mother-to-child transmission (MTCT) of HIV during pregnancy, labour, delivery or breastfeeding if viral load (VL) is not well controlled by antiretroviral therapy (ART). Globally, 90% of pediatric infections occur via MTCT, though lifelong ART is reducing the rate of new infections. Full benefits of ART can only be realized when ART adherence is high. Evidence from South Africa and elsewhere has shown that ART adherence in pregnant and postpartum women is suboptimal. Potential drivers of suboptimal adherence may include poor or inadequate knowledge of HIV and ART. This thesis investigates how HIV-infected pregnant and postpartum women’s knowledge of HIV and ART-related information may be associated with ART adherence as evaluated by HIV VL measures. Components of this thesis include the research protocol, a literature review of previous studies exploring the relationship between knowledge and HIV-related health outcomes in Sub-Saharan Africa and a manuscript describing the results of an investigation into predictors of HIV and ART-related knowledge and the association of knowledge with maternal vireamia (VL>1000copies/mL). This data for analysis came from a cohort of 376 HIV-infected pregnant women, initiating ART during pregnancy, at a primary care antenatal facility in Gugulethu, South Africa. Participants were followed from their first antenatal visit until twelve months postpartum. Knowledge of HIV and ART-related information were assessed at three time points by two knowledge inventories and items were classified as either relating to general knowledge or prevention of MTCT. HIV VL was measured at delivery and twelve months postpartum. Demographic characteristics were surveyed at the first antenatal visit. Analyses included univariable and multivariable regression models to estimate potential predictors of knowledge among demographic and clinical characteristics, as well as to estimate the association between knowledge and maternal vireamia at delivery and twelve months postpartum. We found that HIV and ART knowledge increased marginally over the repeated study visits. Knowledge relating to general HIV or ART information was typically good while knowledge on PMTCT was lacking. Education (OR=-0.52; 95% CI=-0.83- -0.21; P=0.001), previous HIV diagnoses (OR=-0.36; 95% CI=-0.09- 0.63; P=0.009), and weeks on ART at delivery (OR=-0.03; 95% CI=0.00-0.06; P=0.047) were statistically significant predictors of HIV knowledge in adjusted analyses. The associations between the various knowledge outcomes and vireamia at delivery and twelve months postpartum were mixed and generally not statistically significant. In summary, HIV and ART knowledge both increased with increasing time in care and general knowledge was better than knowledge specific to MTCT. Education, timing of HIV diagnoses and time on ART were identified as potential predictors of HIV-related knowledge. Generally, knowledge of HIV or ART was not meaningfully associated with vireamia at delivery or at twelve months postpartum. There remain significant gaps in the knowledge of HIV-infected women, of childbearing age, around how HIV is transmitted and how to reduce the risk of MTCT