• Food, goats & cash for assets in Kenya• SMART anaemia analysis in Bolivia• Cross-sectoral approach to Konzo in DRC• Food security in Afghanistan• Early warning system in Somalia • Integrating IYCF support in Ethiopia• Mitigating soil salinity effects in Bangladesh

ISSN 1743-5080 (print) December 2012 Issue 44

From the EditorContents

Field Articles

3 Food security assessment of high altitude villages of Badakhshan, Afghanistan

35 Integrating anaemia analysis in SMART surveys in Bolivia

38 Food, goats and cash for assets programmes during emergency drought response inKenya

50 Impact of cross-sectoral approach to addressing konzo in DRC

53 2011 famine in South Somalia: the role of the early warning information system

60 Integrating Infant and Young Child Feeding and the Productive Safety Net Programme in Ethiopia

63 Helping homesteads mitigate the effect of soil salinity in Bangladesh

Research

6 Effect of adding RUSF to a general food distribution on child nutritional status and morbidity: a cluster randomised controlled trial

8 Impact of livestock support of animal milk supply and child nutrition in Ethiopia

12 Climate Change as a driver of humanitarian crisis and response

14 A systematic review of obstacles to treatment of adult undernutrition

15 Humanitarian financing for older people and people with disabilities

17 The impact of displacement on older people

18 Nutrition and baseline survey of older people in three refugee camps in Dadaab

19 Disaggregation of health and nutrition indicators by age and gender in Dadaab refugee camps, Kenya

21 Qualitative study of supplementary suckling as a treatment for SAM in infants

23 Patterns of body composition among HIV-infected, pregnant Malawians and famine effects

25 Anthropometry in infants under 6 months in rural Kenya

26 Micronutrient powders v iron-folic acid tablets in controlling anaemia in pregnancy

27 UNHCR feeding programme performance in Kenya and Tanzania

29 Scaling up ORS and zinc treatment for diarrhoea reduces mortality

29 Treatment of undernutrition in urban Brazil

30 Feasibility and effectiveness of preventing child malnutrition with local foods in Kenya

30 Agricultural interventions to improve nutritional status of children

31 Nutritional support services and HIV in sub-Saharan African countries

31 The state of the humanitarian system

33 Concerns on Global HIV/AIDS, TB and Malaria Fund conflicts of interest

34 Double burden of obesity and malnutrition in Western Sahara refugees

News

41 en-net update, August to October, 2012

41 Erasmus Mundus Master Course in Public Heath in Disasters

41 New SMART website launched

42 Urban Humanitarian Response Portal

42 Urban edition of Field Exchange: call for submissions

42 Proceedings of UNICEF/GNC/SC UK workshop on IYCF in development and emergencies

43 Resources to support programming for older people

43 New Global Food Security journal

Views

43 Personal experiences of working during the Horn Africa crisis in 2011

56 Quantity through quality: Scaling up CMAM by improving programmes access

Agency Profile

32 Bread for the World

This is another bumper issue of Field Exchange,with eight field articles and just under 20research summaries. On the one hand, weapologise for the ever-growing volume of our

publication and on the other, we are pleased that somany field practitioners want to write up and shareprogramming experiences. There also seems to be anever increasing volume of research out there thatwarrants dissemination. Usually a theme or two leapsout at us once we have amassed our Field Exchangecontent – not so this time, and we have a broadmixture of material for you to read. So for a change,this editorial doesn’t attempt to link articles by themesbut instead, simply highlights field articles andresearch pieces which we feel are of particular inter-est. Inevitably this will be subjective and we hope thatthis doesn’t stop you looking at all the contents in thisissue.

Four articles get a special mention in this editorial.The piece by Adèle Fox at Concern Worldwidedescribes a programme to build capacity of multi-sector actors at woreda, kebele and community levelsin Ethiopia to deliver effective infant and young childfeeding (IYCF) messages as part of the productivesafety net programme and encourage social andbehaviour change. It went beyond the usual commu-nication channels, engaged and coordinated withcommunity leaders and worked across sectors toaddress barriers to good IYCF practice - policiesaround women’s maternity leave were strengthened,agricultural support addressed food insecurity, andsupport to water and health services aimed to preventcommon illnesses. It appears to be yielding sustain-able and substantial improvements in IYCF practices.

Marie-Morgane Delhoume, Julie Mayans, MurielCalo and Camille Guyot-Bender from ACF- USA havewritten an article about tackling Konzo in DRCthrough introduction of cassava retting techniquesand improving access to water. This is a nice follow onto a baseline study shared in Field Exchange 41 thatfound a 1% incidence in the area. ACF-USA used acommunity outreach and nutrition education andtraining strategy to get the information across at largescale, coupled with interventions to improve dietarydiversification, water access and agricultural process-ing. At the end of the project, households reportedsoaking cassava for longer periods across the inter-vention area (critical to reducing Konzo risk).Impressively, the intervention had at least some partin effecting an 84% reduction in incidence of Konzobetween 2010 and 2011. The greatest reduction innew cases was observed among the under 5 years agegroup.

An article by Geoff Brouwer describes a compara-tive study of four relief and emergency responseactivities – Cash-for-Assets (CFA), Goat-for-Assets(GFA), and two Food-for-Assets (FFA) projects – imple-mented by World Renew (formerly the ChristianReformed World Relief Committee (CRWRC)) inresponse to the 2011 drought in Kenya. The objectivewas to gather a deeper understanding of the variousmodes of asset-exchange and their differences asemergency responses. The authors conclude thatwhile cash transfer programmes may be more effec-tive where markets function, other key factors need tobe considered in selecting an intervention includinggender participation, beneficiary preference, projectownership, adverse impacts and behaviouralresponses. The article shows that while there aremany similarities between different models of assetexchange, they are not interchangeable and cannotbe expected to achieve identical results.

Group of womenparticipating in a

focus groupdiscussion

World R

enew

(CRW

RC),

Kenya,

2012

1

2

A final article to mention is by Kate Sadler of TuftsUniversity. This is a study that follows on from earlierwork about the importance of milk in the diet of chil-dren of pastoralists in the Somali region of Ethiopia.This latest study shows that through targeted live-stock support to milking animals that stay close towomen and children during dry season and/ordrought (overall a relatively small proportion of thewhole herd), milk production and consumptionamong children is improved, and their nutritionalstatus benefits.

We have a very large section of researchsummaries in this issue ranging from clinical trials offood products to reviews of the whole humanitariansystem.

One recent study examines the effect of includinga lipid based nutrient supplement (in the form of aReady to Use Supplementary Food (RUSF)) with ahousehold food distribution in Abeche, capital ofChad, on the wasting incidence during the seasonalhunger gap. Interestingly, the study found thatalthough the RUSF improved haemoglobin statusand linear growth, as well as reducing diarrhoea andfever episodes, there appeared to be no effect onpreventing acute malnutrition. Another study byMartha Mwangome and colleagues in Kilifi, Kenyaassesses the accuracy and reliability of using MUAC,length for age and weight for length in measuringnutritional status of infants under six months of age.They found that community health workers can betrained to take MUAC, weight and length measure-ments accurately and reliably among infants age <6months but that length-based z score indices (lengthfor age and weight for length) are the least reliableanthropometric measures.

There is also a summary of a recent systematicreview on the effectiveness of agricultural interven-tions that aim to improve nutritional status ofchildren. Interventions included were bio-fortifica-tion, home-gardening, aquaculture, small scalefisheries, poultry development, animal husbandryand dairy development. The review found that inter-ventions had a positive effect on production andconsumption of the agricultural goods promoted,but that there was no evidence of a change in totalhousehold income and little evidence of a change inthe overall diet of poor people. Furthermore, therewas no evidence found of an effect on iron intake oron the prevalence of under-nutrition, although therewas some evidence of a positive effect on the absorp-tion of vitamin A.

We also include a series of letters from the WorldHealth Bulletin about corporate sponsorship in thepublic health sector, specifically relating to alcoholgiant SABMiller who received a grant from the GlobalHealth Fund to carry out an education interventionaimed at young adults in drinking establishments inSouth Africa. The letters reflect divergent views aboutthis type of corporate sponsorship which to someextent, rehearse the same arguments around corpo-rate sector involvement in the nutrition sector.

Another review by HelpAge International andHandicap International quantifies the fundingprovided by donors to meet the humanitarian needsof two of the most vulnerable groups: older peopleand people with disabilities. It does so by analysingthe amount of humanitarian funding targeted atthese two groups through the UN ConsolidatedAppeals Process (CAP) and Flash Appeals in 2010 and2011. It is incredible to realise that in 2010 and 2011,only 145 (2.4%) of the 6,003 projects submitted tothe CAP and Flash Appeals included at least one

activity targeting older people or people with disabil-ities, and 61 of these were funded (1%). Notsurprisingly, the authors conclude that if the human-itarian community is to fulfil its commitment to theimpartial provision of humanitarian assistance tothose in greatest need, it must take urgent steps toaddress the needs of these two vulnerable groups.

Finally, we have two pieces that address the stateof the humanitarian system. A recent ALNAP reportpresents a system-level mapping and analysis of theperformance of the international humanitarian assis-tance between 2009 and 2011. One of the majorrecommendations in the report is that with the rise ofthe ‘resilience’ agenda, it is critical that new financinginstruments are considered to provide the long-term,flexible financing that these broader non-relief inter-ventions require. These findings are also reflected in arecent Tufts University published paper whichexplores the relationships between climate change,humanitarian crises, and humanitarian responsethrough a review of published and grey literature.Drawing heavily on the CRED data base, the authorsstate that over the past 11 years, climate-relateddisasters have been killing an average of 33,520people a year, and, as critically, affecting the lives ofover 211 million people. The paper goes on toexplore the likely impact of climate change based ona number of scenarios. Major challenges identifiedinclude the fact that prevention, mitigation, andadaptation policies are not in place, capacity toimplement policies is weak and that systematic datacollection to evaluate and report response resultsdoes not exist. A number of quotes from the conclu-sions of this report are well worth repeating in thiseditorial:

“The evidence also suggests that humanitarianoperations are no longer synonymous with emergencyoperations. Most humanitarian assistance today goesinto operations that have been running for five years ormore. As much as 45% goes into programmes morethan eight years old.”

“In these long-term crisis environments, in Ethiopia,Sudan, Afghanistan, Palestine – all environmentallyfragile states – a major opportunity is being missed touse aid to transform the way communities and theirstates develop the necessary economies and gover-nance for the future.”

“The humanitarian aid system evolved as a Westernbased interventionist endeavour, seeing crises as abnor-mal and responding through exceptional interventions.”

“The old methods of working around governmentsystems, rather than with them, have to be challenged.In many crisis-affected states, aid agencies need to seethemselves as long-term partners of the state, providingresponse services, but must also work to build resilienceinto livelihood systems and the infrastructure of hazard-exposed populations. They need to view recovery fromcrisis as a process of change to a more resilient state.Such change will not be easy. The humanitarianresponse sections of aid agencies have tended to seetheir work in terms of logistics and the impartial, neutralsupply of live-saving aid and have shunned much of thepolitical analysis of the development sector, let alonedeveloped an analysis of complex global processes”.

“International humanitarian aid agencies havegrown to become large, multinational organizations,turning over billions of dollars each year and playing acritical role in the creation of international civil societynorms. They now resemble major transnational corpo-rations and find themselves increasingly challenged bythe risk aversion and inertia that comes with scale andan operational model that is still essentially about orga-nizational control.”

There is much food for thought here. While criti-cism is levied at the humanitarian aid system, it couldbe argued that the system has developed mainly dueto whole scale failure of development actors toinclude the vulnerabilities and risks of ‘fragile states’in their operations. Let’s hope therefore that therapidly emerging resilience agenda proves more thanthe ‘emperor’s new clothes’ and enables a continuumof approach at country level and within agencies(including donors and their funding mechanisms).One marker of progress might be a new profile ofdevelopment-orientated partners stepping up toengage in recovery activities especially. On that note,we welcome new faces to the pages of FieldExchange and encourage our seasoned readership toshare this publication with your colleagues in othersectors and in particular those working in the devel-opment arena.

We hope you enjoy another edition of what is ulti-mately your publication, Field Exchange.

Jeremy Shoham and Marie McGrath

Any contributions, ideas or topics for future issuesof Field Exchange? Contact the editorial team onemail: office@ennonline.net

Dr. I

qbal Kerm

ali,

Afg

hanis

tan,

2009

Assessment team assessing wheat yields in Afghanistan(see field article, p3)

3

Food security assessmentof high altitude villages ofBadakhshan, AfghanistanBy Salim Sumar, Laila Naz Taj and Iqbal Kermali

Dr Salim Sumar heads FocusHumanitarian AssistanceEurope Foundation, which isan affiliate of the Aga KhanDevelopment Network1.Based in London, England,

FOCUS Europe has regional responsibilitiesfor communities in Africa, Middle East,Afghanistan and Europe. Salim has workedextensively in humanitarian emergencies,conflicts and international development inCentral Asia, Africa, Middle East and Europeas well as in academia as Professor of Food,Nutrition & Public Health.

The authors acknowledge the organizationsupport of Focus Humanitarian Assistanceand the Aga Khan Development Network inpreparing this article. They wish also toacknowledge the contributions made bythe staff of Focus Humanitarian Assistance,Afghanistan, in particular Noor Kashani andHesamuddin Hashuri.

Laila Naz Taj is a researcherat Focus HumanitarianAssistance EuropeFoundation. Laila holds anMSc in DevelopmentStudies from the London

School of Economics and has extensiveinternational and field based experience inthe developmental sector, research, policyand practice. She has worked in Pakistanand Afghanistan with a variety ofinternational NGOs.

Dr Iqbal Kermali has astrong internationalbackground with significantexperience in relief,rehabilitation anddevelopment programmes,

particularly Central Asia. His specialisationincludes agriculture development,livelihoods and management of naturalresources. He has worked for FAO and theAga Khan Development Network inAfghanistan. Currently he is the TeamLeader and Training Coordinator of theAsian Development Bank-funded CapacityDevelopment for the Western Basins WaterResources Management Project and isbased in Herat, Afghanistan.

Fie

ld A

rtic

leFocus assessment team with localcommunity members in Zebak

Dr. Iqbal Kermali, Afghanistan, 2009

Afghanistan is a landlocked countrythat is bordered on the north byTurkmenistan, Uzbekistan andTajikistan, on the extreme north-

east by China, on the east and south byPakistan, and by Iran on the west. The coun-try is divided east to west by the Hindu Kushmountain range, rising in the east to heightsof 24,500 feet (7,485 m).

In its recent past, Afghanistan has endureddecades of political instability, protractedconflict and natural disasters, all of whichhave contributed to various insecurities in thecountry. As such, progress with regard tohuman development has been slow, with lifeexpectancy marginally increased to 44.5 yearsfor men and 44 years for women2. Most otherhuman development indicators also remainunacceptable and are among the worst in theworld, including maternal mortality rates(1,600 deaths per 100,000 live births) andunder 5 mortality rates (199 per 1,000 livebirths). Based on the broader set of the UnitedNations Development Programme (UNDP)human development index indicators(health, education, living standards),Afghanistan ranks 181 out of 182 countries.

Almost half of the estimated population of25 million still live below the poverty line andthe nation faces many challenges, particularlywith regards to poor nutrition and foodinsecurity3. Approximately half of Afghanchildren under 5 years are chronicallymalnourished and over a third are under-weight. In addition, three-quarters of allchildren under 5 years suffer from criticalmicronutrient deficiencies, most notably ofiron and iodine. These poor nutritionaloutcomes are closely linked to the state offood security in the country. Food securityand its consequences vary considerablyacross the 34 provinces in Afghanistan,reflecting the diversity of social and economicconditions4.

FOCUS in AfghanistanFocus Humanitarian Assistance (FOCUS) hasbeen implementing emergency response andrehabilitation activities in Afghanistan sincethe 1990s. FOCUS also work in the mostremote (high altitude) regions of easternAfghanistan. As a part of its activities,

detailed food security assessments areconducted at regular intervals. This informa-tion is shared with its partners including thegovernment and international organisations,especially the World Food Programme(WFP). The assessment described in this arti-cle was conducted in Badakhshan, one of the34 provinces of Afghanistan with a popula-tion of 2 million inhabitants.

MethodologyThe study conducted in 2009/2010 profilesfood security in the high altitude regions ofBadakhshan province. The province islocated in the north-east of the country andhas a total area of 44,060 km², most of whichis occupied by the Hindu Kush and Pamirmountain ranges.

Food production is limited by insufficientarable land, extreme winters, a shorter grow-ing season, poor marketing and lack ofinputs.

A team of four trained assessors, accompa-nied by community volunteers, visited 195villages over a period of ten months, startingin mid-June, 2009. The team interviewed keyinformants including members of the villagecouncils, officials of the local and districtgovernments and members of civil societyorganisations. A semi-structured assessmentform was designed, translated into Dari andvalidated to facilitate use at the field level.The team was also trained in conductingsemi-structured interviews and in collectinginformation from communities. Informationcollected included data on demographics,livelihoods, state of food security, causes ofand any coping/management strategiesemployed to deal with food insecurity. Thisinformation was translated from Dari intoEnglish and compiled into a database.

1 http://www.akdn.org/2 Afghanistan Overview, Fighting Hunger Worldwide,

World Food Programme, [online] Available at

http://www.wfp.org/countries/Afghanistan/Overview

> [Accessed] 10 May 20123 Summary Evaluation Report Afghanistan, 2010, World

Food Programme, [online] Available at

http://one.wfp.org/eb/docs/2010/wfp213169~1.pdf

> [Accessed] 16 May 20124 Poverty and Food Security in Afghanistan, 2012, The

World Bank Economic Policy & Poverty Sector South

Asian Region, [online] Available at

http://moec.gov.af/Content/files/FSR_v7.pdf >

[Accessed] 18 May 2012

4

Field Article

neighbouring countries. With sufficient infra-structure and investment, it is clear that thereexists a vast labour pool which can be chan-nelled into productive industries.

Despite popular belief, there is no significantproduction of narcotics in the area. However,trade involving external actors is evident alongthe border with Tajikistan. There is no monetarygain to the villages assessed through narcoticsproduction or through trade. However,narcotics are readily available to the residentsand thus represent a large monetary loss to thefamilies whose members are addicts. The over-all addiction rate in all the villages assessed wasestimated at approximately 4%. This comparesto a national figure of 8% of the populationaged 15 to 64 years. The lower addiction rates inBadakshan are due primarily to improved secu-rity in the region and the population beingmore aware of the negative impact of consum-ing and trading in narcotics. There also appearsto be a correlation with the geographic areas ofproduction that are mainly in the northern andsouthern regions of the country6.

Food security conditionsFood secure villages were also scored based onthe local knowledge of the informants using localDari terminology. The informants scored thegeneral food security conditions of their respec-tive villages in the better year, worst year and theprevious year (2008). Food secure villages werescored either as normal or borderline.

In an average year7, 99 villages are foodsecure (17 villages are normal and 82 villagesare borderline) and 96 are insecure (of which 39are very insecure). The number of very foodinsecure villages in 2008 was 56, i.e. it was avery bad year. The average annual food gapduration in the villages is 58 days, and the aver-age number of households within a village thatare food insecure is 80%.

In a bad year, the percentage of food insecurehouseholds is higher than 90% in all areas andis 100% in 10 of the 13 districts. In the face ofsuch conditions, families employ various(damaging) coping mechanisms; respondentsindicated that they have on different occasionsresorted to selling assets, migrating, purchasing

are at more than 3,000m. This is shown inFigure 1 which plots the altitudes of individualvillages of the nine districts surveyed.

Villages located above 2,000m (86% of thosesurveyed) depend mostly on subsistence farm-ing for food and are generally food insecureand characterised by both low food consump-tion and lack of dietary diversity. Thesemountainous areas have only one crop per yearand there is very little arable land for cultiva-tion. In general, most communities in theseareas are geographically isolated because ofpoor or no roads and they have little or noaccess to seasonal food markets or health facili-ties, particularly during winter. The traditionallivelihood system in these areas is primarilylivestock husbandry.

Overall living conditionsThe overall living conditions for each villagewere assessed by the local village informantsusing simple and local Dari language terminol-ogy (translated into English for analysis). Thesekeywords roughly equated to the village as‘very poor’, ‘poor’, ‘sufficient’, ‘good’ and ‘verygood’. These simple and qualitative scores aretherefore based on informant’s perception rela-tive to other villages in the province. Resultsshowed that 102 villages were below normal(poor or very poor), 33 villages were normal(sufficient), and the remaining 60 villages abovenormal (good or very good). While, the datawas encouraging in identifying a significantnumber of villages above normal, the majorityof villages were below normal, indicating aworrying trend. The geo-political instability inthe province presents a further risk as in mostcases, the local, central and provincial authori-ties remain weak.

LivelihoodsDiversity of income streams outside of theagricultural sector is a key component toensuring food security. The main sources ofon-farm income in the villages were cropproduction (58%) and livestock (30%),followed by fruit production (7%) and forestry(5%) (see Figure 2).

Only 6.5% of the total population isemployed in the off-farm sector. About 52% ofoff-farm employment is in the work force(labourers), followed by government servicemostly in education (28%), military (11%),private service (5%), NGOs (5%), and busi-ness/shopkeepers at (4%).

On average, 11% of the households in thedistricts assessed receive external support,usually as remittance from family membersworking in other parts of Afghanistan or in the

5 Anon, 2011, Afghanistan, World Bank Provincial Briefs,

[online] Available at:<http://siteresources.worldbank.org/

AFGHANISTANEXTN/Resources/305984-1297184305854/

ProvBriefsEnglish.pdf > [Accessed 13 January 2012]6 Drug Use in Afghanistan, Executive Summary, 2009,

United Nations Office of Drugs and Crime, [online],

Available at http://www.unodc.org/documents/data-and-

analysis/Studies/Afghan-Drug-Survey-2009-Executive-

Summary-web.pdf > [Accessed 25 May 2012]7 Average year score is derived from the scores for the

better and worst years.

Figure 1: Altitude of surveyed villages

4000

3500

3000

2500

2000

1500

1000

500

0

Alti

tude

s

Finally, the information collected for all thevillages at district level were consolidated,reviewed and validated by the local representa-tives and the non-governmental organisations(NGOs) and UN agencies working in the area.A copy of the information collected wasprovided to the respective village leader forfuture reference.

ResultsThe assessment focused on 195 villages at highaltitudes and examined both agro-ecological(e.g. altitude, wheat yields) and socio-economicfactors (e.g. market access, infrastructure, andlivelihoods) in relation to food security.

A substantial number of the 195 villagessurveyed were found to be food insecuredespite various efforts to rehabilitate agricul-tural systems, improve livelihoods and militateagainst disasters. These efforts includedhumanitarian and rural development activitiesby various UN and other international agen-cies, such as food for work and fertiliserdistribution, since the 1990s.

Overall, results of this study indicated insuf-ficient support, infrastructure and access tomarkets and frequent shocks from floods, land-slides, and droughts. Furthermore, demands onnatural resources and recent (sharp) increasesin the price of food, fuel and transportation hadall contributed to increasing food insecurityand the vulnerability among the rural popula-tions. A recent World Bank survey indicatedsimilar concerns with regards to food insecurityand vulnerability among the population, with70% of the local population being calorie defi-cient and 61% considered to be in poverty5.

Despite sharing international borders withPakistan, Tajikistan and China, internationaltrade links barely exist outside of narcotics.Official trade exists only with Tajikistan acrosstwo border posts, both of which are difficult toaccess (especially in winter). There is no majorelectricity grid and the first paved main road(linking to neighbouring Takhar) came to theprovince in 2010. Badakhshan’s inaccessibilityand lack of available agricultural land hasresulted in a slow and generally underdevel-oped local economy.

AltitudeBadakhshan province is largely mountainous(reaching as far up as 7485 m), allowing for littleagriculture except animal husbandry, usually ofsheep and goats. The population surveyed ismostly settled in higher altitude areas: 49% at2,500m to 3,000m range followed by 24% at2,000m to 2,500m range. Fourteen per cent ofthe villages are at less than 2,000m while 13%

0 10 20 30 40 50Villages

Maimai

Nusai

Shukai

Yomgan

Warduj

Arghanj

Khaw Wakhan

Kuran Wa

Munjan Zebak

Jurm

Darayem

100%90%80%70%60%50%40%30%20%10%

0%

Argh

anj K

Dara

yem

Ishka

shim

Jurm

KwM

Mai

mai

Nusa

i

Shug

hnan

Shuk

ai

Wak

han

War

duj

Yom

gan

Zeba

k

Over

all

Crop Production Livestock Fruit production Forestry

Figure 2: On farm livelihoods in surveyed districts

Field Article

5

no local markets and the population dependson markets in the neighbouring big towns tobuy this staple. Moreover, availability variesfrom month to month.

Yield of irrigated wheat is a good indicatorof food productivity in these areas. Overall, the2008 average yield in the villages (742 Kg/ha)was lower compared to 907 Kg/ha in 2007 and888 Kg/ha in 2006. The 2008 rain-fed wheatyield (408 Kg/ha) was on average 85% of the2007 yield (481 Kg/ha) and 70% of the 2006yield (587 Kg/ha).

On average, the rain fed yield was 47% of theirrigated wheat yield in the villages assessed. Asimilar trend was observed in previous years,with significantly lower yield in rain fed areas(34% in 2006, 47% in 2007 and 45% in 2008). In2011, the combination of falling yield in bothrain-fed and irrigated lands in Badakhshan,coupled with a decrease in the geographicalarea available for farming, has seen a significantwheat deficit arise (-86,100 metric tons). Wheatprices, in turn, have significantly risen since2008, with prices of wheat and wheat flourhaving increased by 79% on the year before.The July 2011 price of all cereals increased by afurther 60% above pre-food price crises (Jan-Oct 2007, FEWS, 2011)8.

Conclusions and recommendationsThe study conclusion reveals a population

locked into a cycle of food insecurity with veryfew avenues of escape. The increased incidenceof natural disasters coupled with a dependencyon subsistence farming has meant thatBadakhshan is a region blighted by poverty andvery low living standards.

On the basis of information collected fromthe food security assessment and from discus-sions with community leaders, authorities andofficials of civil society institutions, a number ofinter-linked measures were recommended forresponding to the repeated cycles of food crisisamong the communities in Badakhshan. Theseinclude:• Empower the Community Development

Councils (CDCs) (see Box 1) in the target villages to play a leading role when planningand implementing any food security relatedactivity at the village level.

• Further develop and implement a compre-hensive food insecurity surveillance and response mechanism for the target areas, with the CDCs playing a major role.

• Strengthen support to create or increase the existing food stored in the stockpiles ownedand managed by the communities.

• Continue emphasis on multi-dimensional agricultural approaches such as agriculture diversification, agro-enterprises and value added agriculture to reduce the food gap.

• Improve access to appropriate agricultural inputs through microfinance or other innovative credit programmes with emphasis on sustainability and economic viability.

• Cash transfers to the most vulnerable during food crisis can be made more sustainable by linking them to community managed revolving food stockpiles.

• Implement additional disaster mitigation projects that can also contribute to the local food economy by improving market access

food on credit, increasing child labour, andbegging.

Overall causes of food insecurityThe villages surveyed identified a number ofdifferent causes of food insecurity. In 186villages, lack of agricultural inputs is the maincause of food insecurity followed by damagefrom hazards (138), poor markets (133) and lackof employment (118). The effects of naturaldisasters are exacerbated by the lack of infra-structure and the isolation of villages strung outalong high mountain valleys, causing disrup-tion to food production. Transportation to andaccess to agricultural inputs and products forthe isolated and remote villages is always diffi-cult due to lack of formal roads and/oravailability of inputs. This is particularly chal-lenging during winter, given the volume ofsnow or at times of landslides and floods, whenassistance is most needed. In addition, lack ofcapital or unusually high price increases forfood results in communities having to cope byborrowing, selling or exchanging other assets topay debts and/or to get more cash to buy foodduring shortages or emergencies. Addiction todrugs was also another cause of food insecurityamong the communities.

Wheat yields and pricesWheat is the main staple food for mostAfghans. In many villages assessed, there are

Box 1: Community Development Councils (CDCs)

Community Development Councils (CDCs) inAfghanistan were first established under the NationalSolidarity Programme (NSP) which is the largestcommunity development programme in the historyof Afghanistan. Known in the local Afghan languageof Dari as ‘Hanbastagi Milli’ and in Pashtu as ‘MilliPawastoon’, it is based on the Afghan traditions of‘Ashar’ – community members working together on avolunteer basis to improve community infrastructureand ‘Jirga’ – councils comprised of respectedmembers of the community; and Islamic values ofunity, equity and justice.

The primary role of the CDCs is to serve as a consulta-tive decision-making body that includes men,women, and traditionally marginalised members ofthe community. It is selected by the communitythrough fair and open elections. Through participa-tion in NSP and other programmes, communities willacquire or strengthen the skills and attitudes neces-sary to define, manage, and govern theirdevelopment.

CDC/NSP consists of four core elements: 1. Community mobilisation – facilitating elections to

establish CDCs and helping CDCs identify priority sub-projects, prepare Community Development Plans, and implement approved subprojects.

Box 2: National Rural Access Programme (NRAP)

The National Rural Access Programme (NRAP),formerly called the National EmergencyEmployment Programme (NEEP), was launched in2002. Its aims were to increase access to ruralinfrastructure for local communities and toprovide employment opportunities for rurallabours. As a result of a joint assessment of NEEPconducted by the World Bank and the govern-ment of Afghanistan in early 2005, theprogramme was restructured to make a smoothtransition to more development-centred assis-tance in Afghanistan. The renamed ‘National RuralAccess Programme (NRAP)’ has a more strategicfocus on the provision and support of enhancedlivelihoods. The Programme provides for thedevelopment of quality rural access infrastructureand capacity to maintain it (enabling rural roadaccess to services such as markets, health careand schools) and in doing so, a mechanismwhereby temporary employment will provide asafety net for vulnerable rural people.

NRAP is a joint national priority programme thatis executed by Ministry of Public Works (MoPW)and the Ministry of Rural Rehabilitation &Development (MRRD). Under the two ministries,UNOPS is responsible for implementation andalso provides technical support and capacitydevelopment of the ministries.

2. Building the capacities of CDC and community members (both men and women) in participation,consensus-building, accounting, procurement andcontract management, operations and mainte-nance, and monitoring.

3. Providing direct block grant transfers to fund subprojects

4. Linking CDCs to government agencies, NGOs, and donors to improve access to services and resources.

Two types of subprojects may be financed under NSP:public infrastructure (such as water supply and sani-tation, irrigation, transport, hospital and schoolbuildings) and human capital development (generaleducation including health and hygiene, child devel-opment training, training on birth attendant trainingand productive skills training, e.g. kitchen gardens,animal husbandry, bee-keeping, food processing, andvocational education.

CDCs are encouraged to link with other programmesand take on additional responsibilities as their capac-ity evolves (and as allowed under Afghan law).Discussions are currently underway within theGovernment to identify an appropriate strategy forlegally recognizing CDCs as a local government bodyand establishing mechanisms to link CDCs to district,provincial, and national government bodies.

8 Anon, 2011, Afghanistan Food Security Outlook, Famine

Early Warning Systems Network (FEWS), [online], Available

at:http://www.fews.net/docs/Publications/Afghanistan_OL_

2011_01.pdfm [Accessed 13 January 2012]

Focus staff conducting interviewswith community informants

Dr. I

qbal Kerm

ali,

Afg

hanis

tan,

2009

6

Research

and increasing water availability, in addition to reducing the hazard risks.

• Participate actively in processes such as the Afghanistan Humanitarian Action Plan9 that are a potential source of funding for disasterresponse including food insecurity.

In addition to the above, four broadrecommendations also emerge from thestudy for Badakhshan:

There needs to be recognition thatfood programmes do not represent thebest method of achieving food security.Income-generating programmes aregenerally more efficient, such as theNational Rural Access Programme(NRAP) formerly called the AfghanNational Emergency EmploymentProgramme (NEEP) and micro-financeprogrammes (see Box 2 for NRAP/NEEP). Regular wage incomes smoothconsumption patterns, therefore alsoensuring health and education security.These incomes can also be reinvestedback into purchasing agricultural inputs,thereby ensuring a diversity of incomesand food sources within a single house-hold.

It is important to appreciate thatextreme weather events represent a majorthreat to food security in Afghanistan,and in particular in mountainous areas ofBadakhshan where natural hazards areprevalent.

An approach that not only emphasisesfood security, but also outlines a path tofood sovereignty should be considered.National programmes which engagelocally governing CDCs are best situatedto tackle this issue successfully.

Ensuring sustainable food production,resolving disputes related to land tenure,and stewardship of natural resourcesneeds to be encouraged. Water conserva-tion, food harvesting techniques based onindigenous knowledge, and low techconstruction methods need to be intensi-fied through basic technology transferand extension work by agricultural aidagencies. Drought-resistant high marketvalue crops should be identified andtested for long-term productivity.

In conclusion, only an integrated (andadaptive) approach will ensure a sustain-able food security future in AfghanBadakhshan with improved healthoutcomes, especially among high altitudevulnerable communities. This is based onthe recognition that extreme weatherevents, disasters, land degradation, lackof investment in infrastructure, poortrade links may all be a consequence andcause of poverty and therefore food inse-curity in this province.

For more information, contact: Laila NazTaj, email: Lailanaz.taj@focushumaitarian.org orSalim Sumar, email:salim.sumar@focushumanitarian.org

The authors of a recent study hypoth-esized that including a daily dose of46 g of Lipid based Nutrient

Supplement (LNS) as Ready to UseSupplementary Food (RUSF), for consump-tion by children between 6 and 36 monthsof age as part of a household food distribu-tion programme would reduce cumulativewasting incidence during the seasonalhunger gap (June to October).

The study was conducted in the city ofAbeche, the capital city of the Ouaddaı¨region in eastern Chad. Since 2006, this partof the country has been characterised bychronic political instability, which has ledto the decline in nutritional status of chil-dren under 5 years of age. Recently, thenon-governmental organisation ActionContre la Faim – France (ACF-France)conducted two cross-sectional nutritionalsurveys of children under 5 years inAbeche. A first survey was conducted atthe beginning of the rainy season (June2009) and revealed a prevalence of wastingof 20.6%, with 3.2% severe wasting, usingthe National Centre for Health Statistics(NCHS) growth reference. A second surveywas carried out during the post-harvestperiod (January 2010) and reported a wast-ing prevalence of 16.8%, with 2% severewasting

The study was conducted in sevenvulnerable sectors of the city of Abeche thatwere preselected from a total of 45 admin-istrative sectors. These seven sectors wereidentified based on data from a communitynetwork organised by ACF-France involv-ing a set of socio-economic, sanitary andnutritional (proportion of children admit-ted to ACF-France nutrition rehabilitationprogrammes) criteria. The selected sectorswere subdivided into 14 geographical clus-ters using main roads and rivers as clusterboundaries. The cluster was the unit ofrandomisation andrandom assignmentwas conducted through an official ceremo-nial gathering with officials andcommunity members. Seven clusters wereassigned to the intervention group andseven to the control group.

The researchers designed the study as atwo-arm cluster-randomised controlled prag-matic trial, targeting children from 6 to 36months of age from vulnerable households. Ahousehold was considered to be ‘‘vulnerable’’when it met one of the following criteria: (1)household head being disabled, pregnant orlactating, or (2) an economic dependencyratio of 4:1 or more (number of economicallyinactive versus active household members).The inclusion criteria for the study werebeing non-wasted (weight-for-height >80% ofthe NCHS reference median, and lack of bilat-eral pitting oedema) and being from a‘‘vulnerable’’ household.

The study used the NCHS growth refer-ence for enrolment in the study, to conformwith the Chad national protocol for themanagement of malnutrition. However, theresearchers opted to analyze the data usingthe WHO international growth standards tomake the results more comparable to recentstudies. None of the study conclusions werealtered when analysing the outcome datausing the NCHS growth reference.

The primary study outcome was the cumu-lative incidence of acute malnutrition orwasting defined as weight-for-height Z-score(WHZ), or presence of bilateral pittingoedema. In order to detect a 50% reduction inthe cumulative incidence of wasting over aperiod of 4 months, with a statistical power of80%, a sample size of 1,220 children was calcu-lated to be needed. Taking into account astudy dropout of 15%, a total sample size of1,435 children was projected. Secondaryoutcomes included mean WHZ change overtime, prevalence of stunting at end pointdefined as height-for-age Z score (HAZ),mean HAZ change over time, mid upper armcircumference (MUAC) change over time,mean haemoglobin concentration at endpoint, prevalence of anaemia at end point(haemoglobin <110 g/l).

Effect of adding RUSF to ageneral food distributionon child nutritional statusand morbidity: a clusterrandomised controlled trialSummary of research1

1 Huybregts L et al (2012). The Effect of Adding Ready-to-

Use Supplementary Food to a General Food Distribution

on Child Nutritional Status and Morbidity: A Cluster-

Randomized Controlled Trial. PLOS Medicine.

www.plosmedicine.org. September 2012, volume 9,

issue, 9, e1001313, pp 1-11

Child during appetitetest at a health facility

offering treatment inMonrovia, Liberia

Ben A

llen,

Lib

eria,

2011

9 Ministry of Rural Rehabilitation and Development,

access at: http://mrrd.gov.af/en/page/69

Re

se

arc

h

7

Research

Study interventionsHouseholds of both the intervention andcontrol groups received a monthly food pack-age representing a daily ration of 425g ofsorghum, 25 g of legumes, 25 g of bleachedpalm oil, 20 g of sugar, and 5 g of iodized salt.This ration was estimated to cover approxi-mately 86% (<1,800 kcal) of the daily energyrequirements for a population at risk of anemergency. The number of food rations distrib-uted per household was proportional to its size.Children from the intervention group receiveda monthly quantity of RUSF (Plumpy’Doz,Nutriset) representing a daily ration of 46g(<247 kcal/d). Recommendations on dosageand frequency of consumption by targeted chil-dren were made to caretakers and repeated ateach follow-up visit. The intervention lasted 4months (June 2010 to September 2010). Prior tothis intervention, an acceptability test wasconducted in a convenience sample of 30 non-wasted children.

The study team encouraged the mothers tobring their children with them to the monthlyfood distributions, regardless of their interven-tion status. Children were enrolled from earlyJune to mid-July 2011 and scheduled to comefor four follow-up visits. At each visit, childanthropometric measurements and morbiditywere recorded. Children who were classified asmoderately wasted (weight-for-height ≥70-<80% of NCHS reference median) or severelywasted (weight-for-height <70% of NCHSreference median) were discharged from thestudy and referred to a community-basedmanagement of acute malnutrition programmelocated within the city’s nutrition rehabilitationcentres. All participating mothers were given afamily food ration as described above.

The study team used a pretested question-naire to collect data on socioeconomic anddemographic characteristics of enrolled house-holds. Child age at screening was estimatedusing a locally adapted event calendar if a birthcertificate was unavailable. Anthropometricmeasurements were conducted monthly.Episodes of diarrhoea, respiratory tract infec-tion, and fever were recalled for 1 week beforethe monthly interview. Respiratory tract infec-tion was diagnosed through reports by themother/ caregiver of persistent cough or diffi-culty in breathing during the last week(yes/no). A diarrheal episode was defined ashaving at least three loose stools within a day.Fever episodes were diagnosed bymother/caretaker during the last week (yes/no). In case of death, the team carried out averbal autopsy adapted from WHO standards.Haemoglobin concentration was measured atbaseline (June) and at the end of the interven-tion (November) or when a child wasdischarged from the study, using a daily cali-brated electronic device HemoCue Hb 201+.Standardised forms were used to collect data.

