Prevention of Mother-To-Child Transmission of HIV: Cost-Effectiveness of Antiretroviral Regimens and Feeding Options in Rwanda

BACKGROUND: Rwanda's National PMTCT program aims to achieve elimination of new HIV infections in children by 2015. In November 2010, Rwanda adopted the WHO 2010 ARV guidelines for PMTCT recommending Option B (HAART) for all HIV-positive pregnant women extended throughout breastfeeding and discontinued (short course-HAART) only for those not eligible for life treatment. The current study aims to assess the cost-effectiveness of this policy choice. METHODS: Based on a cohort of HIV-infected pregnant women in Rwanda, we modelled the cost-effectiveness of six regimens: dual ARV prophylaxis with either 12 months breastfeeding or replacement feeding; short course HAART (Sc-HAART) prophylaxis with either 6 months breastfeeding, 12 months breastfeeding, or 18 months breastfeeding; and Sc-HAART prophylaxis with replacement feeding. Direct costs were modelled based on all inputs in each scenario and related unit costs. Effectiveness was evaluated by measuring HIV-free survival at 18 months. Savings correspond to the lifetime costs of HIV treatment and care avoided as a result of all vertical HIV infections averted. RESULTS: All PMTCT scenarios considered are cost saving compared to “no intervention.” Sc-HAART with 12 months breastfeeding or 6 months breastfeeding dominate all other scenarios. Sc-HAART with 12 months breastfeeding allows for more children to be alive and HIV-uninfected by 18 months than Sc-HAART with 6 months breastfeeding for an incremental cost per child alive and uninfected of 11,882 USD. This conclusion is sensitive to changes in the relative risk of mortality by 18 months for exposed HIV-uninfected children on replacement feeding from birth and those who were breastfed for only 6 months compared to those breastfeeding for 12 months or more. CONCLUSION: Our findings support the earlier decision by Rwanda to adopt WHO Option B and could inform alternatives for breastfeeding duration. Local contexts and existing care delivery models should be part of national policy decisions.