Supplementary Feeding in Mandera: The Right Intervention?

View of the Mandera Camp

Lourdes-Vazquez-Garcia worked for MSF Spain in the Mandera feeding centres during the period covered by this article. She subsequently qualified with an MSc in Human Nutrition at the London School of Hygiene and tropical Medicine in September 1998 writing this article shortly after finishing her summer dissertation which was based on her analysis of the MSF Spain intervention.

Background

Central Mandera is located in northeastern Kenya, 2 km west of Somalia and 5 km south of the Ethiopian border. The estimated population is 37,900, distributed in eight main locations, within a 12 km radius. The population is mainly ethnic Somali, practising traditional nomadic-pastoralism. Central Mandera has been a settlement for refugees since 1991 when many people in the west of Somalia fled to Kenya in search of security and food. About 10,000 refugees still live there in the former refugee camps of Shafshaefi and Handadu, although these are now considered official locations. During the severe drought of 1991-1992 animal holdings were drastically reduced in some population groups in the district rendering many families destitute, and forcing them to migrate to Central Mandera. These displaced pastoralists increased the numbers of urban poor as the limited economic growth in Mandera has been insufficient to absorb the 'unskilled' pastoralist labour.

Evolution of MSF Spain intervention

In March 1996 Northern Aid (a local NGO) contacted MSF Spain out of concerns for the pastoralist population of Mandera district who had been badly affected by the prolonged drought affecting the area over the previous two years. An assessment of the general situation and a nutrition survey were carried out in May 1996 in Central Mandera, as this was considered the most affected area where the majority of the district population were settled.
The nutrition survey results showed 32.4% global acute malnutrition (GAM) (weight for height <-2SD) with 4.6% severe acute malnutrition (SAM) (<-3SD). The assessment also showed food price inflation occurring from December 1995, an increase in animal morbidity and mortality rates, a rising number of female-headed households, reduced purchasing power, large numbers of destitute people around the market and a small migratory movement to the town. All these phenomena were indicative of an acute and severe food crisis amongst the population. However, another assessment in June 1996 concluded that the situation was not quite as bad as originally thought in terms of food security in the district; although the situation in Central Mandera was found to be more critical.
In response to the situation in Central Mandera, MSF Spain began a nutritional intervention in the form of feeding centres with the objectives of:

• Decreasing malnutrition-related infant mortality.
• Preventing a deterioration of the nutritional situation and reducing the malnutrition rate to half that found in the May nutritional survey.

Supplementary Feeding in Mandera

Based on the assumption that twenty percent of the whole population were under five (7580 children) and the survey results, five Curative Supplementary Feeding Centres (SFC) were opened (three on-site and two take-home). The take home feeding centres were established in order to facilitate accessibility for those in more distant locations (see Table 1). In addition, a Therapeutic Feeding Centre (TFC) was opened in Mandera town.
Between September-October 1996 a subsequent food assessment in conjunction with rising readmission rates in the feeding centres confirmed a deterioration in the situation, and led to the implementation of a Targeted Food Distribution (TFD) to vulnerable families in Central Mandera, and a General Food Distribution (GFD) in Fino division. These additional programmes were also implemented by MSF Spain.
In Central Mandera families were targeted through the SFCs (any family with a malnourished child was considered at risk). The family ration consisted of 1100kcal/person/day and the commodities dispensed were beans, maize and oil which were distributed on a monthly basis. A total of 3,000 families benefited from the TFD, accounting for 18,000 people in the area (six members per family).
Heavy rains finally came in April 1997. It was at this point that the Kenyan government recognised the emergency situation and applied for help to the international community (it was coincidentally election time in Kenya). MSF handed over the TFD to the government in July 1997 which had started to receive food from the World Food Programme (WFP). But continuing monitoring confirmed that much of this food did not reach the target population although it should be noted that food aid pledges to WFP were limited so that the amounts of food would in any case have only lasted for a short period of time. At the time of writing this article (August 1998), MSF still continue to operate SFCs in Mandera. Although there were substantial improvements in the nutritional situation in the first six months of the programme, malnutrition rates have not fallen below 20% since then.

What follows is a summary of my personal evaluation of the programme and discussions with the MSF team about the programme that started on my arrival in Mandera.

The Supplementary Feeding Programme

When children are discharged from SFCs they receive Unimix (1000kcal/day) for the first two months. As a way of targeting vulnerable groups, every pregnant or lactating woman having a malnourished child also receives the Unimix.

Performance of the Feeding Centres

The number of children who attended the SFCs from July 1996 up until March 1998 was 11,250. Evaluation of feeding centre performance showed that targets (using MSF reference values) were generally met over the 21 months period studied, except for some months in which aggravating factors (cholera outbreak, heavy rains and floods, malaria epidemic) influenced the outcome or when certain operational problems occurred (Figure 1 & 2).

