RESUMEN
INTRODUCTION: In 2011, Malawi implemented "Option B+," a test-and-treat strategy for the prevention of maternal to child transmission of HIV (PMTCT); however limited data on viral load (VL) suppression exist. We describe VL suppression in HIV-infected women at four to twenty-six weeks postpartum, factors associated with VL suppression and the impact of VL suppression levels on MTCT. METHODS: HIV-positive mothers at four to twenty-six weeks postpartum were enrolled in a nested cross-sectional study within the "National Evaluation of Malawi's PMTCT Programme" cohort study between October 2014 and May 2016. HIV-exposed infants received HIV-1 DNA testing and venous samples determined maternal VL, classified as unsuppressed (>1000 copies/mL), low-detectable (40 to 1000 copies/mL) or undetectable (<40 copies/mL). Socio-demographic and PMTCT indicators were collected. Suboptimal adherence was defined as self-reported ≥2 days missed ART in the month prior to visit. RESULTS: Of the 1274 women, 1191 (93.5%) knew their HIV status and 1154/1191 (96.9%) were on ART. VL was available for 1124/1154 (97.4%) of women on ART: 988/1124 (87.9%) had VL suppression of whom 86 (8.7%) had low-detectable and 902 (91.3%) undetectable VL. Suboptimal adherence was associated with unsuppressed VL (vs. suppressed VL; aOR 3.1, 95% CI 2.0 to 4.9; p < 0.001). Women with low-detectable VL were more likely to be adolescent (vs. undetectable VL; aOR 3.0, 95% CI 1.4 to 6.6), on ART <6 months (aOR 4.4, 95% CI 2.3 to 8.6), report suboptimal adherence (aOR 2.1, 95% CI 1.1 to 3.8; p = 0.02), and less likely to have primary or secondary education (vs. none; aOR 0.3, 95% CI 0.2 to 0.7 or aOR 0.3, 95% 0.1 to 0.6). MTCT ratios among women on ART who had undetectable VL, low-detectable VL and unsuppressed VL were 0.9% (8/902; 95% CI 0.3 to 1.5), 7.0% (6/86; 95% CI 1.5 to 12.5) and 14.0% (19/136; 95% CI 8.1 to 20.0). Unsuppressed VL and low-detectable VL (vs. undetectable VL) increased the risk of MTCT 17-fold (aOR 17.4, 95% CI 7.4 to 41.1; p = 0.002) and ninefold (aOR 8.5, 95% CI 2.9 to 25.2; p < 0.001). CONCLUSIONS: Unsuppressed and low-detectable VL was strongly predictive of MTCT among women on ART and associated with suboptimal adherence. This urges further consideration of optimal VL monitoring and target levels to reach elimination of paediatric infection.