Effective Sampling Methodology for Program Evaluation in Developing Countries

Mass Drug Administration (MDA) is a key intervention implemented to control and treat five diseases classified by the World Health Organization as ―Neglected Tropical Diseases‖ or ―NTDs‖ which affect more than one billion people worldwide. This approach involves the annual administration of safe and effective drug treatments to atrisk populations in endemic countries. Each national NTD control program collects a range of data including drug distribution and MDA coverage data. The Post MDA Survey serves to validate the accuracy of the country reported MDA coverage rates. There are unique challenges to designing and implementing household surveys in developing countries. This paper will present the design effect and the precision of the estimated drug coverage rates of multi-stage surveys from Niger in 2008 based on our experience working with the U.S. Agency for International Development (USAID) NTD Control Program. We will identify effective sampling methodology for program evaluation in developing countries. Specifically, we will focus on the stratification and clustering strategies and identify effective methods to reduce the sampling variance.