Methodology for a Nutritional Survey among the nomadic population of northern Mali
By Eva Vicent and Núria Salse
Eva Vicent has a background in nursing studies in the University of Valencia and is currently working for Action Against Hunger as Nutrition Coordinator in Swaziland. She has been involved in health and nutrition programmes since 2000, working in several countries including Angola, Guinea Bissau, Indonesia, Malawi and Mali.
Núria Salse is the Health and Nutrition Advisor in Acción Contra el Hambre. Previously she spent several years working on nutrition and medical programmes in Angola, Guinea Conakry, Niger and Argentina.
The authors acknowledge Thierry Métais who contributed to developing the methodology, and Waltraud Novak and Mathias Grossiord who undertook the field work and contributed to improving the method.
Acción contra el Hambre España (ACFE) has been working in Gao and Kidal, two regions in northern Mali, since 1996. The population comprises largely farmers and herders in the valley area of the Niger River. Those living in the remaining vast expanse of the two regions are nomadic or transhumant. Both regions have similar annual rainfall ranging from 100 millimeters (mm) in the north to 350mm in the south. The nomadic population tends to practice an opportunistic type of livestock farming. They are continuously on the move following a set route that only varies when faced with a bad year that requires them to adapt in some way.
The ecological, geographical and social environment of this nomadic population does not allow for the application of standard methods for monitoring and assessing their vulnerability. Particular challenges for monitoring are that it is a large area of land with a low population density, it is hard to assess available resources at the end of the rainy season and the movement of the population and herd. These factors also make it very difficult to assess the nutritional situation. On previous occasions, nutritional assessments were mainly conducted on sedentary nomadic populations1 in accessible camps2. However, these survey findings could not be extrapolated to the whole population of the region - in some cases, more than half of the population lives outside camps, is on the move, and has a very different set of problems to the sedentary residents. The challenge, therefore, is how to randomly select a representative sample of the whole population in the region.
The census carried out in Mauritania in 1965 was the first large scale survey of a primarily nomadic population. For this census, the 'fraction'3 was taken as the survey unit and lists of all people belonging to a fraction and its geographic area were drawn up. Heads of each fraction were also recorded. The problem with this methodology was that it required a lot of preparation time and not all of the tents belonging to the different fractions could be identified4.
Exhaustive sampling methodologies had been carried out previously in Mali, using watering points and sedentary areas as starting points. However, given the wide geographic dispersion of the population and the difficulty in establishing the exact location of the camps, this method and ensuing results were also flawed.
Five different group sampling methods for nomadic populations have been put forward by Graft-Johnson (1979) based on a revision of the methods proposed by the United Nations (1977)5:
The group assembly method: The difficulty with this approach lies in trying to get all of the population to gather at a given point.
Camp method: In order to employ this method there has to be good population census data or lists of those in the camps with exact locations known.
Social structure methodology: This is only viable if the head of each fraction can draw up a list of the exact composition and whereabouts of each camp. There is a risk of omitting or double counting subjects, and finding the subjects can also be problematic.
Numbering by areas: As with other methodologies, there is an inherent problem with coverage, as it is not possible to cover an entire area (even if small). There is also the risk that subjects could move from one area to another.
Watering point method: This method can only be applied if the subjects of the survey are the same people who go to the watering holes.
Given the options, in November 2005, Acción Contra el Hambre designed a pilot study to assess nutritional status in the study area. This was based on the area numbering method and on the cluster sampling approach advocated by SMART (Standardised Monitoring and Assessment of Relief and Transition6). In the SMART methodology, if the settlements are small there is the option to have more clusters, with fewer children per cluster, to ensure there will be a sufficient number of children at each site.
Pilot study methodology
The study used weight/height indicators expressed as z-scores7 and oedema. The number of clusters was calculated taking account of the number of children that a team could test per day. In a nutritional survey pilot test, carried out in the Anderamboukan district (Gao region), 14 children were selected per cluster. The sample size was calculated on the basis of the population in each area (as indicated by official statistics), the expected prevalence, the precision and the homogeneity of the sample (design effect).
Population estimates were reached in conjunction with the local authorities, given that there are no official figures per town, only per fraction or family. An estimate of the populations of the different sedentary areas and nomadic camps was made assuming the nomadic area had a homogenously distributed population throughout the survey area and based on discussions with different members of the community. The Emergency Needs Assessment computer programme was used to analyse the clusters (example in Table 1).
|Table 1: Distribution of clusters, population size and corresponding grid|
|Geographical unit||Population size||Assigned cluster|
In the nomadic area, each cluster is represented by a square8. Clusters were only retained if the central point was inside the survey area. Figure 1 shows the grid distribution in Anderamboukan for the 11 selected clusters that correspond to the nomadic area.