ResultsThe overall sample size was 1,038 children in784 households: 598 children in the interventiongroup and 440 children in the control group.

Table 1 details the effects of preventive RUSFon child anthropometry. Compared to baselinevalues, mean WHZ for both control and inter-vention groups were slightly higher at endpoint, while mean HAZ was slightly lower.There was no difference in the incidence of

wasting (incidence rate ratio: 0.86; 95% CI: 0.67,1.11; p= 0.25) or mean change in WHZ (20.002WHZ/month; 95% CI: 20.032, 0.028; p =0.89)between the arms. The difference in weightincrease between groups was 0.02 kg/month(95% CI: 20.01, 0.04; p=0.10). Children in theintervention group had a significantly higherlinear growth velocity of 0.03 HAZ-score/month (95% CI: 0.02, 0.05; p<0.001)compared to the control group . This observeddifference was equivalent to a small differencein height gain of 0.09 cm/month (95% CI: 0.04,0.14; p<0.001).

Identical ponderal (weight)growth in controland intervention groups was confirmed by alack of difference in MUAC growth. Age atinclusion was not found to modify the interven-tion effects on child anthropometric meas-urements. After adjustment for age, sex, socio-economic status, and morbidity status atinclusion, the study found that children fromthe RUSF group had lower risk of self-reporteddiarrhoea by 29.3% (95% CI: 20.5, 37.2; p<0.001)and fever by 22.5% (95% CI: 14.0, 30.2; p<0.001),compared to the control group. RUSF signifi-cantly increased mean haemoglobin concen-tration at end point by 3.8 g/l (95% CI: 0.6, 7.0;p= 0.02), resulting in significantly lower odds ofanaemia (OR: 0.52; 95% CI: 0.34, 0.82; p= 0.004)for children in the intervention group.

The absence of an effect on wasting inci-dence could have multiple explanations. First,the energy contribution of RUSF may have been‘‘diluted’’ by the general food distribution,which mainly provided a supplement of energyand protein. Second, the energy dose of 46 g(<247 kcal) daily RUSF and the duration of thesupplementation could have been insufficientto support ponderal growth, particularly for theolder children in the cohort. Furthermore, it ispossible that the RUSF may have been sharedwith other children in the household. However,if this were done to a large extent, the observedintervention effects on secondary outcomes likeHAZ and haemoglobin concentration are hardto explain.

The absence of an effect on ponderal growth,but modest effects on morbidity, linear growth,

and, most of all, haemoglobin could suggestthat a multiple micronutrients (MMN) effect isat play. For example, zinc, one of the micronu-trients added to RUSF, is currentlyrecommended as adjunct therapy by the UnitedNations Children’s Fund and WHO for thetreatment of diarrhoea.

These observations could lead to the specu-lative hypothesis that supplementation withMMN supplements like powders or tabletsmight result in the same effects as RUSF, if basicfood rations were provided.

One important additional benefit that LNSoffer is the lipid component. In addition toproviding a small amount of essential fattyacids, which hold a potential to support childgrowth, the lipid component serves as an essen-tial matrix that ensures that fat-soluble vitaminslike vitamin A, D, and E are properly absorbed.Particularly when the child’s diet is poor in fat,this would provide leverage to increase the effi-cacy of supplemented fat-soluble vitamins.Therefore, more mechanistic studies arerequired to elucidate the additional contribu-tion to the efficacy of the MMNs by thefunctional fat fraction of the RUSF.

The study had a number of limitations. Theprojected sample size was not attained, limitingthe study’s statistical power. The study partici-pantswere not blinded with respect to theintervention assignment because of the type ofsupplement (paste) provided to children.Clusters were not always geographically sepa-rated from each other. And finally, childmorbidity was recorded through caretaker recall,which could have resulted in underestimation.

In conclusion, adding child-targeted RUSFsupplementation to a general food distributionresulted in increased haemoglobin status andlinear growth, accompanied by a reduction indiarrhoea and fever episodes. However thestudy did not find clear evidence that addingRUSF to a household food ration distribution ofstaple foods was more effective in preventingacute malnutrition. Other context-specific alter-natives for preventing acute malnutritionshould therefore be investigated.

Outcome Control Arm (n=440) Intervention Arm (n=598) p-Value

Wasting

End point mean WHZ (SD) –1.09 (0.95) –1.05 (0.93)

Intervention effect (95% CI), Z-score/moa Reference –0.002 (–0.032, 0,028) 0.89

Cumulative episode WHZ<–2 174 241

Number of observed child-months 1,427 2,199

Number of episodes per child-months (95% CI)b 0.12 (0.10, 0.14) 0.11 (0.09, 0.14)

Incidence rate ratio (95% CI)C Reference 0.86 (0.67, 1.11) 0.25

Stunting

End point mean HAZ (SD) –2.06 (1.39) –1.79 (1.46)

Intervention effect (95% CI), Z-score/moa Reference 0.03 (0.01, 0.04) <0.001

End point prevalence of stunting percent (n) 52.3 (230) 46.2 (276)

OR of end point stunting (95% CI)d Reference 0.69 (0.45,1.07) 0.099

MUAC

End point MUAC, cm (SD) 14.1 (1.2) 14.3 (1.1)

Intervention effect (95% CI), cm/moa Reference 0.01 ( –0.02, 0.04) 0.49a Analyzed using a linear mixed model with random effects cluster, household, and child, adjusted for child’s ae at baseline, child’s

sex, SES, and baseline value.b Confidence intervals are estimated from a Poisson model adjusted for clustering.c Analyzed using a mixed Poisson regression model with random effects cluster, household, and child, adjusted for child’s age at

baseline, child’s sex, SES, and baseline value.d Analyzed using a mixed logistic model with random effects cluster and household, adjusted for child’s age at baseline, child’s

child’s sex, SES, and baseline value.doi: 10.1371/journal.pmed.1001313.t003

Table 1: Effects of preventative RUSF on child anthropometry

8

Impact of livestock support on animal milksupply and child nutrition in EthiopiaSummary by Kate Sadler and Emily Mitchard

Kate Sadler is an Assistant Professor of Nutritionat the Feinstein International Centre, TuftsUniversity and the Research & DevelopmentManager at Valid International. A public nutri-tionist with over 15 years of experience in thedesign, management and evaluation of nutritioninterventions in sub Saharan Africa, shecompleted an MSc in Public Nutrition at theLondon School of Hygiene & Tropical Medicinein 1997. She went on to complete her PhD,which focused on the early development ofCTC/CMAM, in 2008.

Emily Mitchard is a food security and nutritionconsultant currently based in Bangkok, Thailand.

Previously, she worked with FeinsteinInternational Centre on the Milk Matters projectsin Somali Region, Ethiopia and Karamoja, Uganda.She holds an MSc in Agriculture, Food Securityand Nutrition and an MPH in Biostatistics andEpidemiology from Tufts University.

This work was made possible with funding fromthe Office of Foreign Disaster Assistance (OFDA)United States Agency for InternationalDevelopment (USAID). The authors would like tothank Axel Weiser, Girma Tadesse, MohammedMusse, and Nasteha Ahmed of Save the ChildrenUSA and Hassan Ibrahim and Mohammed Ali ofSave the Children UK for providing considerable

BackgroundAnimal milk has long been recognised as animportant component of pastoralist diets acrossthe world1. As a nutrient-dense food, milk isknown to contribute a high proportion of thenutrients, such as high quality protein andmicronutrients, to the pastoralist diet2,3,4.Previous research in the Somali region ofEthiopia has found that Somali pastoralistsconsume, on average, between 20 and 50percent of their energy requirement as milk andanimal products in a normal rain year5,6,7. ForSomali children in particular, when milk isplentiful it was found to constitute a centralpillar of the diet, providing two-thirds of thedaily energy requirement and 100 percent of thedaily protein requirement7. Yet, in recentdecades, levels of global acute malnutritionamong young children in the Somali region areregularly reported to rise above 15%, the leveldefined as a nutritional emergency by theWorld Health Organisation (WHO)9. Thesesurveys identified a seasonal aspect to child

malnutrition, with particularly high rates ofacute malnutrition occurring during the dryseason and periods of drought. Seasonal varia-tion in livestock milk production has also beenwell described in the literature on pastoralismin Africa, with milk supply falling as the dryseason advances10,11,12. Whilst this work indicatesthat the main risk period for child malnutritionoccurs routinely in the late dry season in manypastoralist areas, nutrition and humanitarianprogramming continues to take a reactive emer-gency approach emphasizing the delivery offood aid and therapeutic treatment of severeacute malnutrition (SAM). Moreover, food aidis typically characterized by the delivery of afood aid basket that rarely includes a protein orfatty acid source suitable for infants and youngchildren. Finally, little has been done to under-stand the potential role of milk in preventingmalnutrition in pastoral regions.

As part of Save the Children’s AfricanRegion Pastoral Initiative13, the ‘Milk Matters’

1 Sadler, K., C. Kerven, et al. (2010). The fat and the lean: a

review of production and use of milk by pastoralists in

press. Pastoralism: research, policy and practice2 Galvin, K. A., D. L. Coppock, et al. (1994). Diet, nutrition,

and the pastoral strategy. African pastoral systems: An

integrated approach. E. Fratkin, K. A. Galvin and E. A. Roth.

London, Lynne Rienner: 113-132.3 Fratkin, E., E. A. Roth, et al. (2004). Pastoral sedentarization

and its effects on children's diet, health, and growth among

Rendille of Northern Kenya. Human Ecology 32(5): 531-559.4 Barasa, M., A. Catley, et al. (2008). Foot-and-mouth

disease vaccination in South Sudan: benefit-cost analysis

and livelihoods impact. Transboundary and Emerging

Diseases 55:339-351.5 Webb, P. and J. Braun (1994). The Pastoral Experience.

Famine and food Security in Ethiopia: lessons for Africa.

Chichester UK, Wiley & Sons Ltd: 85-98.6 SCUK and DPPA (2002). Filtu-Dolow Pastoral Livelihood

Zone: a HEA baseline study. Addis Ababa, Save the Children

UK and Disaster Prevention and Preparedness Bureau.

support and hard work throughout the imple-mentation of the study, as well as for providingvaluable feedback for this report. Thanks also goto Adrian Cullis of the Food and AgricultureOrganisation (FAO) for important input on initialconcepts and intervention design. Finally, manythanks go to the dedicated data collectors andcommunity members who made the studypossible.

The full report of the research summarised here,including additional methodology and the resultsfrom the other three intervention sites, can befound on the Feinstein International Centre websitehttp://sites.Tufts.edu/Feinstein/2012/Milk-Matters

7 SCUK (2007). The causes of malnutrition in children under

3 in the Somali Region of Ethiopia related to household

caring practices: Preliminary Report. A.-M. Mayer. Ethiopia,

Report on research findings from Somali Caring Practices

research project in Shinile and Dambal districts of Shinile

zone, Somali region, Ethiopia.8 Sadler, K. and A. Catley (2009). Milk Matters: the role and

value of milk in the diets of Somali pastoralist children in

Liben and Shinile, Ethiopia. Addis Ababa, Feinstein

International Center. See also footnote 7.9 Ethiopian Health and Nutrition Research Institute, UNICEF,

et al. (2009). Final Report from Nutrition and Mortality

Surveys conducted in Seven Mega Livelihood Zones in

Somali Regional State, Ethiopia. Addis Ababa, Ethiopian

Health and Nutrition Research Institute.10 Chell, D. and A. Chell (1979). A detailed evaluation of

the primary health care programme among nomadic

tribes of North Kordofan Province, Sudan. El Obeid,

Nomadic Health Project, Euro-Action Accord.

project is a joint venture between the FeinsteinInternational Centre at Tufts University, Savethe Children USA (SC US), and Save theChildren UK (SC UK) in Ethiopia. The MilkMatters project aims to improve the nutritionalstatus of children in pastoralist/semi pastoral-ist areas of Ethiopia through demonstrating anexplicit link between livestock health, milkavailability and access, and child nutrition14. Italso aims to contribute to improvements inpolicy and programming for child nutrition byproviding scientific evidence in support ofmore holistic and preventative approaches tomalnutrition in pastoral regions. This articlebriefly describes the findings from two cohortstudies conducted in two Zones in the Somaliregion of Ethiopia from July 2010 to July 2011.The studies were designed to assess the impactof small-scale livestock interventions to sustainaccess to and availability of animal milk at thehousehold level over the dry season on thenutritional status of children less than five yearsof age. The study objective, design, and find-

11 Arhem, K. (1985). Pastoral Man in the Garden of Eden:

The Maasai of the Ngorogoro Conservation Area, Tanzania.

Uppsala Research Reports in Cultural Anthropology.

Uppsala, Sweden, Uppsala University.12 Catley, A. (1999). The Herd Instinct: Children and

Livestock in the Horn of Africa. Save the Children Working

Paper 21. London, Save the Children UK.13 The goal of Save the Children’s Africa Region Pastoral

Initiative is to “deepen and replicate innovative approaches

to improve access to basic services and reduce vulnerabil-

ity to drought in pastoralist populations in order to create

positive change for children in this unique and harsh

environment.”14 Previous work under the Milk Matters project, described

extensively in the report, Milk Matters: The Role and Value

of Milk in the Diets of Somali Pastoralist Children in Liben

and Shinile, Ethiopia, laid the foundation for the results

discussed in this article and can be found at

http://sites.tufts.edu/feinstein/2009/milk-matters-phase-one

Dry riverbed in Shinile

Kate

Sadle

r/Tu

fts

Univ

ers

ity,

Eth

iopia

, 2010

Research

ings from one intervention site, aswell as discussion and recommenda-tions, are summarised below.

Study objectiveThe objective of this study was toevaluate the impact of community-defined livestock interventions onchild nutritional status during the dryseason. In doing so, the study askedthe following primary research ques-tion: What is the impact of livestock inter-ventions on children’s consumptionof animal milk and nutritional statusover one calendar year, particularlyduring the dry season?

In addition, the study sought tocompare the costs of an early interven-tion focusing on livestock health andmilk production with an emergencyfeeding programme for children.

Study designTwo cohort studies, each involvingthree sites, were implemented overone calendar year, from July 2010 toJuly 2011. In each study, two siteswere designated to receive livestockinterventions and one site served as acontrol. All children aged 6 to 59months in both intervention andcontrol sites were enrolled in themonthly surveillance system thatcollected data on child health andnutrition. Overall, the surveillancesystem aimed to follow 940 children,610 living in intervention sites and330 living in control sites.

Study site, household, and milkinganimal selectionSix sites were selected to participate in thestudy, three from two separate Zones in theSomali region. Site selection criteria included: 1.Pastoral livelihoods2. Targeted for assistance under SCUS

Protective Safety Net Programme (PSNP) and SCUK Revitalising Agriculture/ Pastoral Incomes and New Markets (RAIN) programme in Liben and Shinile respectively

3. Population size greater than 200 households15, and

4. Vulnerability to elevated rates of child mal- nutrition during the dry season, as indicatedthrough regional nutrition assessments16.

Site selection was completed in December 2009in consultation with local government officials,SCUS (Liben) and SCUK (Shinile), and localcommunities. As far as possible, sites with asimilar level of access to basic resources such aspasture, water, health care, and education serv-ices were selected. Selection of households wasalso done in consultation with local officialsand community members, with a focus oninclusion of all households with children underthe age of five years. Cows and small ruminantswere designated as the targeted livestockspecies for support under the intervention17. Inthe four designated intervention sites, house-holds were asked to choose either one milkingcow or three to four milking goats, a ratio basedon estimated milk off-take per species. Theparticular milking animals at each householdwere selected for support according to thefollowing set of criteria: (1) recently lactating,(2) with a normal milking yield and (3) in over-all good health with no problems that couldcompromise milk production.

The livestock interventionsTwo livestock interventions were designed. Intwo of the four designated intervention sites(Waruf in Shinile and Biyoley in Liben), themilking animals were given a daily ration ofsupplementary feed over the dry season. In the

other two intervention sites (Ayilisoin Shinile and Washaqabar in Liben),the milking animals were given adaily ration of supplementary feedplus a package of vaccinations andde-worming medications at theoutset of the dry season18. The tworemaining sites were designated ascontrol sites and received no inter-vention. While the dry season, Jilaal,typically falls between February andApril in the northern Shinile Zoneand between January and March inthe southern Liben Zone, the inter-ventions were extended in bothZones due to the onset of a drought in2010/1119. In total, the interventionsspanned 146 days in Shinile and 135days in Liben, but due to certain site-specific occurrences, the total days ofactual livestock feeding varied by siteand ranged from 47 to 73 days.

The supplementary feed rationsize for each animal species wasbased on the livestock feeding guide-lines for drought set by the Ethiopiangovernment20 and was meant tosupplement natural browse.However, the ration size was subse-quently raised as a result of thedrought (Table 1). Local agriculturalcooperatives with irrigation capacitywere contracted to grow and trans-port Sudan grass in Liben, while aprivate contractor was ultimatelycontracted to provide Rhodes grass inShinile after an unexpected frost

destroyed the Sudan grass crop. The care andshelter of the grass hay upon delivery to thesites was tasked to community members ineach intervention site. Both SCUS and SCUKprovided assistance in building the shelters forthe hay. Participating households were respon-sible for providing the milking animals withsufficient water.

For the two sites receiving the livestockhealth package, vaccinations were procuredthrough the Ethiopian Government Bureau ofLivestock, Crop and Rural Development, whileprivate vendors of veterinary drugs werecontracted to maintain a consistent supply ofthe designated prophylactic drugs throughoutthe dry season (see Table 2). The drugs werethen collected as needed by trained communityanimal health workers using the vouchersystem established by SCUS and SCUK in therespective regions.

9

Research

15 It was determined that a community with at least 200

households would ensure the desired sample size of 150

children/site.16 See footnote 9.17 The decision to support small ruminants and cows was

based on the results of previous work under Milk Matters

which found that households kept a cow or a few goats at

the household during a typical dry season while the men

took the larger stock far from the settlement site in search

of pasture and water.18 Animal feed was provided in all sites based on the fact that

animal health care alone would do little to improve milk

off-take if the animal remained malnourished, but that

extra health provisions in addition to feed may improve

milk off-take compared to animals receiving only feed. 19 The Somali Region is characterized by a bimodal rainfall

pattern with one long and one short rainy season, and one

long and one short dry season. Jilaal is the Somali term for

the long dry season.20 Ministry of Agriculture and Rural Development (2008).

National Guidelines for Livestock Relief Interventions in

Pastoralist Areas of Ethiopia. Addis Ababa, Ethiopia,

Ministry of Agriculture and Rural Development

Liben

Sudan grass

Shinile

Rhodes grass

ExtendedFeeding (Shinile only)Wheat bran

Adult cow 9 6 3

Adult goat 3 2 1

Calf 3 2

Kid 2 1

Table 1: Final feed rations (kg/day) for milking species

Care Type Cattle Goats

Vaccination Anthrax, blackleg, contagious bovine pleuro-pneumonia (CBPP), Lumpy Skin Disease (LSD)

Contagious caprine pleuro-pneumonia (CCPP),capri pox, peste des petites ruminants (PPR)

Prophylactictreatment

Ivermectin (internal and external parasites) Ivermectin (internal and external parasites)

Other curativetreatment

Treatment provided with diagnosis through-out the dry season

Treatment provided with diagnosis throughoutthe dry season

Table 2: Vaccinations and medications for milking species

Weighing a child in Biyoleyintervention site

Kate

Sadle

r/Tu

fts

Univ

ers

ity,

Eth

iopia

, 2010

intervention (Table 3). This increase was attrib-uted to project inputs. The primary non-projectfactor that could have resulted in increasedmilk off-take was rainfall leading to betterbrowse and water for livestock, but droughtconditions and lack of browse over the inter-vention period rule out this factor21. Asecondary factor, household purchase of feedfor livestock, could also have affected attribu-tion, but was similarly ruled out. A high level ofration was delivered by the study to studyanimals relative to recommended levels andfeed intakes. This suggests that very limited, ifany, feeding of study animals with privately-purchased feed is likely to have taken place.

Milk availabilityMore children received milk and children onaverage consumed greater quantities of milkper day throughout the intervention inWashaqabar than in Makinajab (Figure 1). FromDecember to May, the percent of childrenreceiving milk in Washaqabar increased by 31%compared to 20% change in Makinajab, and theaverage daily amount consumed increased by241ml compared to gradual decline inMakinajab.

Nutritional statusPre-intervention nutritional status was rela-tively stable in both the intervention and thecontrol site (Figure 2). However, with the startof the dry season in January 2011, the averageWAZ of children in the Makinajab control sitestarted to decline, with an overall decrease of0.31 points between January and May. Incontrast, the average z-score fell by only 0.07points from January to May in the Washaqabarintervention site.

DiscussionThe results of this study demonstrate thattargeted livestock support can significantlyincrease daily milk off-take during very dryperiods with direct links to improved milkconsumption among young children and posi-tive implications for nutritional status. To ourknowledge, this is the first study that presentsquantitative data demonstrating this link foryoung children in pastoralist areas. There wereseveral factors, however, that affected the rela-tionship between milk off-take, child milkconsumption, and child nutritional status,summarized in Box 2. These factors are impor-tant to consider as they have implications forboth the impact of the intervention seen here

and the design of similar inter-ventions in the future.

Where the interventionworked well and interventioncoverage of households washigh, such as in Washaqabar, theincrease in milk consumptionseen (1050 mL/day compared to650mL/day in the control site)translates into an additional264kcal, 12.8g of protein, andconsiderably higher intakes ofessential fatty acids, vitamins,and minerals per child each day.For a young child of two years ofage, this increase in nutrientintake would meet circa 26

10

Research

significantly greater during the 2011 dry season with the intervention as compared tothe 2010 dry season with no intervention.

• Given the drought conditions and lack of rainfall, the increase in milk off-take is attributed to the project interventions.

• By the end of the intervention, a greater proportion of children were consuming milk in the intervention sites as compared to the control sites.

• Those children who received any milk in the intervention sites consumed, on average,more milk than children in the control site.

• There was an overall trend towards stabilizednutritional status among young children in the intervention sites over the course of the intervention compared with a decline in nutritional status in the control sites.

• Within the intervention sites, those children who continued to consume some milk throughout the intervention time period maintained higher average WAZ scores than those who did not receive any milk, a difference that was frequently significant.

Case Study: Results from Washaqabar,Liben ZoneKey findings from Washaqabar, Liben Zone aresummarised in Box 1 and detailed below.

Milk off-takeMilk off-take increased significantly inWashaqabar, by as much as 4775%, from the2010 dry season to the 2011 dry season with the

21 Please refer to the original publication for

a detailed presentation of rainfall data

for the referenced time period provided

by the Ethiopian National Meteorology

Agency, Addis Ababa, Ethiopia.

Milk off-take Livestock milk off-take in Washaqabar was significantly greater during the dry season/drought in2011 with the intervention, compared to the previous year dry season in 2010 with no intervention.

Milk availability Milk was more available to young children in Washaqabar compared with the control site, with 94percent of children receiving milk compared to 56 percent in the control, and each child consum-ing on average 366 mL more milk per day than in the control.

NutritionalStatus

The nutritional status of children in Washaqabar remained relatively stable during the intervention,compared with declining nutritional status among children in the control site.

Box 1: Key Findings from Washaqabar

Table 3: Milk off-take in Washaqabar

LivestockType

Stage ofLactation

Mean Daily Milk Off-take (ml) PercentChangeDry season, 2010, no intervention (95% CI) Dry season,b 2011, with intervention (95% CI)

Goata Early 224 (190.5, 257.6) 628 (473.8, 782.9) 280c

Middle 54 (24.5, 84.2) 567 (428.3, 706.6) 1050c

Late 8 (0.0, 20.2) 382 (317.6, 446.6) 4775c

a Because a goat yields less milk than a cow, three goats were considered equivalent to one cow during the intervention. The aboveyield estimates are for one goat.b The 2011 dry season became a drought (see Table 11 of original publication).c Significant at the 95% confidence level.

Figure 1: Milk availability in Washaqabar vs. Makinajab(control)

1500

1300

1100

900

700

500

300

100

90

80

70

60

50

40

30

20

10

0

Mea

n D

aily

Milk

Con

sum

ptio

n (m

l)

Prop

ortio

n (%

) of C

hild

ren

Rece

ivin

g M

ilk

Aug Sept Oct Nov Dec Jan Feb Mar Apr May Month

Children Receiving Milk in WashaqabarChildren Receiving Milk in MakinajabChild Milk ConsumpMon in WashaqabarChild Milk ConsumpMon in Makinajab

Figure 1: Nutritional status of children in Washaqabar vs.Makinajab (control)

Mea

n W

AZ–

WH

O

Aug Sept Oct Nov Dec Jan Feb Mar Apr May

Month

VillageWashaqabarMakinajab

Intervention Start-0.50

-1.00

-1.50

-2.00

The surveillance systemThe surveillance system was designed to moni-tor selected anthropometric measurements,milk consumption patterns, and infection statusof enrolled children less than 5 years of age.Thirty-two data collectors were selected fromwithin the communities because of their uniqueability to track down and follow children asthey moved with their families. The data collec-tors were trained in May 2010 and wereresponsible for following between 25 and 30children each on a monthly basis. The datacollectors administered three questionnairesper child, designed to collect a range of healthand nutrition data, including age and weight,from which weight-for-age z-scores (WAZ)were derived, 24 hour recall of milk consump-tion (plain milk and milk in tea), and infectionstatus over the last two weeks.

Results The data were evaluated by intervention sitecompared with the control site in the sameZone and then analysed according to theimpact of the interventions on three key vari-ables: (1) animal milk off-take, (2) milkavailability – measured as a function of both theproportion of children in each site consumingmilk and the average amount consumed bythose children receiving milk, and (3) childnutritional status measured by WAZ. Insummary, the results seen across all four of theintervention sites were: • Milk off-take in the intervention sites was

Intervention Start

WAZ: Weight for age z score Error bars: 95% CI

percent of energy and 98 percent of proteinrequirements. Whilst we don’t see a dramaticimprovement in weight gain among interven-tion children, we do see an overall pattern ofstabilized WAZ among these childrencompared with a general decline in children incontrol sites over the intervention period.Together, these impacts indicate that interven-tions such as those tested here hold potential tomaintain weights of young children in times ofdrought and to prevent a deterioration to acutemalnutrition. It is well documented in thescientific literature that preventing acutemalnutrition is crucially important for a child’ssurvival and overall mental and physical devel-opment22. Moreover, the results of this studyalso suggest that it could be more cost-effective,with the cost of the livestock interventionstested here ranging from 34 to 81 USD per childto prevent weight loss versus the cost of145 to 200 USD per child to treat severeacute malnutrition in a CMAMprogramme.23,24

In addition to the impacts seen onmilk consumption and nutritionalstatus in young children, the interven-tions had several positive livelihoodoutcomes for participating households,including:• More free time for women as a result

of reduced workload – research on maternal and child health suggests that increasing maternal well-being frequently translates to improved infant and young child feeding practices (such as perceived ability to exclusively breastfeed) with important outcomes for child nutrition25,26,27.

• Protection of critical assets during drought conditions – many house-holds reported a high survival rate of dams and suckling calves and perceived improved rates of repro-duction in some of their animals as compared to previous dry seasons and periods of drought.

• Improvement in milk off-take for local Somali indigenous breeds – this result indicates the potential to enhance production in local breeds, which are well adapted to

and nutrition outcomes in these areas with thepotential of creating substantial aid cost savingsby preventing the need for large selective feedprogrammes such as CMAM.

Recommendations therefore include: • Applying a nutrition lens to the drought

management cycle by, for example, support-ing preservation of milk surplus during the rainy season, community level feed produc-tion/storage, feed interventions for repro-ductive/milking stock, and designing cash/food-for-work activities that do not nega-tively impact on women’s time and ability to maintain their own or their children’s nutrition status.

• Scaling up similar feed interventions as thatconducted in this project, but with adjust-ments to control costs from feed supply andtransport. This could be done through the provision of vouchers or other financial mechanisms where market supply for feed is adequate29.

• If private sector provision of feed is not adequate, investigating opportunities to support local livestock feed production30.

Finally, food security and livelihoodprogrammes must start monitoring moresystematically their impact on nutritionoutcomes. This need not involve frequentcollection of anthropometric data and the meas-urement of nutritional status itself, which, asdiscussed above, can be challenging in theseenvirons. But simple tools for measuring nutri-tion impact, such as participatory impactassessment31 and the dietary diversity index32,need to be used as standard if we are to createthe momentum for investment in food securityinterventions to prevent increases in rates ofmalnutrition where these have been shown tobe effective.

For more information, contact: Kate Sadler,email: kate@validinternational.org

11

Research

Milk off-take and consumption

Challenges ofimplementationdelivery

Feed procurement and delivery challenges in the two Zones included: unexpected frost late inthe growing season in Shinile that killed the Sudan grass crop and necessitated a new contractfor Rhodes grass that was located a greater distance from the study sites, high transportationcosts, occasional poor quality of feed delivered to study sites, preference of some participants tofeed animals from home instead of at designated feeding centres, long distances to feedingcentres that impacted milk supply and availability for young children, movement of a smallproportion of targeted cattle reduced access to feeding centres.

Independentpurchasing offeed

While this happened to varying extents in all sites, the provision of vouchers used for feedpurchase by an external agency to households in one control site affected the validity of thecontrol; this event also demonstrates the difficulty of implementing field research in settings noteasily controlled.

Drought condi-tions

Limited natural browse due to drought necessitated an increase in ration size. Water and pastureshortages forced families to move further than planned resulting in some loss to follow-up.

Sharing of milk In sites where household coverage by the interventions was lower, participating householdsreported sharing milk with non-project households.

Prioritization ofmilk

With less overall milk available, milk was prioritised to children under three years resulting in lessconsumption amongst the older children enrolled in the study.

Milk consumption and nutritional status

Infection status Several misunderstandings between mothers and data collectors led to widespread underreport-ing and underestimation of the prevalence of childhood infections.

Activity levels Data to reflect this was not collected or measured by the study.

Box 2. Factors affecting milk off-take, child milk consumption and nutritional status

A mother and child

Kate

Sadle

r/Tu

fts

Univ

ers

ity,

Eth

iopia

, 2010

22 Bhutta, Z. A. (2009). Addressing severe acute

malnutrition where it matters. Lancet 374: 94-96.23 Puett, C., K. Sadler, et al. (2012 ). Cost-effective-

ness of the community-based management of

severe acute malnutrition by community health

workers in southern Bangladesh. Health Policy

and Planning in press.24 Please refer to original document for more in-

depth presentation of the cost comparison.25 Shell-Duncan, B. and S. A. Yung (2004). The

maternal depletion transition in northern Kenya:

the effects of settlement, development and

disparity. Social Science and Medicine 58(12):

2485-2498.26 Nyaruhucha, C., J. Msuya, et al. (2006).

Nutritional status and feeding practices of under-

five children in Simanjiro District, Tanzania.

Tanzania Health Research Bulletin 8(3):162-167.27 See Footnote 728 Levine, S. and C. Chastre (2011).Nutrition and

food security response analysis in emergency

contexts. London, Humanitarian Policy Group.29 In the participatory impact assessments for this

project, households stated that they frequently

purchased animal feed from local sources as a

“normal strategy” during drought, a very positive

trend that could be supported.30 The role of aid projects in supporting local live-

stock feed production is not clear and this recom-

mendation is dependent on good analysis of

existing private sector production and apparent

growth of this activity, as well as on trends in

private enclosures and commercial fodder

production. Please refer to full report for additional

information: http://sites.tufts.edu/feinstein/2009/

milk-matters-phase-one 31 Catley, A., J. Burns, et al. (2008). Participatory

Impact Assessment: A guide for practitioners.

Medford, MA, Feinstein International Centre.32 Drescher, L. S., S. Thiele, et al. (2007). A New

Index to Measure Healthy Food Diversity Better

Reflects a Healthy Diet Than Traditional Measures.

Journal of Nutrition 137:647-651.

harsh environmental conditions and diseaserisks in the pastoralist areas targeted here.

Conclusions and recommendations This study has demonstrated that throughtargeted livestock support to milking animalsthat stay close to women and children duringdry season and/or drought (overall a relativelysmall proportion of the whole herd), milkproduction and consumption among children isimproved, and their nutritional status benefits.There is some consensus in the programmingliterature at present that the humanitariancommunity tends to spend much more timebefore humanitarian disasters preparing totreat acute malnutrition rather than trying toprevent it28. The interventions presented hereprovide us with the opportunity to change thisfocus and reconnect food security interventions

12

Research

Climate Change as a driver ofhumanitariancrisis andresponse

Tufts University has recently published apaper which explores the relationshipsbetween climate change, humanitarian

crises and humanitarian response through areview of published and grey literature.According to the authors and based on thedatabase of disasters worldwide maintained bythe Centre for Research on the Epidemiology ofDisasters (CRED) at the Université Catholiquede Louvain, Belgium over the past 11 years,climate-related disasters have been killing anaverage of 33,520 people a year, and, as criti-cally, affecting the lives of over 211 millionpeople. Seven categories of disasters on whichdata are gathered can be attributed directly tometeorological phenomena and thus directlyaffected by climate change: drought, extremetemperature, floods, mass movement becauseof drought, mass movement because of flood-ing, storms, and wildfires.

In a detailed statistical analysis of the deathtoll from recent natural disasters between 1980and 2002, it has been shown that there is noevidence that rich nations experience fewerdisaster-events per se; rather, the events haveless impact. It has been shown that less demo-cratic nations and nations with larger incomeinequalities suffer proportionally larger deathtolls from disaster.

However, geography is still important. Anation in Asia is 28.5 percentage points morelikely to experience a disaster in a given yearthan one in Africa. The UNISDR (UN Office forDisaster Risk Reduction) in its 2011 reviewpoints out that the populations of people at riskfrom weather-related disasters, and particularlytropical cyclones, has almost tripled since the1970s. This is because the number of peopleliving in vulnerable coastal cities has increased,with most of this increase being in low-income,shanty-town like developments.

Furthermore, it appears that the intensity,and cost of natural disasters are increasing, andthe twenty-first century holds the possibility formuch greater levels of destruction than previ-ously experienced. To date, increases in thenatural disaster burden can be attributed, atleast in part, to development forces, includingpopulation growth, endemic socio-politicalinequities and the failure of governancesystems to avoid human settlement on danger-ous terrain. However, in its Fourth Assessment

Report in 2007, the International Panel onClimate Change (IPCC) found that naturalhazards themselves were likely to increase infrequency and intensity during the course ofthe century. The report concluded that rain-fedagricultural yields in Africa could drop by up to50% by 2020, and that as many as 1.25 billionpeople in Africa and Asia could be exposed towater shortages and stress by 2050.

The International Food Policy ResearchInstitute (IFPRI) projected that climate changewill in fact cause a decline in the availability ofcalories per capita in developing countries, rela-tive to 2000 levels. The result will be a 20%increase in the number of malnourished chil-dren, relative to a world without climatechange.

A more recent 2012 IPCC report on climateextremes found that several hazards – with thepotential to spawn humanitarian crises – werelikely to emerge in the twenty-first century –more intense cyclones, food and water short-ages, major flooding, droughts, degradedecosystem goods and services, and changes inthe frequency and patterns of disease. Thisincreasing future risk is partially due toprospects for an expanded range for diseasevectors and partially due to the impact thatclimate change is expected to have on foodproduction and on flooding – malnutrition andflood events being aggravating factors in thespread of infectious disease. Since least-devel-oped countries, by definition, suffer fromdeficiencies in food, water, sanitation, andhealth care, these excess stresses are likely toplace disproportionate harm in those places.Potential effects of increased temperature maybe further intensified by the demographics ofaffected populations; a growing proportion ofelderly people in many countries will besusceptible to heat waves.

The humanitarian implications of suchimpacts are clear, yet the true costs are not. Atleast one organisation, however, suggests theprice of aid missions might jump by anywherefrom 32% to 1600% due to climate change.

A landmark report from FORESIGHT in 2011found that migration was likely to be a majorfeature of human societies in the comingdecades. However, the authors argued that, forunderlying social reasons, this movement was

equally likely to be towards areas of environ-mental stress as away from such areas—as in theriver deltas and mega-cities of coastal Asia.Meanwhile, the authors distinguished betweenmigration (which may have a positive role inrisk-reduction) and displacement (which is likelyto be universally negative). Preventing migra-tion out of stressed areas may lead to anincrease in displacement.

A second school of thought focuses on thedislocation related to natural events. Climatechange will create more environmental pres-sures and natural disasters that displace peoplefrom traditional homelands, particularly inareas of extreme exposure, such as coral atolls.These mass migration events will cause humansecurity challenges and may serve as the basisfor humanitarian crises.

The number of people on the move due toclimate change is deeply contested. Commonlycited figures range from 50 million to 1 billionpeople by 2050 (UNFPA, 2009). Critics arguethat such numbers are “deterministic,” failingto account for human agency and strategies inthe face of climate change.

Moreover, major policy biases against migra-tion (especially in its international forms) mayimpede movement for many affected people.Migration calculations are also subject to uncer-tainties about the intensity of climate change.Mild climate change, a less than two degreesCelsius rise in pre-industrial levels, would yieldmigration flows “virtually indistinguishable”from existing patterns of migration.Meanwhile, moderate levels of climate change,two to four degrees Celsius, would lead to moremigration, especially displacement, projected at250 million people. Catastrophic global warm-ing, above four degrees Celsius, could lead toenvironmental destruction and social disloca-tion displacing untold numbers of people.

There is also the issue of government-spon-sored migration in the face of climate change.Directly or indirectly, policy decisions not onlyrespond to environmental migration but alsocontribute to the actual flows of people, espe-cially within national borders. Government

1 Walker.P et al (2012). Climate Change as a Driver of

Humanitarian Crisis and Response. Feinstein International

Centre, Tufts University. June 2012.

Summary of published research1

Thousands of Somalis havebeen displaced by what isdescribed as the worst floodsin the country in 10 years.

©M

anooch

er

Deghati/I

RIN

13

development policies may influence whetherpeople build and settle on ecologically sensitiveareas, such as flood plains and hillsides.Governments may resettle populations fromareas of perceived natural hazard, or in thewake of a natural disaster. Alternatively,governments may take climate-related actions,especially water or food security projects, suchas hydroelectric dams, that either attracts work-ers to an area, displace local populations, orboth. These adaptive actions could reducecertain vulnerabilities, but may create others ofconcern to humanitarians.

Many of the world’s large and fast growingcities are located at lower than ten metres abovesea level along coastlines (so-called ‘low-eleva-tion coastal zones’, or LECZs, which aresusceptible to all manner of ‘seaward threats’).

Rural-to-urban migration has been (andremains) a key strategy for resilience in manydeveloping country contexts, where rural liveli-hoods are particularly susceptible to climatevariability. However, the nature of rapid urban-ization is critical. With the number of slumdwellers set to rise from one billion to twobillion in the coming decades, and “urbanisa-tion [becoming] synonymous with slumformation”, the potential for urban humanitar-ian crisis is rapidly expanding. The low-costand informal nature of slum settlements alsomeans that they are located on highly marginalland, such as flood plains and steep slopes.

Many slums are highly vulnerable to flood-ing. Data from Dhaka, Bangladesh, show thatnearly two-thirds of the country’s urban slumsflood once or more per year. Similarly, a studyin Gorakhpur, India found that parts of the citywere water-logged for five to six months out ofthe year, due to waste management anddrainage problems. This reconfiguration ofwhere and how people live must be accountedfor in any humanitarian planning aroundclimate change.

Climate change is itself a product of globali-sation; in many ways it is the world’s first fullyglobalised environmental risk. Climate emis-sions cause atmospheric warming, no matterwhere on earth they are produced. Indeed, thepoor people in poor countries who will be mostburdened by climate-related disasters areamong the world’s smallest emitters of green-house gasses.

