Personal experiences of working during the Horn Africa crisis in 2011

Interview with Mélody Tondeur and Sarah Style

The following is based on a face to face interview conducted by Carmel Dolan (ENN) with Mélody Tondeur and Sarah Style, two team members of an ENN technical assistance project to UNHCR , who oversaw surveys in Dadaab camps in Kenya during the height of the crisis in 2011. It describes their personal experiences in carrying out the surveys and their views on approaching large scale emergency refugee situations. Both Mélody and Sarah now work with UNHCR.

Background

Drought conditions in the Horn of Africa, particularly in Somalia, decimated the 2011 crop and livestock production causing dramatic increases in food prices. Prolonged insecurity in Somalia also led to a significant reduction in international food assistance and in July and August 2011, the United Nations (UN) declared famine in certain regions of the Horn of Africa. During the build-up, large numbers of Somali’s moved across the Kenyan border to seek assistance in the long-established Dadaab refugee camps. At the peak of the crisis, over 1000 refugees arrived in the camp each day. In August 2011, it was agreed that a nutrition survey was needed to assess the magnitude of the nutrition and mortality situation. Coincidentally, August was the scheduled time of the routine annual survey in Dadaab.

Settlements of new arrivals in the outskirts of Dadaab

The ENN has been collaborating with UNHCR over the last two years providing technical support and developing guidance for conducting standard nutrition surveys. Two members of the ENN technical assistance team at the time (Mélody Tondeur and Sarah Style) went to Dadaab for a 6-8 week period to support the design and implementation of the surveys and at the same time, help strengthen the capacity of UNHCR and implementing partner’s (IPs) in executing surveys. This article describes their more personal experiences shared in their ENN interview.

What were you expecting and what were your early impressions of Dadaab?

Dadaab was the first exposure to a large scale refugee situation for both Mélody and Sarah and, for Sarah, her first refugee camp experience. Prior to departing, Mélody had regular contact with the UNHCR Senior Nutrition Associate (Gloria Kisia) based in Dadaab and both were able to get useful insights from reading situation and survey reports, as well watching the news and seeing images of the camps and the camp outskirts. For Sarah, the main camps were as she had imagined but the environment in the outskirts of the camp was harsher than she had expected in terms of the strong winds, dust, land contamination by human faeces and limited latrines.

How would you describe the preparations and process you undertook for carrying out the surveys in Dadaab?

Four weeks prior to their arrival, preparations were undertaken by Gloria on the ground with the assistance of Mélody, including putting together a pre-survey schedule, compiling survey questionnaires, and organising survey logistics. Through other project activities, Mélody had already met and discussed the situation via email and in person with the UNHCR nutritionist and other survey staff with whom they would be closely working. Once in Dadaab, another two weeks of planning was undertaken by Mélody, Gloria and Terry Theuri (a UNHCR Nutrition and Food Security consultant deployed to Dadaab during the crisis) with key staff from implementing partners (Millicent Kavosa from IRC). They were joined for two weeks by a third ENN team member from the UNHCR technical assistance project (Andy Seal) to help design and plan the survey. Together, the team went to the three main camps which make up Dadaab to understand the lay-out (for sampling purposes) and the setup of the outskirts, talk to the nutrition programme staff to understand food distribution and feeding schedules so that the hours could be modified during the survey to avoid absentees, to look at equipment and quality of equipment and to discuss the referral process for children and women needing treatment.

Using GPS handset to find cluster during survey of outskirts

Mélody and Sarah felt that an enormous advantage in Dadaab with respect to the survey was that some of the key staff (2 from UNHCR and 1 from IRC) had considerable survey capacity and experience and had received prior training on the survey guidelines as part of the UNHCR/ENN collaboration. In this respect, Dadaab was well able to accommodate emergency survey work of this type. In addition, the staff were completely open to strengthening their skills in this area. Even though the demands on the survey teams were enormous, capacity strengthening was carried out (5 days of training involving 4 days in the classroom and 1 day in the field). Ensuring quality was a major focus during the training sessions and as data were collected. In total, 80 surveyors were trained.

A survey was conducted in each of the three main camps as is usually done every year in Dadaab. This year, a fourth survey was also conducted in the outskirts of one of the camps. Each of the four surveys had five survey teams of 4 people each (1 team leader, 2 enumerators and 1 translator), supported by survey team coordinators and supervisors. There were daily feedback sessions in the mornings. The evenings were spent entering data and checking questionnaires so that the UNHCR/ENN coordination team could inform the survey teams of any improvements needed throughout the survey.

