Involving communities in nutritional screening in Ethiopia

By Ken Bailey

Ken Bailey worked on a contractual basis for WHO Geneva from 1991 to 1997 and for a considerable part of that time had responsibility for the "nutrition in emergencies" sub-programme. Since leaving Geneva Ken has worked for Tearfund/UK in Southern Sudan, the United Mission to Nepal and another INGO in Ethiopia.

I began working as a volunteer nutritionist for an international non-governmental organisation (INGO) in July 2000. The INGO had been supporting rural development work in 3 districts (total population 518 000 in year 2000) of North- West Ethiopia (Amhara region) for 15 years. In 1999 there was near-total crop failure in both rainy seasons (belg and meher) and in the first half of 2000 the belg rains came too late for normal planting. This meant that there was almost no harvest for three successive seasons.

In response, the INGO provided a general ration (wheat, CSB and cooking oil) for approximately half the households from July to November 2000 inclusive. Targeted households were selected on the basis of wealth ranking and identified through repeated consultations between Farmers Association leaders, the district administration and the INGO.

• This response effectively averted famine.
• There was virtually no displacement of households or consumption of famine foods.

Tumultuous scenes at food distribtuion sites Ken Bailey

The organisation wanted to monitor the impact of the food relief on nutritional status through sample surveys repeated monthly. Because of lack of manpower it wasn't possible to do this monthly so one survey was carried out in August (before the food distribution in late July could have had any significant impact) and a second in mid-October.

August survey findings

In the August survey two districts had a moderate rate of wasting in children <5 years (approximately 7% were below- 2 z-scores WfH and WfL) while in the third district the rate of wasting was far higher at 16% (<-2z-scores). This implied that there were about 2500-3000 wasted children in each district.

Adult nutritional status was also measured. The mean BMI was low - near 20.0 kg/m2 combined for men and women. But there was no correlation between nutritional status of children and adults in the same households. (This was interpreted to mean that other factors e.g. 'young-child' feeding practices and incidence of diarrhoeal disease played a more significant role than food availability in the prevalence of wasting in young children.)

Highlands of NW Ethiopia (showing highland, middle highland and lowland levels)- Ken Bailey

Two months on... In October the prevalence of wasting was more or less the same although the situation in one district improved while in another it deteriorated - apparently due to spread of dysentery. Apart from overall food shortage, poor feeding practices (identified in surveys on young-child feeding) appeared to have an important role in continued malnutrition amongst this population.

As the new harvest was expected in November, and the rainfall and agricultural conditions up to then were fairly good, an improvement in the situation was expected in subsequent months. However, my view was that i) the level of wasting was likely to remain above 5% - with aggravating factors (especially diarrhoeal disease) and ii) the appropriate response should include at least health and nutrition education, focusing on the prevention of diarrhoea and improved feeding practices for young children. With current practices most children begin solid feeding after 12 months and breast-feeding is very prolonged. Meals are given to children about 3 times daily.

The conventional approach at this point in an emergency project cycle would have been to continue with sample nutrition surveys e.g. at 3-monthly intervals. However, this would not have readily paved the way for action other than another round of relief distribution when/if the prevalence of malnutrition rose above a certain point.

Surveillance versus surveys

I recommended that it would be more useful to implement a total community based nutritional screening with a nutrition education programme targeted at households with wasted children rather than repeating cross-sectional nutrition surveys. This would involve training volunteers at the level of each hamlet (gott) which consists of about 50 households. At present community health agents (CHAs) are found only at the farmers' association level (5,000- 10,000 people).

A programme was subsequently prepared to train volunteers at gott level to:

1. Measure weight and length and record it on a specially designed WfL chart - one chart for each gott with measurements repeated every 3 months.
2. Carry out education on prevention of diarrhoea and optimal feeding practices for young children in households where a wasted child was found.

Supplementary foods could also be provided for these children if found necessary.

Involving the community

Meetings were held with all leaders of farmers' associations to explain how child malnutrition contributed to child mortality, and how to identify malnourished children and counsel the households. The leaders of the associations would therefore effectively manage the total community screening of young children on a continuing basis. They welcomed this approach and accepted the challenge.

The survey must go on despite the rain! - Ken Bailey

The volunteers at gott level were to be trained by health staff of the INGO, the CHAs and community development agents who operate in each farmers' association.

It seemed to me that it would be much more productive to involve the communities themselves in a survey process thereby achieving full community coverage, rather than to carry on with traditional sample nutritional surveys. In this way the long-term nutritional problems would be more comprehensively tackled from within each community.

I also came to believe that sample nutrition surveys of emergency-affected populations may often yield results that are in the range where educational action and perhaps targeted supplementary feeding are appropriate responses. It would be good under these circumstances to consider one further step - organising community-based total screening of vulnerable groups - as a regular "development" phase of the initial relief effort. This could then be planned from the beginning as a follow-up response. Otherwise large numbers of wasted children may remain in the communities with no effective action taken unless/until the prevalence rises again to the level where universal supplementary feeding (i.e. blanket) or generalised ration distribution are seen as the appropriate responses.

For more information contact Ken Bailey at baileykv@ozemail.com.au