GPS assisted coverage survey in DRC

By David Rizzi

David Rizzi graduated in Pharmacy and later took an MSc in Development at the Rome University La Sapienza, Italy. He holds a second MSc in Public Health Nutrition from the LSHTM in UK. He has been volunteering and working for NGOs in Tanzania, Palestine, Angola, Tchad, Burundi, Uganda and DRC, mainly focusing on nutrition survey and therapeutic feeding programmes in emergency contexts.

The author would like to thank Jean-Laurent Martin, GIS officer at OCHA in Goma for his valuable assistance with GPS and maps, and the ECHO office in Goma for their financial support. The author also acknowledges the COOPI field staff and the Binza's population for making this work possible.

This article describes a modification of the centric systematic area sampling method using GPS, to overcome field constraints related to mapping and active case finding.

A child being weighed during the DRC survey.

COOPI is an Italian non-governmental organisation (NGO) that has worked in the Democratic Republic of the Congo (DRC) since 1970. Its main sectors of intervention nowadays are health, nutrition, psychosocial care and water/sanitation. As a leading NGO in the nutrition sector in eastern DRC, COOPI was asked by the Ministry of Health and UNICEF to run a pilot phase of a Community-based Therapeutic Care (CTC) programme in Bunia (Ituri) and to provide technical assistance in the preparation of the national CTC protocol.

In late 2007, ECHO (European Commission Humanitarian Office) funded COOPI to provide nutritional care (or support) to Binza's population (over 100,000 people) in North Kivu province. The target population included many returnees who were subject to a high degree of insecurity, due to national army and rebel groups' dispute and conflict over the area. Displacements were frequent due to continuous attacks and pillaging by the warring factions. The intervention goals were to:

By January 2009, the traditional inpatient therapeutic programmes have progressively been replaced by CTC programmes throughout the country. In 2008, the national CTC protocol had been validated by health authorities, international organisations and NGOs, and was awaiting final approval and publication. The COOPI CTC programme in Binza started in January 2008 and included two SC and ten SFC/OTP (Outpatient Therapeutic Programmes). All the health centres in the zone hosted a SFC/OTP providing good geographical coverage of the zone. After seven months, 652 severely malnourished children had been admitted (averaging 93 per month); the cure rate was above 90% and mortality less than 2%. An anthropometric and coverage survey was undertaken at this point to monitor the progress of malnutrition rates and programme coverage. The anthropometric survey showed a significant fall in most indicators of malnutrition (see Table 1) compared to the previous survey conducted in November 2007. However, it was not possible to infer that this was due to the COOPI intervention or the new CTC approach: seasonality effect, improvement in security, food aid and better crops were likely to have been at least partly responsible for this positive trend.

Table 1: Comparison of malnutrition indicators from November 2007 and July 2008 surveys
  Nov 2007 July 2008 Difference in % p value*
Global acute malnutrition (z score <-2) 10.5% [7.9-13.0] 4.9 % [3.2 - 6.7] - 53.3% <0.0001
Moderate acute malnutrition (-3? & <-2 z score) 6.1% [4.2-7.9] 4.7 % [2.9- 6.5] -23.0% 0.01
Severe acute malnutrition (<-3 z score) 4.4% [2.3-5.5] 0.2 % [0.0 - 0.5] -95.5% <0.0001
MUAC** <125 & ?110mm 6.85% [4.0-9.6] 6.5% [4.9-8.1] -5.1% 0.6
MUAC < 110mm 0.46% [0.29-2.21] 0.3% [0.0-0.7] -34.8% 0.0001

*Calculated from the standard error of the differences
**MUAC: Mid upper arm circumference

There were no baseline data available on coverage of the therapeutic feeding programme so that the CTC programme coverage survey was, to the best of our knowledge, the first assessment of this type in the whole province.

Coverage survey methodology and constraints

The coverage survey was undertaken in July 2008. The CSAS methodology (centric systematic area sampling), which is based on active case-finding, was used2. The first step in the implementation of this methodology consisted of drawing or overlaying a grid on a map of the area under investigation. The communities closest to the centre of each square were then the first to be surveyed.