In order to select the children for the survey, the central point (centroid) of each cluster was taken as the starting point. This point was identified by using the 'MapInfo' programme and then transferring it to the Global Positioning System (GPS). On arrival at this point, the survey team followed the Epi method, which consists of throwing a pen into the air and following the direction it is pointing in when it falls to ground until arriving at a camp or villages. If no village was arrived at by the time the limit of the cluster had been reached or if the direction the pen was pointing in proved to be inaccessible, the team would return to the central point and throw the pen again. Once the team had reached the first settlement it then moved on to the next, and so on until the cluster was fully covered (See Figure 2).
If the centre of the cluster proved to be in an inaccessible place (for example, a lake), the starting point became the closest accessible place to the western edge of the central point. If, having set off in eight different directions, the cluster was still not complete, the survey team proceeded to the cluster directly to the right of the one they were working in.
If, following the direction indicated by the pen, a town or settlement was reached that had already been tested, It was skipped by stopping 3 km before the town or settlement and restarting 3 kilometres from the other side of the settlement or village in the same direction.
In previous years, Acción contra el Hambre had elected to only test sedentary areas or to use an exhaustive methodology to assess nutritional status of the population in the region. This time, by applying different methods, we have obtained data on both the sedentary and nomadic populations. Also, by predetermining the sample and including the most isolated areas, as opposed to using an exhaustive approach, we have obtained a highly representative sample.
At the same time, the reduction in the number of subjects per cluster has allowed us to complete the surveys in each cluster. On average, only one cluster per survey has had to be complemented and therefore completed with children from a subsequent cluster.
This method also ensures the random selection of subjects, although two limitations did have to be taken into account. First, the 'border effect'. Selecting only those clusters that have their geographical centre inside the district and testing them in full means that sometimes the area surveyed does not coincide exactly with the existing administrative boundaries. Secondly, sampling in nomadic areas does not take into account the fact that the population distribution is not homogenous and that some areas are more densely populated than others.
The use of GPS technology during the survey did not pose any difficulties as it is very simple to use and was included in the team training from the start. Sometimes, however, geographical circumstances and difficult terrain made finding the central points, and the teams' w o rk in general, more difficult.
Following the first pilot test, it was decided to carry out exploratory trips before the surveys in order to find the central points and identify any possible geographical difficulties. This meant that the teams had received precise instructions on how to get to the central points and were able to modify the methodology according to the terrain, if necessary. This exploratory work made the teams' work considerably easier although it also increased the preparation time and cost of the survey.
Conclusions and recommendations
Using a group methodology for nutritional surveys in pastoral areas where the population is scattered presents problems in terms of locating and measuring certain groups of the population. The method of selecting groups that was proposed for the pastoral areas and described in this article facilitates the principle of random selection of children. Reducing the number of children per group also ensures that all the required measurements for each group were completed. The methodology used can contribute towards assessing the nutritional situation of the pastoral population of Mali.
If further nutritional surveys are to be carried out using this methodology, certain problems encountered should be addressed as follows:
- Reduce further the number of subjects per cluster in order to reduce the daily work load.
- Plan the time and resources required for identifying nomadic clusters, as this will greatly facilitate implementation of the survey.
- Make necessary improvements in epidemiology statistics to the methodology, in order to reduce possible bias in the selection of the target population and to increase the repre sentativeness of the sample.
Acción contra el Hambre recommends ongoing research into appropriate assessment methods for pastoral areas. For example, recent studies published by Mark Myatt9 indicate that if Weight/Height indicator and Mid Upper Arm Circumference (MUAC) are used in surveys they lead to different findings in terms of prevalence of malnutrition as a result of differences in body shape. Future research on pastoral populations should include the use of MUAC to gauge the prevalence of acute malnutrition.
With this in mind, Acción contra el Hambre will be busy in 2008 improving and consolidating pastoral survey methods through new research involving its international network of agencies. A new publication on this subject will be coming out in the second quarter of 2008.
For further information, contact: Núria Salse, email: firstname.lastname@example.org tel: 00 34 91 391 53 00
1Save the Children UK. Nutrition Assessment in Gola Oda Woreda, East Haraghe, Ethiopia. March 26 April 6, 2001
2Susanne Jaspars and Hisham Khogali. Oxfam Food and Nutrition Group. Oxfam's Approach to Nutrition surveys in emergencies, May 2001
3Fraction: Administrative unit used in Mauritania as a method of social organisation within a tribe.
