Field Exchange Nov 2002: The need for nutrition research in emergencies

Summary of published paper1
With the exception of the management of severe malnutrition, there has been little research specifically conducted to support emergency food and nutrition programmes, according to a recently published article. This has contributed to a lack of empirically based guidance regarding how to accurately identify problems and select the most effective means to achieve desired objectives. As a result programming decisions have often been based on extrapolations, anecdotal evidence and intuition as a result. Consequently, donors, implementers and affected governments (who frequently hold contradictory opinions) expend time and resources on controversies about when and how to act. These controversies reveal inadequate knowledge for which research is urgently needed.

The authors provide examples of past controversies where research was urgently needed, such as the sale of food aid by refugees in Uvira, Zaire in 1996. Here there were radically different views regarding what motivated these sales, which had drastically different implications regarding how interventions should change. Food aid donors and the WFP assumed that the massive sales were a sign that the refugees had adapted to their new environment, were acquiring food on their own and thus needed less food aid. UNHCR and NGOs, on the other hand, felt certain that sales signaled other problems. They felt that refugees were not familiar with the whole grain maize, that food aid commodities took a long time to cook requiring a lot of scarce fuel, and that refugees needed money to buy other foods to balance their meals and to purchase essential non-food items.

In the event, subsequent research showed that both interpretations were correct. A small proportion of the refugees did not need all the food aid and so sold the excess. However, a much larger part of the population sold food aid, although they sorely needed it, to satisfy their daily energy requirements.

The authors argue that the existence of such contradictions is one way of recognising urgent research needs. They provide examples of issues and questions where insufficient knowledge leads to conflicting views and where research is needed.

i) What is the best way to reduce food aid resources in recovery, while ensuring the most food insecure receive adequate food aid? Some believe in limiting food aid to only the poorest households while others advocate reducing the general ration to everyone, thus relying on subsequent redistribution within the community to ensure the poor receive what they need.

ii) Some believe that long-term recipients choose to depend on food aid rather than work while others argue that emergency affected persons immediately initiate coping strategies to obtain other food.

iii) Is under-five wasting the best indicator of population nutritional status when there is evidence, in some situations, of normal levels of wasting in under fives but malnutrition amongst older children and adults?

iv) In resource scarce situations, admission criteria for therapeutic feeding centres may be lowered to less than 60% weight-for-height. Would more lives be saved if those between 60-70% weight-for-height were admitted?

v) In Africa, UNHCR spends about 11 cents per day per refugee. In the Balkans, the figure is 1.23 dollars. What are the nutritional implications of such disparities in spending? Is the international community violating African refugees' basic human right to food? What institutional changes are needed to ensure equitable nutritional assistance to all emergency-affected populations?

The authors suggest that when pressed by the urgency of emergency circumstances to decide, decisions end up being based on anecdotal evidence, common wisdom, logical assumption or simply the hope of doing good. Such decisions create precedents that become institutionalised without adequate verification.

Propositions and conclusions

It is proposed that new techniques and institutional changes are needed to promote research to define and validate emergency programmes. Traditional research approaches need modification since the approaches, tools and methodologies developed for nonemergency settings often do not suit emergency settings. For example, controlled experiments may be unethical as well as impossible to implement. Large surveys are often impractical. Participatory approaches may be difficult where communities have been torn apart and the vulnerable are not represented by the leadership. Emergency affected populations may exaggerate severity of circumstances to gain additional help and qualitative data collection must consequently incorporate numerous cross-checks to ensure validity.

The case is then made that existing structures of international assistance are not geared to promote research or accommodate learning from experience. Relief agencies fear that they will be blamed for shortfalls revealed by evaluations. Currently, few organisations maintain professional emergency nutritionists on staff so that there is limited institutional learning.

The paper concludes that there is a need for qualified observers in the field to capture more information. This would be invaluable in ensuring that programmes remain effective as conditions change. Data collected could also be used for research purposes. However this will require the development and mobilisation of specially trained professional emergency researchers, who can monitor and record information to support research during emergencies and also continue analyses between emergencies. There must also be a means of rapidly mobilising resources for research for these professionals.

Finally, to ensure the quality of emergency nutrition research and its application, a system of peer review is needed.


1 Reed B, Habicht J. P and Garza C (2002): Translating nutrition research into action in humanitarian emergencies. Journal of Nutrition, vol 132: pp 2112S-2116S.