Mortality data was also collected (not done in recent surveys in Dadaab due to a well functioning Health Information System (HIS) system monitoring mortality).

Were there specific methodological challenges?

Mélody and Sarah described how one of the main challenges during survey planning was to decide which data to obtain from the new arrivals and which from the long-standing refugees, as agencies needed information on both populations to aid decision making. Because the new arrivals were in the main camps (thought to represent 10-20% of the main camp population) and in the outskirts (where the entire population was new arrivals), doing one survey covering both new arrivals and long-standing refugees would have required a stratified survey design which would need a lot of work. Hence the decision was taken to survey one outskirt area in addition to the three main camps to obtain data on new arrivals.

The decision to survey the outskirts, where new arrivals from Somalia would congregate waiting to be further assimilated into the newly established camps, created its own sampling challenge, since there were no structures (like the camp blocks) to delineate clusters. The team innovated by using spatial random sampling with GPS and to their surprise, this ended up being easier to administer than anticipated. For sampling the outskirts, the perimeter was mapped using a GPS and the tracks and waypoints uploaded and visualised using Google Earth (version 6). Random spatial clusters were then allocated to the outskirt area map using a grid overlay. Cluster start points were subsequently located by survey teams using GPS handsets and second-stage sampling proceeded using proximity sampling.

A third challenge was the decision to include children from 5 - <10 years of age using MUAC when there is no international agreement on the cut-off point to use for defining malnutrition. In this case, a MUAC cut-off of 14cm and the presence of oedema were used for referral during the survey and then weight and height were taken at the health posts for further screening.

Further, Mélody and Sarah added, the population data wasn’t reliable. Due to the influx of refugees, there was a significant backlog of unprocessed registrations at the time of the survey planning with refugees settling in the outskirts and in the main camps. This meant it was difficult to gather accurate population figures and derive demographic data for survey planning, including the % under five and average household size. Quota sampling was therefore used and this presented some of the teams with difficulties, as they had been used to fixed household sampling (e.g. surveying 13 households a day and not counting the number of individuals in each target group surveyed).

Another challenge was in keeping the household questionnaires reasonably manageable to ensure quality of the information collected. Nonetheless, the questionnaires were quite long and complex compared to previous years (largely because of the addition of the mortality data and the measurement of the children 5-<10 years) making it difficult for some of the survey teams to administer.

Were there other unforeseen challenges?

Unfortunately an MSF staff strike coincided with the survey timing and forced last minute changes to the survey so that two surveys had to run concurrently followed by another two. This put the team under considerable time and logistical pressure.

Also, the survey teams found it hard to keep track of quotas of individuals for the questionnaires meaning that some teams needed a lot of support for data collection. Coordinators had to spend entire days with some teams to keep them on track to complete one cluster in a day. Mélody and Sarah described how there was little leeway to absorb any delays for several reasons including budget, logistics and because the surveyors had to resume their regular jobs at the health clinics and at the community level. The pressure was on with the outskirts survey especially, because of the new arrivals being relocated into newly established camps right as the survey was happening.

How did you involve key stakeholders in the survey?

A great deal of work went into ensuring that the key stakeholders (UNHCR, WFP, UNICEF, IPs and other agencies working in the camps) were fully part of the decision making about the survey design, choice of indicators and questions contained in the household questionnaire. Survey meetings and correspondence was coordinated by the Dadaab Camp UNHCR Senior Nutrition Associate (Gloria).

In Dadaab, an efficient model for nutrition surveys had been implemented in previous years and hence the same model was used. Each IP was designated to survey the camp they work in and had clear roles and responsibilities, survey supervisors, etc. Some partners took on the role of helping with the pre-survey training, supervision and data entry/data quality checks. Overall, the process was successful and there was considerable good will and coordination to enable the surveys to take place as smoothly as possible.

How did you record and store the data?

ENA for SMART software (Delta version, June 20th 2011) was used by the supervisors and coordination team to enter and analyse the anthropometric and mortality data, and a system of double checking was used to verify all data entered. All other data were doubled entered by four clerks using Epi Info Software (CDC, version 3.5.1) and data analysis was done using ENA for SMART and Epi Info software. Double entry was essential since data entry clerks could not be supervised during the day by the coordination team while the survey was on-going.

How quickly were you able to analyse the data and feedback the findings?

This was a major priority so the data entry personnel entered the data in parallel to the data collection to speed up the process of feedback. Three nutritionists (2 from UNHCR and 1 from ENN) were involved in analysing the data for different indicators and a fourth from ENN, for training on data analysis and management, as well as oversee and double check all the results. Further, the survey report was split into separate sections and the writing/ reporting allocated to the four main nutritionists, with one of the nutritionists having the responsibility to oversee the whole report writing process.