Globalisation also changes the profile ofpopulations vulnerable to climate change. Forexample, food systems are more integratedtoday than ever before in human history.Disruptions in one part of the world can thusaffect accessibility thousands of miles away.Food shortages in 2007–2008 triggered foodriots in more than 30 countries. Meanwhile, theRussian forest fires of 2010 - fueled by abnormalheat and drought - led to that country imposingan export ban on grain. Globalisation is alsocreating incentives for states to establish indus-trial production zones at close proximity to thesea, for easier access to intercontinental ship-ping. However, these developments are often atvery low elevation, jeopardising the people andinvestment in property should sea levels rise inthe coming century.

Another hallmark of globalisation has beenthe recent trend of Asian countries, such asChina, investing heavily in trade with Africa

nations, in order to obtain sufficient food andraw materials for their large and industrialisingpopulations. Recently, this investment hasextended to the purchase of large tracts of landin the region, for the purposes of growing food.In doing so, the supply of arable land in theregion is facing a squeeze, just as climatechange will begin to accelerate drought anddesertification processes in a region where agri-cultural land is already under heavy pressure.The implications for land and food security arepotentially alarming.

Conversely, globalisation may impact thehumanitarian response to climate change inmore constructive ways. For example, the IPCChas noted that humanitarian response is oftenrequired as the result of a failure in disaster riskreduction (DRR). By expanding global flows ofcapital, products, and know-how, globalisationmay improve DRR capacity in vulnerablelocales, mitigating the need for humanitarianresponse. Technological innovations, rangingfrom early warning systems in Bangladesh todisaster-resilient schools in Thailand can bedeveloped, shared via the Internet, andaccessed around the globe. Moreover, the glob-alised development and dispersion ofcommunity-based participatory methods maybolster DRR and reduce future vulnerabilitiesto climate change.

State-mediated vulnerability is also a prob-lem. A good example of this comes from themajor waterways in South and Southeast Asiathat run through China. Chinese policymakershave reason to be concerned about future waterand energy scarcities. They have taken adaptiveaction such as the hydroelectric damming ofrivers and the routing of water from southern tonorthern China, where it can be put to use inthe country’s traditional agricultural belt.However, these projects may create watershortages for countries downstream, elevatingthe potential for scarcities and reduced agricul-tural yields, and triggering human securitycrises. Moreover, these projects have ratchetedup political tensions between China and itsneighbours, further raising the humanitarianstakes.

Four major challenges exist with respect tostate response to climate change-related disaster:• Prevention, mitigation, and adaptation

policies are not in place. • Capacity to implement policies is weak. On

a larger scale, while 168 countries subscribe to the landmark Hyogo Framework for Action (HFA) that guides disaster response planning, not all participants are fully able to meet its goals.

• Systematic data collection to evaluate and report response results does not exist

• Lack of data to evaluate and inform the nature of the DRR efforts.

Although widespread support of internationaldisaster guidance signals a shift in support ofstandardised methods of managing disasterresponse, the gap between policy and actionremains a major hurdle. One basic issue thatcontributes to this dilemma lies in the tenure ofpolitical representatives versus the need forlong-term disaster planning. While a politicianmight be in office for two years, comprehensivedecision making requires a much longer timehorizon and commitment that is not subject tothe ephemeral oscillations of election cycles

Agencies typically are very good at measur-ing and tracking the inputs to theirprogrammes (the finance, personnel, andsupplies) and the processes these inputs feed(logistics systems, the supply of water, health-care, food, and so on). However, they becomeprogressively poorer at measuring and moni-toring as they move downstream to programmeoutputs and outcomes and, at the end of theline, hardly ever measure or evaluate impact. Ifhumanitarian aid programmes are to becomeconcerned with enhancing system resilience aswell as basic survival, then they will have toboth adopt the methodologies of impact assess-ment and create the financial and managementsupport needed to ensure such evaluationsbecome routine.

The evidence also suggests that humanitar-ian operations are no longer synonymous withemergency operations. Most humanitarianassistance today goes into operations that havebeen running for five years or more. As much as45% goes into programmes more than eightyears old. Whilst this no doubt saves lives, italso condemns the victims to an endless state ofpurgatory, beholden to others, the agencies orthe state, for their survival. Building resilienceis not part of this mind-set. In these long-termcrisis environments, in Ethiopia, Sudan,Afghanistan, Palestine – all environmentallyfragile states – a major opportunity is beingmissed to use aid to transform the way commu-nities and their states develop the necessaryeconomies and governance for the future.

The humanitarian aid system evolved as aWestern based interventionist endeavour,seeing crises as abnormal and respondingthrough exceptional interventions. If we face afuture where crises are more pervasive, manymore states will have to repeatedly meet theneeds of their crisis affected populations.Increasingly, Southern states- Indonesia,Philippines, Mozambique, for example - arereforming their own disaster response systems,seeing this as a normal part of sovereignresponsibility.

External aid agencies need to adapt to andsupport this change. The old methods of work-ing around government systems, rather thanwith them, have to be challenged. In manycrisis-affected states, aid agencies need to seethemselves as long-term partners of the state,providing response services, but must alsowork to build resilience into livelihood systemsand the infrastructure of hazard-exposed popu-lations. They need to view recovery from crisisas a process of change to a more resilient state.Such change will not be easy. The humanitarianresponse sections of aid agencies have tendedto see their work in terms of logistics and theimpartial, neutral supply of live-saving aid andhave shunned much of the political analysis ofthe development sector, let alone developed ananalysis of complex global processes.

International humanitarian aid agencieshave grown to become large, multinationalorganisations, turning over billions of dollarseach year and playing a critical in the creationof international civil society norms. They nowresemble major transnational corporations andfind themselves increasingly challenged by therisk aversion and inertia that comes with scaleand an operational model that is still essentiallyabout organisational control.

Research

14

This research dissertation reviewedsome of the obstacles to treatment ofadult undernutrition in adults – partic-ularly in a humanitarian context,

where inclusion of adults in nutritionalresponse programmes remains limited. It isimportant to identify what are the commonobstacles experienced by humanitarian healthprofessionals around the provision of care foradults with undernutrition during emergencyresponse programmes.

This issue is not new. In 1999, Salama andCollins reviewed the humanitarian response inBahr a Ghazal in South Sudan (1999) whereadult undernutrition had presented itself as alarge challenge2. The highlighted key issuesunderpinning why international non-govern-mental organisations (INGOs) were notconsidering adults for nutritional treatmentprojects:• Diagnosis and treatment of malnourished

adolescents and adults still in its infancy with a weak evidence base and limited guidance.

• Inexperience at both individual and agency levels to manage adult malnutri-tion, particularly relevant given the lack of guidance.

• Complexity of adult feeding programme design where health complications and food choice are just two issues that make interventions less clear cut than with young children.

• Media and public relations, where images of ‘starving children’ attract more attention(=resources) than starving adults.

The review concluded that despite gaps inguidance and capacity, it is certain that infamine situations, adult malnutrition is a public

health problem and that successful adult treat-ment is possible and is essential for the survivalof individuals and communities. The SouthSudan response in 1999 failed in this regard.

MethodologyFor this latest review, reports and documentswere collected using the words ‘adult’ ‘malnu-trition’ and ‘humanitarian’ from a variety ofsources and websites. Literature in both Englishand French were selected. The inclusion criteriawere literature reviewing nutritional responsesin the context of a humanitarian emergency andrelating to adolescents and adults includingthose older than 65 years of age. Given the widerange of types of papers and methods ofresearch involved, the Critical Appraisal SkillsProgramme (CASP) tool (2006)3 was selectedand adapted to screen the literature.

A key limitation was that much of the litera-ture existed as grey literature, e.g. on agencywebsites and accessed through internetsearches. There were very few peer reviewpublications. Also, the majority of reports onundernutrition focus solely on children underthe age of 5 years. Occasionally, reports includea token reference to adults in relation to theimpact childhood chronic undernutrition haslater in life or on women in relation to infantnutritional status (PLW).

ResultsA total number of 2,280 reports were identifiedand screened. Of these 2,280 reports, approxi-mately 107 were eligible for the final detailedscreening. Of the 107 reports, 105 were writtenin English and only 2 (1.5%) in French.

All 107 documents were screened a secondtime using the detailed questions on the CASPtool, narrowing down the final selection to 38

Summary of MMedSci research1

By Claire Bader

Since graduating as a Registered Nurse in 1995, Claire has spent over 12 yearsworking in a variety of health and nutritional programmes throughout Africa. Sheis an appointed member of the Isle of Man Overseas Aid Committee and hasrecently joined the Save the Children team in Sierra Leone as their Health Advisor.

Claire would like to acknowledge her dissertation supervisor, Dr Tony Blackett, for his supportthrough the process and also to Dr Peter King for inspiring her to share my work.

reports all published or released betweenJanuary 2000 and February 2011. These final 38reports met the proposed selection criteria forthe review, all involved an actual type ofresearch methodology and showed a higherquality of robust research methods or reflectionon programmatic practice.

Fourteen (37%) of the selected reportsfocused on reviews of emergency responseprogrammes and 24 (63%) were nutritionalreviews or focused on nutrition research.Analysis of the 38 selected reports revealed that19 (47%) were written by individual authors orteams of researchers, 13 (37%) by INGOs and 6(16%) by international organisations (UN,donors and institutions). Reports reflectedprogramming/research in Kenya, Sudan,Ethiopia, Niger, Chad, Malawi and Guinea,Haiti and Myanmar.

The author assumed that most of the obsta-cles noted by Salama and Collins in 1999 wouldemerge in this review. However, on screeningthe 38 documents, the findings were surpris-ingly different. The main obstacles identified inthis review were 'weak coordination”, “limitedresources”, “poorly integrated response” and“weak human capacity”. These obstaclescombined to a total of 72.5% of all the pointsnoted. “Weak coordination” was noted mostfrequently (23.5%) and “weak human capacity”was rated 4th (14%). Other obstacles noted werecontinuing lack of standardised adult treatmentguidelines and diagnostic criteria, lack of focusand interest in adults with overall prioritygiven to children under 5 years of age, andadults regarded as a national governmentresponsibility, not one for INGOs. Each of theseobstacles scored 7% or less.

DiscussionPrior to this systematic review, it was presumedby the author that the lack of recognised inter-national adult treatment guidelines orassessment tools that was directly affecting theimplementation of nutritional care for this agegroup. However the results suggest that guide-lines and tools are widely available andaccessible, but it is the training and subsequentability of health care workers to regularly usethem that is not adequate. Signs of poor nutri-tion in adults are easily confused with thesymptoms and presentations of chronic illnessand old age, as such it is often felt that theseadults should be cared for within the nationalhealth system rather than as part of emergencyresponse facilities. However, national nutritiondepartments remain under funded and understaffed in many developing countries and mayhave challenges coping with the extra burden inan emergency context.

Identification of limited resources as anobstacle relates not only to the limitation offunding to enable the inclusion of adults, butalso to the type and quantity of nutritionalresources available. Access to treatment prod-ucts for adult undernutrition itself will remain a

A systematic review of obstacles totreatment of adult undernutrition

Older people, not onlychildren, need support

Help

Age I

nte

rnational, P

akis

tan,

2011

1 Bader C (2011). Obstacles to treatment of adult undernu-

trition: A systematic review submitted (June 2011) as part

requirement for the MMedSci in Advanced Nursing Studies

from Sheffield University. Copies of the full dissertation are

available on request. 2 Salama, P and Collins, S. (1999). An Ongoing Omission:

Adolescent and Adult Malnutrition in Famine Situations.

Field Exchange. ENN. (Issue 6). p19. [Online] Available

from: http://fex.ennonline.net/6/ongoing.aspx [Accessed

30th March, 2010].3 http://www.casp-uk.net/

Research

15

Research

challenge until funding opportunities areincreased and specific nutritional productsfor adult undernutrition are developed.

The review suggests that there is still alack of information and understandingaround the scale, depth and causes of adultundernutrition and that this is still notbeing addressed through early warningsystems and nutritional surveys. A numberof reports advocate for greater funding toenable both the wider collection of infor-mation and the implementation of essentialpackages and community safety nets.

Thirteen out of the 38 reports reviewedidentified weak coordination as an obsta-cle, with many raising issues aroundcluster coordination and sharing of infor-mation between cluster partners.

The obstacles highlighted above areinterlinked and impact on each other, e.g.without effective coordination, sharing andmaximising of resources and human capac-ity development may not be needs basedand may be inequitable.

RecommendationsThere are five main recommendations thatemerge from this review:

It is important that clarity on the inclu-sion of adults within Nutrition Clustercoordination is reached and agreed, lead-ing to strengthened coordination aroundprogramming and information sharing.

Increased flexibility of funding forINGOs will enable them to implement abroader integrated response package toinclude vulnerable adults and interven-tions that address the dietary copingmechanisms that often lead to micronutri-ent deficiencies in women, teenagers andolder adults in particular.

Greater efforts and resources are neededto help develop an adult version ready touse therapeutic food (RUTF) that wouldenable treatment regimens to be adultspecific. Under the present recognisedtreatment guidelines, the volume of RUTFor fortified milk required by adults to beconsumed per day is large and often leadsto non-compliance.

A key priority is to enhance capacitybuilding programme in treatment andprevention of undernutrition for adultsand the elderly. It is important to build theconfidence of health workers in differenti-ating symptoms of acute weight loss ormicronutrient deficiencies from otherchronic health problems or general effectsof old age.

Finally, as the majority of nutritionalassessments focus on the under 5 agegroup, it is often difficult to estimate thenutritional challenges faced by oldermembers of the communities. Until moresystematic collection and analysis of dataon teenagers, adults and older people isundertaken, an accurate overall picture ofthe spectrum of hidden vulnerability willcontinue to elude humanitarian organisa-tions and donors.

For more information, contact: ClaireBader, email: clairebader@gmail.com

Summary of study1

Humanitarian financingfor older people andpeople with disabilities

"I can now go out in the sun"– words of Ghulam Mohammad(60 years old) who received awheelchair from HelpAgeInternational and who alsoprovide a monthly cash grantto his family.

Help

Age I

nte

rnational, P

akis

tan,

2011

Arecent study quantifies the fundingprovided by donors to meet thehumanitarian needs of two of the most

vulnerable groups: older people and peoplewith disabilities. It does so by analysing theamount of humanitarian funding targeted atthese two groups through the UN ConsolidatedAppeals Process (CAP) and Flash Appeals in2010 and 2011. A similar study was conductedby HelpAge International on the humanitarianfunding provided for older people from 2007 to2010. That study found a significant disparitybetween the needs of older people as a vulner-able group and the humanitarian assistancedelivered to meet those needs. Of the total CAPand Flash Appeal funds analysed in 2007 to2010 in selected countries, just 0.2% was allo-cated to projects that included an activityspecifically targeted at older people. Thecurrent research and the previous HelpAgestudy are not directly comparable as theylooked at different countries.

Demographic change means that thenumber of older people affected by crises anddisasters is growing fast. By 2050, the numberof people aged 60 and over will almost triple,reaching 2 billion (22% of the world’s popula-tion). More than 1 billion people in the worldlive with some form of disability (15.3%), ofwhich 200 million experience considerabledifficulties carrying out daily tasks. Many ofthose categorised as having a disability are alsoolder people. Rates of disability are also muchhigher among the 80-plus age group. Whileolder people and people with disabilities makean invaluable contribution to society, they areparticularly vulnerable to external shocks andupheaval in their daily environment.

This research study examined the CAP for14 countries and four Flash Appeals that tookplace in 2010 and 2011. All the projects(totalling 6,003) submitted to these appealswere examined, the majority (5,330) under theCAP. The primary tool for collecting data wasthe Financial Tracking Service (FTS) managedby the UN Office for the Coordination of

Humanitarian Affairs (UNOCHA). The FTSproject sheets were analysed to identify projectsthat targeted older people, or people withdisabilities, or both. The FTS captures all infor-mation on projects in the CAP, including anyactivities targeting specific groups, and donorfunding contributed to the CAP. However,reporting on whether specific projects werefunded is done on a voluntary basis either bythe donor, the aid recipient, or both. It is recog-nised that donors’ funding is not limited totheir contributions to the CAP and FlashAppeals, and in this sense, the research doesnot provide a complete picture. However, thestudy is considered to be a sufficient proxyindicator of the levels of official funding allo-cated to older people and people withdisabilities. CAP and Flash Appeal documentsare used as planning tools for donor support;they are approved by the Resident orHumanitarian Coordinator and serve as thebasis for funding applications to the UNCentral Emergency Response Fund (CERF),country-level pooled funds (such as theCommon Humanitarian Fund), and otherdonors.

The study is based on CAP data down-loaded from the OCHA FTS from 9–16November and 12–13 December 2011. Thedonor-specific information was downloadedfrom the OCHA-FTS on 5 January 2012.

Key findingsIn 2010 and 2011, 145 (2.4%) of the 6,003 proj-ects submitted to the CAP and Flash Appealsincluded at least one activity targeting olderpeople or people with disabilities, and 61 ofthese were funded (1%). Of these 145 projects,most of the other activities focused on generalassistance and/or other vulnerable groups.There was an increase between 2010 and 2011in the overall number of projects submitted,and there was also an increase in the number of

1 HelpAge International and Handicap International (2012).

A study of humanitarian financing for older people and

people with disabilities, 2010–2011. Published by HelpAge

International, London and Handicap International, Lyon

projects targeting older people or people withdisabilities. However, few agencies submittedprojects with an activity targeting older peopleor people with disabilities in both years.

In 2010 and 2011, 47 projects (0.78%)included at least one activity targeting olderpeople, and 18 of these were funded (0.3%). Inabout half of these projects (21), the targeting ofolder people accounted for less than 25% oftotal project activities. In both years, 22 agenciessubmitted projects that included at least oneactivity targeting older people. Only threesubmitted projects in both years. Twenty-two ofthe projects for older people in 2010 and 2011(46%) were submitted by one NGO (HelpAgeInternational).

In 2010 and 2011, 98 projects (1.6% of submit-ted projects) included at least one activitytargeting people with disabilities, and 43 ofthese were funded (0.7%). In both years, 50agencies submitted projects including at leastone activity targeting people with disabilities.Only 10 agencies put forward projects in bothyears. Among the 98 projects submitted in bothyears, 29 exclusively targeted people withdisabilities, of which 18 were submitted by oneNGO (Handicap International).

Only 19 projects (0.3%) included one ormore activities targeting both older people andpeople with disabilities. Of these, four werefunded. Four of the 19 projects were submittedby Handicap International and focused exclu-sively on older people and people withdisabilities. The other 15 projects addressed anumber of vulnerable groups in addition toolder people and people with disabilities, suchas women and children. On average, 9% of thetotal activities in these projects were targeted atmeeting the needs of older people and peoplewith disabilities.

In 2010 and 2011, projects including at leastone activity targeted at older people or peoplewith disabilities were submitted in 15 of theappeal countries (94%). Projects with activitiestargeting older people or people with disabili-ties were put forward in 11 out of the 12Inter-Agency Standing Committee (IASC)sectors (there were none in the security sector).While this appears to be a strong spread of proj-ects, the figure masks substantial differencesbetween the two groups in terms of their repre-sentation in projects within sectors andcountries.

Most projects were in threesectors: health, protection, andshelter and non-food items(NFIs). There were only fourprojects that targeted an activityfor older people in theeconomic recovery sector, andonly one of those was funded.In 2010, projects including atleast one activity targeted atolder people were put forwardin appeals in eight countries.This rose to nine in 2011. In 21countries, there were no proj-ects with activities targetingolder people in any sector, in2010 or 2011: Chad, CentralAfrican Republic, the Republicof South Sudan, West Africa(comprising 16 countries),Yemen and Zimbabwe.

In 2010, projects that included at least oneactivity targeting people with disabilities weresubmitted in 10 of the 12 IASC-recognisedsectors (projects in eight sectors were funded).This fell to nine sectors in 2011 (projects inseven sectors were funded). Most projects werein three sectors: health, water and sanitation,and education. There were no projects in thesecurity sector in either year. In 2010, projectsincluding at least one activity targeted at peoplewith disabilities were submitted in appeals in10 countries, nine of which received funding.This rose to 12 in 2011, 10 of which werefunded. In only one country, Yemen, were thereno projects that included any activities target-ing people with disabilities in any sector, in2010 or 2011.

In 2010 and 2011, US$10.9 billion wascontributed by official donors to the CAP andFlash Appeals. A total of US$73 million wasallocated to projects that included at least oneactivity targeted at older people or people withdisabilities (0.7% of overall funding). A total ofUS$27.6 million went to projects targeted exclu-sively at older people or people with disabilities(0.3%).

In 2010, US$2.6 million was allocated toprojects that included at least one activitytargeting older people (0.04 % of all funding).This increased to US$6.7 million (0.13 %) in2011. Most of this increase is accounted for by asmall number of large projects rather than amore consistent coverage across humanitarianresponses. In 2010, US$40.6 million was allo-cated to projects that included at least oneactivity targeting people with disabilities (0.7 %of all funding). This fell to US$22.3 million (0.43%) in 2011.

In 2010 and 2011, a total of 235 bilateral andmultilateral donors contributed funds to theCAP and Flash Appeals. Just over 60 % of all thefunding was provided by 10 donor countries.More donors funded projects that included atleast one activity for people with disabilitiesthan projects that included at least one activityfor older people. In 2010 and 2011, seven donorsprovided CAP or Flash Appeal funding forprojects that included at least one activitytargeting older people. The number rose fromthree in 2010 to five in 2011. Only one donor, theEuropean Commission’s Humanitarian Aiddepartment (ECHO), funded such projects inboth years. Two of the ten biggest donors to

CAP and Flash Appeals provided no CAP orFlash funding for projects that included activi-ties targeting older people (USA and UK).

The number of donors that provided fund-ing to projects that included at least one activitytargeting people with disabilities decreased,from 18 in 2010 to 14 in 2011. Thirteen donorcountries allocated funds to such projects inboth years. All of the 10 biggest donors to CAPand Flash Appeals allocated some funding toprojects that included activities targetingpeople with disabilities. Of the three biggestdonors, funding from the USA decreased from2010 to 2011, while funding from ECHO and theUK remained constant.

Conclusions and recommendationsThe total amount of projects and funding forboth the elderly and people with disabilitiesremains extremely low, highlighting the signifi-cant disparity between the needs of these twovulnerable groups and the humanitarian assis-tance delivered to meet those needs. A highproportion of the projects submitted to supportolder people and people with disabilities werepresented by two specialist NGOs, HelpAgeInternational and Handicap International (28%). Many of the other projects submitted (62 %)are in countries where these two NGOs arepresent and advocating for better inclusion ofthese two vulnerable groups in humanitarianresponse.

Only 312 of the 6,003 projects analysed(5.2%) mentioned older people and people withdisabilities alongside other vulnerable groups.In other words, thousands of projects made nomention of the potential vulnerabilities of olderpeople and people with disabilities, and howthe crisis affected them. Clearly, there is still along way to go to ensure that the humanitariansystem responds to the needs of older peopleand people with disabilities.

If the humanitarian community is to fulfil itscommitment to the impartial provision ofhumanitarian assistance to those in greatestneed, it must take urgent steps to address theneeds of two of the most vulnerable groups:older people and people with disabilities

Humanitarian agencies must ensure thatneeds assessments provide accurate informa-tion on all vulnerable groups by collecting dataon older people and people with disabilities,and disaggregating the data by age and gender.Greater efforts should be made to document

and share examples of good prac-tice on inclusion of vulnerablegroups so that these can beconsistently applied during proj-ect design and implementation.Cluster Lead Agencies,UNOCHA and HumanitarianCoordinators must provide lead-ership on this issue to ensureadequate accountability to allbeneficiary populations andconsistency across countries andsectors. Bilateral and multilateraldonors have an essential role toplay in encouraging and enablingan appropriate and inclusivehumanitarian response, byproviding flexible, timely fund-ing that is allocated in proportionto need and on the basis of thor-ough needs assessment.

16

Research

Safo Bonaya Boru, Galma Roba Village, Ethiopia

Help

Age I

nte

rnational, 2

011

17

Research

Help

Age I

nte

rnational, P

akis

tan,

2011

Arecent report shines the spotlight on theexperience of displacement for olderpeople in an effort to increase under-

standing of its particular characteristics.Entitled, The Neglected Generation, it drawsprimarily on contexts of internal displacementrather than refugee contexts. Nevertheless, thefindings and policy recommendations areapplicable to both refugees and IDPs.

BackgroundThe world is ageing at a staggering andunprecedented rate. By 2012, 12.5 per cent ofthe world’s population was over 60 – the UNdefinition of an older person. By 2050 there willbe more people over 60 years than children,including significant numbers of people over 80years, who constitute the fastest-growing agegroup. This translates into an average annualglobal increase of 29 million older peoplebetween 2010 and 2050, of whom 80 per centwill be in developing countries. Women willcontinue to live longer than men.

At the same time the number of internallydisplaced persons (IDPs) is increasing. At theend of 2011, the global number of people inter-nally displaced by armed conflict, generalisedviolence or human rights violations stood at26.4 million. Older persons form a significantproportion of groups of IDPs and refugees, as35-65 per cent of them may be over 60 years.

A recent UN paper on the social situation,wellbeing and rights of older persons world-wide states: “it should be noted that whilemuch data and analysis are available on popu-lation ageing, data and information about thelives and situation of older persons are strik-ingly lacking”. This lack of understanding andanalysis of the concerns and rights of olderpersons is prevalent in all areas of develop-ment, but it is especially stark in humanitariancrises. To date, the attention of internationalinstitutions, national governments and thoseresponding to displacement crises has beenfocused almost exclusively on children, ratherthan on supporting both of societies’ mostdependent age groups.

The exclusion of older people in situations ofdisplacement begins with registration to access

assistance, assessments and monitoringsystems. Data collection is often inadequate andthe numbers of older persons in IDP camps,where data are more likely to be available thanfor IDPs living in host communities, oftenremains unknown even if data are collected forother age groups within the population.

When population data are disaggregated bysex and age, disaggregation often stops at age49, reflecting a form of latent discrimination.Nutrition surveys commonly focus exclusivelyon children under five and HIV data usuallystops at age 45, when reproductivity ceases.Even where age-disaggregated data are gath-ered, information on older people will notnecessarily be captured.

Of the 50 countries reviewed by the InternalDisplacement Monitoring Centre (IDMC) for itsglobal IDP survey, only 11 had updated sex-and age-disaggregated data. In only six out ofthe 50 countries had national policies makespecific reference to older people, though threeof the six countries had not gathered any infor-mation on older people. As a result, olderpeople often only become visible when thereturn process is under way and their numbersonly become apparent as they are left behind incamps many years later.

To date, practitioners and policy makershave devoted scant attention to the impact ofdisplacement caused by human rights viola-tions, conflict and natural hazards on older menand women.

Key issues and challenges for older IDPsThe report identifies actual concerns of olderIDPs based on programme evidence from morethan 10 country contexts and includes examplesfrom different locations to inform practice. On this basis, the report identifies the followingkey issues and challenges for older IDPs:

Older persons form a significant proportionof IDP and refugee groups – sometimes as highas 30-65 per cent in contexts where there arehigh numbers of older people in the populationand where younger or more able-bodiedmembers of the IDP population have migratedelsewhere, returned home or integrated intolocal communities.

The United Nations High Commissioner forRefugees (UNHCR) recognises older IDPs asamong the most at-risk individuals, characteris-ing them as “persons with special needs”,1alongside the chronically ill or disabled, andthose who have experienced very high levels oftrauma.

Each stage of the displacement cycle – theflight, the period of displacement and theprocess of return, resettlement or local integra-tion – confronts an older IDP and serviceproviders with specific challenges that need tobe addressed. Prolonged displacement canhave a particularly devastating impact onfamily ties and the community support avail-able to older persons; it cannot be assumed thatcommunities will always assist their old. Inmany cases, families have had to make painfulchoices leading to abandonment of olderpersons in order to survive.

HelpAge and IDMC research and data showthat older people are consistently neglected inhumanitarian operations and policy, thatgeneral programming does not integrate theirneeds and that they are rarely consulted withinIDP operations. Sex- and age-disaggregateddata are rarely collected, contributing to invisi-bility.

Older people have a range of skills, capaci-ties and roles. They often contribute tohousehold income, support household manage-ment through childcare and play a role ascommunity leaders, decision makers or media-tors. The degree to which these roles arerecognised and supported during displacementhas a significant impact on the challenges olderpeople face and their ability to survive andrecover.

Issues of limited mobility, visual and hearingimpairment, and reduced muscle strengthamplify the challenges of living in displacementcamps and accessing services such as food,health care, and water and sanitation. Specificnutritional needs, chronic health disease andmental deficiencies may require further tailored

The impact ofdisplacement onolder people Summary of report1

1 HelpAge International and Internal Displacement

Monitoring Centre (2012). The Neglected Generation – the

impact of displacement on older people.

An older womanin Pakistan

Class Case definition 4.63% (95% CI=3.3%; 6.0%)

Global acute malnu-trition (GAM)

MUAC <210mm oroedema

4.63% (95% CI=3.3%; 6.0%)

Moderate acutemalnutrition (MAM)

185mm ≤ MUAC<210mm withoutoedema

2.71% (95% CI=1.5%; 3.9%)

Severe acute malnu-trition (SAM)

MUAC <185mm oroedema

1.91% (95% CI=0.9%; 3.0%)

Table 1: Prevalence of acute malnutrition in older people inDadaab camps

Nutrition and baselinesurvey of older peoplein three refugee campsin Dadaab

An older man in Dadaab campsduring the HOA crisis in 2011

This short summary of the findings of a nutrition surveyconducted by HelpAge International in a refugee populationin Kenya lend support to two other pieces in the researchsection of this issue of Field Exchange where lack of fundingfor older people caught up in humanitarian crises as well andneglect of the needs of older people amongst displaced popula-tions are highlighted. (Ed)

Between the 3rd–13th of October 2011, HelpAgeInternational conducted a nutrition survey amongrefugees aged 60 years and above in the three main

camps of Dadaab (Ifo, Dagahaley and Hagadera), north-eastern Kenya. The aims of the survey were to estimatethe prevalence of acute malnutrition, health and socialstatus of people aged 60 years and above. The surveyused two-stage cluster sampling.

A total of 629 older people were interviewed, andtheir measurements taken (height, half arm span, andmid-upper arm circumference (MUAC)). Body massindex (BMI) was also calculated. The team used ques-tionnaires to collect data on eating habits, disabilities,social and health status, and access to water and sanita-tion. The prevalence of malnutrition, usingMUAC-based case definitions, is given in Table 1.

The survey found that between 460-840 older peopleliving in Dadaab camps were in need of some form ofnutritional support. Risk factors associated with malnu-trition were not being included in the general ration,having a low dietary diversity and eating less than threedifferent foods items and less than two meals per day.

1 HelpAge International (2011). Nutrition and baseline survey of older

people in three refugee camps in Dadaab, October 2011. Published

by HelpAge International, www.helpage.org

18

Research

Help

Age I

nte

rnational, K

enya,

2011

assistance not usually included in pack-ages provided to displaced populations.

Older women require specific atten-tion - due to increased life expectancy,they are more numerous and morelikely to be living alone. Protection risksare thus increased for women, who arenot necessarily afforded equal status insociety. In addition, in many IDP andrefugee camps, older women take onthe huge responsibility of supportingchildren whose parents have died ormigrated elsewhere.

Some of the main challenges toprotecting the rights of older displacedpersons include obtaining access tovulnerable older persons who are leftbehind when more able-bodied flee,securing identification and documenta-tion; ensuring land and housing rights;providing for basic needs; reunitingthem with families and other individu-als; providing appropriate health careand ensuring access to social supportand income. Without an adequateanalysis of their needs informing everystage of decision making duringdisplacement, older persons willcontinue to be marginalised withinprogrammes and policy intended tosupport the displaced. Crucially, theywill continue to form the majority ofIDPs and refugees left behind in campsor collective centres while youngerpeople begin new lives for themselves.This is a fundamental breach of olderpeople’s basic human rights.

The report makes a number ofrecommendations targeted at actorswith specific protection responsibilities,including national governments, theUN, and humanitarian and develop-ment partners working in displacementcontexts.

RecommendationsIn order that older IDPs access theirrights and entitlements and in recogni-tion of their growing global numbers,protection actors should:• Address older people’s specific

needs and their active contributions within national IDP legislation, policy and assistance and the devel-opment of regional and human rights law.

• Ensure that national disaster risk reduction plans and frameworks recognise and address the risks that older people face, such as by cover-ing their potential evacuation from areas affected by natural disasters.

• Establish adequate systems for the collection of sex- and age-disaggre-gated data on numbers and locationsof IDPs, including older people.

• Assist older IDPs during flight and while displaced to secure their basicneeds and provide adequate levels of protection.

• Pay sufficient attention to the facilitation of durable solutions for older people, including return, reset-tlement and local integration. Steps may include facilitating adequate

social support for older IDPs, such as by ensuring that they are inte-grated into any existing national pension and health care schemesand that these incorporate greater sensitivity to the needs of older adults.

• In collaboration with key stakehold-ers – including older people them-selves and service providers such asthe UN, national and international NGOs – make efforts to develop and support further research on older IDPs and related issues of concern, such as older people in urban displacement, inter-genera-tional relationships and roles duringdisplacement, and options for durable solutions for older people.

• Take active steps to consult with older people and ensure their participation in the decisions that affect their lives, recognising their capacities as well as the risks they face.

• Integrate older person’s concerns into sectoral and multi-sectoral assessments.

• Collect registration and monitoring data disaggregated by sex, age and location to inform programming.

• Ensure that older people have access to information concerning their rights and entitlements. In this context, consider the mobility, visualand aural challenges that older people may have.

• Where populations are fleeing from conflict and natural disasters, liaise with relevant actors to enable access to the most vulnerable remaining behind; to assist with transportationand movement for the most vulnerable; and to support family tracing and reunification for older people.

• Promote programmes to involve older adults and ensure that community centres include “older-friendly” spaces for meaningful social interaction and informal support groups.

• Develop appropriate community-level identification and referral systems for older persons and moni-tor their access to services.

• Ensure that support for older peopleis integrated into strategies to assist other age groups. For example, childprotection strategies must include their older carers, and familiesshould be supported to take care of their old.

• Ensure that livelihood support for IDPs, returned IDPs and those seek-ing to integrate locally – including training and small business loans – does not discriminate against older men and women who are still able and willing to work.

• Include and specifically target older people in cash transfer schemes.

• Facilitate access to identification documents that enable older people to obtain social support and provideassistance with administrative processes for the most vulnerable.

Summary of research1

19

Research

Disaggregation of health and nutrition indicators by ageand gender in Dadaab refugee camps, Kenya

Henry recently graduated with a BSc in Food andHuman Nutrition from Newcastle University. He hasconducted research in The Gambia and interned withUNHCR at the regional office in Nairobi, Kenya.

Adequate nutrition in the early years oflife is vital for optimal growth anddevelopment. However, within the

under 60 month (less than 5 years) age group,the nutritional needs vary widely. Despite this,health and nutrition data are often not disag-gregated throughout this age range.

The United Nations High Commissioner forRefugees’ (UNHCR) Health InformationSystem (HIS) was developed to be a standard-ised tool with the aim of helping to guide,monitor and evaluate all public healthprogrammes within UNHCRs operations. Thedata derived from this system are essential informing evidence based decisions on projectsand interventions1. HIS is operational in morethan 90 refugee camps in 18 countries world-wide, providing monthly data on a number ofcritical indicators across several sectors (e.g.camp population, growth monitoring, nutritionprogrammes).

The Dadaab refugee camps in the North Eastof Kenya were originally opened in 1991. Thecamps have remained open since, due primarilyto continued insecurity, violence and on-goingfood crises in Somalia. In Dadaab, most data aredisaggregated in terms of broad age groups(less than 5 years, 5 years and older) andgender. However, it is uncommon for healthand nutrition programmes in Dadaab to disag-gregate health and nutrition indicator data forthe under 60 month age group. In an exceptionto routine practice, for a 16 month period span-ning 2007-2008, data for a number of key healthand nutrition indicators were disaggregatedinto sub groups for children less than 5 years.This article summarises the results of examin-ing thus disaggregated data for children under60 months in Dadaab in terms of mortality (all-cause), morbidity (watery diarrhoea only) andadmissions to feeding programmes.

ObjectivesThe objectives of the research were:• To determine whether gender- and age-

disaggregated data in children under 60 months of age were more effective at meas-uring morbidity and mortality rates than data aggregated across the age range.

• To investigate how representative aggregateage data for children less than 60 months of age compared with each disaggregated age group individually.

• To investigate whether disaggregated data have the potential to be a more accurate tool to guide and monitor health and nutri-tion programmes.

MethodsDuring the 16 month period, data werecollected by UNHCR and their implementingpartners using HIS protocols. The under 60month age range was disaggregated into four

(p=0.003). However, one-way ANOVAconducted on data grouped by age suggestedthere was no significant difference (p=0.401).

Mortality variations between age groupsFigure 1 shows the mean mortality rate for eachage group in Dadaab. One-way ANOVA wasconducted to assess the variation between eachindividual age group and the aggregate agemortality data, within each gender. The find-ings of the analysis are summarised in Table 2.One-way ANOVA was also conducted to assessthe mortality rate variation between each of theage groups individually.

The mortality rate for female infants under 6months of age was significantly higher than theaggregate age mortality data. A significantdifference also existed between the ≥12 < 24months and ≥24 <60 months age groups and theaggregate age mortality data in both genders.Both the ≥12 <24 months and ≥24 < 60 monthsage groups had a significantly lower mortalityrate than the aggregate age data.

There was no significant difference in themortality rate between the <6 month and ≥6<12 months age groups. However, there was asignificant difference in mortality rate betweenboth of these age groups and the ≥12 <24 monthand ≥24 < 60 month age groups observed inboth genders (see Figure 1).

Morbidity (watery diarrhoea) variationsbetween gendersGLM ANOVA was conducted to assess the vari-ation in morbidity (watery diarrhoea) ratesbetween genders. No significant variation wasdetected (p=0.578).

Morbidity (watery diarrhoea) variationsbetween age groupsFigure 2 shows the mean morbidity rate foreach age group in Dadaab. One-way ANOVAwas conducted to assess the variation betweeneach individual age group and the aggregateage mortality data, within each gender. One-way ANOVA was also conducted to assess themorbidity rate variation between each of theage groups individually.

All of the individual age groups were signif-icantly different to the aggregate age morbiditydata, as summarised in Table 3. The only twoage groups that did not have a significantlydifferent morbidity rate compared with eachother were infants <6 months and ≥6 <12months of age. Figure 2 indicates that themorbidity rate was significantly higher in thetwo youngest age groups, with a sharp decreasein morbidity rate in children ≥12 month of age.

sub-groups for age (< 6 months, ≥6-<12 months,≥12 -<24 months, ≥24 -<60 months) and aggre-gated as < 6 months.

Morbidity and mortality data were classifiedby gender as is standard practice in all UNHCRprogrammes utilising the HIS. Feedingprogramme admission data was not disaggre-gated by gender (standard practice also).

Monthly data for each indicator wereprovided for all three of the camps that wereoperational at the time of data collection:Dagahaley, Hagadera and Ifo. Data across theage range were aggregated retrospectively(referred to as ‘age aggregate’ data) for compar-ison with the individual age group data in thisstudy. The aggregate data depict how data forthese indicators would have appeared haddisaggregation not been conducted during thistime period.