At all feeding centres, attendance by beneficiaries has been above 80%, and estimated coverage has been 90%. Also, 80% of those discharged monthly (450 to 943 children) had reached 85% of the reference median WFH, reflecting the overall effectiveness of the feeding centres. In spite of the overall adequate performance of the feeding centres there were a number of issues and questions about the programme which concerned the MSF team:

1. were the same children continuously being readmitted and if so, why?
2. why did malnutrition rates not improve following the TFD?
3. did the high prevalence of malnutrition reflect a chronic problem and, if so, is the MSF strategy of establishing feeding centres the most effective use of intervention resources?

Are the same children continuously being readmitted?

Re-admissions

Some re-admissions are to be expected due to poor feeding practices at home or recurrence of disease. Figure 3 shows the percentage of re-admissions over a period of months. Readmission were defined as any child who was re-admitted to the programme within two months of discharge. However, we suspected that these figures underestimated the extent of re-admissions as a survey in June 1997, found that 33% of malnourished children had previously been in the programme. Under-estimates may have occurred for two reasons: The definition itself (which only accounts for children enrolled within the previous two months), and the reluctance of mothers to tell the truth because of fear of not being admitted again or being 'punished' by the staff. These high levels of re-admissions in conjunction with the evidence for 'double registration' (large numbers of double registrations were discovered during a follow up survey of defaulters in June 1997) suggests that the feeding centres may be used to provide an income/food transfer to the entire household. In other words children may be intentionally under-fed so that households can gain access to programme food resources. Once enrolled children may then remain under-fed as a means of ensuring increased food intake for other family members and in order to release or create income for household expenditure. Gaining access to the supplementary feeding programme effectively becomes part of the overall household coping strategy.

Why malnutrition rates did not improve following the TFD

Although the TFD was implemented for a six month period in Mandera malnutrition rates remained the same. However:

• beneficiary families were selected on the basis of having a malnourished child at SFCs, while studies suggest that such households may not necessarily be the most food-insecure. This targeting method was used as it was felt to be the only feasible targeting strategy, i.e. it would limit the degree of 'leakage' and corruption and would be accepted by leaders.
• the distribution of family cards only took place over a one month period so there was only partial coverage of vulnerable groups.
• only a half ration was distributed (based on the perception that a full ration was not needed).
• rations were shared through community social networks.

Furthermore, although the primary objective of the TFD in Central Mandera was the alleviation of malnutrition in children it does not follow that beneficiaries were using that food to tackle problems of malnutrition. They might well have had other priorities, such as preserving animals or buying seeds, or maintaining the work capacity of the adult income-earning members of the household. It was not unusual to see "MSF food" (mainly Unimix and beans) in the markets.

Did the high prevalence of wasting reflect a chronic problem and is the MSF strategy of establishing feeding centres the most costeffective use of intervention resources?

A more detailed analysis of the causes of malnutrition led MSF to the conclusion that chronic food insecurity and poverty are the key factors determining the high prevalence of wasting with the situation further aggravated by environmental conditions (cholera outbreaks have occurred every year for the past 3 years while this year there has been a malaria epidemic). Lack of availability/accessibility to health services aggravated conditions still further. In Central Mandera high wasting rates are a permanent feature. Given this fact, what are the solutions? Reducing food aid now without a parallel expansion in genuine development activities will serve to increase further the vulnerability of the population. At the same time it needs to be appreciated that there may be negative long-term effects of food aid provision, e.g. the diversion of agency resources which might be used for other priority activities and the possibility of free food handouts becoming a disincentive to self-help activities.
In this case supplementary feeding programmes are acting more as a food/income transfer equivalent to families. My view is that interventions should address economic as well as food needs at the family level. At the present time, SFCs in Central Mandera are seen by MSF as a 'holding' operation whereby under-five mortality is reduced, by keeping malnourished individuals at a weight associated with minimum risk of illness or death and as an efficient form of damage control preventing a further deterioration of the nutritional status.
However, in this type of situation nutrition interventions need to combine a number of activities. The association of poor nutritional status with poverty, indicates long-term development needs, including income-generating activities and public health measures.
In order to come up with specific and appropriate solutions, on site research into the root causes of the problem (malnutrition being the symptom) needs to be conducted in conjunction with the community. This would involve collaboration between beneficiaries, Government, and humanitarian agencies.

Guidelines should be expanded to include different 'famine' scenarios as the type of chronic emergency found in Mandera is becoming increasingly prevalent in many parts of Africa. Guidelines need to acknowledge this and provide alternative and new response options based more on an analysis of underlying problems and chronic/structural poverty rather than by advocating the use of trigger levels of wasting to stimulate automatic implementation of emergency selective feeding programmes.

Post-Script by Amaia Esparza

A nutritional survey was carried out in Mandera between the 16th and 20th of November and found a GAM of 20.4%. Food distribution was subsequently changed so that children received 4kg of premix/week instead of 2kg. Children above 75% were being discharged and got 8kg of unimix instead of 4kg.

The programme was closed in mid-December '98.