4Jaques Brenez. L'óbservation démographique des milieux nomades. L'enquête de Mauritanie ». Population. Année 1971, Volume 26, Number 4 p 721-736
5William D. Kaslbeek, University of North Carolina, Anne R. Cross, Westinghouse Health Systems. Problems in sampling nomad populations.
7These were calculated using the NCHS (National Centre Health Statistics) tables and the recent WHO 2005 standard tables. NCHS tables were used to compare with previous surveys and for programme planning (as NCHS were used as a criteria of admission). The low acute rate prevalence found in this population did not allow for detecting major differences between the NCHS references and WHO standards.
8The sampling area was represented as a grid as per the Centric Systematic area sampling (CSAS) method
9Report concerning the analysis of data collected for the MUAC/weight-for-height/body-shape research study. Mark Myatt, Institute of Ophthalmology, University College London. 30th May 2007.
More like this
By Anne-Marie Mayer, Mark Myatt, Myriam Ait Aissa and Nuria Salse Anne-Marie Mayer is a technical consultant for this project and carried out the first field test in Mali with...
en-net: SMART HHs Sampling
Dear Colleagues, We are planning to conduct a SMART survey in one area in Sudan. This area is a mixture of IDPs camps and host communities with no specific arrangement of HHs...
By Thierry Metais, Amador Gomez and Carole Lambert, ACH Thierry Metais is Technical Coordinator of Acción Contra el Hambre in Mali and Niger. He has been working in the Sahel...
Dear All, This is to request your advise and experience regarding the following topic Population scattering and mobility are the two main issues encountered in sampling in...
By Susana Moreno, Brigitt Olagivel and Elisa Dominguez Susana Moreno Romero is the Nutrition Programme Manager for Acción contra el Hambre (ACH) (Action Against Hunger -...
en-net: When sampled cluster doen't have enough HH demanded by sample size because of urgent and sudden population movement
Experienced from my current SMART survey, we faced one issues as HH from one sample cluster doesn't have enough HH to met the sample size demanded for each cluster. The target...
Hello, Id like to know what is the best recommended methodology for SMART Survey in Urban Setting with large population (total population: 500,000p, pop/cluster >2000...
By Adrian Díaz, Amador Gómez, Nuria Salse and Gabriela Cormick Adrian Díaz is a physician and Technical Coordinator for ACH in Argentina. He has worked with MSF in Honduras on...
Few questions 1. To what extent should the results of ENA plausibility tests be used to reject surveys which collect more than anthropometric data,in the first place was...
Now days, the term "Small scale SMART survey, SSSS" is becoming common. In principle, is it different from that of normal SMART survey procedure? To make it clear, does SSSS...
FEX: Stunting and wasting in children under two years old in a semi-nomadic pastoralist population in Kenya
By Amelia Reese-Masterson, Masumi Maehara and Mark Murage Gathii Amelia Reese-Masterson is Research Advisor in the Nutrition, Food Security and Livelihoods Unit at...
The current approach is to have single two-stage cluster survey by agro-ecology. If a given district has two agro-ecological zone, this means there should be two independent...
I know the SMART Methodology recommends that the minimum number of clusters in a survey should be 25. However, I have a scenario where the number of clusters sampled were 32...
FEX: Evaluation of the Supplementary Feeding Programme carried out in Marsabit District, Kenya, April-November 1997
Vincent Brown (Epicentre), Sylvia Carbonell (Médecins Sans Frontières) MSF implemented a nutrition rehabilitation programme in Marsabit District between April...
By Damien Helleputte and Julien Jacob Damien Helleputte is technical coordinator of the Accion Contra el Hambre mission in Guinea since 2008. Julien Jacob is currently...
By Leah Richardson, Moazzem Hossain and Kevin Sullivan Acknowledgements go out to the survey teams as well as to those responsible for the close partnership between UNICEF,...
en-net: Representative sample
What is the smallest/least representative sample or number of children/households one can assess in a SMART survey. Thank you Mark for this comprehensive answer. Indeed, you...
I would like to know the most appropriate and cost effective sampling system which is relevent to apply the pastoral communities like Somalia. The article describing the Mali...
By Elena Rivero, Núria Salse and Eric Zapatero Elena Rivero is currently working for Action Against Hunger as Surveillance Advisor in the Malawi Integrated Nutrition and Food...
By Amador Gomez and Elisa Dominguez Amador Gomez is Technical Director of Acción Contra el Hambre (ACH). Previously he spent several years working on nutrition and medical...
Reference this page
Eva Vicent and Núria Salse (2008). Methodology for a Nutritional Survey among the nomadic population of northern Mali. Field Exchange 33, June 2008. p14. www.ennonline.net/fex/33/mali