Morbidity data were collected at healthcentres, within each of the camps. Mortalitydata were collected in health centres and bycommunity health workers throughout thecamps. Accurate monthly population data wereprovided by UNHCR for each of the individualage groups and genders allowing mortality andmorbidity rates to be calculated. The datadisplayed for both mortality and morbidityrepresent deaths (all causes) or cases of waterydiarrhoea per 1000 of the specific populationper month respectively. All data represented inthis article are an aggregate of the three campsand are referred to as ‘Dadaab’.

Nutrition programmes available in thecamps are supplementary feeding programmes(SFP) for moderate acute malnutrition (MAM)and therapeutic (TFP) and outpatient program-ming (OTP) for severe acute malnutrition(SAM) management. Feeding programmeadmissions were collected on a monthly basisby health workers in each of the camps andwere disaggregated for each of the four agegroups, but not for gender. Data for the threecamps were compiled retrospectively toprovide aggregate ‘Dadaab’ data. Feedingprogramme admission rates were calculated asnumber of admissions per 1000 of the specificpopulation per month.

ResultsSample characteristicsTable 1 shows the mean sample characteristicsfor the data collection period. The sex ratioshows that within each age group, the propor-tion of males to females was comparable.

Mortality variations between gendersGeneral Linear Model (GLM) ANOVA wasconducted on total pooled data to assess theoverall gender variation in mortality. Theanalysis suggested there was a significantdifference in mortality rate between genders

1 UNHCR. Health Information System. 2011 [cited 2012

09/03/2012]; Available from:

http://www.unhcr.org/pages/49c3646ce0.html. See also

research piece on HIS in this issue of Field Exchange.

The author would like to give special thanks to UNHCR staff AllisonOman, Senior Regional Nutrition and Food Security Officer andCaroline Wilkinson, Senior Nutrition Officer, for their technical inputand guidance. With thanks also to Ismail Kassim and Gloria Kisia.

By Henry Mark

20

Research

Feeding programme admission ratesAdmission rates to all of the nutritionprogrammes provide an insight into the nutri-tion situation within each of the age groupsduring the study period (see Figure 3). There islow availability of treatment for MAM and SAMin infants <6 month of age that affects theadmission rate for this age group which is verylow (rate/1000 of population/month is SFP 2.9,TFP/SC 7.6 and OTP 1.7). The lack of treatmentoptions for this age group is a concern, given thehigh rates of morbidity and mortality observedin this age group (as highlighted above).

With the exception of the <6 month agegroup, Figure 4 shows a clear trend in decreasingadmissions to all nutrition programmes with agefrom 12 months. This trend mirrors the decreaseseen in the mortality and morbidity data for thesame age groups. There was a significant fall inthe admissions rates between the ≥6 <12 monthand ≥12 <24 month age groups, for both SFP(p=0.021) and TFP/SC (p=0.001).

These variations mean that aggregate datafor the <60 month age group provides a signifi-cant underestimation of the admission rates forthe ≥6 <12 months and ≥12 < 24 month agegroups for all of the feeding programmes.Equally, the aggregate data provide an overesti-mation of admissions to all of the feedingprogrammes in the ≥24 <60 months age group.

DiscussionAlthough the refugee camp context is uniqueand may not reflect the situation in which manyhealth programmes operate, the findingspresented here pose interesting considerationsthat apply to the wider humanitarian sector, aswell as to other refugee settings.

When mortality data were pooled for age, asignificant difference in mortality rate wasdetected between genders. However, this associ-ation was not seen when data were grouped forage, suggesting that significant amounts of datamay be needed in order to detect such differ-ences. This trend has also been noted inprevious research in non-refugee populations inKenya2. To determine if gender disaggregatedindicators for mortality are beneficial, datacollection over a longer time period is needed.This would ascertain if there are any significantgender differences in mortality within each agegroup. There is also a need to determine if anyscenarios, such as a quick onset emergency, mayimpact more significantly upon a specificgender, perhaps through a change in healthseeking behaviours of caregivers.

In female infants <6 months of age, therewas a significant difference in mean mortalityrate compared with the aggregate age data. Themortality rate in male infants < 6 months of agewas not significantly different from the aggre-gate age mortality data. However, the analysisshowed there may be a weak association(p=0.087) and this is worth further investiga-tion. Furthermore, the aggregate age data gavea significant overestimation of the mortalityrates in all age bands for all children over 12months of age for both genders.

A significant difference was seen in themorbidity rate between all of the individual agegroups and the aggregate age data, for bothgenders. In this instance, the aggregate morbid-ity data gave a significant underestimation ofmorbidity in the youngest two age groups anda significant overestimation in the eldest twoage groups.

Aggregate data may detect differences in theoverall morbidity and mortality trends in apopulation. However this analysis stronglysuggest that aggregate age data for infants andchildren less than 60 months of age does notaccurately represent morbidity and mortalitythroughout the age range. Perhaps most impor-tantly in relation to health and nutritionprogrammes, the aggregate data fail to reflectthe high morbidity and mortality rates ininfants <12 months of age and the strong, clini-cally meaningful decrease in morbidity andmortality rate with increasing age. This trend ismirrored in nutrition programme admissionsthat decrease with increasing age.Disaggregated data may also allow for moreaccurate and timely responses to changes in thehealth and nutrition status of specific popula-tion groups, guide and monitor specific healthand nutrition interventions and also helpimprove targeted resources allocation.

Public health and nutritional interventionsmay target specific age groups (e.g. comple-mentary feeding interventions to children≥6-<24 months of age, exclusive breastfeedingsupport to infants < 6 months) but detection ofimpact is masked by aggregate data. Given theurgent need to strengthen the management ofacute malnutrition in infants <6 months, thecollection and presentation of health and nutri-tion data in infants < 6 months helps to definethe caseload and spotlight the need.

The age categories used for disaggregationin this study are closely matched to significantchanges in the nutritional requirement anddevelopment of infants and young children. Insome situations it may be suggested that the>12 month age groups could be aggregated.However, given the significant difference inboth the morbidity rate between the ≥12 <24and ≥24 <60 month age groups and the admis-sion rates to the feeding programmes, the fourage groups used in this study would seemappropriate for future data collection.

This study presents a strong case for agedisaggregation of data in children <60 months.Further research is needed to determine thecost:benefit (financial, capacity, time) of suchdisaggregated data collection, the additionalburden this would place on existing health andnutrition programmes and their implementingpartners and how to address these. The implica-tions will partly depend on prevailing practicethat does vary. For example, some implement-ing partners in Dadaab routinely record onlyaggregate data at data collection point (e.g. achild is only classified as <5 years), while otherscollect data by smaller age groups (e.g. < 6months) that are aggregated for reporting. Anagreed standardised approach to age disaggre-gation would greatly enable comparisonsbetween UNHCR programmes, as well as withnon-UNHCR programmes. There is a need toconduct further research with a larger samplesize and within different settings in order tobetter understand the dynamics and strengthenefficacy of gender and age disaggregated datacollection in this age range in different humani-tarian contexts.

For more information, contact: Allison Oman,email: oman@unhcr.org or Henry Mark, email:hmark3483@gmail.com

2 Hill, K. and D.M. Upchurch, Gender differences in child health

- evidence from the demographic and health surveys.

Population and Development Review, 1995. 21(1):p.127-151.

Age Group Mean population

Male Females Sex ratio

<6 months 1,034 1,031 1.003

≥6 <12 months 1,457 1,429 1.020

≥12 <24 months 3,073 3,020 1.018

≥24 <60 months 10,077 9,568 1.053

< 60 months (aggregate) 15,641 15,048 1.043

Table 1: Sample characteristics

Age group comparison Male Pvalue

Female Pvalue

<6 months versus < 60 months 0.087 0.023

≥6 <12 months versus < 60 months 0.563 0.154

≥12 < 24 months versus < 60 months 0.015 <0.001

≥ 24 < 60 months versus < 60 months 0.004 <0.001

Table 2: Differences in mean mortality rate betweenage groups for males and females in Dadaab

Age group comparison Male Pvalue

Female Pvalue

<6 months versus < 60 months 0.004 0.006

≥6 <12 months versus < 60 months 0.002 0.001

≥12 < 24 months versus < 60 months 0.001 <0.001

≥ 24 < 60 months versus < 60 months <0.001 <0.001

Table 3: Differences in mean morbidity (watery diarrhoea) rate between age groups formales and females in Dadaab

Figure 1: Line graph of mean mortality rates formales and females in all age groups inDadaab, with 95% confidence intervals

mor

talit

y ra

te p

er m

onth

<6m ≥6 <12m ≥12<24m ≥24<60m <60m (agg)Age group

Male

Female

5

4

3

2

1

0

Figure 2: Line graph of mean morbidity (watery diarrhoea) rates for males and females in allage groups in Dadaab, with 95% confidenceintervals

mor

talit

y rat

e per

1000

per m

onth

<6m ≥6<12m ≥12<24m ≥24<60m <60m (agg)Age group

Male

Female

140

120

100

80

60

40

20

0

Figure 3: Admission rates to supplementary, therapeutic and out-patient feedingprogrammes in Dadaab

mor

talit

y rat

e per

1000

per m

onth

<6m ≥6<12m ≥12<24m ≥24<60m <60m (agg)Age group

SFP admission rateTFP admission rateOTP admission rate

250

200

150

100

50

0

21

By Natasha Lelijveld

Natasha Lelijveld has recentlycompleted her MSc in InternationalChild Health at UCL. She is currentlyinterning in the Nutrition departmentat ACF UK and continues to work withDr Marko Kerac developing MAMI-

related and Malawi-focused projects.

The author would like to thank Ajib Phiri, Neil Kennedyand Ellah Mawerenga, along with many of the otherstaff at Queen Elizabeth Hospital in Blantyre.Acknowledgement is also due to ChawanangwaMahebere-Chirambo for his contribution during datacollection. Thanks are also due to members of theMAMI project (ENN, ACF and UCL) for highlighting theneed for research in this area, and lastly, special thanksand acknowledgment to Dr Marko Kerac for his guid-ance and input as supervisor of this project.

Research

BackgroundThe focus of most acute malnutrition researchand programming to date has been on themanagement of children aged 6 to 59 months;this is slowly changing. A recent study esti-mated that of 20 million children under 5 yearswith severe acute malnutrition (SAM) world-wide, 3.8 million are infants under 6 months(infants <6m)2, dispelling a common miscon-ception that malnutrition in this age group israre. Further, the final report from theManagement of Acute Malnutrition in Infants(MAMI) Project3 stressed the paucity ofresearch regarding management of this agegroup, in particular highlighting the insuffi-cient evidence and lack of a ‘gold standard’treatment method.

The prevalence and immediate/longer termconsequences of growth faltering during thefirst 6 months of life have received greaterattention of late, as further evidence of theimportance of nutrition between conceptionand 2 years (the ‘critical 1000 days’) has come tolight4,5. This period of 1000 days is also referredto as the ‘window of opportunity’ as nutritioninterventions have the potential to prevent last-ing damage and are most cost-effective duringthis time6.

Three distinct stages of nutrition affectgrowth during these ‘critical 1000 days’: mater-nal nutrition during pregnancy, exclusivebreastfeeding for 6 months and the complemen-tary feeding period from 6 -< 24 months. Eachstage is different and requires unique knowl-edge and research, care and interventions.Nutrition interventions during pregnancy andfrom 6 - 24 months are well studied and largelyunderstood. However it is a very different storyfor interventions between birth and 6 months,the least studied and understood stage of the‘1000 days’.

The motivation for this research project wasin response to the MAMI Report, to build theevidence base regarding the use of supplemen-tary suckling (SS) in infants <6 months (see Box1 on SS).This qualitative research explores thebarriers and facilitating factors to implementingSS as a treatment for SAM in infants <6m. Theproject also explored the acceptability of usingSS with orphans who are malnourished, andinvestigated perceived underlying causes ofSAM in infants, in order to improve use andacceptability of SS across all settings.

Supplementary suckling in the literatureA review of international and national guide-

lines7 found that SS is recommended by themajority (97%) of guidelines as the mainmethod of treatment for SAM in infants <6m,however in practice it appears to be rarely used.Malawi is a typical example of this and hencewas the setting for this research.

A literature review of SS also revealed adisparity between its regular mention innational policy documents and very little eval-uation, or even mention, in the scientificliterature. Just two studies, previouslypublished by ENN’s Field Exchange, have eval-uated SS as a treatment for malnutrition ininfants8,9. One further study assesses effective-ness of SS at preventing hospital-inducedmalnutrition in infants10. Other studies haveconsidered SS in relation to situations besidesmalnutrition, such as initiation of lactation inadopting mothers or feeding of Low BirthWeight (LBW) infants following a period ofhospitalisation.

Treatments available to infants <6mThe management of infants <6m suffering fromSAM is challenging as the treatments generallyon offer (Ready-to-use Therapeutic Food(RUTF)) are not indicated in this age group.SAM management in infants <6m requiresspecialist staff knowledge and time-intensiveskilled input to treat and restore effective,exclusive breastfeeding whenever possible.There is currently no evidence-based treatmentfor infants <6m with SAM as SS has not beenformally evaluated. Other treatment options forinfants at present include administering thera-peutic milk via a cup, a spoon or an NG tube. SSis, however, thought to be the best treatmentoption for these infants as it is unique in itspotential to re-establish exclusive breastfeedingin the context of therapeutic treatment.

1 Lelijveld N (2012) “Barriers and Facilitating Factors to

Implementing Supplementary Suckling as a treatment for

Severe Acute Malnutrition in Infants: a Qualitative Study”.

Unpublished MSc thesis, UCL2 Kerac M., Blencowe, H, Grijalva-Eternod C, McGrath, M,

Shoham, J, Cole, T. J. & Seal, A. (2011). Prevalence of

wasting among under 6-month-old infants in developing

countries and implications of new case definitions using

WHO growth standards: a secondary data analysis.

Archives of Disease in Childhood, 96, 1008-10133 MAMI. (2010). Management of Acute Malnutrition Project

Report. Emergency Nutrition Network, UCL Centre for

International Health and Development, Action Contre la

Faim [Online]. Available: http://www.ennonline.net/

research/mami [Accessed 03/01/ 2012].4 Victora, C. G., De Onis, M., Hallal, P. C., Blössner, M. &

Shrimpton, R. (2010). Worldwide timing of growth

faltering: revisiting implications for interventions.

Pediatrics, 125, e473-e480

Qualitative study of supplementarysuckling as a treatment for SAM ininfants This article summarises key findings of an MSc thesis1

5 S.U.N. (2010). 1000 Days, Scaling Up Nutrition [Online].

Available: http://www.thousanddays.org/about/ [Accessed

03/04/ 2012].6 World Bank (2006). Repositioning Nutrition as Central to

Development; a strategy for large scale action. Washington

DC, USA.7 Kerac M, Trehan I, Lelijveld N, Onyekpe I & Manary M

(2012). Inpatient Treatment of Severe Acute Malnutrition in

infants <6 months: An AGREE appraisal of national protocols

and GRADE review of published literature. Unpublished;

Presented at WHO Nutrition and Growth Advisory Group

Meeting, 1-3rd Feb, 2012.8 Corbett M (2000). Infant Feeding in a TFP. Field Exchange.UK.9 Oberlin O & Wilkinson C (2008). Evaluation of Relactation by

the Supplemental Suckling Technique. Field Exchange. UK.10 Alves, J., Figueira, F. & Nacul, L. C. (1999). Hospital

Induced Malnutrition in Infants: Prevention by Relactation.

Indian Pediatrics, 36, 484-487.

Box 1: What is Supplementary Suckling?

Supplementary Suckling (SS) provides the SAMinfant <6m with therapeutic milk in order to initi-ate rehabilitation and weight gain, while alsoaiming to re-establish exclusive breastfeedingthrough stimulating relactation.

SS is sometimes referred to as NursingSupplementer, LactAid Supplementer orBreastfeeding Supplementer.

Malawi’s National Guidelines on CommunityManagement of Malnutrition recommend SS beimplemented as follows:• The therapeutic milk is given using a size 8

NGT (naso-gastric tube).• Therapeutic milk is put in a cup. The mother

holds the baby in her arms and the cup with the milk in one hand.

• One end of the tube is put in the cup and the other end of the tube is put on the breast at the nipple. The infant is offered the breast in the normal way so that the infant attaches properly.

• When the infant suckles on the breast, with the tube in his/her mouth, the milk from the cup is sucked up through the tube and taken by the infant.

• The cup should be at least 10cm below the level of the breast so the milk does not flow too quickly.

• The mother can hold the tube at the breast with one hand or may find it more convenientto hold the tube in place with a strip of tape.

• It may take one or two days for the infant to get used to the tube and taste of the therapeutic milk.

See also Figure 2

Mothers and children waitingoutside the malnutrition wardat the Queen ElizabethHospital, Blantyre

N L

elijv

eld

. M

ala

wi, 2

012

22

The lack of an evidence-based treatment method for infants<6m, in addition to the limited implementation of SS worldwide,leads to the need to understand better the barriers and facilitat-ing factors for implementing SS, with the ultimate aim ofimproving its usage and the evidence base.

MethodQualitative interviews and focus group discussions (FGDs) wereconducted at Queen Elizabeth (QE) Hospital in Blantyre, Malawiwith mothers, carers and nursing staff (caregivers) using a semi-structured topic guide. Thematic analysis of data using Long TableMethod11 was applied in order to identify themes.

Objectives• To explore the barriers and facilitators to implementing SS, as

perceived by care-givers in Blantyre, Malawi.• To explore the acceptability of using SS as a treatment for

malnutrition in orphans, as perceived by caregivers in Blantyre, Malawi.

• To explore the causes of SAM in infants <6 m, as perceived bycaregivers in Blantyre, Malawi.

The background literature notes that staff support and mothers’confidence appear to be important in the success of SS12, hencethese groups comprise the target populations for this study.Convenience sampling was largely used, however purposivesampling ensured that healthcare staff recruited had some expe-rience of treating SAM and that the ‘mothers’ were a female carerof an infant less than 2 years old, either breastfeeding or non-breastfeeding. See Figure 1 for sample size; it was felt thatsaturation was reached and exceeded.

Interviews and FGDs began with a standardised explanationof SS, including an image used in the Malawi national guidelines(see Figure 2). Participants were then encouraged to ask ques-tions, allowing further clarifications where needed.

The topic guide was initially designed by the researcher, withinput from relevant professionals, including the national InfantFeeding in Emergencies (IFE) committee. It was then piloted onthree mothers, one nurse and one focus group, and discussedwith the translator prior to completion. All participants wereoffered a convenient time for interview and use of a translator.No staff members requested the use of the translator whereas allbut one of the mothers did. Interviews lasted between 20 and 40minutes. FGDs lasted between 1 hour and 1 hour 30 minutes.

Inductive coding and taking of memos took place throughoutdata collection; as saturation was reached themes began to solid-ify and emerge clearly. Following the creation of some keythemes, Long Table Method was used, with the translator, tofurther link, expand and refine themes, until the final resultswere created. The translator’s active role as an additional coderwas important due to his prominence during the interviewingand translating process.

ResultsBarriers, facilitating factors and solutions to SS implementationFive Major Themes and patterns emerged from the data whichrelate to each of the barriers, facilitating factors and solutionsdiscussed by the participants.• Motivation • Breastfeeding views • Practicality • Understanding • Perception of medicine

These Major Themes are generic and will likely affect the imple-mentation of SS in all settings. For the Malawi setting, this studyidentified Sub-themes under each major theme, which are lessgeneric but allowed for tailored solutions to implementing SS inMalawi (see Table 1 for summary of Major Themes and Sub-themes).

“On my side it is not difficult to follow”Mother of infant on Nursery Ward

“Siyovuta” – “It is not a difficult one”Mother in a FGD

“Some would not be happy to use [SS] because of not understanding how it is working”

Mother of child on Malnutrition Ward

“The problem is telling others of this technique; they could be scared that youwould pierce though their breast with the tube or feed the infant through the

nostrils”Mother on Malnutrition Ward

“There are some who could ask, but others won’t, they will just start tellingothers that I have found such-such a mother feeding a child with a tube, I think

she is HIV infected... they will say you have a problem... saying she is justpretending to use the breast, the main feeding channel is the tube”

Mother in a FGD

“I don’t know how much is the milk or what type or who will be offering it. Willwe be buying?”

Mother on Nursery Ward

“I think first of all, the mothers should have a health education about the tube,yes, how to take care of it... and maybe there should be somebody to help them

with the milk because a child could be crying and needs to be breastfed... Maybethere should be somebody to help so that the mik should be readily diluted”

Nurse working on Nursery Ward

“Of course it can be difficult but if you have discussed it with the family first, itis fine”

Nurse on Malnutrition Ward

“Due to this disease [HIV] one is supposed to be tested first; you should knowthe HIV status, if she is ok she could take the child and breastfeed it, but if HIV

infected then it is not right for her to breastfeed the child”Mother in FGD

“They must take the carer from the particular ‘family line’ that the mother hasdied...if the sister is not breastfeeding then the grandmother must breastfeed”

Nurse on Malnutrition Ward

Figure 2: Images used to explain SS toparticipants

Figure 1: Sample size

44 participants invited

5 declined toconsent/take part

39 participants

6 nurses 33 carers

All individualinterviews

17 individualinterviews

16 across4 FGDs

SS well received following thorough explanation:

‘Understanding’ as a barrier to implemeting SS:

Extra staff support needed:

Concerns for using SS with orphans:

11 Krueger RA & Casey M (2000). Focus Groups: A practical guide for applied research, California USA, Newbury Park: Sage.12 MAMI. (2010). Management of Acute Malnutrition Project Report. Emergency Nutrition Network, UCL Centre for International Health and Development, Action Contre le Faim [Online]. Available: http://www.ennonline.net/research/ mami [Accessed 03/01/ 2012].

Mothers attending oneof the focus groups

N L

elijv

eld

. M

ala

wi, 2

012

Research

Breastfeeding Supplementer as depicted in‘Infant Feeding in Emergencies: a guide formothers’, WHO, 1997

This research found that SS is gener-ally received positively by caregivers inBlantyre, Malawi, with 82% of partici-pants saying they would use SS and 74%saying it looked easy.

However, ‘understanding’ of a varietyof aspects of SS stood out particularly asa common barrier but with a relativelysimple solution. A tailored explanation ofSS, particularly of aspects which werefrequently asked about such as the place-ment of the tube, HIV-related fears andpracticalities such as hygiene, wouldaddress many of the barriers perceivedby caregivers in this setting. As a result ofthis research, a tailored explanation of SS,addressing FAQs, has been created andwill be used by QE Nursery Ward as theybegin to implement SS. Other settingsshould consider conducting a smallnumber of interviews with mothers andnurses in order to tailor this explanationof SS to their specific location.

Staff also identified that their work-load is over-stretched and thereforeworried about the amount of timeneeded to address mother’s understand-ing of SS and other practicalities. Thesolution suggested for this was theacquisition of a dedicated staff memberfor this task, which has been acknowl-edged by the hospital. Other healthfacilities need to consider strongly theirhuman resource requirements beforeimplementing SS as it is thought to befairly time-intensive.

Use of SS for orphansThe application of SS for orphans (wherea wet nurse is identified for an orphanedinfant <6m with SAM) was also exploredby this study, with mixed results.Participants were divided as to whetherbreastfeeding of an orphaned infant isacceptable in the culture; the usual barri-ers and facilitating factors also appliedhere. HIV was a particularly contentiousissue in this case. Some participants didfeel that SS for an orphan is acceptableprovided the carer is a direct familymember, is HIV negative and a discus-sion has been held with the family. It istherefore advised that, at this time,application of SS for orphans should beassessed on a case-by-case basis inMalawi and in settings with similarlyhigh prevalence of HIV.

Perceived causes of SAM in infants <6mLastly, causes of SAM in infants in thissetting were found to be largely related

23

to breastfeeding problems or choices,such as perceived breastmilk insuffi-ciency, consistent with causesidentified in previous studies. Onecause worth highlighting is ‘maternalinfection’, which was frequentlymentioned by participants as the maincause of malnutrition in infants. Therelationship between child infectionand malnutrition is widely recognised,however the relationship betweenmaternal infection and child malnutri-tion far less so. The results of this studysuggest that this may be a major causeof infant malnutrition which is in needof further exploration.

These results are having immediatepolicy and practice implications for theNursery Ward at QE hospital, whichnow plans to implement SS and hire aLactation Support nurse in order toaddress notable barriers within themesof Understanding, Breastfeeding viewsand Practicality. In addition, the wardwill be using the ‘Explanation of SS’created from these results. Thesechanges to local policy and practice arefirst aimed at having an immediatepositive impact on infants on the ward,but also aim to contribute to existingresearch findings. Conducting alarger-scale acceptability trial, post SSintervention, is necessary to considerwhether caregiver’s acceptance of SS isstill present having actually used thetechnique.

Other inpatient health facilitiesacross the world should consider theprevalence of SAM in infants <6m intheir setting and then consider imple-menting SS. Locating specific views ofcaregivers within the five major themesoutlined should help to plan the bestmode of implementation andmaximise chances of success. As aresult of this research, further work inQE Hospital now has the potential torevolutionise treatment of SAM ininfants <6m across the world, givinginfants a much better outlook, bothduring and after those critical ‘1000days’. It is hoped that health facilitiesin other inpatient settings will alsofollow suit.

For more information on this study,contact: Natasha Lelijveld, email: natasha.lelijveld.11@ucl.ac.uk

If you are planning studies aroundinfants <6m, let us know, email: marko.kerac@gmail.com

Patterns of bodycomposition amongHIV-infected, pregnantMalawians and famineeffects

Few studies have examined maternal anthropome-try and its predictors during pregnancy insub-Saharan Africa in the context of HIV/AIDS. In

resource-limited countries, exposure to inadequatedietary intake, frequent reproductive cycles, infectiousdisease, and demanding physical labour may alterwomen’s body composition dynamics during pregnancycompared to what is seen among women in countrieswhere these stresses are mostly absent. The importanceof seasonality to reproductive health is emphasized byseveral studies across Africa. Among predominantlyHIV uninfected populations in Africa, lower gestationalweight gain and increased maternal morbidity have beenassociated with the rainy season, which is often charac-terized by increased food shortage, physical labour, andmalarial infection rates. Because loss of lean body masshas been shown to be a predictor of HIV survival, inde-pendent of CD4 count, it is important to describe thechanges in body composition throughout pregnancy andto identify its key determinants in HIV-infected popula-tions.

A recently published study hypothesized that mid-upper arm circumference (MUAC), arm muscle area(AMA) and arm fat area (AFA) would decrease in HIVinfected populations, given the tendency for maternalwasting combined with the increased effects of foodshortage, physical labour, and infection rates during therainy seasons. The researchers analyzed repeated anthro-pometrics from HIV-1 infected pregnant Malawianwomen to evaluate two primary objectives: (1) todescribe patterns of change in maternal weight, MUAC,AMA, and AFA among HIV infected pregnant women,and (2) to identify potential seasonal, CD4 count, andsocio-demographic factors, including age, parity, maritalstatus, wealth, and the educational level and occupationof both the mother and her partner, as predictors ofmaternal anthropometrics in this population.

The study was based on a secondary data analysis,including prenatal women who consented and met pre-delivery screening criteria between April 2004 andAugust 2006 for the Breastfeeding, Anti-retrovirals, andNutrition (BAN) Study, a postnatal clinical trial. The data

1 Roshan T et al (2012). Patterns of body composition among HIV-

infected, pregnant Malawians and the effects of famine season.

Maternal and Child Health Journal, volume 15, No 1. Doi10.1007/

s100995-012-0970-6

Dr. C

harles

van d

er

Hors

t/BAN

Stu

dy

Mothers attendingantenatal clinic

Major Themes Motivation Breastfeedingviews

Practicality Understanding Perception ofintervention

Sub-themes

Barriers Child’scomfort

Choosing notto breastfeed

Hygieneconcerns

Lack of understanding

Dislike of intervention

Disinterest Time pressures

Facilitatingfactors

Doing thebest forInfant

Strong breast-feedingculture

Own responsibility

Empathy forlactation issues

Abide bydoctors

Better thanthe alternatives

Table 1: Summary of major themes and sub-themes affecting SS implementation

Research

Summary of research1

24

Research

for the analysis were derived from surveysconducted to screen women who consented tobe in the BAN Study and to collect baselinedemographic, anthropometric, dietary, andhealth status data during antenatal care and atdelivery. Study participants were recruitedfrom four sites with outreach to all pregnantwomen in Lilongwe, Malawi.

By August 2006, 1,745 women met initialantenatal screening criteria: > or =14 years ofage, no prior antiretroviral medication use, < 30weeks gestation, and no serious complicationsof pregnancy. Of these, 1,336 women returnedfor the second antenatal screening visit approx-imately 1 week later and met eligibility criteriabased on blood test results: HIV positive status,CD4 count ≥200 cells/dL, haemoglobin ≥7g/dL, and normal liver function tests (≥2.5times the upper limit of normal). At the secondantenatal visit (the baseline visit), eligiblewomen completed a baseline interview andphysical exam. Of the 1,336 eligible women atthe baseline visit, 168 were missing a heightmeasurement and 38 were missing another keybaseline factor. Therefore, there were 1,130women available for this analysis. The samplesize of 1,130 was determined to be sufficient todetect a 1 unit difference in the main outcomeswith 90% power.

None of the women took antenatal anti-retrovirals during this or prior pregnancies. Inaccordance with national guidelines fromMalawi’s Ministry of Health on initiation ofantiretroviral treatment, these women did notqualify for treatment due to their high CD4count levels. Additionally, at the time, Malawi’snational guidelines on prevention of mother tochild transmission of HIV did not include anti-retroviral prophylaxis during pregnancy. Datawere collected on interim malarial infectionsbut were not available for all participants. Allwomen received iron and folate supplements,screening for anaemia, malaria prophylaxis,and mosquito nets.

The BAN Study protocol was approved bythe Malawi National Health Sciences ResearchCommittee and the institutional review boardsat the University of North Carolina at ChapelHill and the U.S. Centres for Disease Controland Prevention.

Gestational ages at baseline and subsequentprenatal visits were derived from the date oflast menstrual period (LMP) or if LMP wasunknown, the first available fundal height (FH).

Weight, height, MUAC and triceps skinfoldthickness were measured at each visit bytrained BAN nutrition staff. For MUAC andskinfolds, quality assurance checks wereperformed periodically to ensure consistentinter- and intra-reliability of measurements.Weight was measured to the nearest 100 g ateach visit. MUAC was measured to the nearest0.1 cm. MUAC and triceps skinfold thicknesswere used to AMA, an indicator of musclemass, and AFA, an indicator of fat mass. CD4count was measured as a cross-section duringthe first screening visit.

In Malawi, food availability, malnutrition,and infectious disease morbidity vary substan-tially by season due to cycles of rainfall andagricultural production. The famine season,locally referred to as the ‘Green Famine’,extends from August to March and includes therainy season prior to the harvest. This time

period is marked by limited food availability asstores of the previous year’s crops are depletedand incidence of infectious diseases peaks.Exposure to the famine season was measuredas the number of days during the month priorto each measurement that were spent in thefamine season.

Basic socio-demographic information wascollected during the baseline interview: age,parity, marital status, household characteristics,and the educational level and occupation ofboth the mother and her partner. Wealth wasdefined based on a wealth index, derived fromprincipal components analysis (PCA) of house-hold characteristics. The different categories orlevels of wealth were represented by indexquintiles. Depending upon the estimated gesta-tional age at baseline (range: 12–30 weeks),women were asked to return for follow- upprenatal care at approximately 28, 32, and 36weeks gestation. The average gestational age ateach follow-up visit was 29, 33, and 36 weeks,and the number of participants at these visitswas 694, 868, and 703, respectively. More than90% of the sample had at least 2 visits. Theaverage time between visits was 4.5 weeks(Standard Deviation (SD) = 2.8). The averagetotal time between the baseline and last follow-up visits was 10.4 weeks (SD = 5.1). The timebetween the last antenatal visit and deliveryranged from 0 to 10 weeks because in a fewcases, the last visit coincided with the deliverydate. The analyses therefore reflected changesduring the later stages of pregnancy.

and a decline was noted for AFA. In multivari-ate analysis, weight increased at a rate of 0.27kg per gestational week and AFA decreased at arate of 0.06 cm2 per gestational week. Exposureto the famine season was also associated withdecreased weight gain and loss of AFA perweek of pregnancy. CD4 count, as a continuousvariable, was directly associated with MUAC,AFA, and weight. Each 100 cells/dL increase inCD4 count was associated with an increase of0.08 cm (95% CI 0.01, 0.15) in MUAC, 0.15 cm2(95% CI -0.01, 0.30) in AMA, 0.21 cm2 (0.01,0.41) in AFA, and 0.24 kg (0.03, 0.45) in weight.

Wealth, occupation, education level, age,and parity were associated with anthropomet-ric changes in MUAC, AMA, AFA, or weightduring pregnancy. Women with greater wealthhad increased MUAC and AFA while occupa-tion, but not wealth, among women had asignificant direct positive association withAMA. Women who completed primary educa-tion had higher MUAC due to increased AMAthan those who completed secondary or higherlevel education. Age and parity were positivelyassociated with MUAC and AMA, but only agewas associated with weight. AFA was unrelatedto age and parity.

Over two-thirds of the women in the studyhad some exposure to the famine season in themonth prior to an anthropometric assessment.Exposure to the famine season negativelyimpacted weight gain in this study, which isconsistent with a report of highest pregnancyweight gain among Malawian women who

deliver in July–September (mean gain 0.25–0.30kg/week) and lowest gain among those whodeliver in January-May (mean gain 0.10–0.20kg/week).

The rate of change in MUAC and AMAduring pregnancy was modified by exposure tothe famine season. Women with no exposure tofamine during the previous month experienceda subtle increase in MUAC and a significantincrease in AMA. In contrast, women whospent the entire preceding month in a famineperiod had a significant decrease in MUAC andAMA per week of gestation.

In conclusion, the findings among HIV-infected, pregnant women are similar to thosereported for uninfected women in sub-SaharanAfrica. However, effects of the famine seasonamong undernourished, HIV-infectedMalawian women are of concern. Strategies tooptimise nutrition during pregnancy for thesewomen appear warranted.

ResultsA crude analysis of weight change over the2,338 intervals between consecutive prenatalvisits indicated that 17.3% of intervals showedweight loss. About half of intervals also had aloss in AMA (48.7%) and AFA (53.1%). Of thoseintervals in which muscle stores were lost,34.3% lost both muscle and fat stores. Theprevalence of wasting (MUAC<22 cm) was1.9%, 3% of women gained no weight, 8.4%showed weight loss, and 59% had low fat stores(AFA<20 cm2). Most women in this samplewere young with low parity. The median base-line CD4 count was 439 (inter-quartile range:319–592) per 100 cells/dL. Average maternalweight, MUAC, AMA, and AFA at baseline was58.7 kg (SD = 8.2), 26.5 cm (SD = 2.7), 36.7 cm2(SD = 6.5), and 19.7 cm2 (SD = 8.1), respectively.

In bivariate linear analysis, maternal weightincreased at a rate of 0.24 kg per gestationalweek. There was no evidence of change inMUAC, while an increase was noted for AMA

Dr. C

harles

van d

er

Hors

t/BAN

Stu

dy

25

Research

Summary of research1

Anthropometry in infants under6 months in rural Kenya

It is currently estimated that worldwide 8.5million infants under 6 months are wasted.In poor communities, low rates of exclusive

breastfeeding and the introduction of mixedfeeding before the age of 3 months exposeinfants <6 months to risks of microbial contam-ination and malnutrition. In Kenya, 9.7%infants below 6 months are wasted (weight-for-length z score (WFLz) < -2) and 11% are stunted(length-for-age (LFA) z score < -2).

The Government of Kenya has proposed astrategy in which Community Health Workers(CHWs) are trained to deliver communityhealth services, including basic primary healthcare, growth monitoring (GM) and referral ofcritically ill patients to hospital. CHWs will beexpected to undertake door-to-door anthropo-metric screening of children and provide basicnutrition education and counselling.

At rural health facilities, weight iscommonly measured in infancy. However,weight-for-age (WFA) alone does not differenti-ate wasting from stunting and is typicallyaccessible only to those attending mother andchild health clinics (MCH). WFL is recom-mended for the diagnosis of acute malnutritionin this age group, but is rarely routinelyassessed because length boards are not usuallyavailable and length measurement is poten-tially unreliable. Among children aged 6–59months, the Mid Upper Arm Circumference(MUAC) may be used to diagnose severe acutemalnutrition (SAM). MUAC is a better predic-tor of mortality than WFA and within this agerange, is age-independent. In rural communi-ties, MUAC could be a valuable tool for use byCHWs for early detection of acute malnutritionin infants. However, reliability of MUAC meas-urement in early infancy is unknown, andcut-off values to determine intervention thresh-olds have not been defined. To address the firstof these questions, a recent study set out todetermine the inter-observer variability andaccuracy of MUAC and WFLz measurementstaken by CHWs among infants under 6 monthsin rural Kenya.

The study was conducted from February2008 through August 2009 in Kilifi District, arural district on the Kenyan Coast. Kilifi is thesecond poorest district in Kenya with an esti-mated 67% of the people living in poverty. Thestudy recruited three cadres of participant: (i)an expert in anthropometry with more than 30years of experience of anthropometry trainingand conducting nutritional assessments inKenya, (ii) health professionals (HPs) compris-ing nurses and public health officers in-chargeof Mother and Child Health (MCH) clinics, and(iii) CHWs based at health centres throughoutthe district.

1 Mwangome. M et al (2012) Reliability and accuracy of

anthropometry performed by community health workers

among infants under 6 months in rural Kenya. Tropical

Medicine and International Health. Doi.10.1111/j1365-

3156.2012.02959.

Study methodsThe research employed a cross-sectionalrepeatability design using the term ‘reliability’in a statistical sense to mean the inter-observervariability. For applicability within the healthsystem, the research team used measuringequipment normally in use in government facil-ities in Kenya, with an exception of theinfantometer, which is usually not available.Weight was measured to the nearest 100 g usinga ‘hanging’ scale costing 103 USD. The machinewas quality controlled every morning usingstandard weighing stones certified by theKenya Bureau of Standards. Length was takenusing a professional infantometer calibrated tothe nearest 1 mm and costing 443 USD. MUACwas measured on the left arm of the child usingTALC insertion tape marked to the nearest 2mm costing 0.25 GBP. All measures followedprocedures indicated in the United Nations(1986) guidelines.

Sample sizes were calculated separately forinfants older and younger than 90 days.‘Complete unreliability’ was defined as anintra-class correlation (ICC) of <0.4 and esti-mated the number of infants required for 90%power to distinguish ICC values of 0.6, whichwere defined as minimum reliability, from 0.4.This gave a required sample size of 71 infants ineach age group for three observers. To allow forpossible dropout from the study, the aim was torecruit at least 75 infants for each age group.Sample size for accuracy was 15 infants forevery 75 infants recruited.

The study was then undertaken in threestages. First, to establish intrinsic reliability, theexpert anthropometrist measured weight,length and MUAC among infants visiting theMCH at Kilifi District Hospital. Measurementswere repeated after each cohort of 10 children.First and second set of measurements wererecorded on separate forms. At the secondstage, a training manual was produced fortraining HPs and CHWs following guidelinesfrom the United Nations 1986 on anthropome-try. The expert and the first author trained sixHPs on anthropometry, safety procedures whenhandling infants and quality control of themeasuring equipment. After the training, theHPs were divided into two groups of three eachand repeatedly measured weight, length andMUAC of 150 infants (75 infants above and 75below 90 days old). Each child was measuredonce by each of the three HPs. For every 5thchild, the expert took measurements to deter-mine accuracy. Further training was given toaddress issues arising in the second stage toestablish HPs as trainers for the CHWs.

