Field Exchange Nov 1999: Evaluation of the Impact of the ACF Therapeutic Feeding Programme in Burundi

ACF and ECHO (European Commission Humanitarian Office) undertook a joint evaluation of the ACF therapeutic feeding programme in Burundi in November 1998. The nature of the emergency in Burundi meant that there were a number of factors which had a critical bearing on the outcome of the programme. Although the situation had improved greatly by the time the evaluation was undertaken, the constraints were primarily related to the very insecure period of 1996/7. These included the following:

  • humanitarian staff and beneficiaries regularly faced security incidents
  • humanitarian organisations were denied access to beneficiaries in the community
  • Therapeutic Feeding Centres (TFCs) only provided day care due to insecurity at night and as staff did not want to stay
    the embargo on Burundi had caused large price increases and restricted imports of key commodities for humanitarian agencies and the general population
  • ACF had problems in locating well trained and qualified staff
  • food aid flows were restricted and irregular
  • the vast majority of the population were farmers - yet had limited access to land.

The main findings of the evaluation were as follows.


  • treatment in the centres was of high quality
  • mortality rates were acceptable (less than 10%) although it had taken more than one and a half years to achieve these rates
  • weight gain was on average more than 8gm/kg/day
  • between June and October an estimated 5000 lives had been saved in the TFCs
  • local staff had been trained and would be able to work in national structures following ACF's eventual departure


  • it was impossible to obtain information on the coverage of the programme as there were no population data or nutritional survey information
  • there was little knowledge or information about what happens to individuals following discharge and reasons for relapse
    for most situations the degree of integration with national structures would be considered insufficient. However, in the Burundi context this may not be an appropriate judgement
  • donor funding for the programme has been on short-term contract (3 months) making it very difficult to plan ahead
  • the targets established for therapeutic feeding programmes in the Sphere guidelines are not wholly achievable in the Burundi context and may not be appropriate
  • the degree of participatory planning with the community has been insufficient


The following were amongst the recommendations in the evaluation:

  • to strengthen integration with national health structure, e.g. establish withdrawal criteria and division of responsibilities during phase out
  • improve capacity to treat adult malnutrition
  • to re-evaluate the roles of the Sphere guidelines* within TFCs and their limitations in highly insecure environments.


*Humanitarian charter and minimum standards in disaster response, ISBN 92-9139-049-6 (First edition 1998. Geneva)