Estimating program coverage in the treatment of acute malnutrition using population-based cluster survey methods: Results from surveys in Burkina Faso, Chad, Democratic Republic of the Congo, and Niger

Grace Heymsfield ^1* Elizabeth Radin ¹ Marie Biotteau ² Suvi T. Kangas ³ Zachary Tausanovitch ¹ Casie Tesfai ¹ Léonard Kiema ⁴ Wenldasida Thomas OUEDRAOGO ⁵ Badou Seni Mamoudou ⁶ Mahamat Garba Issa ⁷ Lievin Bangali ⁸ Marie Christine Atende Wa Ngboloko ⁹ Balki Chaïbou ¹⁰ Maman Bachirou Maman ¹¹ Eva Leidman ¹² Oleg Bilukha ¹²

• ¹ International Rescue Committee, New York, United States
• ² International Rescue Committee, Dakar, Senegal, Dakar, Senegal
• ³ International Rescue Committee, Brussels, Belgium, Brussels, Belgium
• ⁴ International Rescue Committee, Ouagadougou, Burkina Faso
• ⁵ Ministère de la Santé et de l'Hygiène Publique, Ouagadougou, Burkina Faso
• ⁶ International Rescue Committee, N'Djamena, Chad
• ⁷ Ministère de la Santé Publique et de la Prévention, N'Djamena, Chad
• ⁸ International Rescue Committee, Goma, Democratic Republic of Congo
• ⁹ Programme National de Nutrition (PRONANUT), Kinshasa, Democratic Republic of Congo
• ¹⁰ International Rescue Committee, Niamey, Niger
• ¹¹ Ministère de la Santé Publique, de la Population et des Affaires Sociales, Niamey, Niger
• ¹² Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, United States

Background: Despite their utility for program planning, acute malnutrition treatment coverage estimates at the national and sub-national levels are rarely available. Prior work has identified methodological concerns with current approaches.We estimated the point prevalence and treatment coverage of acute malnutrition in 11 districts (or similar subnational areas) across four high-burden countries in Africa using representative cluster-based population survey methods and compared these estimates to those derived from administrative data and other direct methods where available. We also aimed to assess information about risk factors for malnourished children by coverage status.The point estimate of coverage suggests that <20% of eligible children with severe acute malnutrition (SAM) were enrolled in treatment in nine administrative areas. We found that in some contexts, coverage estimates derived using administrative data are useful, while in others, they are not -and that their accuracy can vary by month and year. By comparison, coverage estimates from other direct methods were overestimated and/or outdated, and practitioners tended to overestimate coverage. Coverage did not differ significantly by sex or age of the child but did vary by mid-upper arm circumference (MUAC) at assessment. Measured SAM coverage did not correlate either with measured SAM prevalence or with expected coverage estimated a priori by program staff.Our findings suggest that in the assessed high-burden countries, many more children are eligible for treatment than are enrolled. We present this methodology as an alternative to existing primary methods and a complement to coverage estimates from routine program and population data.