At the third stage, 18 CHWs were recruited,three from each of six sites: a district hospital, a

peri-urban health centre, two rural healthcentres and two rural dispensaries. HPsconducted a 1-day practical training on anthro-pometry, safety procedures when handlinginfants and equipment quality control. Thisaimed to replicate the type of training thatcould be providedoperationally. Then, each group of three CHWsindependently measured MUAC, weight andlength among 150 infants (75 under and 75 over90 days old) at their health facilities. A singleMUAC, weight and length measurement weretaken once by each CHW on each infant.Measures were blinded from each other. Ateach facility, one HP took measures on every5th infant to estimate accuracy.

Key findingsKey findings of the study were that amongCHWs, ICCs pooled across the six sites (924infants) were 0.96 (95% CI 0.95 0.96) for MUACand 0.71 (95% CI 0.68–0.74) for WFLz. MUACmeasures by CHWs differed little from theirtrainers: the mean difference in MUAC was 0.65mm (95% CI 0.023–1.07), with no significantdifference in variance (P = 0.075).

There are few published data on the accu-racy of anthropometry in early infancy. Nonehave used a systematic approach and includedCHWs as used in this study. In this study, therewas better concordance between CHWs andtheir trainers for MUAC than for WFLz.

In other studies, MUAC achieved high speci-ficity (>95%) and varied sensitivity (48–58%) inidentifying infants with severe malnutrition(WFLz <-3) and low-birth- weight (weight <2500g). In such studies, however, there is atrade-off between specificity and sensitivity,and therefore reported levels of sensitivity andspecificity should be interpreted within thestudy context and the cut-off used.

Absolute measures of MUAC, weight andlength were more reliable than calculated zscores. Length was the least reliable whenmeasured by HPs (pooled ICC was 0.82).Overall, WFLz was the least reliable anthropo-metric index (overall ICC was 0.71, and at onesite, met the ‘completely unreliable’ criteria).The most likely explanation is that WFLz isvery sensitive to changes in the absolute meas-urements, e.g. a 1cm change in length of a6kg/65cm infant results in a 21% change inWFLz. Errors in calculating z scores were notstudied.

Reliability study participants

M.

Mw

angom

e,

Kenya,

2011

26

Research

Summary of research1

Micronutrient powders v iron-folic acidtablets in controlling anaemia in pregnancy

The major cause of anaemia in pregnancy is iron defi-ciency, which is preventable. It is estimated that 56%of pregnant women in developing countries and

18% of pregnant women in industrialised countries areanaemic. The prevalence of anaemia during pregnancy inSouth Asian countries is among the highest in the world. Astudy conducted in rural Bangladesh reported a 54% preva-lence of anaemia among pregnant women during thesecond trimester. Similar high rates of anaemia have beenreported among lower-and middle-class pregnant womenin Dhaka. Although a number of national programmes havebeen implemented attempting to address this problem,continuing high rates of anaemia among pregnant womensuggest that traditional strategies such as iron–folic acid(IFA) tablets are not effective. This limited effectiveness maybe attributable to inadequate supply, low coverage, and/orpoor adherence. In order to address poor adherence tostandard iron drops in young children, the micronutrientpowder ‘Sprinkles’, a single-serving sachet containing ablend of micronutrients in powder form, was developed.Powdered micronutrient supplements can be easily incor-porated into foods prepared in the home. Several clinicaltrials have demonstrated the efficacy of micronutrientpowders in lowering the burden of anaemia in children inthe developing world. However, no study to date has exam-ined the efficacy of this intervention against anaemiathroughout pregnancy.

A recent cluster randomised trial, set out to determinewhether home fortification with a micronutrient powder forpregnancy (MNP-P) is at least as efficacious as IFA tablets forimproving haemoglobin concentration in pregnant women.The trial was carried out in the sub-district of Kaliganj incentral Bangladesh. This is a densely populated rural areathat relies primarily on agricultural labour. Participants wererecruited from 42 community-based Antenatal Care Centresoperated by BRAC (formerly known as the Bangladesh RuralAdvancement Committee), a large non-governmentalorganisation founded in Bangladesh in 1972.

All pregnant women who received services at one of theAntenatal Care Centres and were between 14 and 22 weeksof gestation were asked to join the study. Gestational agewas approximated by using the woman’s recall of her lastmenstrual period. If the number ofparticipants enrolled from one Antenatal Care Centre didnot meet the sample size requirements, the study teamreturned to the centre the following month and repeatedthe recruitment process. For logistical reasons, theresearchers limited the recruitment phase to the first fivemonths of the study period, which started in October 2005and ended in March 2006. Women did not meet the eligibil-ity criteria if they were severely anaemic (haemoglobin < 70g/L), had a haemoglobin concentration greater than 140g/L, were at more than 22 weeks of gestation, or werealready taking iron supplements prior to the start of thestudy. The primary reason for declining to participate in thestudy was unwillingness to provide a blood sample forhaemoglobin assessment. A total of 779 pregnant womenfrom 42 clusters were identified. Out of this population, 478women met the eligibility criteria and were randomized, atthe cluster level, to the MNP-P group (21 clusters, n = 243)or the IFA tablet group (21 clusters, n = 235). Each clusterconsisted of approximately 300 households, and thenumber of women per cluster ranged from 3 to 26. Theprimary reasons for exclusion at baseline were gestationalage greater than 22 weeks (15.8%), recent use of ironsupplements (11.3%), pregnancy loss (4.4%), and plannedmove out of the study area (2.2%). Two women wereexcluded because they had severe anaemia (haemoglobin< 70 g/L), and 12 were excluded because they had haemo-globin levels above 140 g/L.

A single Antenatal Care Centre typically serves onevillage or population of approximately 300 households.Each Antenatal Care Centre was defined as a cluster, andeach cluster was randomly allocated to receive either MNP-P (60 mg of elemental iron, 400 μg of folic acid, 30 mg ofvitamin C, and 5 mg of zinc) or IFA tablets (60 mg ofelemental iron and 400 μg of folic acid). The iron dosecorresponded to recommendations for areas in which theprevalence of anaemia is above 40%, and the dose of folicacid administered was based on the amount of syntheticfolic acid recommended for prevention of neural tubedefects. Both interventions were administered daily, start-ing at the first study visit and lasting until 32 weeks ofgestation. Those assigned to the MNP-P group wereinstructed to sprinkle the full contents of a package intoany semi-solid or semi-liquid food when the food was atroom temperature.

Changes in haemoglobin from baseline were comparedacross groups using a linear mixed-effects regressionmodel. At enrolment, the overall prevalence of anaemiawas 45% (n = 213/478). After the intervention period, themean haemoglobin concentrations among women receiv-ing the micronutrient powder were not inferior to thoseamong women receiving tablets (109.5 ± 12.9 vs 112.0 ±11.2 g/L; 95% CI, –0.757 to 5.716). Adherence to themicronutrient powder was lower than adherence to tablets(57.5 ± 22.5% vs 76.0 ± 13.7%; 95% CI, –22.39 to –12.94).However, in both groups, increased adherence was posi-tively correlated with haemoglobin concentration.

Focus group discussions to explore the reasons for non-adherence found that non-adherent women in the MNP-Pgroup expressed concern that adding Sprinkles to theirdaily food would worsen the nausea associated with preg-nancy. However, a few women also mentioned thatSprinkles increased their appetite. Given that food wasgenerally scarce, increased appetite was not a positiveattribute of the intervention

Overall, the findings suggest that despite lower adher-ence, the relative efficacy of MNP-P and IFA tablets wassimilar when they were started in the second trimester ofpregnancy. However, the interpretation of results is limitedto ‘relative’ comparisons, since the randomized design didnot include a non-intervention group.

Given the limited number of studies on MNP use duringpregnancy, it is also difficult to assess the effectiveness andfeasibility of this approach in a programme setting. In fact,recently published World Health Organization (WHO) guide-lines on the use of MNPs during pregnancy recommendagainst this intervention as an alternative to iron and folicacid supplements due to a lack of evidence on its potentialbenefits and harm. The guidelines also state, however, thatcurrent supplementation and mass food fortificationapproaches do not always work because of difficulties inimplementation or reaching target populations. There hasbeen growing interest in alternative strategies for providingmicronutrients to pregnant women in low- and middle-income countries, and WHO has recommended that futureresearch be conducted to evaluate the effectiveness ofMNPs as one such strategy. Considering the findings fromthis study, the authors suggest that further research onfactors related to MNP adherence that addresses culture-specific food preferences would be of particular benefit. It ispossible that simply having a choice between tablets orMNP-P might improve overall adherence to iron and folicacid interventions among pregnant women.

1 Choudhury et al (2012). Relative efficacy of micronutrient powders

versus iron–folic acid tablets in controlling anaemia in women in

the second trimester of pregnancy. Food and Nutrition Bulletin, vol.

33, no. 2, pp 142-149, 2012

This study evaluated CHWsacross a representative range ofsites in a rural district in a real-istic setting after a typicalpractical training. Such anarrangement is likely to be simi-lar to proposed changes by theGovernment of Kenya’sMinistry of Health when aimingto better identify infants in situwithin their community at riskof malnutrition.

However, evaluation in thecontext of research may haveresulted in a better than normalenvironment for measuringinfants in that there was moretime with fewer interruptionsand greater supervision. Thismay limit the application ofthese findings to rapid assess-ment or a door-to-door visitscenario. Secondly, the measur-ing equipment was in goodcondition and regularly cali-brated. This may not normallybe the case in rural public healthfacilities in Kenya. Thirdly,owing to a delay in equipmentavailability, the expert anthro-pometrist was unable to takeweight and length measure-ments in the same group ofinfants. It was therefore notpossible to calculate and esti-mate expert’s intra-observervariation of WFLz. Finally, themajority of the infants involvedin this study were recruitedfrom the routine GM clinics andnot randomly selected withinthe community, thus they wererelatively healthy.

The authors conclude thatCHWs can be trained to takeabsolute MUAC, weight andlength measurements accu-rately and reliably amonginfants age <6 months.However, the length-based zscore indices, LFAz and WFLz,are the least reliable anthropo-metric measures. Furtherstudies of the generalisability ofthese findings in other settingsand to assess the relationship ofMUAC with mortality andillness to establish appropriatecut-off values for MUAC useamong infants under 6 monthsold are needed.

Routine monitoring data are availablefrom the many nutrition programmesoperating in camps supported by

UNHCR, typically growth monitoringprogrammes, under the Expanded Programmeon Immunisation, MUAC2 based householdsurveillance, Supplementary FeedingProgrammes (SFPs) and community based ther-apeutic feeding programmes (TFPs). However,operational challenges and varied reportingstructures across implementing agencies havelimited the utility of this information for theassessment of changes in population healthstatus or comparisons of trends across regions.In 2005, UNHCR launched a Health InformationSystem (HIS) to enhance the quality and consis-tency of routine health information available inprotracted refugee situations (see Box 1), becom-ing the first source of routine feedingprogramme data collected with standardizedcase definitions and reporting formats acrossrefugee settings.

A recently published paper reviewsUNHCR’s HIS data to assess the coverage andeffectiveness of selective feeding programmesimplemented by partners in Kenya andTanzania from January 2006 to May 2009. Itexamines the extent to which routinely collectedHIS data can be used to inform nutritionprogramme strategies in post-emergency situa-tions. Kenya and Tanzania were chosen as casestudies because of their geographic proximity,similar restrictions on livelihood opportunities,distinct population trends, and the fact thattogether they account for more than half of theglobal refugee population living in campsreported in the UNHCR HIS.

Since the early 1990s, Kenya and Tanzaniahave hosted hundreds of thousands of refugeesfleeing war and insecurity in neighbouringcountries. In Kenya, the Dadaab and Kakumacamps currently host approximately 340,000refugees from Somalia, Sudan, Ethiopia, andother countries. A mid-2006 survey in Kenyashowed that despite widespread distribution ofWFP rations (~2,100 kcal/day) since 2003,

27

UNHCR feeding programme performance in Kenyaand Tanzania Summary of research 1

global acute malnutrition rates remained wellabove internationally recognised emergencythresholds at 22.2% in the Dadaab Camps and15.9% in Kakuma Camp. The nutritional statusof refugees in these camps has improved signif-icantly in the past few years. Recent surveysshow that global acute malnutrition rates havefallen from their peak in 2006 to 5.6%–10.7% inDadaab Camps and 9.2% in Kakuma Camp in2010.

In Tanzania, refugees from Burundi, Rwandaand the Democratic Republic of Congo arehoused in clusters of camps in far westernKigoma, Kasulu and Kibondo districts in theKigoma Region and the Ngara District in theKagera Region. As in Kenya, the Governmentof Tanzania prohibits refugees from seekingformal employment, travelling more than 4 kmoutside their camp, cultivating land outside thecamp boundaries, trading in markets or grazinglivestock. A nutrition survey in 2006 showedthat despite these restrictions and limited foodrations, the rate of global acute malnutritionwas low, with a weighted prevalence rate of3.1% across the three camps. A joint UNICEF,UNHCR, and WFP survey in late 2008 showedfurther improvements in the nutrition situation,with the prevalence of global acute malnutri-tion decreasing to 1.9% and with no severeacute malnutrition (SAM). The surveyconducted in September 2010 showed thatthese improvements have been sustained, with2% prevalence of global acute malnutrition.

Data on camp population, growth monitor-ing, and nutrition programmes were exportedfrom the UNHCR HIS database. The dataincluded camp-specific information about themonth of reporting, total camp population, andpopulation size by age group (less than 5 yearsof age, 5 years of age or older). Nutritionprogramme data included the number of chil-dren admitted to feeding programmes, thenumber remaining enrolled in the programmeat the end of each month, mean length of stay,average weight gain in therapeuticprogrammes, and reason for discharge (recov-

1 Tappis. H et al (2012). United Nations High Commissioner

for Refugees feeding programme performance in Kenya

and Tanzania: A retrospective analysis of routine health

information systems. Food and Nutrition Bulletin, vol. 33,

no. 2 ©2012, The United Nations University2 Mid Upper Arm Circumference

ery, death, default, or referral). Information onnutrition programme admissions and enrolmentwas combined with population data to calculateadmission rates and enrolment proportionswhere possible. Analysis was limited to childrenunder five years who were admitted to supple-mentary or therapeutic feeding programmes formalnutrition.

ResultsSFPsThere were a total of 39,899 new SFP admis-sions in Kenya and Tanzania between January2006 and May 2009, accounting for approxi-mately 63% of SFP admissions reported in theUNHCR HIS. Of these, 95% of admissions (n =37,741) were in Kenya camps and only 5% (n =2,158) were in Tanzania. Globally, readmissions,defined as moderately malnourished childrenenrolling in SFP within 1 month of successfuldischarge, account for 8.6% of all SFP admis-sions. In Kenya and Tanzania, the proportionsof readmissions were lower at 5% and 2%,respectively.

Although monthly growth monitoring atten-dance was much higher in Tanzania (78%) thanKenya (42%), children in Kenya were 11.60 (95%CI, 10.61–12.68) times more likely to be admit-ted to SFPs and 9.77 (95% CI, 8.49–11.25) timesmore likely to be enrolled in SFPs than childrenin Tanzania. In Kenya, SFP performance consis-tently exceeded all UNHCR performancestandards, with the exception of the averagelength of enrolment in SFPs, which ranged from11 to 20 weeks (standard < 8 weeks). All Kenyancamps had recovery rates above 90% (standard>75%), mortality rates at 0.1% or below (stan-dard <3%), and default rates below 8%(standard < 15%). All camps in Tanzania alsomet overall exit rate standards, with the excep-tion of Kigoma Camps, where an average of 20%of cases were referred for further treatment, andNgara Camps, where an average of 19% of exitswere due to default and 1% to death.

Table 1 summarises SFP performance bycountry and camp. With the exception of lengthof enrolment, the HIS data showed improve-ments over time in all SFP performanceindicators in Kenya camps. Between 2006 and2009 in Kenya, the average recovery rateincreased from 89% to 96% and the averagedefault rate decreased from 7% to 2%. InTanzania, the average recovery rate was lowestat 69% in 2007 (and 92% in 2009), and theproportion of exits due to default increased overtime, with a default rate of 13% in 2008.

TFPsBetween January 2006 and May 2008, there werea total of 4,347 new admissions of children underfive years to TFPs in Kenya and Tanzania camps.

Research

Settlements of new arrivalsin the outskirts of Dadaab

Mélo

dy T

ondeur

and S

ara

h S

tyle

, Kenya,

UN

HCR,

2011

28

Research

defaulting by increasing accessibility,addressing opportunity costs, or improv-ing reporting.

ConclusionsInferences from the comparison of HISadmissions and enrolment data suggestthat increasing admission and enrolmentin SFPs was successful in preventingcases of severe malnutrition in theDadaab Camps. Modifications to routinedata collection and reporting, such asinclusion of the proportion of moderateand severe wasting cases detected atgrowth monitoring, follow-up ondefaulter and referral cases, and addi-tional information on Community-basedManagement of Acute Malnutrition(CMAM), could also shed more light on

Admissions in these two countries accounted for approxi-mately 30% of TFP admissions reported in all countriescovered by the UNHCR HIS during this time period. Ofthese, 79% (n = 3,417) were for acute wasting and 21% (n =930) were for oedema (OR, 3.83; 95% CI, 3.55–4.12). In generalTFP admissions and enrolment rates in both Kenya andTanzania decreased over time, and higher admission andenrolment rates were observed in Kenya than in Tanzania.Acute wasting admissions were substantially higher in Kenya(n = 3,014, 70% of all admissions across the two countries)than in Tanzania (n = 403, 9% of all admissions). The samewas true for oedema, with 690 admissions (16% of all admis-sions) and 240 admissions (5% of all admissions) reported inKenya and Tanzania, respectively.

TFP performance met UNHCR standards for both acutewasting and oedema cases at the country level, with theexception of mean length of stay for acute wasting cases inTanzania, which was 34 days (standard < 30 days). As Table2 shows, the main factor limiting the recovery rate in TFPswas referral. In Kenya, Hagadera was the only camp withaverage recovery rates below the standard of 75%; referralsaccounted for a high proportion of both acute wasting andoedema exits during 2007 in Kenya. In Tanzania, below-average recovery rates in Kigoma Camps were associatedwith an increasing proportion of acute wasting referralsover time and a high proportion of oedema deaths in 2007 (4of 23 cases).

In Kenya camps, the average TFP enrolment periodranged from 22 to 27 days for acute wasting and from 23 to27 days for oedema (standard < 30 days). In Tanzania, theaverage enrolment period was longer, ranging from 25 to 43days for acute wasting and from 21 to 34 days for oedema.Most camps met the > 8 g/kg/day standard or averageweight gain, and average daily weight gain was substan-tially greater in Kenya than in Tanzania for children withboth acute wasting and oedema. Camps that did not meetUNHCR standards included the Ngara Camps in Tanzania(average weight gain of 5 g/ kg/day for children with acutewasting), Dagahaley Camp in Kenya, and the Kibondo andKigoma Camps in Tanzania, all which reported averageweight gains of 7g/kg/day for children with oedema.

DiscussionDuring the 3 and a half year period that was analysed usingHIS data, close to 45,000 malnourished refugee childrenunder five years were treated in UNHCR-supported selec-tive feeding programs in Kenya and Tanzania. With averagerecovery rates of 77.1% and 84.6% in the therapeutic andsupplementary feeding programmes, respectively, mortalityrates of less than 1%, and an average readmission rate below5%, the HIS data suggest that selective feeding programmeshad a beneficial effect on enrolled children. Even the campswith higher SFP enrolment rates had very low TFP enrol-ment rates (≤ 0.5%), which suggests that growth screeningand SFP objectives are being met and that SAM is relativelywell controlled in both the Kenya and the Tanzania camps.Any comparison and analysis of trends, however, must takeinto account the myriad of contextual factors that influencenutritional status, feeding programme size, and coverageand outcomes of targeted feeding programmes.

LimitationsConclusions are constrained by several factors. One chal-lenge is that HIS data on the SFP programme includes localpopulations that are enrolled in feeding programmes but arenot represented in denominators; thus SFP coverage amongrefugee populations may be overestimated. Average weightgain and duration of enrolment for discharged cases are thestrongest indicators of TFP performance. However, thesemeasures are also the most complicated to accurately meas-ure and are prone to omissions and reporting errors, whichdecrease their reliability as performance measures. In addi-tion, defaulters limited recovery rates in both Kenya andTanzania. However, there is no mechanism in the HIS fortracking the reason for default or outcomes of children whodefault. Consequently, it is difficult to determine whetherprogramme strategies should prioritize efforts to prevent

Location No. ofnew SAMadmissions

Averagelength ofenrollment(days)

Averageweightgain(g/kg/day)

Discharge category – mean % (range)

Recovered Death Default Referral

Acute wasting admissions

Kenya camps

Dagahaley 747 22 12 80 (66, 90) 8 (8, 9) 4 (2, 5) 8 (0, 20)

Hagadera 884 24 11 67 (38, 98) 5 (2, 7) 3 (0, 7) 24 (0, 57)

Ifo 999 22 15 77 (57, 92) 6 (0, 10) 4 (2, 7) 13 (1, 31)

Kakuma 384 27 11 87 (84, 90) 3 (2, 4) 10 (7, 12) 1 (7, 12)

Kenya total 3,014 24 12 78 (63, 93) 6 (3, 6) 5 (3, 7) 11 (3, 7)

Tanzania camps

Kasulu camps 79 25 11 84 (80, 90) 2 (0, 8) 8 (0, 15) 6 (3, 10)

Kibondo camps 102 43 14 87 (85, 100) 5 (0, 7) 0 (0, 0) 8 (0, 10)

Kigoma camps 122 29 10 50 (23, 60) 9 (4, 15) 11 (5, 22) 31 (23, 45)

Ngara camps 100 38 5 79 (74, 95) 1 (0, 2) 20 (5, 50) 0 (0, 0)

Tanzania total 403 34 10 75 (65, 81) 4 (4, 5) 10 (5, 18) 11 (10, 13)

Oedema admissions

Kenya camps

Dagahaley 215 24 7 79 (68, 89) 4 (2, 8) 6 (5,8) 11 (0, 25)

Hagadera 302 23 10 57 (27, 79) 5 (0, 7) 7 (0, 14) 31 (1, 63)

Ifo 106 27 16 86 (67, 97) 10 (3, 33) 3 (0, 5) 1 (0, 2)

Kakuma 67 23 9 92 (88, 95) 1 (0, 1) 1 (0, 1) 7 (3, 13)

Kenya total 690 24 11 78 (70, 87) 5 (5, 14) 4 (2, 6) 2 (1, 25)

Tanzania camps

Kasulu camps 56 21 9 80 (75, 84) 6 (0, 14) 0 (0,0) 14 (2, 25)

Kibondo camps 24 31 7 79 (50, 100) 15 (0, 25) 6 (0, 25) 0 (0,0)

Kigoma camps 111 32 6 60 (36, 66) 7 (0, 17) 9 (7, 13) 23 (4, 51)

Ngara camps 49 30 8 86 (75, 94) 7 (0, 17) 7 (0, 25) 0 (0,0)

Tanzania total 240 28 7 77 (70, 83) 9 (1, 15) 6 (3, 9) 9 (6,19)

Table 2: TFP performance, January 2006–May 2008

areas for improved programme quality inboth the Tanzania and the Kenya camps.In addition, linking nutritional surveil-lance data from the HIS with otherrelevant information sources (e.g., nutri-tion surveys, household food economystudies, local market surveys, and jointUNHCR/WFP assessment reports)would allow users to triangulate theresults and generate a more comprehen-sive understanding of the predictors ofincreased admission and enrolment needswithin selective feeding programmes.UNHCR has made the first step towardachieving this aim through the recentlaunch of an online version of the HIS(called ‘webHIS’ http://his.unhcr.org).This integrated platform hosts publichealth data from multiple sources.

SFP indicator

No. of newadmissions

Average length ofenrollment (wk)

Discharge category – mean % (range)

Recovered Death Default Referral

Kenya camps

Dagahaley 10,178 13 94 (91, 99) 0 (0, 0) 5 (1, 7) 1 (0, 3)

Hagadera 16,782 12 93 (92, 97) 0 (0, 0) 3 (1, 6) 2 (2, 3)

Ifo 8,916 11 92 (86, 98) 0 (0, 0) 4 (1, 8) 3 (0, 6)

Kakuma 1,865 20 89 (83, 96) 0 (0, 0) 8 (2, 12) 3 (2, 12)

Kenya total 37,741 14 92 (89, 96) 0 (0, 0) 5 (2, 7) 2 (2, 7)

Tanzania camps

Kasulu camps 519 19 78 (64, 94) 0 (0, 0) 3 (0, 5) 18 (6, 30)

Kibondo camps 413 11 82 (79, 98) 0 (0, 2) 3 (0, 4) 15 (2, 18)

Kigoma camps 399 14 68 (48, 98) 0 (0, 0) 12 (0, 21) 20 (2, 38)

Ngara camps 827 11 77 (66, 94) 1 (0, 2) 19 (2, 34) 4 (0, 5)

Tanzania total 2,158 14 76 (69, 92) 0 (0, 1) 9 (0, 13) 14 (7, 17)

Table 1: SFP performance, January 2006 – May 2009

SFP, supplementary feeding programme

TFP: Therapeutic Feeding Programme, SAM: severe acute malnutrition

29

Research

Scaling up ORS and zinctreatment for diarrhoeareduces mortality

With most countries still not on track to achievemillennium development goal 4 - that of reducingchild mortality by two thirds from 1990-2015, a

recent editorial in the British Medical Journal (BMJ) highlightsthe need to tackle the treatment of diarrhoea. The authorsargue that investment in the treatment of diarrhoea with oralrehydration salts (ORS) plus zinc is one of the best opportunitiesto achieve such rapid impact on mortality. Acute diarrhoea isthe second biggest cause of death in children worldwide caus-ing 1.2 million deaths each year. Treatment with ORS and zinccould rapidly and cost efficiently avert most of the deaths notprevented by vaccines.

A systematic review estimated that universal coverage withORS would reduce diarrhoea related deaths by 94% whileanother review estimated that, in zinc deficient populations,zinc treatment reduces diarrhoea related deaths by 23%. Yetonly about 30% of children with diarrhoea in high burdencountries receive ORS, and less than 1% receives ORS plus zinc.The use of ORS has stagnated globally since 1995.

Scaling up the provision of zinc and ORS could rapidlyreduce child mortality for four reasons:• Although it has been almost eight years since the WHO

recommended combination treatment with zinc and ORS, few countries have implemented basic interventions to increase the currently low use of adjunctive zinc. Such interventions would include marketing zinc to caregivers and distributing it in large volumes through both public and private facilities.

• Children with diarrhoea can be reached and given appropriate treatments such as antibiotics and anti-diarrhoeal agents. Simply switching the treatments children receive, which is less challenging than trying to change caregivers treatment seeking behaviour, could therefore drive substantial increases in ORS and zinc coverage.

• In contrast to treatments for malaria or pneumonia, effec-tive treatment of diarrhoea does not need to be carefully targeted to selected children for whom a definitive diagnosisis made. A strategy of ‘flooding the market’ with ORS and zinc – distributing them through all outlets where caregivers seek treatment – could be pursued safely.

• A full course of zinc and ORS treatment costs less than $0.50, and the marketing, training and distribution necessary to drive product uptake could also be imple-mented at comparatively modest cost. Moreover, public funding for procurement of zinc and ORS in many countrieswould be further moderated by the fact that most treatmentfor diarrhoea is delivered through the private sector and paid for out of pocket.

The authors ask the question – what will it take to scale up thedelivery of ORS and zinc for the treatment of diarrhoea world-wide? They suggest that an essential first step will be to focusattention on the problem. For the first time, the ten countrieswith the highest burden of diarrhoea have developed ambi-tious plans to scale up coverage of effective treatments ofdiarrhoea and pneumonia.

The authors conclude that dedicated resources and practi-cal operational support are now needed to translate thosecountries plans into success. ORS and zinc treatment for diar-rhoea should appeal to any donor seeking a high return oninvestment and the ability to have a rapid effect on childmortality. Contributions from early donors could be leveragedwith other private and public contributions to realise adramatic reduction in child deaths from diarrhoea and a furtherleap towards achieving the millennium development goals.

Summary of editorial1

Treatment of undernutritionin urban Brazil

1 Sabot.O et al (2012). Scaling up oral rehydration salts and zinc for

the treatment of diarrhoea is cheap and effective and could accelerate

reductions in child mortality. BMJ 2012; 344doi:10.1136/BMJ.e940.

Published February 2012

Summary of research1

In Brazil, 6% of children aged lessthan 5 years suffer from height-for-age (HAZ) deficit, while the

prevalence of this condition is higher(8.2%) among the poorest sector of thepopulation. Moreover, the prevalence ofHAZ and weight-for-age (WAZ) deficitsamong children in the 5-9 years agerange is 6.8% and 4.1% respectively. Inaddition, population growth in the lastthree decades has been intense leadingto massive urbanisation, competition forresources and urban poverty. This hasled to increased violence, undernutritionand family breakdown. In order totackle this problem, the FederalUniversity of Sao Paolo created theCentre of Nutrition Recovery andEducation (CREN) in 1994 with theobjectives of developing: i) methodolo-gies for the nutritional recovery ofchildren with respect to weight and, inparticular, stature and ii) approaches fordirect intervention in the complex socio-economic and familial situation ofimpoverished urban areas.

More than 1000 undernourished chil-dren are assisted annually by CREN in aday hospital (n=50), in an out-patientclinic (n=700) or directly in the commu-nity (n=300) depending on the degree ofundernutrition and accessibility of thecentre. Children accepted into the day-hospital regime originate either fromimpoverished communities or frompublic health units and comprise thosewho have been diagnosed with primaryundernutrition and who cannot be reha-bilitated in out-patient clinics or byhome visits. The routine weekdayschedule at CREN comprises five dailymeals together with recreational andeducational activities (including hygienepractices), sleeping time after lunch andperiodic medical examinations. Sickchildren are treated as appropriate.Children receive protein-rich food aswell as vitamin supplements andprophylactic doses of iron until they are24 months old. Children are dischargedand redirected to day nurseries or pre-schools after effective recovery fromundernutrition, i.e. when the HAZ andWAZ parameters are normalised (z-score >-1). Weight-for-height z scoresand mid upper arm circumference(MUAC) are not used.

A recently published study byresearchers from CREN set out to build alife table and determine the factorsrelated to the time of treatment of under-nourished children at a nutritionalrehabilitation centre in Sao Paolo. Thesubjects for the study were undernour-ished children (n=228) from the

southern slums of Sao Paolo who hadreceived treatment at CREN under a dayhospital regime between the years of1994 and 2009. The average age of thechildren at admission was 24.45 months.

Nutritional status was assessed fromWAZ, HAZ and Body Mass Index (BMI)-for-age z-scores, while neuro-psychomotor development was classifiedaccording to the milestones of childhooddevelopment. Life tables, Kaplan-Meiersurvival curves and Cox multiple regres-sion models were employed in dataanalysis. The Kaplan-Meier curves ofsurvival analysis showed statisticallysignificant differences in the periods oftreatment at CREN between childrenpresenting different degrees of neuro-psychomotor development (log-rank =6.621; P = 0.037). Estimates based onmultivariate Cox model revealed thatchildren aged ≥ 24 months at the time ofadmission exhibited a lower probabilityof nutritional rehabilitation (hazard ratio(HR) = 0.49, P = 0.046) at the end of theperiod compared with infants aged upto 12 months. In spite of this, it wasobserved that increases in HAZ weregreater than in WAZ. The positive effectswere observed even in children aged ≥24months at the time of admission,although the magnitude of such recov-ery was inferior to that shown byyounger individuals. Children present-ing slow development were betterrehabilitated in comparison with thoseexhibiting adequate evolution (HR =4.48; P = 0.023). The authors believe thatthe positive results obtained with chil-dren presenting with delayedneuro-psychomotor development areassociated with greater severity of theirnutritional condition, as well as theirlower age at admission compared withthe other group. No significant effects ofsex, degree of undernutrition or birthweight on the probability of nutritionalrehabilitation were found.

The authors concluded that age andneuro-psychomotor developmental statusat the time of admission to CREN arecritical factors in determining the dura-tion of treatment. These findingsdemonstrate the importance of earlynutritional recovery and indicate thatthe treatment should preferentially startbefore the age of 24 months. Linearcatch-up has also been achieved forolder children at CREN, but at a lowerspeed and after an extended period oftreatment.

1 Fernandes M et al (2011). A 15 year study on the

treatment of undernourished children at a nutrition

rehabilitation centre (CREN), Brazil. Public Health

Nutrition: 15 (6), pp 1108-1116

30

Agricultural interventions toimprove nutritionalstatus of children Summary of review1

Asummary of a recent systematic reviewon the effectiveness of agriculturalinterventions that aim to improve

nutritional status of children has recently beenpublished. In order to be included in thereview, studies had to investigate agriculturalinterventions with the explicit goal of improv-ing the nutritional status of children.Interventions included were bio-fortification,home-gardening, aquaculture, small scale fish-eries, poultry development, animal husbandryand dairy development. The review onlyincluded studies from the published andunpublished literature that were producedafter 1990, were written in English, wereconducted in a middle or low income country,reported an effect on at least one of the selectedoutcome indicators, and used a credible coun-terfactual method.

Twenty three studies were selected for thereview. The reviewers investigated the effectsalong all the main steps of the causal chainrunning from the agricultural intervention tothe final goal of reducing under-nutrition,specifically participation in the programme bypoor people, household income, compositionof the diet, iron and vitamin A intakes, andprevalence of stunting, wasting and under-weight.

The studies reported no information onparticipation rates or characteristics ofprogramme participants. The interventionsreviewed had a positive effect on productionand consumption of the agricultural goodspromoted, but reviewers found no evidence ofa change in total household income and littleevidence of a change in the overall diet of poorpeople. No evidence was found of an effect oniron intake but there was some evidence of apositive effect on the absorption of vitamin A.Very little evidence existed of an effect on theprevalence of undernutrition. Of eight studiesreporting undernutrition rates, only one founda statistically significant effect on prevalence ofstunting, whereas three studies found a posi-tive effect on prevalence of underweight andtwo found a positive effect on wasting.

Many of the studies reviewed had method-ological weaknesses. Post hoc powercalculations showed that most studies wereunlikely to find an effect on undernutritionrates if present, because of the small samplesizes used. Furthermore, factors such as healthenvironment and cultural practices may haveaffected the impact of the agricultural interven-tions reviewed independently of their efficacy.

1 Masset E et al (2012). Effectiveness of agricultural inter-

ventions that aim to improve nutritional status of children:

systematic review. BMJ February 2012, volume 433, pp 16

Feasibility and effectiveness ofpreventing child malnutritionwith local foods in Kenya

Summary of research1

The findings of a study to establish theoperational feasibility and effective-ness of using locally available foods to

prevent malnutrition and improve childgrowth in Kenyan children have recentlybeen published. The study took place in ruralvillages in the arid lands of eastern Kenya,which have a high prevalence of child malnu-trition as identified by a survey conducted inMarch 2009. The study sites were in YattaDistrict of Eastern Province, Kenya.

The diet in the area is maize-based andchildren are commonly fed a porridge madefrom maize flour with a low energy densityso that the growth of young children faltersas early as 3-4 months of age. Animal-sourcefoods are rarely fed to children. One of thereasons for carrying out the study was recog-nition that aid programmes using commer-cially prepared products are not sustainableand that distribution of local foods combinedwith an education programme could poten-tially lead to improved local production andresult in better complementary feeding.

The study employed a quasi-experimentaldesign with an intervention group of childrenin all villages in one region and a non-inter-vention group of children in all villages in anadjacent region. Both sub-locations aregoverned by the same local chief and havecommunity health workers who participatein the screening of the households with chil-dren under five years of age for acutemalnutrition. All children who lived in thetwo sub-locations were eligible to participateif they were 6-20 months of age at the onset ofthe study and had a weight-for-height z score(WHZ) ≥-2. Children with WHZ <-2 wereenrolled into a community-based therapeuticfeeding programme and were excluded fromthe study. The intervention was the distribu-tion of a monthly food ration for the indexchild, a separate family ration, and groupeducation on appropriate complementaryfeeding and hygiene.

Children in the intervention and non-intervention groups had similar baselineanthropometric measures. The caregivers in

the intervention group confirmed that theintended amounts of food supplements werereceived and child nutrient intake improved.During the 7-month intervention period therewere significant group differences in pre-postz-score changes between the intervention andnon-intervention groups for weight-for-age(0.82, P<0.001) and WHZ (1.19, P<0.001), butnot for height-for-age (0% v. 8.9%, P = 0.0002)and underweight (6.3% v. 23.0%, P <0.0001).Infectious morbidity was similar in bothgroups.

The likely explanation for the substantialdecline in height-for-age Z-score in the inter-vention group as well as the non-interventiongroup was that the intervention period wastoo short to have an impact. A contributingfactor may have been the age of the childrenwhen the intervention started. Although thestudy was planned with the intent to inter-vene as early as possible after complem-entary foods are introduced, the mean age atstudy entry was 14 months in the interven-tion group and the prevalence of growthstunting was already 27%. As anticipated,food was being shared with other familymembers, mostly other children.

However, intra-household sharing of foodwas taken into consideration in the interven-tion design which allowed for an extra 920Kj/d per study child. Therefore, despite theleakage of the supplements, the index childreceived a sufficient amount of the food tosignificantly improve the nutrient intake andprevent the development of wasting.

The findings suggest that the distributionof locally available foods combined with briefgroup nutrition education sessions is opera-tionally feasible and improves child weightgain and decreases acute malnutrition inKenyan children. However, it remains to beseen whether such an approach has greaterpotential for sustainability through increas-ing production and availability of local foodsin the longer-term.

1 Tomedi.A (2012). Feasibility and effectiveness of supple-

mentation with locally available foods in prevention of

child malnutrition in Kenya. Public Health Nutrition: 15

(4), pp 749-756

ResearchAngelo

Tom

ed,

Kenya,

2010 One of the researchers

(Rebecca Ashton)weighs a child

31

The state of thehumanitariansystem Summary of report1

Arecent report by ALNAP presentsa system-level mapping andanalysis of the performance of

international humanitarian assistancebetween 2009 and 2011 building on anearlier pilot report published covering2007 and 2008. The method for themapping involved a number of compo-nents:• An evaluation synthesis analysing

the findings from recent analytical literature

• A series of interviews with key informants among humanitarian practitioners and policy-makers

• Field studies in Kenya and South Sudan

• A compilation of descriptive statistics,mapping the system’s organisational components

• Analysis of humanitarian financial flows

• A series of global surveys on humanitarian performance indicators that solicited views of humanitarian actors and stakeholders across a rangeof field settings

• A review of data on affected popula-tions to estimate global humanitarian need.

There are now some 4,400 non-govern-mental organisations (NGOs) worldwideundertaking humanitarian actions on anongoing basis. Yet the system is stilldominated in terms of operational pres-ence and resource share by the smallgroup of United Nations (UN) humani-tarian agencies, the InternationalMovement of the Red Cross and RedCrescent (ICRC) and five international‘mega’ NGOs whose combined humani-tarian expenditure in 2010 exceeded 2.7billion dollars. The population of human-itarian workers continued to grow in thepast two years but not as fast as previ-ously – combined growth rate of 4% in2009-10. In 2010 there were an estimated274,000 humanitarian workers world-wide. Controlling for the surge in fundsfor Haiti in 2010 and adjusting for infla-tion, the long-term upward trend inhumanitarian funding continued at 1%on average for 2009-2010.

