Field Exchange Mar 2003: Nutrition assessments in Zimbabwe - a local perspective

By George Kararach

For the past two years, George Kararach has worked as a consultant policy analyst in Zimbabwe - for the last year working with UNICEF Zimbabwe.

The support of UNICEF in preparing this article is gratefully acknowledged. The opinions expressed in this article are solely those of the author, and not of any particular organisation.This article outlines the author's personal experience of nutrition-related assessments in Zimbabwe during the recent emergency crisis, and provides some recommendations for future assessments.

Following deterioration in the food security situation in Zimbabwe at the end of 2001, by December 2002, a number of emergency assessments had been carried out in the country. Some of these have been multi-sectoral, covering water and sanitation, health, nutrition, food security and adolescent reproductive health.

The crisis in Zimbabwe is acknowledged to be a complex emergency and is, at least in part, manmade. However, the argument by the humanitarian and international community that the crisis of governance is a major contributor to the humanitarian crisis, has met with a lot of resistance, if not outright denial, by the government of Zimbabwe. This, in turn, has contributed to what can be described as the politicisation of data1, with the outcome of assessments contested by both donors and government, depending on who conducted the assessment. Attempts to create broad consensus around assessment findings has meant the assessment process has been slow, and in some instances painful, with technical considerations swept away in favour of pragmatic political ones.

In seeking to describe the dynamics around nutrition assessments2 in Zimbabwe during this period, a broad analysis is necessary in order to interpret nutrition assessment data - especially in an environment where data are likely to be 'politicised'.

Methods of nutritional surveillance

The use of nutrition surveillance as a tool for early warning systems is fairly well understood. Many would argue that nutritional status monitoring can provide sensitive and timely information regarding 'human stress'. In particular, it crucially informs when a food crisis does become a famine, or at least begin adversely to affect nutritional status.

Nutrition surveillance has been perceived as crucial in monitoring trends in the humanitarian crisis in Zimbabwe. Findings of assessments conducted in 2002, along with the results from the demographic health survey of 1999, are summarised in table 1.

A number of different methods have been used in Zimbabwe to collect national level nutrition data.

Table 1 Nutritional status in children under 5 years of age in Zimbabwe

Under five population surveyed (n)
Weight/height (<-2 sd)
(<-2 sd)
(<-2 sd)
Demographic Health Survey, 1999
Ministry of Health/ UNICEF, May 2002
VAC, August 2002

Nutrition surveys

In May 2002, the Ministry of Health and Child Welfare, and UNICEF carried out a nutrition assessment. Using a two-stage 30x30 cluster method, 23,400 children were sampled in 28 districts of Zimbabwe and a prevalence of 6.4 % moderate malnutrition (weight-for-height <-2SD) was found. A significant limitation of this assessment was the interpretation of anthropometric data without considering other relevant factors, such as coping strategies, livelihoods and childcare practices.

Emergency food security assessments

This type of assessment was carried out in August 2002, and repeated in November/ December 2002, by the Vulnerability Assessment Committee (VAC) - a conglomeration of government and humanitarian agencies. The aims of the assessment were to determine whether households were able to meet their immediate food needs (i.e. any food deficit), and to identify vulnerable households and their coping capacities.

This approach integrated analysis of the household economy, coping strategies, dietary intake, agricultural inputs, and nutritional anthropometry. Using purposive multi-stage sampling, a map of the livelihood systems in Zimbabwe was developed and used as a basis for defining household food economy groups. Food economy zones were then defined based on expert knowledge within the country - in a large part, reflecting livelihood and land-use systems. Twenty-five food economy zones were identified in Zimbabwe and households were randomly sampled from within each zone. Food economy analysis and anthropometric assessment were conducted at a household level. Fieldwork was carried out in all the districts (57 at the time) and a total of 1,460 households were sampled. Results were scheduled for use in targeting, and to determine the relative importance of food and non-food responses.

Sentinel site surveillance

This form of nutritional surveillance took place through the national Child Supplementary Feeding Programme (CSFP), set up by the Nutrition Working Group3, as well as through community based sentinel sites operated by the Ministry of Public Service, Labour and Social Welfare. The CSFP provides data on weight-for-age through a clinic based growth monitoring system, while the community based system provides weight-for-age and height-for-age data for children under-five from randomly selected households (4,500). Owing to the humanitarian crisis, both of the systems were re-vitalised. In fact, the community based sentinel site system arose from a surveillance system that operated in Zimbabwe in the 1990s, which monitored the impact of structural adjustment policies.