Constraint 1: Lack of a detailed map of villages

No detailed map of the health zone was available for Binza and recent population movement had resulted in many villages being abandoned. When this problem occurred in previous surveys, a 'blank' map had been given to all the field teams so that on the day of data collection, they could locate the village closest to the centre of the square by means of a landmark shown on the map, e.g. river, hills, villages, etc. This method proved to be very time consuming, inaccurate and ultimately, prone to bias; the teams would hurry through this first phase in order to maximise time for the data collection itself. Moreover, where maps didn't show any physical landmarks or no landmarks were present, the task proved even more difficult.

A child being measured during a survey in DRC.

Constraint 2: Absentees at the time of measurement

In previous coverage surveys, another problem was that many children were absent when the team visited the household. Children under 5 years of age are usually taken to the field with their mothers early in the morning and come back at sunset. Where this occurred, an appointment, whenever possible, would be made and a second visit arranged to measure the absent children at a later stage.

Moreover, the active case-finding approach needs the assistance of carers, key informants and local authorities to identify suspected malnourished children and those children enrolled in a programme. However, when those people were contacted on the day of data collection, many complained about not having had sufficient notice to think about the children to include in the survey (and to inform them). This process further reduced the time that COOPI field teams had for data collection.

A new strategy

A two-phase modified methodology was therefore designed to address these problems. In the first phase, Global Positioning System (GPS) receivers were used to select the target villages and key informants were informed about the survey and asked to assist in case finding. In the second phase, the teams focused solely on anthropometric data collection.

Phase 1: GPS aided village selection and preliminary actions

The best map available was kindly provided by the Office for Coordination of Humanitarian Affairs (OCHA) office in Goma. A physical map with a 1:50,000 scale showing the main road and some large villages, it included longitude and latitude data also. An updated list of villages and populations of the intervention area was obtained at the Central Health Office of the Binza's Health Zone (Bureau Central de la Zone de Santé). The provincial doctor, nurses and community health workers were asked to position all the villages from the list onto the map. A provisional, partially hand-drawn, map was then available for the survey.

A grid was drawn on the map, and 23 out of 30 squares were selected on the basis of having at least 50% of their area in the health zone. The centre of each of these squares was found (tracing diagonals) and its latitude and longitude calculated from the data on the sidebars. This was subsequently recorded on four GPS receivers (again kindly provided by OCHA).

Any GPS receiver, including non-cartographic ones, can perform the tasks demanded for this survey. However, it is important to configure the devices so that they all have the same coordinate system as the map. Furthermore the longitude and latitude data from the map had to be converted from degrees/minutes/seconds into decimal degrees, as used by the GPS3.

At this point a first field visit with the following objectives was planned:

Four teams (of two people each) were trained on GPS usage, people to contact and messages to convey. They were also given a list of villages within each square.

The GPS receivers were set on the 'go to' mode, which shows the direction and the distance to a given waypoint.

These activities largely ran smoothly. However there were some reported problems with estimating the GPS waypoint location when these were not reachable (due to thick woods or being far from any road). In these cases, with the aid of local people, the team had to estimate the location and find the nearest village.

Once the village or community closest to the centre was identified, local authorities, health community workers and other key informants were met and informed about the objectives, sample and timetable of the survey. They were also asked to think about children to include, to establish a list with their names and to ask them to stay at home on the day of the data collection. These actions were repeated in the next nearest villages until five villages were visited within one square. Given the average size of the villages, the anticipated number of children to assess and the time available, we estimated that no more than five villages could be visited in one day. The time lag between the first visit to a village and the data collection was seven days, on average. Each team was able to work on two squares per day thereby visiting a maximum of ten villages.

The first-phase teams recorded the names of each village, their relative position within one square (1 being the nearest to the centre and 5 the last visited village) and the name of the most significant authority met. Coordinates of all visited villages were also recorded on GPS. It had been hoped that a more accurate map would be produced once the GPS coordinates of the villages had been handed over to OCHA and that a re-allocation of villages into the 23 squares based on the updated map would have been possible. However, due to technical problems with the large format printer, this was not possible and the provisional map was used until the end of the survey.