Although humanitarian funding hasshown a ten year rising trend, the major-ity of actors surveyed perceived fundingin their settings to be insufficient, espe-cially for the sectors of protection andearly recovery. On average, fundingcoverage against needs (as stated in

Research

Nutritional support services and HIVin sub-Saharan African countries Summary of research1

On the basis of the latest WHO criterion for HIVtreatment initiation (CD4 cell count ≤350cells/mm³), anti-retroviral therapy (ART) was

accessed by an estimated 37% of sub-SaharanAfricans in need of ART by the end of 2009. A signifi-cant proportion of people living with HIV/AIDS in theregion are simultaneously affected by nutritional defi-ciencies. Cross-national population-based datasuggest that the prevalence of mild and moderatemalnutrition among adults living with HIV/AIDS is15.4% and 10.3% respectively, in sub-Saharan Africa,and is also elevated among young children living inAIDS-affected households. The disproportionateburden of HIV/AIDS and nutritional deficiencies insub-Saharan Africa is of particular concern as eachcondition may perpetuate and aggravate the severityof the other.

In order to mitigate the adverse effects of foodinsecurity on HIV outcomes, international organisa-tions such as the WHO, the Joint United NationsProgramme on HIV/AIDS, the World Food Programme(WFP) and major HIV initiatives such as the USPresident’s Emergency Plan for AIDS Relief (PEPFAR)recommend the integration of nutritional supportservices, specifically nutrition assessment, educationand counselling, into HIV/AIDS treatment and careprogrammes. Accordingly, PEPFAR have allocated$US130 million to support the integration of HIV andnutritional support services. Since 2006, PEPFARimplementation guidelines have recommended theprovision of nutritional evaluation and food supportto all orphans and vulnerable children, to HIV-positivepregnant and lactating women and to all adultsreceiving ART with clinically defined malnutrition.However, little is known about the extent to whichnutritional support services are available in HIV careand treatment programmes across sub-SaharanAfrica.

A study conducted in 2008 (but only publishedrecently) set out to examine the availability of nutri-tional support services in HIV care and treatment sitesacross sub-Saharan Africa. The study conducted across-sectional survey of sites providing ART in ninesub-Saharan African countries. Outcomes includedavailability of: (i) nutritional counselling, (ii) micronu-trient supplementation, (iii) treatment for severemalnutrition and (iv) food rations. Associations withhealth system indicators were explored using bivari-ate and multivariate methods.

A total of 336 HIV care and treatment sites, serving467,175 enrolled patients, were surveyed. Of these,303 sites offered some form of nutritional supportservice. Nutritional counselling, micronutrient supple-mentation, treatment for severe acute malnutrition(SAM) and food rations were available at 98%, 64%,36% and 31% of sites respectively. In multivariateanalysis, secondary or tertiary care sites were morelikely to offer nutritional counselling (adjusted OR(AOR): 2.2, 95% CI 1.1, 4.5). Rural sites (AOR: 2.3, 95% CI1.4, 3.8) had increased odds of micronutrient supple-mentation availability. Sites providing ART for >2years had higher odds of availability of treatment forSAM (AOR: 2.4, 95% CI 1.4, 4.1). Sites providing ART for>2 years (AOR: 1.6, 95% CI 1.3, 1.9) and rural sites(AOR: 2.4, 95% CI 1.4, 4.4) had greater odds of foodration availability.

The study authors made a number of observationsabout the results. Although 90% of sites had at leastone form of nutritional service, the type of nutritionalsupport service reported to be available variedsubstantially, including by country, geographicalcontext (urban/rural) and site-level characteristics(type of site, age of programme, reported presence ofa nutritionist on staff ).

They also observed the encouraging fact thatnutritional counselling to adults and to pregnant andlactating mothers was reported to be available at 95%and 91% of sites, respectively. Nutritional counsellingis considered to be a cornerstone of HIV treatmentsupport, recommended by multiple internationalguidelines. It is also important for supporting HIV-affected households. Studies in sub-Saharan Africahave found that HIV-exposed but uninfected childrenhave poorer growth compared with HIV-uninfectedchildren, possibly because of reduced breastfeedingpractices by HIV-infected mothers.

Although some type of micronutrient supplemen-tation was reported to be available at 64% of sites, thismostly included vitamin A and iron supplementation.The fact that multivitamin and mineral supplementa-tion was reported to be available at only 25% of HIVcare and treatment sites suggests that there is roomfor improvement. Studies have found that manypeople living with HIV/AIDS suffer various micronutri-ent deficiencies and have increased micronutrientneeds. However, evidence of the effectiveness ofmicronutrient supplementation on HIV-infectedadults and children remains limited, and findingshave been varied.

Food rations were reported to be available at 31%of HIV care and treatment sites. Studies have notedthat provision of food rations may be associated withimproved medication adherence and retention in HIVcare and treatment in sub-Saharan Africa, which isboth a major and persistent challenge of scale-up inthe region.

The fact that treatment for SAM is more likely to beavailable at sites that had been providing ART formore than 2 years may reflect the increased logisticcapacity of longer established sites. The study alsofound that availability of treatment for SAM hasexpanded significantly over a short period whichsuggests that ART treatment and care sites may bedistributing Ready to Use Therapeutic Foods (RUTF).This is promising given that RUTF use has been asso-ciated with rapid weight gain, improvements inphysical activity and increased uptake of voluntarycounselling and testing (VCT) and ART servicesamong HIV-positive individuals with SAM.

The authors acknowledge that a major limitation ofthe study is that information was only obtained onavailability of different types of service and not on theaccess, quality, comprehensiveness or coverage of theindicated services at the sites. They conclude thatfurther efforts are therefore urgently needed to deter-mine uptake, quality and effectiveness of these servicesand their impact on programme outcomes such asadherence to ART, retention in care and survival.

1 Anema.A (2011). Availability of nutritional support services

in HIV care and treatment sites in sub-Saharan African

countries. Public Health Nutrition: 15 (5), 938-947

32

appeals) stayed effectively static in 2009-10compared to 2007-8, rising to 54% from 53%,with unequal growth across sectors. Most fund-ing continues to go to a small number ofprotracted crises: Sudan, Pakistan, Ethiopia,Somalia, Democratic Republic of the Congo,Afghanistan and the occupied PalestinianTerritories, and to high-profile natural disasters,such as Haiti. The systems poor showing incoverage and sufficiency is largely a conse-quence of financial, human and materialresources not growing fast enough to keep pacewith rising needs. The growth in needs is due toa combination of factors including increases inclimate-related natural disasters, agency effortsto account for the true total of affected popula-tions, and appeals for the humanitarian systemto take on more activities in recovery, prepared-ness and development.

Most survey respondents from the interna-tional system saw a moderate improvement inthe quality of needs assessments in the past twoyears. Where operations were found to be lessrelevant, reviews cited the inability to meet thefull spectrum of need and a weak understand-ing of local contexts as key reasons. Surveysclearly found that humanitarian organisationshad failed to consult with recipients in theirsetting or to use their input in programming.This deficit could be addressed in the nearfuture by technical advances in methods ofneeds assessment. Humanitarian organisationsare increasingly focused on more comprehen-sive inclusive and participatory needsassessments, and this reporting period saw aserious drive from the centre of the humanitar-ian system in this regard. Examples include theInter Agency Standing Committee (IASC)’sNeeds Assessment Taskforce (NATF), resultingin a new set of tools.

Most interventions were found to be effec-tive or partially effective in terms ofachievement against projected goals or interna-tional standards, the avoidance of negativeoutcomes and/or the receipt of positive feed-back from aid recipients. Where overalleffectiveness has been questioned, the keyreasons were time delays and poorly definedgoals. Each major emergency during the report-ing period had a mixed review in terms ofeffectiveness. In particular, the response in theHorn of Africa was found to be abjectly slow ata systemic level, with significant disconnectsbetween early warning systems and response,and between technical assessments and deci-sion-makers.

The key elements of humanitarian reformeffectively came of age during this reportingperiod, as the cluster system, the Central

Emergency Response Fund (CERF) and coun-try-level pooled funding all underwentfive-year evaluations. Each evaluation revealedthat these instruments have become accepted asthe new and generally improved means ofoperation. The lynchpin role of theHumanitarian/Resident Coordinator (HC/RC)continues to be a focus of attention. Questionsremain about skills prioritisation and merit-based selection. Survey responses suggest thesomewhat more nuanced finding that the great-est dissatisfaction within the system is with thelack of overall coordination at country level,rather than with the individual RC/HC.However, both aspects are found to be wantingby different actors in the system.

Most of the aid recipients surveyed felt thatthe foremost way in which humanitarian organ-isations could improve would be to “be faster tostart delivering aid”. At global level, the CERFgets high marks for timeliness in fundingdisbursements, but the timeliness of donorcontributions overall declined between 2007-8and 2009-10. Response rosters have improvedthe capacity of agencies in operations and coor-dination to fill key posts in timely fashion.However, high staff turnover and the ongoingchallenge for humanitarian agencies in prepar-ing staff to understand the political and culturalcontext, often at short notice, greatly under-mine success in this area.

The issue of monitoring, identified as a keyweakness in the pilot study, has still receivedlittle attention and states are still notably absentfrom evaluating their own responses or partici-pating in joint evaluations with counterparts.

A significant rise in capacity of NationalDisaster Management Authorities has rein-forced the system’s engagement withgovernments of affected states and heightenedrecognition of the need to support their priori-ties more effectively in natural disasters.However, much practical work remains to bedone to help strengthen host-country coordina-tion structures and response capacities. Therelationship between donors, non-governmentactors and recipient states can often be strained.For instance, the focus of some governments onsovereignty and self-reliance in the face ofdisasters has seen increasing refusals to issuethe ‘standard emergency’ appeals that havetraditionally triggered the internationalsystem’s response. For their part, host govern-ment representatives are frustrated by the‘artificial’ division between relief and develop-ment aid in the international aid architecture.

Despite the increasing importance of localpartnerships in highly insecure settings, thereremains an underinvestment in the capacities oflocal operational partners. Interviews and eval-uations noted that national organisations areoften working at or beyond their maximumoperational capacity and find the additionalpressure to meet a variety of international stan-dards challenging, if not impossible, givenavailable resources and time. Yet local capacity-building remains one of the hardest areas toraise funds for in non-emergency periods.While there are increased opportunities forlocal organisations to apply for pooled funding,there is very little bilateral funding beingdirected towards them.

Initiatives by DAC and its members, such asInternews, Frontline SMS, BBC Media Action

and others, have greatly increased engagementwith the views of affected populations, as wellas providing information as a vital form of aidin emergencies.

Clusters and country-level pooled funds arecredited with bringing larger volumes of fund-ing and contributing to stronger coordination,but at times can sacrifice speed for inclusive-ness. Some major donors, facing budgetpressures, became increasingly concerned withcost efficiency, or ‘value for money’. Challengesremain, however, in clarifying the concept andmaking meaningful comparisons betweencontexts.

Key innovations in humanitarian actionincluding use of cash and mobile communica-tions technology, reached a transformative scaleduring this period. The subject of innovationitself became a major area for action in thesystem, with new funds and mechanismsdesigned to study and support innovation inhumanitarian programming.

The extent to which actors converge aroundshared principles and goals seems to haveweakened. Many humanitarian organisationsappear to have willingly compromised a princi-pled approach in their own conduct throughclose alignment with political and militaryactivities and actors. The internal coherence ofthe humanitarian system has increasingly beentested, as a gulf widens between strict ‘tradi-tionalist’ humanitarian actors such as ICRC andMédecins Sans Frontières (MSF), and the rest ofthe system which is populated by multi-mandated organisations.

In cyclical and slow-onset disasters, the long-acknowledged disconnect between develop-ment and humanitarian programming hasfailed populations at risk. The concept of resil-ience may offer a basis for increasing coherence.

Major recommendations in the reportinclude the following:i) With the rise of the resilience agenda, it is

critical that new financing instruments are considered to provide the long-term, flexible financing that these broader non-relief interventions require.

ii) There is a need to deepen investments in contextual analysis and to engage aid recipients and local organisations more meaningfully in determining needs and programme design.

iii) Increased efforts must be made to collect and appropriate use data disaggregated by sex and age.

iv) To increase aid effectiveness, there is a need for more effective humanitarian leadership in crisis countries, preparednessand surge capacity for more rapid responseas well as investment in monitoring and the need for greater engagement in evalua-tions on the part of host states.

v) A focus on documenting good practice or achievements in collective, coordinated, principled approaches in crisis contexts would be valuable and would serve to support much-needed learning on the effective operationalisation of humanitarianprinciples.

1 ALNAP (2012). The State of the Humanitarian System. ODI,

London, July 2012. For access to the report as well as video

footage of the key report messages and the launch of the

report, visit:http://www.alnap.org/ourwork/current/sohs.aspx

Research

33

Concerns on Global HIV/AIDS, TB and Malaria Fundconflicts of interest Summary of letter1 and responses2,3

This article raises a number of issues about corpo-rate sponsorship in the public health sector and assuch is of interest to those working in the nutritionsector where ‘corporate capture’ is a muchdiscussed issue (Ed).

Recent letters in the ‘Round Table’ sectionof the World Health Bulletin debate theissues around funding by the Global

Health Fund of an education intervention byalcohol giant SABMiller. A first letter sets thescene in stating that SABMiller is a major beersupplier to approximately 34,000 licensedoutlets in South Africa and through them to anestimated 200,000 illegal outlets called‘shebeens’ that act as de facto distribution arms.The company has established an educationintervention that aims to minimise alcohol-related harm in men and reduce male violenceagainst women and children, as well as reducethe spread of HIV/AIDS. The Global Fund toFight AIDS, Tuberculosis and Malaria hasincluded SABMiller as a sub-recipient of itsRound 9 funding in support of the brewery’sTavern Intervention Programme for Men.

The authors argue that Global Fund supportfor this initiative is cause for concern because itreflects the successful attempt of a highly prof-itable industry to position itself as committed topublic health objectives. In reality, they argue,the liquor industry’s aggressive marketing of itsproducts is irrevocably linked with majorhealth harms throughout the world, in SouthAfrica in particular. Furthermore, as the impe-tus by government and non-governmentalorganisations (NGOs) to address alcohol-related harm in South Africa has increased,there has been an upsurge in efforts by theliquor industry to partner with governmentand public health agencies. Such partnershipslend legitimacy and provide a platform for theliquor industry to lobby against proposals toreduce the availability of alcohol, increase theprice of alcohol, through raising excise taxesand place restrictions on the marketing of alco-hol, despite the global evidence that thesemeasures are the most cost-effective way todecrease alcohol-related harm.

The authors argue that although aprogramme that aims to reduce HIV infectionas well as violence against women and children

is welcome, it is debatable whether men whoattend shebeens are the best target group for theintervention, whether a drinking establishmentis the best location and whether the educationalintervention itself is effective. Their experienceis that the liquor industry is inclined to supportalcohol interventions that have limited impacton drinking at a population level. Also, elicitingfunds earmarked for the global public good notonly provides the liquor industry with freeadvertising and mechanisms to achieve itsgoals, but also reduces available funding forless well-resourced organisations.

The authors state that calls for accountabilityand good governance relating to the GlobalFund and similar agencies have been increasingrecently and that such bodies should not fundorganisations with conflicted interests. Theyalso state that despite submission of a requestfor the Global Fund to reconsider the award toSABMiller, the funding is going ahead. Theyfear that the problem of ‘corporate capture’described in the field of nutritional research,has now spread to one of the largest healthfunders in the world.

The Global Fund were invited to respond tothis letter in the same issue and in so doingmade a number of points. The Global Fund, asa public-private partnership, encourages theprivate sector to engage in all aspects of itswork, ranging from mobilisation of resources,implementation of grants and governance offunds. The Global Fund does not endorse theactions, practices or policies of any corporationor industry beyond the field of the fight againstthese three diseases (HIV/AIDS, TB andMalaria). The model of the Global Fund isbased on the concept of country ownership.Countries determine their own programmaticpriorities and implementation strategies andsubmit requests for funding based on identifiedfunding gaps. One of the central channels ofcountry ownership is the Country CoordinatingMechanism, which is made up of representa-tives from both the public and private sectorsincluding governments, multilateral or bilateralagencies, NGOs, academic institutions, privatebusinesses and people living with the diseases.In its mandate as the Country CoordinatingMechanism, the Resource MobilisationCommittee of South Africa submitted a Round9 HIV proposal focusing on HIV preventionand care and support activities. The proposalwas reviewed by the Global Fund’s TechnicalReview Panel, an independent panel of interna-tional experts on health and development.

A portion of the Round 9 funding is directedtowards a cost-sharing programme that isimplemented jointly by SABMiller, the SouthAfrican Business Coalition on HIV/AIDS(SABCOH) and the Government of SouthAfrica. This programme focuses on the provi-sion of HIV counselling and testing, trainingand peer education in taverns and shebeensand psychosocial support for caregivers. TheGlobal Fund regards this cost-sharing

1 Matzopoulos. R et al (2012). Global Fund collusion with liquor

giant is a clear conflict of interest. Bulletin World Health

Organisation 2012; 90:67 69|doi:10.2471/BLT.11.0914132 Bampoe.V et al (2012). Response from the Global Fund.

Bulletin World Health Organisation

2012;90:70|doi:10.2471/BLT.11.0969903 Gilmore. A and Fooks.G (2012): Global Fund needs to

address conflict of interest. Bulletin World Health

Organisation 2012. 90:71-72|doi:10.2471/BLT.11.098442

Research

programme as an important endeavour tomobilise greater resources for the fight againstHIV and recognises the contribution thatSABMiller and SABCOHA can make in reach-ing at-risk populations with prevention andcare and support activities.

Researchers estimate that more than 85% oflocations where individuals meet new sexualpartners in South Africa are shebeens and otheralcohol-serving establishments. Studies haveshown that even brief interventions in bars andtaverns can result in reduced risky sexualbehaviours. The Global Fund considers menattending alcohol-serving establishments to bea key target population and sees these establish-ments as viable and dynamic locations forintervention. The Global Fund thereforesupports South Africa’s implementation strat-egy to collaborate with SABMiller andSABCOH as an innovative approach to addressthe HIV/AIDS epidemic. As is the case withother programmes it supports, the Global Fundwill monitor the performance of the grant andexpects an evaluation of the programme in linewith its system of performance-based funding.

A final letter in the round-table discussionre-emphasises the conflict of interest elementstating that at the heart of the problem is theapparent failure by both the Global Fund andthe Government of South Africa to recogniseand adequately address the potential conflictbetween corporate interests and public healthgoals. The authors suggest that the initial fail-ure to recognise this conflict of interest lies withthe South African Government, which enteredinto a partnership with SABMiller before theGlobal Fund funded this public-private part-nership. It is further argued that industrieswhose products are harmful to health areincreasingly attempting to enter into such part-nerships as part of their corporate socialresponsibility strategies. Furthermore, evidencesuggests that these corporate social responsibil-ity strategies are intended to facilitate access togovernment, co-opt NGOs to corporate agen-das, build trust among the public and politicalelite and promote untested, voluntary solutionsover binding regulation.

The authors conclude that the Global Fund isbeing naive in simply exempting tobacco andarms producers from its remit. The productssold by these corporations may be unique buttheir conduct is unlikely to be and these twoissues should not be confused. Whether acompany sells cigarettes or alcohol, its maingoal is to maximise shareholder returns.Policies that could reduce such returns aretherefore contrary to its interests.

A typical example of a shebeen inan informal settlement, Namibia

©W

HO

34

Research

Double Burden of obesity andmalnutrition in WesternSahara refugees Summary of published research1

There is growing recognition of a ‘doubleburden’ of malnutrition among popula-tions in both affluent and less-affluent

countries, i.e. the coexistence of undernutrition(e.g. stunting or underweight) with overweight,which has been observed at national and house-hold levels. At present, little is known aboutwhether undernutrition and overweight co-exist among refugees in protracted settings, orabout the proportion of refugee householdsthat may be affected by this double burden. Arecent study of refugees from Western Saharaaimed to use anthropometric data from aroutine UNHCR nutrition survey to investigatethe existence of the double burden of malnutri-tion in a refugee population highly dependentupon food assistance and living in a protractedemergency.

People from Western Sahara (also known asSahrawi) have lived as refugees in camps nearTindouf city, in southwest Algeria, an area witha harsh desert environment. Their situation isconsidered a protracted emergency, as there is astalemate to negotiations, with no sign of immi-nent resolution. Although accurate estimatesare not available, the host country estimatesthat there are 165,000 people living in fourcamps (Awserd, Dakhla, Laayoune and Smara),mostly dependent on food assistance frominternational organisations.

A two-stage household cluster survey withfour strata (one per camp) was conducted inOctober and November 2010 to collect nutritionindicators from children (<5y) and women ofchildbearing age (15–49y).The study sampled2,005 households, collecting anthropometricmeasurements (weight, height and waistcircumference) in 1,608 children (6–59 months)and 1,781 women (15–49y). The prevalence ofglobal acute malnutrition (GAM), stunting,underweight, and overweight was estimated inchildren. Stunting, underweight, overweightand central obesity (waist circumference ≥than80cm) prevalence was estimated in women. Toassess the burden of malnutrition within house-holds, households were first classifiedaccording to the presence of each type ofmalnutrition. Households were then classifiedas undernourished, overweight or affected bythe double burden if they presented memberswith undernutrition, overweight or both,respectively.

The study found the prevalence of GAM inchildren was 9.1%, 29.1% of children werestunted, 18.6% were underweight, and 2.4%were overweight. Among the women, 14.8%were stunted, 53.7% were overweight or obese,and 71.4% had central obesity. Central obesity(47.2%) and overweight (38.8%) in womenaffected a higher proportion of households thandid GAM (7.0%), stunting (19.5%), or under-weight (13.3%) in children. Overall, householdsclassified as overweight (31.5%) were mostcommon, followed by undernourished (25.8%),and then double burden–affected (24.7%).

The authors of the study ask how a popula-tion that was previously nomadic, possiblyexperiencing chronic energy insufficiency,could have developed the observed high levelsof overweight and obesity, while living inrefugee camps in the absence of economicdevelopment. One factor suggested is that theSahrawi were traditionally nomadic and cultur-ally associate larger bodies with wealth andbeauty. Thus fattening practices are commonamong Sahrawi, involving periods of ritualoverfeeding and the use of appetite enhancersand traditional medication (suppositoriescomposed of a mix of dates, seeds, and medici-nal plants that are believed to increaseperipheral fat accumulation). Urbanisation haspossibly created synergy between thesecustoms and the adoption of processed foodsand modern medicines, thereby increasing thelikelihood of obesity. Such a synergy might alsoaffect those living as settled refugees, as theydepend on food assistance and have limitedaccess to local markets in Algeria. Anotherfactor may be the excessive sugar consumptionhabit among the Sahrawi. One example isfound in the frequent and widespreadconsumption of green tea (with an averagereported consumption among refugees of threeservings of 30 ml each, three times a day ),which is usually prepared adding about fiveteaspoons of sugar for each teaspoon of greentea leaves. Lastly, urbanised Sahrawi womenwith high Body Mass Index (BMI) values havebeen found to walk significantly less than thosewith normal BMI values, thereby reducing theirenergy expenditure.

These factors may help to partially explainthe high prevalence of overweight in this popu-lation. However, they are complemented byother factors affecting refugees living in thecamps, which may help explain the high preva-lence of both undernutrition and overweight inthis population. Importantly, some factors thatare associated with undernutrition in early lifeappear to increase susceptibility to overweightin later life. Both nutritional deficiencies andfood insecurity, which as observed in the studyfindings often result in wasting and stunting inearly life, are also associated with subsequentobesity. The underlying mechanisms are stillbeing established. For instance, studies fromshanty towns in Brazil have suggested thatstunted children have impaired fat oxidationcapacity, a risk factor for obesity, although it isnot known if this developmental adaptationoccurs in other populations. Programming ofleptin receptors in the brain is another potentialmechanism receiving attention. There is also agrowing understanding that individuals expe-riencing undernutrition early in life are moresusceptible to developing obesity by subse-quent exposure to refined, carbohydrate-richdiets and high sugar intake, features character-istics of this population’s diet. One crucialaspect is that Sahrawi refugees are dependenton food assistance to cover most of their nutri-

tional needs and thus lack control over theirfood system. A typical food assistance basketfor this population will often be rich in starchyfoods (refined grain cereals, pulses, andblended foods) and sugar. The refugee foodassistance package typically contains low quan-tities, if any, of fresh or dried vegetables andfruit, therefore providing a low diversity diet.Recent evidence suggests that a low diversitydiet is related to obesity and associated co-morbidities, as well as being associated withnutritional deficiencies. In other words, thequality of the diet deriving from the food assis-tance currently provided may be implicated inboth nutritional extremes.

These findings raise numerous challenges.First, the emergence of obesity and the doubleburden of malnutrition have serious implica-tions for how international organisationsshould plan and provide assistance, especiallyfor those populations exposed to conflict ordisplacement of protracted duration. For exam-ple, food assistance policies need to be revisedand adapted, as those currently designed tomeet population minimum needs during anacute emergency will need to consider theirpotential contribution to the later developmentof non-communicable diseases (NCDs).Additionally, efforts are needed to promotelong-term food security and higher nutritionadequacy in protracted emergencies. Theactions needed range from improved food secu-rity assessments, with special focus on diversitywithin food groups, to provision of cash orvouchers, to community involvement insustainable livelihood programmes such asgardening and small-scale businesses. TheSahrawi refugees have been residing in campssince 1975. Generations of adults from birthhave received food assistance as their mainsource of food. Their children are now thesecond or third generation exposed to a consis-tently low quality diet. The intergenerationalimpact of this exposure is of serious concern inthis and similar protracted emergencies.

Second, efforts are needed to evaluate andmonitor the health impact of obesity and thedouble burden in refugee situations. Obesityand NCDs should be routinely included innutrition and health assessment exercises inprotracted refugee settings, and should beincorporated into the UNHCR HealthInformation System database.

Third, the development of appropriate andeffective behaviour change interventions toprevent and tackle obesity in these contexts willneed innovative approaches. These will requirehealth personnel and community participationin the identification of needs and implementa-tion of solutions. Additionally, a detailedeconomic assessment is needed to correctlyevaluate the resources needed for preventionand treatment. Lastly, careful policy and advo-cacy work will be required to convey thecomplexity of the situation, and to ensure thatcontinued support for life-saving food assis-tance programmes and the tackling ofundernutrition and nutritional deficiencies isnot jeopardised as the threat of obesity torefugee health receives the attention it deserves.

1 Grijalva-Eternot. C et al (2012). The Double Burden of

Obesity and Malnutrition in a Protracted Emergency

Setting: A Cross-Sectional Study of Western Sahara

Refugees. PLOS Medicine, October 2012, volume 9, Issue

10, e1001320, pp 1-12

35

Field Article

Integrating anaemiaanalysis in SMARTsurveys in Bolivia

By Susana Moreno, Brigitt Olagivel and Elisa Dominguez

Susana Moreno Romero is the Nutrition ProgrammeManager for Acción contra el Hambre (ACH) (ActionAgainst Hunger - Spain). She has previously worked inSierra Leone, Niger and Argentina. She has a PhD inNutritional Anthropology and is a member of UniversidadComplutense de Madrid Researching Group EPINUT.

Brigitt Olagivel is a Bolivian nutritionist. She has beenworking with ACF in El Chaco region since 2009.

Elisa Dominguez is a medical doctor and the Manager ofthe Health & Nutrition Department at ACF-Spain HQ andnutrition advisor for ACF projects in Latin America. Shehas previously worked in the field (Guinea, Thailand,Liberia, Mali, Angola, Niger) for more than 10 years.

This study was co-funded by the Swedish International DevelopmentCooperation Agency (Sida) and European Commission (through ACH),WFP, COOPI (Cooperazione Internazionale) and UNICEF. The authorswould like to acknowledge the political and health authorities fromregional to community level. Thanks are also extended to the teamsresponsible for the survey implementation and analysis, from surveyors tologistics and administration. A special thanks goes to the families andchildren who participated in the survey and to all those who helped asguides or in providing essential information, without whose collaborationthis survey would not have been possible.

The ‘Chaco’, which is a term derived from the Quechua ‘Chaku’and means ‘hunting land’, is a geographic area that takes inparts of Bolivia, Paraguay and Argentina. The Bolivian Chacohas a surface area of approximately 128,000 km2. It has very

irregular weather patterns affecting temperatures, rains and winds andis characterised by intense rainy summers and dry winters. The regionis especially subject to regular floods and droughts related to El Niñoand La Niña oscillations. The vulnerability of the Chaco rural popula-tion, who subsist on a farming and livestock based economy and havepoverty levels higher than those in many other areas of the country(INE 20011), increases the potential impact of disasters.

Between February 2009 and November 2010, the Chaco region, espe-cially the Bolivian Chaco, suffered an exceptional rain deficit accordingto the Meteorological Hazards and Seasonal Forecasting Group ofBenfield UCL Hazard Research Centre. During the austral (southernhemisphere) summer 2009-2010, rains were 60% less than expected.

Because of this protracted drought, the Bolivian governmentdeclared a National Emergency Situation in Chaco Region in June 2010and a Plan of Emergency Assistance and Farming Recuperation wasestablished. Water consumption, farming and livestock of more than7,600 households were severely affected according to a UNETE evalua-tion report (2010)3.

In October 2010, global acute malnutrition (GAM) prevalence (basedon weight for height z score (WHZ)) in rural children under five yearsof age was above 10% in two of the three Chaco departments (Cordilleraand Chuquisaca). High prevalence rates of acute malnutrition in sevenout of 16 Chaco districts were confirmed by CT-CONAN (ComitéTécnico – Consejo Nacional de Alimentación y Nutrición4).

To explore the nutritional and food security situation further, Accióncontra el Hambre (ACH), in collaboration with WFP, UNICEF andCOOPI, implemented ESAE5 and SMART nutritional surveys includinganaemia prevalence, in the Bolivian Chaco Region, between March-April 2011. Urban populations were excluded as ACH was onlytargeting rural areas. Also, it was considered that a bias could be intro-duced if urban areas were included, given the large social and economicdifferences between rural and urban populations in Latin America.

Box 1 overviews the SMART survey approach. The inclusion ofhaemoglobin (Hb) analysis to determine anaemia status in SMARTsurveys gives a more complete nutritional status assessment, particu-larly with regard to potential constraints for adequate child growth anddevelopment due to iron deficiency. Such analysis is especially impor-tant in countries like Bolivia where iron deficiency anaemia is a majornutritional problem, affecting 61% of Bolivian children between 6-59months old (DHS 2008)6.

This article describes the survey undertaken with a particular focuson the anaemia assessment component.

MethodologyPopulationThe Bolivian Chaco is located in the southeast of the country andextends over five provinces belonging to the departments of Cordillera,Tarija and Chuquisaca (see map). According to the last national census(2001), the 16 Chaco districts have a population of nearly 300,000 inhab-itants, over half (56%) of whom live in rural areas. Districts have askewed urban population distribution, where three out of the nineurban areas in Chaco Region host more than 84% of total urban popu-lation of this area (Yacuiba (51%) and Villamontes (12%) in Gran ChacoProvince, Tarija, and Camiri (21%) in Cordillera province, Santa Cruz).

The population of Chaco is ethnically composed of mainly Guaraníesand Mestizos (rather than Quechuas) and other aborigine groups likeAymara, Carai or Weenhayek. However, the ethnic composition variesbetween the Chaco provinces. More than 80% of households surveyedpractice agriculture, and maize is the main subsistence crop. More than

1 INE 2001. Censo Nacional de Población y Vivienda. Instituto Nacional de Estadística.

http://www.ine.gob.bo/anda/ddibrowser/?id=312 The full assessment report is available in Spanish at: http://www.reliefweb.int/rw/rwb.nsf/

db900sid/VDUX-89GRUC?OpenDocument&query=bolivia3 UNETE. RED HUMANITARIA 2010. Bolivia: Sequía en el Chaco – Reporte de misión de

valoración y validación de información. Report Government of Bolivia, UN Country Team in

Bolivia. http://reliefweb.int/node/3682784 Technical committee of the National Council on Food and Nutrition 5 Evaluacion de seguridad alimentaria de emergencia (food security assessment in

emergencies). More information at: http://documents.wfp.org/stellent/groups/public/

documents/manual_guide_proced/wfp203212.pdf6 Available at: http://www.measuredhs.com

BolivianChaco Region

36

Field Article

95% of households have livestock – in order ofpriority, poultry, pigs and cattle.

Due to a number of political and administra-tive factors and in order to allow comparison, atotal of three SMART surveys, one per depart-ment, were implemented.

SamplingUsing cluster sampling and the latest publishedGAM prevalence rates (UNETE 20107), a repre-sentative sample for each department wascalculated using ENA software for SMART.Values are shown in Table 1.

All rural localities listed in the last census2001, except Mennonite communities9, wereconsidered for random selection using ENA(Emergency Nutrition Assessment) software.The current population for each communitywas calculated applying the 2010 district popu-lation provision of INE (National StatisticsInstitute) based on the 2001 national census.Localities with more than 2,000 inhabitants(considered as urban), were not included in thesurvey. Between 52 and 58 localities weresampled in each department. The final sampleswere 884 children between 6 and 59 months oldin Cordillera (Santa Cruz), 477 in Chuquisacaand 420 in Tarija. Some parents did not autho-rise Hb analysis, so final sample sizes forcollection of anaemia data were slightly lower.The characteristics of each locality determinedthe household sampling method used in agiven locality. The approach employed thesampling decision tree set out as part of theSMART methodology.

EquipmentIron deficiency anaemia was detected throughuse of a Hb analyser HemoCue Hb 201+(Figures 1a and 1b). This portable analyserallows field measurement of the Hb concentra-tion in peripheral blood through photometricdetection. Material for Hb analysis includesmicrocuvettes, lancets, gloves, alcohol, toiletpaper, cotton, batteries and a waste bottle. Theanaemia analysis cost was about 2 dollar/child(excluding the cost of the HemoCue which wasapproximately 1000 dollars/unit).

Microcuvettes do not need refrigeration andas the Hb analysis is implemented right afterthe blood sample is taken, there is no specialrequirement for transport or storage. The mainfield constraint to take into account is theclimatic environment. In order for the chemicalreaction to work properly in the HemoCue Hb201+, temperatures lower than 15ºC or higher

than 30ºC or humidity more than 90% withoutcondensation must be avoided.

TrainingInclusion of anaemia analysis in the SMARTsurvey did not necessitate increasing thenumber or qualifications of staff. However, insome countries, lab officers or nurses may berequired by the Ministry of Health.

Theoretical and practical training for theanaemia test took around four hours. Withgood team organisation, the average time to dothe Hb analysis in the field is about fiveminutes per child, although this does dependon the child’s amenability and team experience.The chemical reaction in the HemoCue Hb 201+usually takes one minute.

The survey was implemented in March andApril of 2011 coinciding with the last twomonths of the rainy season and the first ‘choclo’(green corn) harvest. One to two weeks beforethe survey, local and selected communityauthorities as well as health posts/centres wereinformed and authorisation for the survey wasrequested. Some ethnic groups, or populationsfrom specific geographic zones, were averse toblood tests so this prior sensitisation and mobil-isation stage is especially important foranaemia studies.

Family authorisation was also required. Forthis it is important to distinguish betweenpermission for anthropometrical measurementsand bilateral oedema which is non-invasive, topermissions for blood tests which are invasive.

The fact that results are obtained immedi-ately allows the family to have the Hb status oftheir children confirmed instantaneously. Thisis highly appreciated. A model of the card givento each family with the results is shown inFigure 3. Different stamps were designed andused to represent a child with anaemia, withoutanaemia and borderline and made it easier forcaregivers to understand (Figure 3). Althoughonly 1 to 3 blood drops are needed for the reac-tion analysis, one common belief and reason forresistance to the test was the suspicion thatblood may be sold. One way to reassure the

7 See footnote 38 WFP 2010. Rapid assessment of nutrition and food security

in Chaco municipalities affected by droughts. 9 Mennonite communities were not included as they are not

targeted by ACF programmes. Also, Mennonite communities

have significantly different socio-economic and cultural

backgrounds to the wider rural population in the Chaco,

hence were not included even if they were set in rural

areas.

Box 1: What is a SMART survey?

The Standardised Monitoring and Assessment ofRelief and Transition (SMART) programme is an inter-agency initiative to improve monitoring andevaluation of humanitarian assistance interventions.

The SMART Methodology provides a basic, integratedmethod for assessing nutritional status and mortalityrate in emergency situations. It provides the basis forunderstanding the magnitude and severity of ahumanitarian crisis. The optional food securitycomponent provides the context for nutrition andmortality data analysis.

SMART surveys measure acute malnutrition of thewhole population via estimates of:• Prevalence of Global Acute Malnutrition (GAM) in

children aged 6-59 months.• Crude mortality rate (CMR) in a given population

over a specific period of time.

• Food security assessments, which are used to understand and interpret nutritional and mortality survey data.

The SMART manual deals specifically with nutritionand mortality surveys, including sampling, nutritionalmeasurements, and mortality rates. It describesgeneral survey procedures and provides informationon how to collect data necessary for planning directinterventions in emergency settings or for surveil-lance. It also provides step-by-step instructions foranalysing survey data using Emergency NutritionAssessment (ENA) software and procedures for foodsecurity assessments.

For more information, including access to the SMARTManual, additional resources, ENA software and aSMART forum for practitioner questions, visit:http://www.smartmethodology.org

Figure 1a: Material for Hb analysis

Sampling Cordillera Chuquisaca Tarija

Total population* 57,321 53,641 66,667

GAM prevalence(WFP 20108)

13.5% 19.9% 6%

Estimated precision

3 5 2.5

Design effect 1.5 1.5 1.5

Estimated samplesize

814+5%=855

400+10%=440

348+10%=383

Final sample size 884 477 420

Average household size(No. of peopleper household)

5.23 5 5

Children under 5years (%)**

13% 15% 15%

Households thatdeclined to participate (%)

3% 3% 3%

Estimated no ofclusters

48 32 32

Final no. of clusters

48(49)** 32 33

Total no. of communities

58 52 53

No. of childrenunder 5 years ineach cluster

18 14 12

No. of childrenwith Hb sample 6-59 months(6-24 months)

859 (302)

469(161)

420(154)

Table 1: Sample estimations for SMART nutritionalsurvey by Chaco department

* Total population for each department (excluding populations

over 2000 inhabitants as considered urban and not included

in the survey)

**An additional cluster was included to enable the expected

number of children to be reached.

Figure 1b: Blood drop for Hb analysis

37

Field Article

These results confirm anaemia as one of themajor nutritional problems of rural children inBolivian Chaco. Iron deficiency anaemia preva-lence found in ENDSA 200813 was also above 40%showing that iron deficiency anaemia is anendemic problem with structural causes amongstthe Bolivian Chaco population. Iron deficiencyanaemia in children has adverse effects on physi-cal and psychomotor development, the immunesystem and physical performance14. Its reductionthrough program- mes and policies which impactthe structural causes must therefore be a keygovernment goal.

ConclusionsAnaemia analysis is easily integrated intoSMART surveys without an excessive increase ofbudget or resources required. It provides a moreprofound understanding of the nutritional prob-lems affecting a population than anthropometricsurveys alone and hence the policies, strategiesand programmes that must be pursued to eradi-cate malnutrition. Its application is especiallyimportant in areas or groups with high risk of“hidden hunger” due to micronutrient deficien-cies, i.e. in regions like Latin America or urbansettlements in developing countries. ACF iscurrently evaluating whether the urban popula-tion of the Bolivian Chaco needs to be included infuture anaemia surveys.

If you have any comments regarding this article,please contact: hom-pe@acf-e.org

AnalysisEach child’s information was entered in the ENAsoftware in order to check and correct possiblemistakes. The same software was used to analyseindicators based on weight and height measure-ments and presence of oedema, and to calculateprevalence of acute and chronic malnutrition.Anaemia and MUAC data were analysed withExcel 2007 and SPSS 17.0.