Impressively, the GAM/SAM results in children 6-59 months were fed back ‘informally’ 2-3 days after the surveys (by email). The results on all indicators were fed back officially 2 weeks after the surveys were completed, to allow partners to start working on the recommendations and new interventions. This was done through a full partners meeting. The final and detailed report was disseminated two months following the partners meeting.

Were you surprised by any of the findings and were there any significant unanswered questions around the findings?

In general, the malnutrition results were really high, even in the main camps. The team had predicted these results from rapid assessments and screening results but the results from the outskirts were higher than assumed. Partners were especially alarmed with the GAM prevalence levels (38.3% (32.1-44.8%) in the outskirts, 17.2% Hagadera. 22.4% Ifo, 23.2% Dagahaley). The mortality rates were as expected in the main camps and quite low, however they were very high in the outskirts (Crude death rate 1.23 (0.73-2.06), U5 Death Rate 3.02 (1.72-5.24)). Coverage of the nutrition programmes (SFP and OTP) was found to be very low though the reasons for this were not clear and considering also that nutrition surveys are not the best way to establish levels of coverage. In addition, screening in the community was based on MUAC and oedema whereas admission was based on MUAC, oedema and weight for height z score (WHZ). This meant that estimates of coverage varied depending on if WHZ was included or not.

Mélody and Sarah weren’t surprised by the survey results for children in the 5-<10 year old category in the outskirt survey (MUAC <14 cm was 18.3% in the outskirts compared to 6.8- 12.2% in the other camps) - screening of new arrivals had indicated this and they remembered being shocked by how thin they witnessed the older children to be during the outskirts survey. It was evident that not only children 6-59 months, but also older children had been affected by the crisis. However knowing what to do with these children was more of a challenge given the limited knowledge and research available on this age group.

There was a significant decrease in levels of anaemia in children and women, which was a very positive finding for UNHCR, WFP and the IPs who had been implementing anaemia control measures in Dadaab in previous years.

What new things did you learn from your experience in Dadaab?

For Sarah, this was the first time she had taken part in a large scale survey at the height of such a large scale emergency and two key lessons stood out ; first be flexible as in such situations things are constantly changing and second, being able to switch off and to find ways to relax. Mélody echoed the latter and emphasised the importance of taking time for yourself, getting a good night’s sleep and either read or watch a movie before going to bed to help switch-off.

For both Mélody and Sarah, having the opportunity to work on a daily basis with the UNHCR highly experienced and long standing nutritionist based in Dadaab, as well as the other UNHCR,ENN and IP nutritionists, was vital to understand and deal with the context -both for the survey design, data collection and interpretation of the data. All members of the coordination team complemented each other well, each bringing their own knowledge, skills and experience.

What advice would you give to others going to support large-scale surveys in complex environments such as Dadaab?

Both Mélody and Sarah would recommend to be very organised and ensure that the relevant support within the camp is also there. It is very important to consider the security advice given and to keep track of the security situation on daily basis.

Contact people ahead of time so you arrive and know what you need and know what questions you have. Compile as many background reports as possible as these tend not to be readily available and avoid asking questions on the ground when you already have access to the information. Also, it helps to keep notes of what you’re learning so you don’t have to keep asking the same questions. This reduces the pressure on field staff who are already very busy and can end up ‘repeating information for the benefit of many visitors’

Capacity building is possible in a large scale emergency situation. The ENN team were able to strengthen the capacity of key UNHCR staff and IP staff on how to execute quality surveys. Innovation is also possible, as together the team managed to solve the sampling challenge in the outskirts.

On a personal front, Mélody and Sarah emphasised you need to be fully prepared to live in difficult conditions (live in tents, share rooms with others, don’t have your home comforts), work very long hours during survey implementation (average from 7am until 11pm) and to put aside other work demands so that the only focus is on the survey. Mélody did this by avoiding unrelated emails.

Is there anything else you would want readers of FEX to know about your experiences?

Both concluded that the Dadaab experience is a good example of the importance of preparation for a nutrition survey during an emergency situation, i.e. in this case, waiting and working to ensure there was adequate information from rapid assessments and screening to inform sample size, survey design and the content of the questionnaire.

Despite being a very challenging context to work in - both professionally and personally - it also a rewarding one, reflected in the fact that Mélody and Sarah have continued to work with UNHCR and remain full of enthusiasm!

For more information, contact: Mélody Tondeur, email: melotondeur@gmail.com Sarah Style, email: sarahlstyle@gmail.com