Chaque année 330 000 (dont 300 000 en Afrique) nouvelles infections à VIH surviennent chez les moins de quinze ans – majoritairement par un mécanisme de transmission verticale d’une mère infectée à son enfant. Pourtant, des interventions efficaces pour réduire la transmission mère-enfant (TME) existent mais ne sont pas toujours disponibles en pratique de routine pour les femmes qui en ont besoin. L’agenda international en cours vise une élimination de l’infection pédiatrique de VIH – et le maintien des mères en vie – à l’horizon 2015. A la lumière de la crise économique mondiale, il apparait opportun de pouvoir juger du montant de ressources nécessaires à la réalisation de cet objectif. Ce travail de thèse dresse le panorama du mode de délivrance des interventions pour la prévention de la TME (PTME) et de la prise en charge des enfants infectés par le VIH dans cinq pays à ressources limitées que sont la Chine, la Côte d’Ivoire, la Namibie, le Rwanda et l’Ukraine. Il apporte une estimation de leur coût et renseigne sur leur efficacité et leur caractère abordable. En outre, une estimation du coût de l’implémentation de potentiels scénarios de passage à échelle des interventions est apportée ; la mesure du nombre de cas (et de décès) non évités sur la période de projection permet de se rendre compte du coût de l’inaction. Enfin, sont discutés deux obstacles majeurs à la diffusion des services de PTME que sont le contexte politique ambiant et l’organisation de l’offre de soins (contrainte en ressources humaines qualifiées notamment). / A decade into the implementation of prevention of mother-to-child HIV transmission (PMTCT) programs, countries are making impressive progress towards fulfilling global commitments and achieving global goals. However, achieving a generation free from HIV will entail sustained effort and commitment in the coming years and estimating the cost of this work is fundamental. In light of the global economic downturn, reliable and up-to-date information on the costs of the dedicated services is needed to estimate the amount of additional resources required to scale up those services within the context of the Millennium Development Goals, and to assist resource-limited countries, external funders and multilateral agencies in planning, identify cost drivers and areas of potential savings. This work pursued the aims (1) to produce a reliable estimation of the costs associated with PMTCT services in five low- and middle-income countries, and (2) to ascertain the range and median unit costs needed for scaling up these services across a prevention, treatment and care continuum. Countries were selected given three key criteria related to their economic level and health spending, the national HIV epidemic status and typology, and the coverage of their maternal, neonatal and child health (MNCH) and PMTCT programs. Data were collected in Cote d’Ivoire, China, Namibia, Rwanda, and Ukraine using a health-care provider perspective from October 2009 to April 2010. In-country samples were designed to reflect publicly funded MNCH facilities that were delivering a define range of PMTCT and pediatric HIV services. The package includes HIV testing and counseling, male partner testing, CD4 testing, antiretroviral prophylaxis provision, community-based activities, HIV-exposed infant prophylaxis, and 2-year post-partum family planning. Additional information on prices and volume of services was collected from national government bodies, international donor partners, and non-governmental organizations. Cost data were then analyzed using a bottom-up approach of "micro-costing" where costs are calculated by identifying the actual resources used for each patient. Primary outcomes are costs per specific end-users and nation-wide total program costs, subdivided by major cost categories. After that, we simulated the expected costs over time according to a scale-up analysis estimating the costs needed to increase service coverage under various scenarios of coverage and antiretroviral protocols. These scenarios are incremental in their content: each subsequent scenario adds components and costs to the previous one. Future costs are discounted at a 3% rate. The affordability of each scenario (compared to the previous one) was concomitantly examined through the generation of incremental cost effectiveness ratios. The cost of inaction, measured in both terms of cases non-averted and deaths non-avoided throughout the scaling-up period, was ultimately assessed.