MUAC cut-off points were established at 115mm for severe acute malnutrition and 125 mm formoderate acute malnutrition.

Cut off points for anaemia and the severityscale varies according to authors, age, pregnancy,countries and geographic altitude (INS 2005). Inthis case, and largely to allow comparisons,anaemia classification was based on the same cut-off points used for children between 6-59 monthsold in the previous Demographic and HealthNational Survey (2008) in Bolivia. These areshown in Table 2.

Results and discussionAccording to these first SMART surveys inBolivia, GAM prevalence in Chaco rural popula-tion under five years is less than 1.5%. There isless than 6% underweight and approximately 20%stunting. Malnutrition prevalence in each depart-ment is specified in Table 3. The highest valueswere found in Cordillera province (Santa Cruz).The child nutritional situation after the 2010drought seems to have returned to normal values,when compared with ENDSA 2008 and PMA200611 survey findings. Anthropometric results inconjunction with the ESAE analysis showed thatcauses of malnutrition in Bolivian Chaco aregenerally structural and only change when occa-sional dramatic events like severe droughts orfloods occur, as occurred in the 2009-2010drought.

Based on MUAC, a total of nine cases of acutemalnutrition were detected in all Chaco Region.Out of these nine cases, only one was classified asacutely malnourished using weight for height z-score.

Hb analysis showed that iron deficiencyanaemia in rural children between 6 and 59months old in Bolivian Chaco is above 40%.Based on the FAO (2006) scale12, this signifies asevere iron deficiency problem in this region. Age,sex and district differences were found: anaemiawas higher in children under 24 months of age, inboys and in Cordillera province (Santa CruzChaco) (see Table 4). Anaemia prevalence in 6-23months aged boys from Cordillera was 75.3%,compared to 67.9% in girls. Overall prevalence in6-59 months (both sexes) was highest inCordillera (57%) and lowest in Tarija (48.8%).Approximately 70-80% anaemia cases were mildand less than 1.5% were classified as severe(Figure 6).

population about this fear was to allowthem to keep the microcuvettes used fortheir analysis. However, we do not recom-mend standardising this practice due towaste management difficulties.

The peripheral blood sample was takenfrom the left hand middle finger (Figure 4)and when this extraction was not possible(under 1% of cases), the sample was takenfrom the heel (Figure 5). The specificmethodology used to take the blood sampleand conduct the Hb analysis can be foundin INS (2005)10.

Height, weight, mid-upper arm circum-ference (MUAC) measurements andbilateral oedema check-up, as well as theblood test, were done for each selectedchild following standardised protocols.

10 INS. Instituto Nacional de Salud. 2005. Manual del

Encuestador. Monitoreo Nacional de Indicadores

Nutricionales. Perú. Capitulo II: Manual de diagnóstico de

anemia por hemoglobinómetro y su utilización. Perú.

http://www.ins.gob.pe/portal/jerarquia/5/311/monitoreo-

nacional-de-indicadores-nutricionales-monin/jer.31111 PMA. 2006. Diagnóstico de la Seguridad Alimentaria y

Nutricional en el Chaco Boliviano. Programa Mundial de

Alimentos de las Naciones Unidas.12 FAO. 2006. Indicadores de Nutrición para el Desarrollo.

Organización de las Naciones Unidas para la Agricultura y

la Alimentación. Roma, 200613 ENDSA. 2008. Encuesta Nacional de Demografía y Salud.

INE.14 World Health Organization 2001. Iron Deficiency Anaemia.

Assessment, Prevention and Control. A Guide for

Programme Managers. WHO/NHD/01.3

Table 2: Cut-off points for anaemia classification

Table 3: Underweight, acute and chronic malnutritionprevalence and confidence intervals perdepartment of Bolivian Chaco Region

Source: ENDSA, 2008. Encuesta Nacional de Demografíay Salud. INE.

Cordillera Chuquisaca Tarija

n=884 n=477 n=420

% (n)(95% C.I.)

% (n)(95% C.I.)

% (n)(95% C.I.)

Acute malnu-trition

1.4% (12)(0.8 – 2.4

0.8% (4)(0.3 – 2.3)

0.7% (3)(0.2 – 2.3)

Underweight 5.9% (52)(4.2 – 8.1)

2.5% (12)(1.2 – 5.4)

3.6% (15)(2.1 – 6.1)

Chronicmalnutrition

22.3% (197)(18.9-26.0)

18.8% (90)(15.0-23.4)

16.9% (71)(12.5-22.5)

Age (months) Category upper limit

Mildg/dl

Moderateg/dl

Severeg/dl

0-3 8.9 6.9 4.9

3-6 10.4 8.4 5.4

7-23 10.9 8.9 5.9

24+ 11.4 9.4 6.4

Table 4: Anaemia prevalence in boys and girls aged 6-23 months and 6-59 months per BolivianChaco Departments

Figure 2: Results card for the families, oneper child

Figure 3: Stamps representing anaemiastatus

Figure 4: Puncture zone on middle handfinger

Figure 5: Puncture zone on heel

Age(months)

Sex % anaemia prevalence

Cordillera Chuquisaca Tarija

6 – 23 Girls 67.9 55.5 55.1

Boys 75.3 57.7 67.1

Total 71.9 56.5 61.1

6 – 59 Girls 55.5 49.1 47.0

Boys 58.6 50.3 50.5

Total 57.0 49.7 48.8

80%70%60%50%40%30%20%10%

0%

Mild Moderate severe

6-23months

Cordilera Chuquisaca Tarija

6-59months

6-23months

6-59months

6-23months

6-59months

Figure 6: Prevalence of anaemia by severity in children6-23 months and 6-59 months per BolivianChaco departments

1

19

52

0.5

12

45

1.2

15

40

0.6

11

38

0

18

43

0.2

9

39

Note: Numbers have been rounded for graphic display

38

Field Article

Background and purposeThe vast majority of humanitarian aid world-wide (90%) is provided as in-kind transferssuch as food, livestock, seeds, and shelter ratherthan in the form of cash.1 However, it is beingwidely observed that there is an ongoing trend,exhibited by humanitarian organisations andgovernments alike, moving towards less tradi-tional modes of relief intervention, especiallycash transfers. Correspondingly, there has beenan expanding body of knowledge in the pastdecade on the use of cash as an alternativemeans of relief in a variety of emergencies.

The bulk of existing literature indicates thatthe motive behind this is two-fold. In caseswhere markets are functional, cash is said to bemore cost effective, more efficient and have amore positive effect on local economies.2

Because in-kind food aid requires substantiallogistical support and organisation, it oftenincurs a higher financial cost. Secondly, severalstudies have highlighted the fact that providingcash is more respectful of human dignity as it isless paternalistic and respects the autonomy ofrecipients to make choices and prioritise theirneeds which, in turn, also achieves a higher rateof satisfaction.3

Yet despite the growing literature surround-ing cash transfers, little critical attention hasexamined how and in what ways beneficiariesand communities respond to the mode ofexchange. While claims have been made on theefficiency of a cash intervention compared tofood, it is also important to examine theperspectives and preferences of recipients,investigate the differences of impacts on liveli-hoods, determine how well each met theirobjectives, and assess the challenges and limita-tions faced by each mode of intervention.

MethodologyThis study employs both qualitative and quanti-tative methods. Surveys with projectbeneficiaries were completed before and afteractivities to assess impact. Focus group discus-

sions were completed in order to gather abroader understanding of how projects impactedbeneficiaries and communities. The focus groupswere selected randomly (see Table 1).

A total of 18 focus group discussions werecompleted involving a total of 184 participantscomprised 121 women and 63 men. The aver-age size was 10.2 persons per group, but rangedfrom a minimum of five to 21 in certain cases.Additional participatory methods were usedsuch as filed observations and interviews withkey informants including beneficiaries, localgovernment officers, project coordinators andcommunity leaders.

Project summariesFood-for-Assets (FFA)The FFA projects examined in this study wereimplemented in two different counties inKenya, Mbeere and Kilifi. The two projectswere designed as short-term interventionsduring the hunger months between September2011 and January 2012 in response to failedharvests of the 2011 drought. Both projectstargeted agro-pastoralists in semi-arid areasand aimed to provide a 50% food ration (50 kgmaize, 10 kg split peas) to 1,000 households(HH) and 2,400 HH respectively. Beneficiariesfrom each HH were required to participate inwork activities on community assets such aswater pans in Kilifi but also tree planting, andcommunity farms in Mbeere. The targeting

Food, goats and cashfor assets programmesduring emergencydrought response inKenya By Geoff Brouwer

Geoff Brouwer worked as a research and communicationsconsultant for World Renew-formerly the ChristianReformed World Relief Committee (CRWRC) - in Kenya fortheir International Disaster Response team. After workingbriefly in Canadian politics he took on several researchassignments in Kenya examining various topics including

community water management and the Kenyan Water Act. He is currentlycompleting his Master’s degree in Ottawa at the Norman Patterson Schoolof International Affairs.

World Renew extends thanks and recognition to the Canadian InternationalDevelopment Agency (CIDA), the Canadian Food Grains Bank (CFGB),Integral Alliance and ACT Alliance for their generous support in the Horn ofAfrica. The author also gratefully acknowledges and recognizes the dedica-

1 Kelaher, D and Dollery B, 2008. Cash and in-kind food aid

transfers: the case of tsunami emergency aid in Banda Aceh.

International Review of Public Administration 13(2):117-128.2 World Food Programme. Cash and food transfers for food

security and nutrition: emerging insights and knowledge

gaps from WFP’s experience. October, 2006. 3 M Benjamin, C Donovan, and V Kelly. 2009. Can cash

transfers promote food security in the context of volatile

commodity prices? A review of empirical evidence.

Agricultural Economics. Michigan. Kelaher D, and Dollery B.

2008. Cash and in-kind food aid transfers: the case of

tsunami emergency aid in Banda Aceh. International Review

of Public Administration 13 (2): 117-128. Pierson R. T.

Ntata. 2010. Bridging the hunger gap with cash transfers :

experiences from Malawi. Development in Practice 20 (3):

422 – 427.

Table 1: Sample estimations for SMART nutritionalsurvey by Chaco department

Project No. offocusgroups

Participants (male: female)

Focus group participantsas % of total participants

Food-for-Assets

4 – Mbeere3 – Kilifi

50 (17:33) –Mbeere30 (9:21) – Kilifi

5%8%

Cash-for-Assets

5 45 (15:30) 9.5%

Goats-for-Assets

4 59 (22:37) 9.8%

Total: 16 184 (63:121) --

tion of World Renew’s partner Christian Community Services (CCS) and theircontinued work in communities throughout Kenya. Lastly the author wouldlike to express deep gratitude to the welcoming communities and individualswho participated in the research.

This field article is based on a comparative study of four relief and emergencyresponse activities - one Cash-for-Assets (CFA), one Goat-for-Assets (GFA), andtwo Food-for-Assets (FFA) projects - implemented by World Renew, formerly theChristian Reformed World Relief Committee (CRWRC), in response to the 2011drought in Kenya. The fieldwork included surveys that were collected intermit-tently throughout the project cycles in addition to focus group discussions andinterviews with key informants, which were carried out over several months inearly 2012. The objective was to gather a deeper understanding of the variousmodes of asset-exchange and their differences as emergency responses.

criteria for the two projects were similar: farm-ers who lost 90% of their harvests, herders wholost a significant number of their animals, orfamilies who were eating less than one meal aday. Due to the high level of need and the sever-ity of food insecurity, priority was given to thehighly vulnerable e.g. women- and child-headed households, orphans, elderly, chronic-ally ill and the disabled. The expected outcomesof the project were to increase food consump-tion, improve resilience to drought, access towater and increase crop production.

Cash-for-Assets (CFA)The CFA project was conducted in Kilifi districttargeting 600 HH and shared many similaritiesto the FFA activities. While the identified loca-tions in Kilifi overlapped with some FFAprojects, they targeted different beneficiaries.The cash-transfer projects started later fromOctober 2011 – February 2012 dispensing 2,500KES/month via nearby banking systems inexchange for participation (10 day/month) inthe construction and rehabilitation of waterpans. Initially targeting 600 participant house-holds, the project was only able to retain 475HH due to a variety of constraints that arediscussed below. As with the FFA projects theobjectives of this intervention were to increasehousehold food consumption as well asincrease community resilience to drought.Households were targeted if they had lost 50%of their harvest and were consuming less than 2meals per day.

Women gathering to collect waterfrom a water pan in Kilifi

World Renew (CRWRC), Kenya, 2012

39

Goats-for-Assets (GFA)The GFA projects had distinctly different objec-tives, since the activities followed after the endof the 2011 drought. Targeting 600 HH inMbeere from the months of February-May 2012,families were required to participate in thecompletion of water pans and dams inexchange for participation in a goat re-stockingprogramme. A work programme involving 10-12 work days per month was rewarded at theproject’s completion with a hybrid milk-producing goat for each family, training inlivestock care, and a male goat shared betweenfive families. The project was designed toimprove livestock health, increase householdfood security and to alleviate suffering from thedrought by supporting livelihoods through theavailability and accessibility of water.

FindingsBeneficiary participationA widespread observation made in all projectswas the disproportionate involvement offemale participants to males on the variouswork schemes. In each of the work sites,women made up the majority of the partici-pants, ranging between 70-95%. Targetingcriteria played some role in this, as thoseselected were often widows and single mothers.However, this does not account for the femaleparticipation rates at certain work sites being ashigh as 90-95%. The most common communityexplanation for this was that many men had leftfor larger towns to find casual labour or wereengaged in other activities such as farming orcharcoal burning. It was also suggested that theprovision of food and water were traditionallya woman’s responsibility, and since most of thework projects involved water pans and, at leastindirectly, the provision of food, women wereoften sent on behalf of male household-heads tocomplete work activities.

However, men were more involved in theCFA projects. Group discussions revealed thatmen were more interested in receiving cashpayments than food. Several men describedbeing proud of finding employment andnumerous women explained how men werereturning from nearby towns to find employ-ment at home. One challenge with this type ofemployment was that some participants wouldcease to work on water pans once the cashpayments were finished, indicating a limiteddegree of ownership.

Another challenge with the CFA was thattarget beneficiary numbers fell to 475 due todrop out. Five of the six water pans failed tohave the target number of beneficiaries and inone case nearly 50% of participants left the proj-ect. The reasons given by participants andproject coordinators varied. One major factorwas difficulties in the delivery of cash through

the banking system. A number of beneficiariesremarked that they had not received paymentsfrom the banks, and others had difficultiesestablishing an account due to lack of identifi-cation. Secondly, since some travelled longdistances to access banking services and wereultimately turned down, they were discour-aged. Another factor was the difficulty of thework; as the project progressed the workbecame increasingly difficult leading some toconclude that the pay was insufficient.

Both the GFA and FFA projects were moresuccessful in reaching planned beneficiarynumbers. Project coordinators noted that therewas a higher level of demand for the projectsthan there were food rations available. It wasalso noted that non-direct beneficiaries partici-pated in the work activities of FFA projectsbecause they also recognised the benefit of thewater pans. Similarly, the FFA work activities inKilifi continued beyond the food distributionscontrasting with the lack of participation in thecash-transfer projects in the same region.

Impacts on food security and livelihoodsNearly all respondents highlighted the primarybenefit of being able to send their children toschool. Cash inputs allowed families to pay offgrowing debts to schools and have their chil-dren return to education, while those receivingfood were able to use money for school feesrather than food expenditures. Both pro-grammes also allowed families to send theirchildren to school on full stomachs. An equallyimportant benefit from both projects was theresulting increase in food consumption at thehousehold level. All focus groups describedhaving more energy and increased health dueto the cash or food, which allowed them toengage in productive work activities such asfarming. A third major benefit cited in all buttwo communities was how families were ableto retain their livestock and avoiding sellingthem for poor prices in order to earn cash forfood.

These benefits to food security were alsoevident from the survey results. The two FFAactivities showed substantial increases in thenumber of meals per day. The baseline surveysshowed that in Kilifi and Mbeere, 46.5% and57% of targeted beneficiaries respectively wereconsuming two or more meals per day,compared to 73% and 97% upon projectcompletion. In Mbeere, this included anincrease in those consuming three meals a dayfrom 3.7% to 40%. Reliance on negative copingmechanisms also showed strong signs ofdecline in Mbeere including the proportion ofthose limiting meal portion sizes from 70% to39%, those begging for food on a daily basisfrom 7.3% to 0.5% as well as a decrease in char-coal burning activities.

Similar effects were identified in the CFWproject. Before the implementation of the proj-ect during the hunger months, just 39% offamilies were eating two or more meals per day(average over the month), with the majorityeating just one meal. Upon completion of theproject, consumption of two or more meals perday had risen to 69% showing an increase of30% (see Table 3). This rise was unlikely to bedue to improved seasonal food availability asthe projects ended before the harvest wasmature – projects were designed to carrycommunities through to the harvest after thenext rains, but not into the harvest, as to notdeter farming activities. However, the surveyfound that the benefits were not sustained forlong periods. A common complaint from recip-ients of food relief was the inadequacy of foodrations. Just 34.6% of households reported thatthe rations lasted more than two weeks. Thefood allowances were designed to accommo-date a household of eight persons, yet manyhouseholds stated that ration had to be sharedamongst many more, including neighbours andguests. Consequently, food rations were ofteninadequate in eliminating hunger during theproject months. The CFA projects revealed simi-lar findings.

Some focus group participants explainedhow money was too liquid and they were notaccustomed to having such a large sum of cashat one time. According to surveys, at least 75%of the cash recipients would withdraw the totalamount of cash from their bank account, muchof which was spent within the first week.

Other impactsFocus group discussion helped identify othersignificant impacts which were not identified inthe surveys. One commonly identified benefitwas the effects on family unity and cohesion.Groups from both the FFA and CFA expressedhow payments (cash or food) allowed familiesto solve their immediate conflicts. Certain fami-lies had described how when their householdsbegan lacking food or cash, disagreementsarose between husbands and wives, and occa-sionally between children and parents. Therewere numerous accounts of how wives hadmoved to their parents’ home, or husbands toanother city or live with another woman.Several informants described the healing ofrelationships as coinciding with the interven-tion. Moreover, men and women in both cashand food transfer programmes described howtheir sex drive returned resulting, in somecases, in new pregnancies.

However, in three of the five focus groupsconducted during the CFA programme, severalrespondents described how the influx of cashalso brought about conflicts. A group of womendescribed how the ability to conceal or hide cashwould lead to suspicion and doubt. Spouseswere known to question each other on howmuch money remained or where it was spent.Those who highlighted this conflict also

Table 2: Summary of asset-exchange projects

Project Location Number of house-holds (male: female)

Duration Inputs Outcomes

Food-for-Assets

Kilifi

Mbeere

2,400 HH(1,301:1,290)

1000 HH(412:427)

Sept 2011 – Jan 2012 (5 months)

Oct 2011 – Feb 2012(5 months)

Food grains (maize,split peas), seeds(fruit trees, grass,crops)

Increased food consump-tion, access to water andimproved harvests

Cash-for-Assets

Kilifi 475 HH Oct 2011 – Feb 2012(5 months)

2,500 KES per month(10 work days)

Increased food consump-tion, access to water andimprove livelihoods

Goats-for-Assets

Mbeere 600 HH Jan 2012 – May 2012(5 months)

600 female, 30 malesgoats, rehabilitationwork on 6 water pans

Increased water accessibil-ity, animal support services,increased food security

Table 3: Daily meal consumption

Location Baselinesurvey: 2+ mealsper day

End of projectsurvey: 2+ meals perday

% change

FFA Kilifi

Mbeere

46%

57%

73%

97%

+27%

+40%

CFA Kilifi 39% 69% +30%

Field Article

described how an in-kind transfer, suchas food, would not lead to this type ofchallenge. A small number of individualsalso recounted that certain individualsinvolved in the project had spent portionsof money on alcohol or gambling.

Unique to the CFA projects was howthe high participation of women gavethem a greater control over cash in thehousehold. Despite there being a higherparticipation of males in CFA comparedto GFA or FFA work sites, the majority ofparticipants were still women. Surveyresults showed that 56% of the bankaccounts were registered under femaleparticipants, while just 30% were regis-tered under male (14% were registeredunder the names of neighbours orfriends). Ninety two percent of respon-dents indicated that it was the first timethat they had possessed a bank account.Female participants described howhaving their own income allowed themto attain greater power over makingdecisions about household purchasesthan before. Previously, some marriedwomen described having to ask for money fromtheir husbands to make purchases, whereasnow husbands were asking for spendingmoney. Monthly surveys throughout the projectcycle indicated that on average women madeHH financial decisions 47% of the time,whereas 51% of the time decisions werereported as shared between HH heads. Womencommonly expressed both pride and happinessin being the breadwinners.

Beneficiary preferencesA significant portion of focus group discussionswas spent on dialogue surrounding the benefi-ciary preferences for the different types ofprogramme. In each of the GFA and FFA proj-ects, nearly all recipients chose their currentmode of transfer over the alternative of cash.Occasionally a food recipient would showpartiality towards cash but this was an anom-aly; consensus in nearly every group indicatedsatisfaction with what they were receiving.Conversely, those who were enrolled in CFAprogrammes were split between payment incash or food. Of the five focus groups done inthe CFA locations, four exhibited more than halffavouring in-kind food to cash, one of whicheven unanimously agreed on food. It wasevident, however, that many of those express-ing a desire for cash had few opportunities toearn income, while some preferring food hadalternative means of income (such as casuallabour or charcoal burning). Comparatively,male participants exhibited a greater preferencefor cash than their female counterparts,although the preference for cash was still in aminority.

Those preferring cash-transfers explainedthat being given money allowed them toaddress a number of needs, not just hunger.While hunger was consistently regarded as themost pressing need, it was pointed out thatother pressing needs could not be solved withfood, e.g. medical costs, school fees andpayments of debts. Those preferring cash alsohighlighted their satisfaction with being able tomake choices themselves. Some specificallyexpressed satisfaction with being able to choosebetween different types of food such aspurchasing oil, meat or salt. A less common

conveyed how both food and cash arefleeting and would be quicklyconsumed, while a goat, especially ahigh milk-producing hybrid, canproduce income and food for years tocome. Furthermore, goats would alsoperiodically give birth to offspring andproduce manure for the fertilisation oftheir farms. One respondent in Mbeereindicated that goats acted as a savingsaccount or insurance mechanism, as theycould be sold in dry spells. Lastly, goatshave intrinsic value as a sign of wealthand are often used in traditional customssuch as in the payments of a dowry.

Payment usesThe questionnaires for the CFA projectsindicated that spending habits of benefi-ciaries changed throughout the projectcycle but consistently showed the largestexpenditure was on food. Mid-project,when food prices were highest, 85% ofhouseholds spent at least half of theirincome on food, while 55% spent overthree quarters of the cash on food. Theend of project survey indicated that by

March 2012, after the short rains harvest, only6% of households spent three-quarters or moreof their income on food, while 73% spent morethan half. The end of project survey indicatedthat 14% spent most on school fees while expen-ditures on servicing debts, paying medical costsand livestock care accounted for about a third ofthe cash transfer. These findings were consis-tent with focus group findings. Occasionally,households would have to pay large amountstowards school fees for secondary or collegeeducation or towards an unexpected medicalemergency. While just 19% of beneficiariesmanaged some minimal savings (81% savednone), focus groups discussions were unable tolocate any individual who had any substantialfunds left in their account. While many likedthe idea of having a bank account, they indi-cated that it was merely a mechanism totransfer funds and that it was not possible tomaintain savings for the future without asteady income.

Surveys from the FFA projects revealed thatbetween 88-90% of delivered food wasconsumed by the intended household. Aportion of the food (19.8% in Mbeere and 12.0%in Kilifi) was shared or given to neighbours. Forthe GFA project, focus group discussions werenot conducted after the distribution of goats,and an end of project survey has not yet beencompleted.

ConclusionsCurrent guidelines increasingly stipulate thatwhere markets are functional, CFAprogrammes are superior to food distributionsspecifically in terms of efficiency. While thismay be the case, other key factors includinggender participation, beneficiary preference,project ownership, adverse impacts and behav-ioural responses should also be considered. Theabove case studies carried out by World Renewshows that while there are many similaritiesbetween different models of asset exchange,they are not interchangeable and cannot beexpected to achieve identical results.

For more information, contact: Geoff Brouwer,email:geoffbrouwer@gmail.com orgeoffrey_brouwer@carleton.ca

40

explanation given for a preference of cash washow neighbouring families expected to begiven a portion of food in the FFA programmefood until it was used up. It is uncommon to“hoard” food during a drought period, espe-cially when others are going hungry. Cashrecipients expressed high levels of possessive-ness and ownership over their cash as they feltthey earned it, and described feeling less of aninclination to share.

Those preferring food (approximately halfthose given cash and nearly all of those givenfood), attributed their preference to a variety ofreasons. Most focus group participants agreedthe hunger was the primary need that neededto be addressed. Parents expressed how whenfood was not in the home not only were theyhungry, but also distressed because their chil-dren were weak or sick. This was a commonpoint of tension and grief for families.Individuals further described how not havingfood to serve their family fuelled argumentsand conflicts in the home. Thus, when food wasprovided peace was restored in the home.

Other preferences for food stemmed from acritique of cash. Some of the respondents whohad received cash but preferred food describedhow cash was variable in its purchasing power.During the early months of the project, whenprices were volatile and often inflated (at timesdoubled), they lost purchasing power.Furthermore, others described their inability tolimit spending; none had described previouslymaking savings and were unaccustomed tokeeping cash. Accordingly, a month’s food allo-cation cannot easily be consumed in a week,while cash can. Several individuals describedlack of discipline and being “tempted” intomaking trivial purchases. Nearly 40% of focusgroup participants described how the cash wasfinished within the first week.

Finally, the vast majority of GFA participantsexpressed preference for goats over cash andfood during focus group discussions. It isimportant to note the GFA project was specifi-cally implemented after the dry season andcorrespondingly food needs were less pressing.Explanations from community members

Field Article

World R

enew

(CRW

RC),

Kenya,

2012

Women and Men in a Kilificommunity working on alarge water pan 

ACF-Canada has released a new version of the SMART website.Training material and website content is now available in threelanguages: English, French and Spanish.

Content includes and is organised as:Capacity building toolbox that comprises• The SMART methodology manual (manual and sampling paper)• The Standardized Training Package (STP) (modules and video)• Emergency Nutrition Assessment (ENA) software (ENA software and

hybrid Epi-Info/ENA software)

Technical FAQsSMART forum (online bulletin board where members can post questionsand comments)SMART calendar (where you can view or add SMART events)

The site also includes jobs available, a list of SMART trainees by country,and key resources.

Membership involves a simple online registration. Visit:http://www.smartmethodology.org/

41

New SMARTwebsitelaunched

Erasmus Mundus MasterCourse in Public Heathin Disasters

News

The Erasmus Mundus Master Course in Public Heath in Disasters(EMMPHID) is a joint university educational programme organisedby the Universidad de Oviedo (Spain), the Karolinska Institutet

(Sweden) and the Université Catholique de Louvain (Belgium). It isdesigned to deliver high quality education and training in the field ofPublic Health in disaster situations.

The Programme is a one year course sponsored by the European Unionand involves thirteen associated universities and institutions around theworld.

The Master has three components: • Common Disaster Core Component• Elective Track Component with two possible options (public health

response to disaster or disaster epidemiology and research)• Internship in one of the associated master centres and final Master’s

Thesis.

To be eligible for the EMMPHID master programme, applicants musthold a recognised primary degree in areas related to Health,Management and Administration, or Social Sciences (a minimum ofthree years’ study at a university and of a minimum of 180 ECTS, orequivalent). Applicants must be proficient in English.

A number of scholarships (contribution to tuition fees and maintenance)are available.

All information and the online application are available at.www.pubhealthdisasters.eu or contact: Universidad de OviedoInternational Graduate Centre, Plaza de Riego s/n, E-33003 Oviedo,Spain, tel: +34 98 510 49 17/+34 98 510 49 18, email: infopostgrado@uniovi.es

The deadline for submitting applications to the 2013-2014 programme is30th November 2012

Forty-seven questions were posted on en-net in the three monthsAugust to October 2012 inclusive, eliciting 152 replies. In addition, 18job vacancies and five notices were posted.

Recent discussions have included: appropriate treatment protocolsusing ready-to-use therapeutic food (RUTF) for older people with severeacute malnutrition (SAM), discharge criteria for Outpatient TherapeuticProgrammes (OTPs), treatment of SAM in the absence of RUTF, how to inter-pret very low weight-for-height z-scores (WHZ) in nutrition surveys andwhether to include them in analyses, assessing nutritional status of disabledchildren and establishing their nutritional requirements.

The Coverage assessment forum area has been busy since its inceptionalmost a year ago. A particularly interesting dilemma was raised recentlyaround handling cases that have been admitted to a community-basedmanagement of acute malnutrition (CMAM) programme with WHZ criteria,but a mid-upper arm circumference (MUAC) above 115mm and no oedema.In the situation in question, a SAM case was defined for the Semi-Quantitative Evaluation of Access and Coverage (SQUEAC) survey as MUAC< 115mm and/or presence of bilateral pitting oedema, while admissioncriteria to OTP included presence of bilateral pitting oedema, WHZ < -3 orMUAC < 115mm. The questioner was unsure as to whether a child admittedto the OTP using WHZ criteria should be considered a ‘recovering case’ or nota case at all in the coverage survey.

While responses presented different scenarios and reasoning for bothinclusion and omission of the child from point coverage and period cover-age calculations, there was consensus that the number of such children islikely to be very small in most surveys and therefore unlikely to greatly affectthe overall coverage figure. However, an issue was raised as to whether it isappropriate to use a different case definition for admission into aprogramme as that for estimating coverage, as the coverage estimate doesthen not necessarily measure the same group of children who are enrolledin CMAM in the area. In programmes where nearly all admissions come fromthe community, assessed by MUAC and/or oedema, and there are very fewWHZ admissions, there would probably be little impact on the coverageestimate. But in others it might be worthwhile confirming with programmedata first that the proportion of children admitted to a CMAM programmebased on WHZ is very small before excluding WHZ as a criterion in coverageassessment.

The discussion concluded with the observation that SQUEAC is a flexiblemethodology that can be adapted to different contexts, and a reminder thatthe main objective of SQUEAC is not to give a precise coverage figure but toexamine and understand ‘barriers and boosters’ to achieving good cover-age. Experience from coverage assessments to date suggests that coverageis generally low and this is not because of missing out the children admittedto programmes by WHZ, but because the community component ofprogrammes is still weak. More attention and initiatives are needed to effec-tively address the community mobilisation aspects of programming.

To view the full discussion, go to http://www.en-net.org.uk/question/846.aspx

In response to growing interest and experience in the area of urban foodsecurity and nutrition, en-net has launched a new forum area on urbanprogramming, http://www.en-net.org.uk/forum/14.aspx.

The first question focuses on: Urban malnutrition – what do we know?

Urban malnutrition data is very difficult to ascertain, because surveys do notdifferentiate between urban and rural, or because there is a view that ruralareas are more vulnerable. In cases where urban malnutrition rates are reportedthere is often a failure to disaggregate data by poor areas and those that arebetter off, so that the specific vulnerabilities of disadvantaged groups areconcealed. Please share your views on this, experiences or research findings athttp://www.en-net.org.uk/question/873.aspx

Discussions will feature in a special ‘urban edition’ of Field Exchange plannedin the first half of 2012 (Issue 46), see news piece this issue.

To join any discussion on en-net, share your experience or post a question,visit www.en-net.org.uk

Contributions from Freddy H, Mark Myatt, Lio Fieschi, Gwyneth, Jose LuisAlvarez, and Ernest Guevarra.

en-net update,August toOctober 2012By Tamsin Walters, en-net moderator

42

ALNAP and UN-Habitat have recently launched an UrbanHumanitarian Response Portal - a knowledge sharing platform,with a focus on disaster preparedness, relief and early recovery

in urban crises. It fulfils a recommendation made by the ALNAPmembership at the 27th Annual Meeting to bring ‘scattered learning’about urban humanitarian response together in one place. It not onlycontains a growing collection of reports, evaluations, meeting docu-ments and lessons papers related to humanitarian action in cities, butalso information on upcoming events and a forum to share questions,opinions and ideas.

Those interested in urban humanitarian action are strongly encour-aged to contribute to the Portal by submitting new documents or sharingtheir thoughts via the dedicated forum ‘Humanitarian Response inUrban Crises’ to ensure that the latest learning and information is avail-able to the humanitarian community as a whole. The hope is that thisPortal will help humanitarian practitioners in their work and fosterlearning around humanitarian response in cities.

Visit the Portal here: www.urban-response.org

About ALNAPThe Active Learning Network for Accountabil ity and Performance inHumanitarian Action (ALNAP) was established in 1997. ALNAP aims toimprove humanitarian performance through increased learning andaccountability. It incorporates most of the key humanitarian organisa-tions and experts from across the humanitarian sector. Members aredrawn from donors, NGOs, the Red Cross/Crescent, the UN, independ-ents and academics.

To keep up to date with ALNAP’s work, visit www.alnap.org

About UN-HabitatThe United Nations Human Settlements Programme, UN-HABITAT, isthe United Nations agency for human settlements. It is mandated by theUN General Assembly to promote socially and environmentally sustain-able towns and cities with the goal of providing adequate shelter for all.

Visit http://www.unhabitat.org

Urban HumanitarianResponse Portal

News

Proceedings ofUNICEF/GNC/SC UKworkshop on IYCF in development and emergencies

The report is now available of a 5 day workshop on infant andyoung child feeding (IYCF) in development and in emergencies(IYCF-E) held from 25th-29th June, 2012 in London. The work-

shop was organised by Save the Children UK, funded and in partnershipwith UNICEF (IYCF and Emergency Units) and the Global NutritionCluster (GNC).

The overall objective of the workshop was to support agencies toimprove non-emergency IYCF programming (essential emergencypreparedness) and to inform development-sensitive IYCF-E program-ming. The workshop was attended by 67 participants representinginternational non-governmental organisations (INGOs), United Nations(UN) agencies, donors, academics, government, and independentexperts. The workshop brought practitioners working on IYCF in bothemergencies and development contexts together.

The specific aims of the meeting were to share experiences and chal-lenges in IYCF and IYCF-E programming in different contexts, todisseminate IYCF and IYCF-E policies and capacity development tools,and to discuss recent developments in programming, assessment/M&Eand coordination. Two days each were devoted to IYCF and IYCF-Erespectively, with half a day spent on discussing the necessary “hand-shake” needed to bridge the emergency/development ‘divide’ that oftenexists. Action plans were developed through the course of the workshopand finalised in Day 5 (these will be followed up in 3-6 months).

The workshop opened with a review of the global situation on IYCFand IYCF-E and was followed by a participatory gap analysis onIYCF/IYCF-E programming and ‘infrastructure’. There were common areasof IYCF/IYCF-E concern in terms of poor understanding of M&E and indi-cators, lack of programme funding, lack of policy/poor implementation,lack of understanding of what constitutes ‘good’ IYCF/IYCF-E program-ming and few trained frontline staff.

The workshop highlighted that for IYCF, key policies, strategies,design and implementation tools for IYCF recommended interventionsare now available. The Programming Guide on Infant and Young ChildFeeding (2012) by UNICEF was highlighted as a key resource for IYCFprogramme planning and implementation. While IYCF-E is alsosupported by a variety of policies, strategies, guidelines and resources, itstill lacks key tools to support implementation.

Four key issues arose through the workshop with a call for immediatecommitment and action: • There is a need to explore how to better link IYCF and IYCF-E in

practical terms that includes lesson learning from different contextsand developing linkages between emergency and development programming.

• There is a need for practical ‘how to’ guidelines – including agreed core M&E indicators – for IYCF-E programming in different contexts. This gap is limiting programming and funding

• There is a need to better ‘package’ IYCF-E to encourage more internal (agency) buy-in from senior management and communications/advocacy teams, as well as from donors. As an action from the workshop, a letter was sent to a number of bilateral donors high-lighting this (and included in report annex)

• There needs to be clarity about role and responsibilities for moving the IYCF/IYCF-E agenda forward taking into account existing UN mandates, agency expertises and capacities.

Finally, the report recommends the development of a strategy on IYCF-Esetting out the key actions needed and timeframe to aid in coordinatingefforts and to serve as an advocacy tool. It concludes with a recommen-dation for an urgent meeting between key players in order to discussoutcomes of this workshop and to frame this strategy.

For more information, contact: Ali Maclaine, SCUK, email: A.Maclaine@savethechildren.org.ukAccess the full report at: http://www.unicef.org/nutritioncluster/

Urban edition of FieldExchange: call forsubmissionsWe are planning a special edition of Field Exchange in the first halfof 2013 focused on nutrition and food security emergency program-ming in urban contexts. The ‘emergency’ scope includes chronicsituations, preparedness and recovery and we will consider relevantexperiences, research and news from development contexts.

We are gathering material for this now. Please send us anyideas/outlines/research or news that you would like to feature.Please also share this email with your colleagues or send us contactsthat we can follow up.

Submit ideas, articles, research or news to: Marie McGrath,marie@ennonline.net or call: +44 (0)1865 249745

43

Resources to supportprogramming forolder people

New Global FoodSecurity journal

HelpAge International have produced a variety ofresources to support good practice regarding theelderly in emergencies including guidance onincluding older people in emergency shelterprogrammes, and a checklist on preparing for anemergency.

In addition, Help Age’s new Helping OlderPeople in Emergencies (HOPE) trainingprogramme provides humanitarian practitionerswith the tools and knowledge to ensure olderpeople are included in their programmes. Trainingis targeted at experienced humanitarian practition-ers and policy makers who are responsible forsupporting, planning or managing humanitarianresponse. Trainings have taken place in Nairobi(June 2012), Gaza city (July 2012), Goma, DRC,Bangkok, Thailand, Madrid and Barcelona, Spain(October 2012) and are scheduled for Dhaka,Bangladesh (14 &15 November, 2012) and London,UK (29 & 30 November, 2012

For more resources and information, contact:Pascale Fritsch, email: PFritsch@helpage.org orvisit: http://www.helpage.org/resources

A new journal – Global Food Security – is launch-ing in 2012.The journal aims to provide readerswith:• Strategic views of experts from a wide range of

disciplinary perspectives on prospects for ensuring food security, based on the best available science, bridging the gap between biological, social and environmental sciences.

• Reviews, opinions and debates that synthesize, extend and critique research approaches and findings on global food security.

The journal will publish research on the followingelements of food security:• Availability (sufficient quantity and quality)• Access (affordability, functioning markets and

policies)• Nutrition, safety and sanitation • Stability and environment (resilience and

ecosystem services)

Issues of the publication will contain severalpapers that address specific, timely topics ofimportance to food security. Authors will berecognised authorities in their field. It will focuson interdisciplinary aspects of food security chal-lenges at national to global scales, and will look tochallenge current paradigms, seeking to provide‘out-of-the box’ thinking on global issues.

The first edition (October 2012) features 11 articleson lessons learned from the Somalia 2011 famine.

Visit: http://www.sciencedirect.com/science/journal/aip/22119124

News Views

Personal experiences of workingduring the Horn Africa crisis in 2011

BackgroundDrought conditions in the Horn of Africa,particularly in Somalia, decimated the2011 crop and livestock production caus-ing dramatic increases in food prices.Prolonged insecurity in Somalia also ledto a significant reduction in internationalfood assistance and in July and August2011, the United Nations (UN) declaredfamine in certain regions of the Horn ofAfrica. During the build-up, largenumbers of Somali’s moved across theKenyan border to seek assistance in thelong-established Dadaab refugee camps.At the peak of the crisis, over 1000refugees arrived in the camp each day. InAugust 2011, it was agreed that a nutritionsurvey was needed to assess the magni-tude of the nutrition and mortalitysituation. Coincidentally, August was thescheduled time of the routine annualsurvey in Dadaab.

