Challenges to the Monitoring and Evaluation of Large Nutrition Programs in Developing Countries: Examples from Bangladesh

Most large-scale nutrition surveys carried out in developing countries have been conducted to compile or update national or regional data that are independent of project activities. At the same time, a growing number of large scale projects are developing monitoring and evaluation systems that also build on periodic baseline and evaluation surveys to assess whether project objectives are being met, and whether improvements in nutritional status observed are indeed attributable to the interventions. Donors and governments are demanding increasingly rigorous evaluations to assess whether their operational investments are cost-effective. Such information also guides the direction of future activities. One intervention, the Bangladesh Integrated Nutrition Project (BINP), has been using baseline and evaluative surveys since 1996 to assess its performance and to inform decision making. The project is widely acknowledged to be one of the most impressive large-scale, community-based nutrition programs in the world. At the same time, a careful analysis of its own survey data has revealed problems in collecting, analyzing and interpreting data--difficulties which are not unique to Bangladesh. This discussion paper uses the BINP survey experience to explore three data-related issues: a) the role and validation of control groups; b) problems in interpreting anthropometric data (specifically, weight-for-height); and c) challenges in the collection of consistently comparable baseline and evaluation data. It is hoped that this discussion might inform the design and interpretation of future baseline and evaluative surveys.