Phase 2: Data collection

A child and his mother at the Kitsharo Stabilisation Centre, one of the two SCs of the CTC programme.

In the second part of the fieldwork, the full survey team (20 people divided into five groups) visited the survey area. No GPS or maps were needed during this stage. Each team was handed the list of villages to visit in each square (until 3 p.m. or until the completion of the five villages) and the person to contact. The list of suspected malnourished or in-programme children was also given to them so that they could focus exclusively on data collection. As expected, the vast majority of the children's caretakers were aware of the survey and present.

Results

The coverage rates were worked out for both the OTPs and the SFPs. The results are set out in Table 2.

Table 2: Point and period coverage
Programme Coverage %
OTP Point 47%
  Period 72%
SFC Point 36%
  Period 62%

 

Given the relatively short duration of the CTC programme, and the upward trend in admissions at the OTP and SFP, we consider these results to be acceptable, although improvement is possible. The rise in programme admission in spite of a lower prevalence of moderate and severe malnutrition (a 53% reduction since the last nutrition survey) is probably partly due to better coverage. Further investigation into programme coverage and anthropometric status of the population in 4-6 months time has been recommended, to achieve better understanding of programme performance and its evolution.

Conclusions

David with a local nutritionist, Matthieu Umba, using GPS.

The two-phase coverage survey methodology, based on the CSAS procedure, was developed to get around the problem of lack of a detailed map of the area under investigation, as well as other constraints, e.g. lack of notice to key informants and population. This method facilitates identification of the villages to survey by means of GPS devices. Other advantages include less bias in village selection, a lower absentee rate, a reduced number of missed children (suspected malnourished children or inprogramme who are not referred to the survey team), more time (and energy) available for the teams for the data collection. However, this methodology does require some additional resources, i.e. the first-phase necessitated eight people working for four days. A normal coverage survey on average involves 20 people working for 7-10 days for the entire data collection process. Other extra costs for the new method were only incurred for vehicles running costs, as GPS devices were borrowed from OCHA.

Although not achieved in this case due to technical problems, the use of GPS to record village coordinates at the time of the visit should allow the production of more detailed maps for future use and a more accurate calculation of coverage in each square.

For more information, contact: David Rizzi, email: david.rizzi@gmail.com

Show footnotes

1BDOM (Bureau de Développement des Oeuvres Médicales) is part of the CARITAS association.

2Myatt M, Teshome F, et al. (2005). A field trial of a survey method for estimating the coverage of selective feeding programmes. WHO Bulletin 2005; 83(1): 20-26.

3Although doable on the field through simple calculation, free online tools are available for such conversion: http://www.fcc.gov/mb/audio/bickel/DDDMMSS-decimal.html (accessed on January, 5 2009)

More like this

FEX: New Method for Estimating Programme Coverage (Special Supplement 2)

By Mark Myatt This article gives an overview of the coverage estimation method developed for the Community Therapeutic Care (CTC) Research Programme in Malawi1. Coverage is...

FEX: New Method for Estimating Programme Coverage

By Mark Myatt Mark Myatt is a consultant epidemiologist and senior research fellow at the Division of Epidemiology, Institute of Opthalmology, University College London. His...

FEX: Applying GIS to Nutrition Surveys

By Filippo Dibari, Andrew Seal and Paolo Paron Filippo Dibari is a Food Technologist with long term community based experience in developing countries and consultancies with...

FEX: Implications of a Coverage Survey in Ethiopia

By Simon Kiarie Karanja Simon Karanja is currently the regional nutrition advisor with GOAL in East Africa. Previously he worked as the CTC Coordinator for GOAL Ethiopia and...

