ED Sociétés, Politique, Santé Publique
Antenatal care coverage within the framework of the 2016 WHO recommendations: recent patterns, individual and contextual determinants, and impact of a community-based intervention in Burkina Faso and Mali
by Dofinissery BOGNINI (Bordeaux Population Health Research Center)
The defense will take place at 9h00 - Florence Nightingale (E112) 1er étage de l'ISPED, Bâtiment TP, Université de Bordeaux, 146 rue Léo Saignat CS 61292, 33076 Bordeaux Cedex
in front of the jury composed of
- Valérie BRIAND - Directrice de recherche - Université de Bordeaux - Directeur de these
- Smaïla OUEDRAOGO - Professeur - Université Joseph Ki-ZERBO - Rapporteur
- Anne CHANTRY - Professeure - Université Paris Cité - Rapporteur
- Renaud BECQUET - Directeur de recherche - Université de Bordeaux - Examinateur
- Raquel GONZALEZ - Professeure - Universitat de Barcelona - Examinateur
- Catherine DENEUX-THARAUX - Directrice de recherche - Université Paris Cité - Examinateur
- Halidou TINTO - Professeur - Université Nazi Boni - CoDirecteur de these
Background Antenatal care (ANC) is essential for monitoring pregnancies. In 2016, the WHO recommendations for antenatal care evolved from 4 so-called "focused" ANC visits to eight "contacts" between the pregnant woman and a qualified healthcare provider. These latest recommendations are poorly adopted or implemented in sub-Saharan Africa (SSA). Although 90% of pregnant women in SSA seek ANC at least once, only 52% benefit from 4 ANC visits. In Burkina Faso and Mali, coverage of 4 or more ANC visits (ANC4+) remains insufficient due to several individual and contextual barriers, highlighting the need to find innovative strategies to increase this coverage. Despite the efforts of health systems, ANC coverage and perinatal mortality remain unsatisfactory. This work addressed 3 objectives: 1) to evaluate the determinants and 2) the evolution over time of ANC and the effect of a community intervention on ANC coverage, 3) pregnancy outcomes and the incidence of malaria during pregnancy in Burkina Faso and Mali. Methods Our work is part of the INTEGRATION project, which is a study of the implementation of a community intervention to improve coverage of intermittent preventive treatment during pregnancy for malaria prevention, carried out in the health districts of Boussé (Burkina Faso) and Kaganba (Mali). The intervention was evaluated through a 2-arm cluster randomized trial (community distribution vs. community + health center). It also aimed to encourage and sensitize women about ANC and refer women to ANC if necessary. This was provided at home, during the seasonal malaria chemoprevention for children, every month from July to October (2022 to 2023). Several datasets were used for our analyses through the following models: -1) an interrupted time series model (monthly aggregate data of the total number of ANC4+ and deliveries extracted from the District Health Information Software version 2, from 2015 to 2023 in Burkina Faso and 2019 to 2023 in Mali) : objective 1; -2) a logistic regression (INTEGRATION baseline household survey data from 2022): objective 2; and -3) a difference-in-differences model (individual routine data from health centers): objective 3. A significance threshold of 5% was set. Results In Burkina Faso and Mali, the 2016 WHO recommendations were adopted without being implemented on a large scale. ANC4+ coverage evolved from 55.5% to 69% between 2015-2023 in Burkina Faso, and from 45% to 50% between 2019-2023 in Mali. In both countries, we showed that the factors associated with not receiving ANC4+ were: perception of the high cost of ANC, long distance and long travel time (≥1h) to reach the health center, and lack of perception of the need to seek ANC multiple times. In Mali, the factors associated with receiving ANC4+ were: history of stillbirth, time spent in ANC (≥1h), and high socioeconomic level of the woman. In Burkina Faso, a moderate and temporary significant increase in ANC4+ coverage was observed, but not in Mali. Finally, no effect of the intervention was found on adverse pregnancy outcomes and the cumulative incidence of malaria during pregnancy. Conclusion The community intervention had a moderate and short-term effect on ANC4+ coverage, with no overall effect on pregnancy outcomes and gestational malaria. The intervention addressed the geographical and cost-related barriers to ANC that were identified in the study area. Determinants of ANC4+ utilization have been identified in both countries, which can be addressed to improve ANC4+ coverage and perinatal morbidity and mortality.