Responses to: ‘An evaluation of the impact of a US$60 million nutrition programme in Bangladesh’

The article by Hossain et al. (2005) in the January issue of Health Policy and Planning uses a flawed data set to provide a biased and unfortunate portrayal of the Bangladesh Integrated Nutrition Project (BINP), a genuinely innovative public health programme reaching huge numbers of poor women and children in that country. First, while the field methods utilized are only partially divulged (even the original 70 page report which includes the 2002 survey does not include the questionnaires, nor detailed sampling methodology), a methodology which uses as controls sub-district areas adjacent to project areas is inherently flawed. Spillover of project effects would certainly be expected, particularly after as many as 5 years of project implementation. More fundamentally, the authors readily admit: ‘This ex-post study design is inherently weak as it is not possible to control for any differences in the rates of malnutrition between project and non-project areas at the start of the intervention’ (p. 36). The importance of such weakness can be seen by differences among subdistricts at BINP baseline with height-for-age z-scores ranging from –1.9 to –2.7. There is tremendous variation in nutritional status within and between communities in Bangladesh.