FEX: The Challenge of Applying CSAS in DRC

By Jason Stobbs, AAH-USA Jason Stobbs is a long-serving member of Action Against Hunger - USA, volunteering over the past four years in diverse contexts such as Afghanistan,...

en-net: How to Maintain the Quality of CSAS Method

As we all know, coverage is an internationally recognized indicator of performance for selective feeding program (SFP) and CSAS is the current gold standard for assessing...

FEX: Area sampling for rapid population assessment

Summary of a published paper1 In the initial phase of an emergency, an immediate assessment of population size is vital to provide relief workers with the necessary data to...

en-net: Determine carers willingness to walk to access services

In Myatt, Mark et al. 2012. Semi-Quantitative Evaluation of Access and Coverage (SQUEAC)/Simplified Lot Quality Assurance Sampling Evaluation of Access and Coverage (SLEAC)...

FEX: Using SLEAC as a wide-area survey method

By Ernest Guevarra, Saul Guerrero, and Mark Myatt Ernest Guevarra leads Valid International's coverage assessment team. He has formal training as a physician and a public...

FEX: Comparison of LQAS and 30-by-30 two-stage cluster sampled survey method for the assessment of coverage indicators

By Asrat Dibaba, Charles Chimombo, Ari Uotila, Whitney King and Mark Myatt Asrat Dibaba is the Maternal, Newborn and Child Health Advisor for World Vision East Africa Region...

FEX: SQUEAC: Low resource method to evaluate access and coverage of programmes

By Mark Myatt Mark Myatt is a consultant epidemiologist and senior research fellow at the Division of Ophthalmology, Institute of Ophthalmology, University College London....

FEX: SQUEAC in routine monitoring of CMAM programme coverage in Ethiopia

By Lily Schofield, Selome Gizaw Lalcha and Terefe Getachew Lily Schofield has worked in many countries in Africa and Asia as a nutrition consultant. She has been involved in...

FEX: Boosters, Barriers, Questions: an approach to organising and analysing SQUEAC data

By Andrew Prentice (VALID), Balegamire Safari Joseph (VALID), Esther Ogonda McOyoo (Concern), Faith Manee Nzidka (ACF), Hassan Ali Ahmed (Mercy USA), Jackson N Chege (Islamic...

FEX: Remote monitoring of CMAM programmes coverage: SQUEAC lessons in Mali and Mauritania

By Jose Luis Alvarez Moran, Brian Mac Domhnaill and Saul Guerrero Jose Luis Alvarez Moran is a Medical Doctor with a PhD in International and Public Health. He works as an...

FEX: CTC Approach (Special Supplement 2)

by Steve Collins (Valid International) 2.1 Main principles of CTC Community Therapeutic Care (CTC) is a community-based model for delivering care to malnourished people. CTC...

en-net: SQUEAC Stage

As per the SQUEAC investigation manual/guide, the purpose of the exercise in Stage 2 is to confirm hypothesis of homogeneity/heterogeneity of coverage in the program area. if...

FEX: CTC in North Darfur, North Sudan: challenges of implementation (Special Supplement 2)

By Kate Sadler (Valid International) and Anna Taylor (SC-UK) People waiting at a clinic in Darfur, North Sudan. Child eating plumpynut® in Darfur, North Sudan. North Darfur...

FEX: Case Studies (Special Supplement 2)

3.1 CTC in Ethiopia- Working from CTC Principles Isolated village in the highlands of South Wollo, Ethiopia. By Kate Golden (Concern Ethiopia) and Tanya Khara (Valid...

en-net: Stage 3 - Sample size issue - Help

Dear all, We are currently conducting a SQUEAC investigation in two townships of the Northern Rakhine State in Myanmar. We have been doing well so far but we are now facing a...

FEX: Socio-Anthropological Aspects of Home Recovery from Severe Manutrition

By Adrienne Daudet and Carlos Navarro-Colorado Morlaï, cured at home, at last weekly visit. Adrienne Daudet is an agronomist and has an MA in anthropology. She has four years...

Close

Reference this page

David Rizzi (2009). GPS assisted coverage survey in DRC. Field Exchange 35, March 2009. p35. www.ennonline.net/fex/35/gps