Abstract
Objective: Describe changes in sexual behaviour and determinants of unsafe sex among HIV-infected women in the 24 months after delivery.Design: Cohort analysis nested within a prevention of mother-to-child transmission trial in Burkina Faso (n ¼ 339) and Kenya (n ¼ 432).Methods: Women were followed during pregnancy and until 12–24 months after delivery. At each visit, structured questionnaires were administered about sexual activity and condom use, and risk-reduction counselling and condoms were provided.Results: At study entry, a median 2 months after HIV testing (interquartile range ¼ 1–4), 411/770 (53.4%) of women reported partner disclosure, increasing to 284/392 (71.9%) at the final visit. Although most partners were supportive following disclosure, between 5 and 10% of disclosed women experienced hostile or unsupportive partner responses during follow-up visits. At each visit, about a third of sexually active women reported unsafe sex (unprotected sex with HIV-uninfected or unknown status partner). In multivariable logistic regression, unsafe sex was 1.70-fold more likely in Kenyan than in Burkinabe women [95% confidence interval (95% CI) ¼ 1.14–2.54], and in those with less advanced HIV disease or aged 16–24 years. Compared with women who disclosed their status to partners and others, unsafe sex was over six-fold higher in nondisclosers (95% CI ¼ 3.31–12.11), the effect size reducing with increasing disclosure.Conclusion: HIV-infected women who recently delivered have a high potential for further HIV transmission, especially as HIV discordance is common in Africa. Longitudinal care for women, including positive-prevention interventions, is needed within new services providing antiretroviral prophylaxis during breastfeeding – this repeated interface with services could focus on reducing unsafe sex. Much remains unknown about how to facilitate beneficial disclosure.