Background: The wide implementation of the PMTCT Option B+ approach in South Africa warrants a closer examination of postpartum antiretroviral therapy (ART) outcomes. In this study we examine the association between timing of the first pregnancy with the risk and predictors of postpartum ART failure and disengagement from HIV care in South Africa.
Methods: This is a retrospective cohort study of 5780 HIV positive women 15 to 49 years of age, initiated on ART between 2004 and September 2014 in Johannesburg, South Africa. The incidence and predictors of ART failure (two consecutive viral load>1000 copies/ml) and loss to follow up (LTFU, >3 months late for a scheduled visit) during 24 months post-delivery/equivalent time were assessed using Cox proportional hazards modelling.
Results: Compared to non-pregnant women (Rate 5.6 per 100PY), women were more likely to be LTFU after a prevalent (Rate 13.7 per 100PY; HR 8.2, 95%CI: 6.3-10.6) or an incident pregnancy (Rate 10.1 per 100PY; HR 5.0, 95%CI: 4.0-6.2). The risks of ART failure following an incident pregnancy (Rate 5.9 per 100PY; HR 2.2, 95% CI: 1.6 - 2.9) and in the risk in the non-pregnant group (Rate 7.6 per 100PY; HR 1.9, 95% CI: 1.4 - 2.6) were higher than the risk after a prevalent pregnancy (Rate 4.9 per 100PY). Predictors of postpartum ART failure were being anaemic at delivery (HR 1.25, 95% CI: 1.01 - 1.54), having a low CD4 (< 350 cells) (HR 1.9, 95% CI: 1.6 - 2.4) and meeting the definition for LTFU (HR 1.4, 95%CI: 1.1-1.9). When stratified by CD4 count, among women with low CD4 (< 350) at delivery, the hazard of failure in the incident pregnancy group remained higher than in the prevalent pregnancy group (HR 2.5, 95% CI: 1.8-3.5). There was no difference in the high CD4 strata.
Conclusions: The risk of HIV treatment failure remains high among postpartum women, particularly those who conceive while on ART. The results highlight the importance of strengthening retention and monitoring efforts for postpartum women to sustain the benefits of the PMTCT program.