Abstract Body

Background: No study has assessed the proportion of HIV-positive pregnant or breastfeeding (PBF) women virally suppressed at population level. Furthermore, critical data assessing the cascade of care among PBF women are needed to inform program policy. We used data from three population surveys to assess these key indicators.

Methods: Three household based surveys of individuals age 15 to 59 took place between September 2012 and November 2013 in Ndhiwa (Kenya), Chiradzulu (Malawi) and Mbolongwane/Eshowe (South Africa). Following consent, all individuals were interviewed and tested for HIV. Women reported their pregnancy and breastfeeding status and the result of their last HIV test. All individuals found HIV-positive were tested for viral load and CD4, regardless of the ART status. At the time of the survey, PMTCT option A was implemented in Kenya, B+ in Malawi and B in South Africa. Infection during PBF was defined as a positive-HIV test among a PBF woman who was unaware of her status and reported a negative HIV test during Ante Natal Care (ANC).

Results: Among the 21,782 individuals eligible, 12,461 were women and of them, 11,550(92.7%) were included in the surveys. More women were PBF in Kenya (37.8%, 1,413/3,760) and in Malawi (33.8%, 1,444/4,275) than in South Africa (12.5%, 439/3,515). Among them, HIV prevalence ranged from 13.4% in Malawi to 22.2% in Kenya and 23.0% in South Africa. The proportion of PBF women with viral load<1,000copies/ml was higher in South Africa (63.4%;95%CI 65.5-72.3) and Malawi (72.3%; 95%CI 65.5-78.2) compare to Kenya (27.3%; 95%CI 22.5-72.3). Of the breastfeeding women with viral load >1,000copies/mL, 58.6% (95%CI 52.0-65.0) were not diagnosed for HIV at the time of the survey. This proportion was similar across sites (p=0.16).
A total of 103(4.1%) breastfeeding women were infected during pregnancy or breastfeeding. This proportion was higher in Kenya (6.5%) than in South Africa (4.3%) and Malawi (1.9%). These new infections accounted for 37.5% (95%CI 31.5-44.4) of the HIV-positive breastfeeding women with viral load>1,000 copies/ml.

Conclusions: Viral suppression at population level among PBF women ranged from 27 to 72%. Our analysis showed that the majority of HIV+ PBF women with viral load>1,000cp/mL were undiagnosed which could be partly due to infection after HIV testing at ANC. To identify those women and to prevent transmission or at least to ensure early diagnosis of their infants, repeated HIV-testing should be implemented until the end of breastfeeding.