Abstract
The modelling of HIV trends in Tanzania uses surveillance data from antenatal clinics after adjusting for the reduction in fertility of women living with HIV (WLWH). The rollout of HIV care and treatment services has enabled many WLWH to start on antiretroviral treatment (ART) earlier and are counselled on the options to prevent HIV transmission to their children. The assumption that being HIV positive leads to lower fertility needs to be revisited. This study aims to quantify the effect of ART program expansion on the fertility rate of WLWH in Tanzania. We used Cox regression model to estimate fertility rate and associated factors among WLWH of reproductive age (15-49 years) who enrolled in HIV care and treatment at 57 centers in Tabora from 2008 to 2014. A decomposition of Poisson regression was used to explore the reasons for fertility rate differences observed among WLWH. A total of 6,397 WLWH aged 15-49 years were followed for a median time of 2.0 years. The total fertility rate of 48.8/1,000 person years (95%CI: 44.6 to 52.9/1,000) was inversely proportional to age and WHO clinical staging. WLWH on ART had higher fertility compared to those not started on ART (aHR = 1.5, 95%CI: 1.2-1.9). Being married or cohabiting, having higher CD4 cell count and not using contraceptives were associated with higher fertility rate. The fertility rate after post-ART initiation was 54.95/1,000 and among pre-ART users was 40.52/1,000, a difference of 14.43/1,000 in fertility rate between the groups. In the decomposition analysis, proximate determinants of fertility rate among WLWH on ART accounted for a 93.8% smaller increase than expected. In an era of ART expansion in Tabora region, fertility rates of WLWH increased. Higher fertility rates in women on ART may alter the estimation of HIV prevalence and incidence.