While data from the CSFP sites are not nationally representative, they are perceived by many as suitable for providing information on the impact of the blanket feeding of under-fives programme, in the context of the emergency. The community-based sentinel site data provides more nationally representative data, since the sites were selected randomly across the various land-use systems throughout Zimbabwe.


There have been a number of constraints experienced, both in implementation of assessments and utilisation of the data collected.

Politics: Only three major nutrition assessments took place in 2002 due to competing priorities amongst the various stakeholders, most notably between the government and the humanitarian community. Resisting pressure for more assessments, the government argued that the priority was getting food assistance to the communities (particularly when it wasn't being delivered) and that there was a risk of survey fatigue. The government's position was undermined by 'hearsay' amongst members of the humanitarian community, suggesting that results of the May 2002 assessment (involving the Ministry of Health and Child Welfare) were unreliable. However, no hard evidence was produced to substantiate claims that the methodology was flawed, or that there was 'tampering' during the data cleaning process.

Political dynamics also affected security during surveys. Field worker teams, conducting surveys outside the state system, attracted suspicion with threatened attacks by vigilantes in rural communities.

Lack of context for interpretation of results: Some technicians have suggested that since severe malnutrition levels could be classified as 'normal' for the area (average prevalence 1.5% severe malnutrition), nutrition interventions, especially therapeutic feeding, were not justified. However, analysis of anthropometric data in conjunction with other factors such as coping strategies, makes it apparent that preventative nutritional interventions are urgently required.

Preparedness is also important. Currently, therapeutic feeding is carried out within hospitals and, although there is no increase in the percentage of severely malnourished, the absolute numbers of severely malnourished are still significant and could well rise. There is, therefore, a need to prepare for this, such as training of health workers and improvement of existing facilities.

Inter-agency co-ordination: Poor inter-agency coordination has impeded both the speed of survey implementation and the subsequent use of results by agencies. Since May 2002, the Nutrition Working Group (NWG) has operated as the co-ordination secretariat for nutritional surveys/ surveillance. Despite its existence, there were instances when agreements on the need for a survey were not readily secured. There were also differences between agencies over the most appropriate survey methodology. This lack of consensus affected acceptance or endorsement of results by stakeholders.

Limited government capacity: Government capacity to carry out large surveys has been gradually eroded since the mid-1990s. A' brain-drain' of key workers out of Zimbabwe, the impact of the HIV/AIDS epidemic, and a weak economy have all been significant undermining factors. This has, in part, given impetus to the development of multi-agency assessments/ large-scale surveys, the first of which was carried out in August 2002.

Lack of geographical/area based assessments: In Zimbabwe, there have been significant population movements as a result of the land reform process. Considering this, the various assessments carried out in Zimbabwe would have benefited greatly from geographical information on population densities. A better understanding of the new resettlement areas would have complemented findings from the various assessments and strengthened service delivery to the newly resettled populations.

Recommendations for future assessments

There is a need to strengthen the assessment/ surveillance capacity in Zimbabwe, particularly in the context of emergency situations. Existing systems need to be reviewed and constraints associated with the use of the data addressed.

Furthermore, a system is required which marries the strengths of the various survey/ monitoring methodologies currently in use. Such a system should be based on multiple indicators, which allows monitoring of the key factors that influence nutritional status, as well as demonstrating nutritional trends. This inevitably requires an analysis of livelihoods and coping strategies. Critically, means must be identified to reduce the potential for political agendas to influence when surveys are implemented and how results are interpreted.

Finally, any progress will require a strengthening of government capacity to plan and implement surveys and monitoring, and an improvement in the operation of co-ordination mechanisms between humanitarian agencies.

For further information, contact George Kararach at email: gkararach@unicef.org


1 The term 'politicisation of data' is used here to mean that data are employed to further political aims.

2 The term 'nutritional assessment' is used in the broadest sense in this article, i.e. it encompasses assessments of livelihood systems and food security. Furthermore, the term assessment may be used interchangeably with surveillance, since part of the nutrition assessment data in Zimbabwe is based on on-going (routine) data collection (surveillance).

3 The Nutrition Working Group - this was made up of various agencies working on nutrition issues including government, NGOs and UN agencies, chaired by UNICEF.