The ENN has been collaborating withUNHCR over the last two years providingtechnical support and developing guid-ance for conducting standard nutritionsurveys. Two members of the ENN techni-cal assistance team at the time (MélodyTondeur and Sarah Style) went to Dadaabfor a 6-8 week period to support thedesign and implementation of the surveysand at the same time, help strengthen thecapacity of UNHCR and implementingpartner’s (IPs) in executing surveys. Thisarticle describes their more personal expe-riences shared in their ENN interview.

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

Dadaab was the first exposure to a largescale refugee situation for both Mélodyand Sarah and, for Sarah, her first refugeecamp experience. Prior to departing,Mélody had regular contact with theUNHCR Senior Nutrition Associate(Gloria Kisia) based in Dadaab and bothwere able to get useful insights from read-ing situation and survey reports, as wellwatching the news and seeing images ofthe camps and the camp outskirts. ForSarah, the main camps were as she hadimagined but the environment in theoutskirts of the camp was harsher than shehad expected in terms of the strong winds,dust, land contamination by human faecesand limited latrines.

How would you describe the prepara-tions and process you undertook forcarrying out the surveys in Dadaab?

Four weeks prior to their arrival, prepara-tions were undertaken by Gloria on theground with the assistance of Mélody,including putting together a pre-surveyschedule, compiling survey question-naires, and organising survey logistics.Through other project activities, Mélodyhad already met and discussed the situa-tion via email and in person with theUNHCR nutritionist and other surveystaff with whom they would be closelyworking. Once in Dadaab, another twoweeks of planning was undertaken by

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 assis-tance project to UNHCR , who oversaw surveys in Dadaab camps in Kenya during theheight of the crisis in 2011. It describes their personal experiences in carrying out thesurveys and their views on approaching large scale emergency refugee situations. BothMélody and Sarah now work with UNHCR.

Settlements of new arrivalsin the outskirts of Dadaab

Mélo

dy T

ondeur

and S

ara

h S

tyle

, Kenya,

UN

HCR,

2011

waiting to be further assimilated into the newlyestablished camps, created its own samplingchallenge, since there were no structures (likethe camp blocks) to delineate clusters. The teaminnovated by using spatial random samplingwith GPS and to their surprise, this ended upbeing easier to administer than anticipated. Forsampling the outskirts, the perimeter wasmapped using a GPS and the tracks andwaypoints uploaded and visualised usingGoogle Earth (version 6). Random spatial clus-ters were then allocated to the outskirt areamap using a grid overlay. Cluster start pointswere subsequently located by survey teamsusing GPS handsets and second-stage samplingproceeded using proximity sampling.

A third challenge was the decision to includechildren from 5 - <10 years of age using MUACwhen there is no international agreement on thecut-off point to use for defining malnutrition. Inthis case, a MUAC cut-off of 14cm and the pres-ence of oedema were used for referral duringthe survey and then weight and height weretaken at the health posts for further screening.

Further, Mélody and Sarah added, thepopulation data wasn’t reliable. Due to theinflux of refugees, there was a significant back-log of unprocessed registrations at the time ofthe survey planning with refugees settling inthe outskirts and in the main camps. This meantit was difficult to gather accurate populationfigures and derive demographic data for surveyplanning, including the % under five and aver-age household size. Quota sampling wastherefore used and this presented some of theteams with difficulties, as they had been used tofixed household sampling (e.g. surveying 13households a day and not counting the numberof individuals in each target group surveyed).

Another challenge was in keeping the house-hold questionnaires reasonably manageable toensure quality of the information collected.Nonetheless, the questionnaires were quitelong and complex compared to previous years(largely because of the addition of the mortalitydata and the measurement of the chil-dren 5-<10 years) making it difficult forsome of the survey teams to administer.

Were there other unforeseen challenges?

Unfortunately an MSF staff strike coin-cided with the survey timing and forcedlast minute changes to the survey so thattwo surveys had to run concurrentlyfollowed by another two. This put theteam under considerable time and logis-tical pressure.

Also, the survey teams found it hardto keep track of quotas of individuals forthe questionnaires meaning that someteams needed a lot of support for datacollection. Coordinators had to spendentire days with some teams to keepthem on track to complete one cluster ina day. Mélody and Sarah described howthere was little leeway to absorb anydelays for several reasons includingbudget, logistics and because thesurveyors had to resume their regularjobs at the health clinics and at thecommunity level. The pressure was onwith the outskirts survey especially,because of the new arrivals being relo-cated into newly established campsright as the survey was happening.

How did you involve key stakeholders in thesurvey?

A great deal of work went into ensuring that thekey stakeholders (UNHCR, WFP, UNICEF, IPsand other agencies working in the camps) werefully part of the decision making about thesurvey design, choice of indicators and ques-tions contained in the household questionnaire.Survey meetings and correspondence was coor-dinated by the Dadaab Camp UNHCR SeniorNutrition Associate (Gloria).

In Dadaab, an efficient model for nutritionsurveys had been implemented in previousyears and hence the same model was used.Each IP was designated to survey the campthey work in and had clear roles and responsi-bilities, survey supervisors, etc. Some partnerstook on the role of helping with the pre-surveytraining, supervision and data entry/data qual-ity checks. Overall, the process was successfuland there was considerable good will and coor-dination to enable the surveys to take place assmoothly as possible.

How did you record and store the data?

ENA for SMART software (Delta version, June20th 2011) was used by the supervisors andcoordination team to enter and analyse theanthropometric and mortality data, and asystem of double checking was used to verifyall data entered. All other data were doubledentered by four clerks using Epi Info Software(CDC, version 3.5.1) and data analysis wasdone using ENA for SMART and Epi Info soft-ware. Double entry was essential since dataentry clerks could not be supervised during theday by the coordination team while the surveywas on-going.

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

This was a major priority so the data entrypersonnel entered the data in parallel to thedata collection to speed up the process of feed-back. Three nutritionists (2 from UNHCR and 1

Mélody, Gloria and Terry Theuri (a UNHCRNutrition and Food Security consultantdeployed to Dadaab during the crisis) with keystaff from implementing partners (MillicentKavosa from IRC). They were joined for twoweeks by a third ENN team member from theUNHCR technical assistance project (AndySeal) to help design and plan the survey.Together, the team went to the three maincamps which make up Dadaab to understandthe lay-out (for sampling purposes) and the set-up of the outskirts, talk to the nutritionprogramme staff to understand food distribu-tion and feeding schedules so that the hourscould be modified during the survey to avoidabsentees, to look at equipment and quality ofequipment and to discuss the referral processfor children and women needing treatment.

Mélody and Sarah felt that an enormousadvantage in Dadaab with respect to the surveywas that some of the key staff (2 from UNHCRand 1 from IRC) had considerable surveycapacity and experience and had received priortraining on the survey guidelines as part of theUNHCR/ENN collaboration. In this respect,Dadaab was well able to accommodate emer-gency survey work of this type. In addition, thestaff were completely open to strengtheningtheir skills in this area. Even though thedemands on the survey teams were enormous,capacity strengthening was carried out (5 daysof training involving 4 days in the classroomand 1 day in the field). Ensuring quality was amajor focus during the training sessions and asdata were collected. In total, 80 surveyors weretrained.

A survey was conducted in each of the threemain camps as is usually done every year inDadaab. This year, a fourth survey was alsoconducted in the outskirts of one of the camps.Each of the four surveys had five survey teamsof 4 people each (1 team leader, 2 enumeratorsand 1 translator), supported by survey teamcoordinators and supervisors. There were dailyfeedback sessions in the mornings. Theevenings were spent entering data and check-ing questionnaires so that the UNHCR/ENNcoordination team could inform the surveyteams of any improvements needed throughoutthe survey.

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

Were there specific methodological challenges?

Mélody and Sarah described how one of themain challenges during survey planning was todecide which data to obtain from the newarrivals and which from the long-standingrefugees, as agencies needed information onboth populations to aid decision making.Because the new arrivals were in the maincamps (thought to represent 10-20% of the maincamp population) and in the outskirts (wherethe entire population was new arrivals), doingone survey covering both new arrivals andlong-standing refugees would have required astratified survey design which would need a lotof work. Hence the decision was taken tosurvey one outskirt area in addition to the threemain camps to obtain data on new arrivals.

The decision to survey the outskirts, wherenew arrivals from Somalia would congregate

44

Using GPS handset tofind cluster duringsurvey of outskirts

Views

Mélo

dy T

ondeur

and S

ara

h S

tyle

, Kenya,

UN

HCR,

2011

Saul Guerrero is the SeniorEvaluations, Learning andAccountability (ELA)Advisor at ACF UK basedin London. Prior to joiningACF, he worked for Valid

International Ltd. in the research, develop-ment and roll-out of CTC/CMAM. He hasworked in over 18 countries in Africa andAsia.

Maureen Gallagher is theSenior Nutrition & HealthAdvisor ACF USA based inNew York. She has workedfor the last 10 years innutrition, food security

and hygiene promotion programming inNiger, East Timor, Uganda, Chad, DRC,Burma, Sudan and Nigeria.

The authors would like to thank nutritionadvisors across ACF International for theirdirect contribution, in particular Anne-Dominique Israel, Sandra Mutuma, CecileBasquin, and Silke Pietzsch. We would alsolike to thank all ACF nutrition advisors ontheir contributions in discussions on theissue including Elisa Dominguez, MarisaSanchez, Nyauma Nyasani, Marie-SophieWhitney, Olivia Freire, Fabienne Rousseauand Maria Masferrer. Our gratitude alsogoes to all the ACF International field coor-dinators and MoH staff for their on-goinginvolvement in reviewing the past, presentand future of CMAM programmes.

45

relax. Mélody echoed the latter and emphasisedthe importance of taking time for yourself,getting a good night’s sleep and either read orwatch a movie before going to bed to helpswitch-off.

For both Mélody and Sarah, having theopportunity to work on a daily basis with theUNHCR highly experienced and long standingnutritionist based in Dadaab, as well as the otherUNHCR,ENN and IP nutritionists, was vital tounderstand and deal with the context –both forthe survey design, data collection and interpre-tation of the data. All members of thecoordination team complemented each otherwell, each bringing their own knowledge, skillsand experience.

What advice would you give to others going tosupport large-scale surveys in complex environ-ments such as Dadaab?

Both Mélody and Sarah would recommend to bevery organised and ensure that the relevantsupport within the camp is also there. It is veryimportant to consider the security advice givenand to keep track of the security situation ondaily basis.

Contact people ahead of time so you arriveand know what you need and know what ques-tions you have. Compile as many backgroundreports as possible as these tend not to be read-ily available and avoid asking questions on theground when you already have access to theinformation. Also, it helps to keep notes of whatyou’re learning so you don’t have to keep askingthe same questions. This reduces the pressure onfield staff who are already very busy and canend up ‘repeating information for the benefit ofmany visitors’

Capacity building is possible in a large scaleemergency situation. The ENN team were ableto strengthen the capacity of key UNHCR staffand IP staff on how to execute quality surveys.Innovation is also possible, as together the teammanaged to solve the sampling challenge in theoutskirts.

On a personal front, Mélody and Sarahemphasised you need to be fully prepared to livein difficult conditions (live in tents, share roomswith others, don’t have your home comforts),work very long hours during survey implemen-tation (average from 7am until 11pm) and to putaside other work demands so that the only focusis on the survey. Mélody did this by avoidingunrelated emails.

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

Both concluded that the Dadaab experience is agood example of the importance of preparationfor a nutrition survey during an emergency situ-ation, i.e. in this case, waiting and working toensure there was adequate information fromrapid assessments and screening to informsample size, survey design and the content ofthe questionnaire.

Despite being a very challenging context towork in – both professionally and personally – italso a rewarding one, reflected in the fact thatMélody and Sarah have continued to work withUNHCR and remain full of enthusiasm!

For more information, contact: MélodyTondeur, email: melotondeur@gmail.comSarah Style, email: sarahlstyle@gmail.com

from ENN) were involved in analysing thedata for different indicators and a fourth fromENN, for training on data analysis andmanagement, as well as oversee and doublecheck all the results. Further, the survey reportwas split into separate sections and the writ-ing/reporting allocated to the four mainnutritionists, with one of the nutritionistshaving the responsibility to oversee the wholereport writing process.

Impressively, the GAM/SAM results in chil-dren 6-59 months were fed back ‘informally’2-3 days after the surveys (by email). Theresults on all indicators were fed back officially2 weeks after the surveys were completed, toallow partners to start working on the recom-mendations and new interventions. This wasdone through a full partners meeting. The finaland detailed report was disseminated twomonths following the partners meeting.

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

In general, the malnutrition results were reallyhigh, even in the main camps. The team hadpredicted these results from rapid assessmentsand screening results but the results from theoutskirts were higher than assumed. Partnerswere especially alarmed with the GAM preva-lence 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 themain camps and quite low, however they werevery high in the outskirts (Crude death rate1.23 (0.73-2.06), U5 Death Rate 3.02 (1.72-5.24)).Coverage of the nutrition programmes (SFPand OTP) was found to be very low though thereasons for this were not clear and consideringalso that nutrition surveys are not the best wayto establish levels of coverage. In addition,screening in the community was based onMUAC and oedema whereas admission wasbased on MUAC, oedema and weight forheight z score (WHZ). This meant that esti-mates of coverage varied depending on ifWHZ was included or not.

Mélody and Sarah weren’t surprised by thesurvey results for children in the 5-<10 year oldcategory in the outskirt survey (MUAC <14 cmwas 18.3% in the outskirts compared to 6.8-12.2% in the other camps) – screening of newarrivals had indicated this and they remem-bered being shocked by how thin theywitnessed the older children to be during theoutskirts survey. It was evident that not onlychildren 6-59 months, but also older childrenhad been affected by the crisis. However know-ing what to do with these children was more ofa challenge given the limited knowledge andresearch available on this age group.

There was a significant decrease in levels ofanaemia in children and women, which was avery positive finding for UNHCR, WFP andthe IPs who had been implementing anaemiacontrol measures in Dadaab in previous years.

What new things did you learn from your expe-rience in Dadaab?

For Sarah, this was the first time she had takenpart in a large scale survey at the height of sucha large scale emergency and two key lessonsstood out ; first be flexible as in such situationsthings are constantly changing and second,being able to switch off and to find ways to

Quantity throughquality: Scalingup CMAM byimprovingprogrammesaccessBy Saul Guerrero & Maureen Gallagher

Views

A child being assessed(appetite test) at a healthfacility offering treatmentin Monrovia, Liberia

46

BackgroundACF International’s strategy for scaling upnutrition programming is underpinned by onesimple idea: we are not reaching enough of theaffected population. According to the 2008Lancet series, only between 5%1-10%2 of chil-dren suffering from acute malnutrition arereceiving nutritional care. To deal with this,ACF International has committed to a gradualprocess of scaling up interventions to reach anestimated 500,000 children per annum by theyear 2015. In 2010, the organisation reached justunder 225,000 children through itsprogrammes. To meet its objectives, the organi-sation must effectively double its currentcaseload. To achieve this, ACF has placedconsiderable emphasis on securing the neces-sary political will and civil society participationat national and international level to enablegrowth and expansion of services3. Internally, ithas also placed great emphasis on partnershipsand capacity building as a means of enablingprogrammes to expand and reach newgeographical areas.

All in all, this represents an outward visionof growth, an approach that favours expansionover consolidation; the replication of existingapproaches on the assumption that more of thesame will deliver results. The implicit confi-dence in the performance of existingprogrammes is undoubtedly linked to the greatstrides made over the last decade with thedecentralisation of care. By shifting fromTherapeutic Feeding Centres (TFC) to CMAM,the organisation has laid the foundation for asignificant increase in programme uptake. Yet,like many nutrition organisations around theworld, ACF is gradually coming to the realisa-tion that the shift in treatment approaches is noguarantee for success. There is now an increas-ing body of evidence showing that offeringservices, even closer to the communities, is nottantamount to improving access. Whilst theefficacy of the CMAM model and protocol is

now firmly established, its effectiveness is stilldependent on the quality of programme imple-mentation by (or with the support of) NGOs, asoften Ministries of Health (MoH) have limitedresources for nutritional activities.

Scaling up must therefore start by consoli-dating our work, by finding ways of reachingthose that we are consistently excluding. Thepotential benefits of consolidating our work, orshifting to an inward approach for scaling up,has been conclusively established. If we were toincrease our current coverage by 30% we wouldnot only meet international minimum stan-dards, but we could reach our target of 500,000children per year without opening a singleadditional programme. We do not yet have ageneric recipe for effective CMAM program-ming, what we do have is sufficient evidence tostart developing a new approach that focusesmore closely on ensuring programme effective-ness. The aim of this paper is to use some of theavailable evidence to propose a clear andspecific definition of effective CMAM program-ming. But it also seeks to go further. Byreviewing results of current trends in program-ming, it sets out to identify key steps forconsolidating ACF’s existing programmes, andto outline the programmatic and organisationaltransformations that these would have for ACF.

What’s in a number? Defining quality innutrition programmesThe aim of public health nutrition interven-tions, such as CMAM, is to meet the needs ofthe largest possible proportion of an affectedpopulation4. There are two sides to the questionof needs met. On the one hand, there is thequality of care or efficacy of treatment. This isgenerally assessed through standard nutritionindicators including cure, death, defaulter andnon-responder rates. Although these indicatorsvary from context to context and according to anumber of factors (including severity at presen-tation, level of compliance, implementationapproach, etc), there is now ample evidence tosupport the idea that the efficacy of CMAMtreatment protocols is close to 100%5 in acontrolled environment. These efficacy indica-tors, however, give us only a partial view of theimpact of the programme on the childrenreached. The other part of the needs met equa-tion, the one often missing in our analysis,concerns the number of affected cases thatprogrammes do not reach. The quality of nutri-tion programmes must be determined by acombination of the treatment efficacy outcomesvis-a-vis the proportion of the affected popula-tion being reached (coverage).

Coverage has been an integral component ofhumanitarian evaluative frameworks, withboth OECD-DAC and ALNAP6 recommendingthat humanitarian agencies “…present an esti-mate of the proportion of those in need covered,expressed as a percentage, rather than an absolutenumber”. The importance consistently attrib-uted to coverage is exemplified in Figure 1.

When programmes with coverage of 30% (A)successfully cure a high rate of children (B), theproportion of needs met is still low (C). Whenprogrammes with high coverage (X) cure onlyhalf of the admitted cases (Y) the proportion ofneeds met is higher (Z).

Since they first appeared in 20048, theSPHERE Standards have included specificcoverage indicators for nutrition interventionsin rural (50%), urban (70%) and camp environ-ments (90%). The timing of the introduction ofthese new standards was important. Prior to theintroduction of the CTC model in the late 1990s,inpatient programmes rarely reached over 25%of the affected population9. But by 2001, as CTC(and its successor, CMAM) became morecommon, NGOs consistently showed that itwas possible for programmes to reach up to70% coverage. This was taken as a sign of theintrinsic quality of the approach, rather than asa partial reflection of the active and directinvolvement of NGOs in its implementation.

Great expectations: ACF programmecoverage performance against international standardsACF has a strong history in supporting cover-age estimations (see Box 1). Between February2010 and February 2012, ACF has carried out 15coverage surveys in 11 different countries, withmore planned for the short and mid-termfuture. Whilst the contexts have varied signifi-cantly, regular coverage surveillance hasprovided us with two valuable areas of infor-mation: coverage diagnosis and programmediagnosis. Together, these two areas of informa-tion provide clear ideas of where the problemsand the solutions lie, offering practical recom-mendations for improving/ensuring thequality of programmes.

Based on the coverage monitoring datacollected since 2007 (Table 2), on average ACFprogrammes achieve coverage of around 30%of the total affected (SAM) population10; forevery three acutely malnourished children inour areas of operation, only one receives treat-ment through our programmes. None of theprogrammes surveyed so far has met theSPHERE minimum standards (>50% for ruralareas, >70% for urban areas, >90% for camps).The comparison against SPHERE standards mayonly be relevant to some of these programmeswhich were implemented (directly by ACF)

1 Horton et al (2010). Scaling Up Nutrition: What Will It Cost?

(World Bank, Directions in Development/Human Development,

2010, p.19). This is based on the estimations of the authors

that only 1 million of the total 19 million children suffering

from SAM (c. 5%) are receiving treatment.2 Based on UNICEF’s more recent estimations (PD-Nutrition

Section E-Bulletin, Issue 1, October 2012, p.2) the actual

number of children receiving treatment is closer to 1,961,772

which suggest that the proportion of cases receiving treat-

ment could be closer to 10%.3 ACF International, (2010). ACF International Strategy 2010-2015

Figure 1: Coverage vs. effectiveness in nutritionprogrammes

Coverage = 70%

Coverage = 30%Met Need = 35%

Met Need = 27%Effectiveness = 90%Perc

enta

ge (%

)

100

80

60

40

20

0

X

Y

ZA

B C

Effectiveness = 50%

4 Needs based on SAM caseload as defined by National

Nutrition Protocols5 In controlled settings, in uncomplicated incident cases with

MUAC at or just below admission criteria/mild oedema. 6 ALNAP (2006) Evaluating humanitarian action using OECD-

DAC criteria: an ALNAP guide for humanitarian agencies

(Over-seas Development Institute, London, March 2006, p.

38-39)7 Adapted from Sadler, K, Myatt, M, Feleke, T and Collins, S

(2007). A comparison of the programme coverage of two

therapeutic feeding interventions implemented in neigh-

bouring districts of Malawi (Public Health Nutrition, April

2007, 10(9), p.912)8 The SPHERE Project (2004). Humanitarian Charter and

Minimum Standards in Disaster Response (First Edition,

2004, London, p.39) 9 Vautier, F (1998). Selective Feeding Programmes in Wadjir:

Some Reasons for Low Coverage and High Defaulter Rate

(Field Exchange, Emergency Nutrition Network, Issue 5,

p.17). Coverage calculated indirectly (using prevalence

data against population estimates).10 All figures refer to point coverage.

Views

Ben A

llen,

Lib

eria,

2011

47

Views

under emergency conditions. The remainder ofthese programmes have been implemented in part-nership with (i.e. indirectly through nationalMOH. The SPHERE standards represent a set ofbenchmarks that have proven difficult to attain byintegrated programmes led by MoH. This hasresulted in the commonly held belief that inte-grated CMAM programmes run by MoH areintrinsically limited by infrastructure andresources, and are thus naturally unable to meetprevious CMAM outcomes and standards. Whilstintegrated CMAM programmes are different,focusing on the need for new standards for inte-grated programmes (or provision of externalinputs to support achieving standards) representsan easy way out. What this argument effectivelydoes is remove the pressure from NGOs to under-stand the factors affecting the performance ofintegrated CMAM prog- rammes. This in turnoften leads to the creation of artificial programmeconditions – including additional staff, financialincentives and supply systems – that help boostperformance, but does little to strengthen localhealth systems.

The challenges faced by integrated CMAMprogrammes are very real: beneficiary popula-tions, for instance, have a pre-conceived idea ofwhat health facilities can and cannot offer, aboutthe kind of (staff-patient) treatment they are likelyto receive there, including official and/or unoffi-cial costs of treatment. The location (often limitedand sparse) of health facilities in a given areameans that reaching those in need is generallydetermined by the proximity offered by the MoHinfrastructure chosen as service delivery units forCMAM services. In addition, the resources(human and financial) available to MoH to carryout the supporting functions needed by anysuccessful nutrition programme are very oftennot there. ACF has sought to address this, but thecoverage performance of these programmessuggests that the current allocation of technicalsupport and resources to support integratedprogrammes is based less on needs and gaps ofhealth systems and more on the traditional organ-isational clinical expertise and focus on treatmentefficacy. Meeting a higher proportion of the needs,and reaching the expected 500,000 SAM childrena year, will require that ACF looks beyond clinicaloutcomes and addresses key factors for achievinghigh quality programmes.

Key factors for achieving high qualitynutrition programmesPart of the answer to the question of wheresupport is needed to improve the quality andperformance of our programmes is provided bythe coverage assessments themselves. In 2007, theACF Uganda nutrition team set out to prove thatprogramme coverage could be increased throughthe strengthening of community mobilisationactivities, including sensitisation, case-findingand follow-up. The programme succeeded inincreasing coverage by more than 12% in 12months12. More recently, in December 2010, acoverage assessment carried out in Kanem (Chad)once again highlighted the need for improvedcommunity engagement. By taking on board therecommendations from this assessment, theprogramme was able to increase its coverage byover 9% in 10 months. Together, these two experi-ences prove that within a relatively short period,programmes can positively influence their cover-age by addressing some of the bottlenecks13

affecting access and that community mobilisationcan play a pivotal role in achieving this.

Since then, our understanding of the factorsaffecting programme performance has increasedthrough available data from within and withoutthe organisation. A review of 12 CMAMprogrammes published in 2010 concluded thatprogramme coverage was directly affected by 1)the degree of rejection amongst referred children,2) the level of awareness (about the condition andservices) amongst the population, and 3) thedistance between targeted communities and serv-ice delivery points14. The recognition andincorporation of mid upper arm circumference(MUAC) as part of national nutrition protocolsfor admission in many countries has significantlyreduced rejections in programmes15. The otherfactors, however, continue to negatively impactprogramme performance. A pilot study carriedout in 23 health centres by Concern Worldwide inEthiopia identified lack of awareness about theprogramme as the single most important barrieraffecting the performance of the integratedCMAM programme, preventing any of the facili-ties evaluated from reaching more than 50% ofthe affected population16.

ACF’s recent surge in coverage surveillancehas created a body of evidence that corroboratesthese conclusions (see Table 3). In all coveragesurveys carried out by ACF since February 2010,awareness about programme and/or malnutri-tion has been identified as the primary reason(s)for non-attendance. Simply put, the large major-ity of people in the communities where ACFworks remain unaware of the existence of CMAMservices, or do not perceive it as the solution tothe condition affecting their children. Theevidence suggests that the current approach toimplementing and supporting integrated-CMAMprogrammes is inappropriate to deliver thepromise of greater access that the CMAM modelwas built on. As the CMAM approach continuesto be scaled-up and rolled out, it is more pressingthan ever to revisit and review the current modelinvolving health systems, communities andnutrition organisations. This does not mean start-ing from scratch; the wider public health sectorhas been tackling these issues for years, gatheringvaluable lessons and experiences that can bebrought into the fold.

Rethinking the CMAM service deliverymodel and the role of nutrition organisationsThe role of humanitarian agencies has shiftedsignificantly since the direct interventions of the1970s and 1980s. Organisations involved inpublic health programmes have gradually scaledup by working in partnership with local authori-ties. For nutrition organisations like ACF, this hasrepresented a shift from direct implementation to“focusing on strengthening health systems’ own

Box 1: A little history of ACF and coverage estimation

The introduction of international coveragestandards for nutrition programmes raisedthe need for a reliable means by which tomeasure coverage. Indirect methods, usingpopulation estimates and prevalence wereunreliable, and a more direct method wasneeded. Since direct means for measuringprogramme coverage were first developedby Valid International and their partners inthe early 2000s, ACF has increasinglysupported their testing, development andintroduction into regular programming. Thefirst method, the Centric Systematic AreaSampling (CSAS) approach was first used byACF in 2007 to determine programme cover-age11 in Burundi, Uganda, and Sudan. In2008, ACF became one of the first organisa-tions to support Valid International inpiloting the Semi-Quantitative Evaluation ofAccess & Coverage (SQUEAC) method,designed to make coverage surveillanceeasier, faster and less-resource intensive.Since 2010, the use of the new SQUEACmethodology has enabled the organisationto systematically monitor and diagnoseprogramme coverage.

11 As defined by National Nutrition Protocols and corresponding

admission criteria.12 Doledec, David (2008). Impact of community mobilisation

activities in Uganda (Field Exchange, Emergency Nutrition

Network, Issue 34, October 2008, p. 15)13 Not all bottlenecks can be addressed rapidly or through

community mobilisation. For a more detailed discussion, see

Tanahashi, T (1978). Health service coverage and its evaluation

(Bulletin of the World Health Organisation, 56 (2):295-303.14 Guerrero, S et.al (2010). Determinants of coverage in

Community-based Therapeutic Care programmes: towards a

joint quantitative and qualitative analysis (Disasters,

Overseas Development Institute, April 2010, 34(2); 571-585)15 In countries where MUAC has not been incorporated as an

admission criterion, such as Burkina Faso, rejection continues

to be an important barrier to access. 16 Schofield, L et.al (2010) SQUEAC in routine monitoring of

CMAM programme coverage in Ethiopia (Field Exchange,

Emergency Nutrition Network, April 2010, 38: p. 35)17 ACF West Africa Strategy, 2011-2015, p.6.18 CMAM Integration Guide draft, Feb 2011

Table 2: Coverage of ACF-supported nutrition programmes (2010-2012)

Country Location Date Coverage

Burkina Faso Tapoa Feb 2010 21.8%

Chad Kanem Dec 2010 27.1%

Burkina Faso Tapoa Mar - Apr 2011 17.6%

Mauritania Guuidimaka Mar - Apr 2011 33.0%

Liberia Monrovia Feb - Apr 2011 24.8%

Mali Gao Jul 2011 35.4%

Nigeria Yobe Aug 2011 33%

Chad Kanem Sep 2011 36.4%

Chad Bahr el Ghazal Oct 2011 34.1%

South Sudan Gogrial West Oct 2011 44.7%

South Sudan Aweil East Nov 2011 45.5%

South Sudan Twic Dec 2011 27.3%

Myanmar Maungdauw Nov - Dec 2011 40.7%

Haiti Haut Artibonite January 2012 12.4%

Sierra Leone Moyamba February 2012 12.1%

A communityvolunteer screens

children in thecommunity in Yobe

State, Nigeria

Saul G

uerr

ero

, N

igeria,

2011

Views

48

capacities to treat severe acute malnutrition”17.The health system strengthening approach –with nutrition as an entry point – often variesbetween nutrition organisations as well asbetween ACF missions18. Generally speaking,however, there are some fundamental areas forsupport:• Supporting the coordination & creation (or

review) of technical frameworks including national nutrition policy, protocols, guide-lines, and training manuals, as members of National Technical Working Groups.

• Supporting strengthening capacity efforts (i.e. training, coaching, in some cases addi-tional human resources (HR)) for staff involved in the management and imple-mentation of CMAM activities, including national/regional/local managers, health facility staff and outreach workers.

• Strengthening/supporting supply chainmanagement, including systems for forecast-ing, requesting and distributing essential drugsand/or ready to use therapeutic food (RUTF). Though these areas remain key to supportingthe integration of CMAM into national healthsystems, the performance of integrated-CMAMprogrammes continues to falter. Addressingthis requires a rethink of the present and futureof the CMAM service delivery as part of thehealth system and the role of nutrition organi-sations like ACF in this process.

Consolidating experiences in healthsystem strengtheningWhat is ACF’s CMAM health systems’strengthening approach? With so many vary-ing integration definitions and approachesbetween ACF missions, ACF is in the process ofdefining a clear position and model on the keyfactors that ensure ownership, performanceand sustainability of integration of CMAM intoroutine health services. As health systems varyfrom country to country, one model cannot fitall, and it becomes important to learn from ourexisting experiences to see where we are, and todefine a CMAM institutionalisation frameworkthat is adaptable and replicable within differenthealth systems. The emphasis of such a frame-work should consider all key elements forsuccessful treatment yet maintain the existinghealth system as central to decision making.

Since 2009, a CMAM Integration Guide hasbeen under development by ACF. The success

of such a guide rests on its abilityto capture experiences from acrossa wide range of contexts, and inparticular, on its capacity to iden-tify successful approa- ches andprovide practical guidance insome key areas, including:1) Operational Planning (e.g.

what is the case load? How is it managed? How many days of treatment are provided andwhy? What incentive systems are in place and why?)

2) Human Resources (e.g. who is managing treatment? How are health workers involved in treatment? Are additional staff supported? How? Why?)

3) Logistics (e.g. who does the supply management? What is the medicine provision system?)

4) Training (e.g. what is the training approach? Who conducts training? How is training impact measured?)

5) Monitoring & Evaluation (M&E) (e.g. whodoes data collection and how? What is the health system personnel involvement and understanding of M&E?)

In collecting such information in a systematicmanner, ACF can consolidate experiences, iden-tify areas for further research/analysis, anddevelop an operational framework/key princi-ples for integration including timeframes andexit strategies. In this process, factors will beaddressed through the health system strength-ening lens, in exploring how differentcomponents fit into the six health systemstrengthening blocks – service delivery, supply,health workforce, financing, health informationsystems, leadership & governance19. This frame-work will provide a clear position andstrengthen the capacity of ACF in the shift fromdirect implementer to advisor on healthsystems’ strengthening. This requires looking atnutrition as a specific treatment in a largerpublic health setting.

Revising and prioritising communitymobilisationRaising awareness, sensitisation and socialmarketing have long been recognised as keycomponents of successful public health inter-ventions. The challenge for programmesoperating through health structures is the

limited or complete absence of financialresources allocated by most MoH to sensitisa-tion and/or outreach activities. As a result,nutrition organisations like ACF have sought toreplicate the same strategies used in NGO-implemented CMAM programmes, butwithout the financial compensation given tooutreach workers. Instead, integrated CMAMprogrammes increasingly rely on existingcommunity volunteers (generally linked toMoH) to deliver community mobilisation activ-ities. Whilst this avoids creating parallel (andunsustainable) structures, this approach consis-tently faces operational challenges thatultimately define the (poor) performance ofintegrated CMAM programmes.

Improving community mobilisation to fosteroptimal programme coverage is less aboutaddressing the individual challenges associatedwith working with volunteers20, and moreabout redefining the overall paradigm thatplaces individual community members (volun-teers) and sustainability at the heart of acommunity mobilisation strategy. The currentworking model behind community mobilisa-tion in integrated CMAM programmes is basedon two fundamental assumptions: communityvolunteers are the primary means by which toidentify and refer cases, and the activeness ofthese volunteers must be maintained withoutincurring payments which cannot be sustainedby the local health systems. The biggest weak-ness of this model is that it overestimates theimportance of individual volunteers andunderestimates the importance of collectivecommunity involvement. The proportion ofadmissions in CMAM programmes betweenthose referred by individual volunteers andself-referred by communities themselves issuch that a different approach is possible.

Integrated CMAM programmes should aimto create a critical mass capable of triggering amore sustainable dynamic between thecommunity and the services provided (seeFigure 2). At the start of CMAM activities,efforts should be placed on large-scale commu-nity sensitisation and case-finding. This wouldlead to volunteers referring the majority of

Table 3: Top reasons for non-attendance (by country)

ACF-supported CMAM Programmes Reasons for non-attendance

Awarenessabout theprogramme

Awarenessaboutmalnutrition

Distance SFP-OTPInterface

Rejection CarerBusy

WaitingTimes atOTP

RUTFStock-Outs

Husband'sRefusal

Burkina Faso Tapoa Feb 2010

Chad Kanem Dec 2010

Burkina Faso Tapoa Mar - Apr 2011

Mauritania Guuidimaka Mar - Apr 2011

Liberia GreaterMonrovia

Feb - Apr 2011

Mali Gao Jul 2011

Nigeria Yobe Aug 2011

Chad Kanem Sep 2011

Chad Bahr el Ghazal Oct 2011

South Sudan Gogrial West Oct 2011

South Sudan Aweil East Nov 2011

South Sudan Twic Dec 2011

Myanmar Maungdauw Nov - Dec 2011

Haiti Haut Artibonite Jan 2012

Sierra Leone Moyamba Feb 2012

19 WHO (2010). Monitoring the Building Blocks of Health

Systems: A Handbook of Indicators and their Measurement

Strategies.20 The first challenge relates to volunteer’s workload. Because

of how important they are, many public health interventions

incorporate the same cadre of volunteers into their outreach

strategies. The result is an increasingly overburdened work-

force capable of dedicating increasingly less time to each

activity. The second challenge relates to the motivation of

volunteers. The issue of motivation is linked to the issue of

workloads and the issue of compensation. The tendency has

been to motivate volunteers through the ad hoc provision of

incentives, ranging from in-kind items (e.g. soap, sugar,

t-shirts and bags) to cash payments. Whilst this commonly

raises questions about sustainability, the bigger and more

relevant issue is whether volunteers should be placed at the

heart of community sensitisation and case-finding in the

short, medium and long-term.

Training session with health staffin Port-au-Prince, Haiti

Sam

uel H

auenst

ein

Sw

an,

Haiti, 2

011

49

health interventions tackling TB, HIV/AIDS,Malaria and Family Planning have turned tothis tier for the delivery of support and care.Many of the reasons that have led to the decen-tralisation of care reflect the same challengescurrently faced by integrated-CMAMprogrammes: weak health facilities with limitedand overworked staff, high caseloads leading tolong-waiting times, stigma associated with thecondition and high opportunity-costs linked toattendance.

There is encouraging evidence that commu-nity case management (CCM) of acutemalnutrition is not only possible, but can effec-tively deliver high quality results. Existingevidence from Malawi has shown that theoutcomes of treatment delivered by CHWs arecomparable to treatment delivered at healthfacility level21. Available studies have concludedthat “home-based therapy with RUTF adminis-tered by village health aides is an effectiveapproach to treating malnutrition during foodcrises in areas lacking health services”.22 Similarlarge-scale research also carried out in Malawi23

concluded that “home-based therapy withRUTF yields acceptable results without requir-ing medically trained personnel”.

In 2011, Save the Children with the supportof Tufts University, GAIN and Pepsico, carriedout operational research on CCM of SAM inSouthern Bangladesh. Unlike previousresearch, the Save the Children project meas-ured both the efficacy of treatment and thecoverage of the intervention. As in Malawi, theproject achieved high recovery rates (92%), andlow defaulting and mortality rates (7.5% and0.1% respectively). The coverage of theprogramme (89%)24 was found to be one of thehighest ever recorded by a CMAM programme.Subsequent research has also shown that theproject delivered high quality of care.

The Malawi and Bangladesh experiencesshow that a CHW-based service delivery modelcan be effective, but the evidence so far hasbeen largely based in contexts with robustCHW networks or where additional resourceshave been made available to support these. Theexistence of a professional cadre of paid CHWsis no guarantee in itself. In 2005, the EthiopianGovernment introduced the Health ExtensionProgramme (HEP) designed to bring togetherall basic maternal and child health interven-tions, including nutrition. Yet, CMAMprogrammes have often struggled to incorpo-rate nutrition activities (even case-findingalone) into a workload that initially included 17different health packages, from HIV/AIDS tocontrol of insects and rodents.

The evidence from Malawi and Bangladeshis promising, and the improved access andproximity offered by this approach could offer away of decreasing pressure on health facilities(in high prevalence areas in particular),decreasing defaulting and improvingprogramme coverage. More evidence is needed,from larger interventions, over longer periodsof time and outside of the controlled environ-ments (where there is large scale resourceinvestment) of the Bangladesh and Malawiexperiences. However, the real success of thistype of CCM model may ultimately rest on theability to implement such programmes as partof the much broader process of strengtheninghealth systems and successfully linking CHWs,health facilities and the communities whichthey serve.

ConclusionThe shift away from centralised, inpatient caretowards a community-based model wasarguably one of the most important paradigmshifts in the history of public health nutrition.This shift, however, is far from complete; asnutrition interventions enter a new phase char-acterised by the integration of nutrition servicesinto national health systems, the coverage andimpact of these interventions is decreasing.Turning this around is possible, but to do so,nutrition organisations must adapt to thechanging demands linked to health systemstrengthening and the prioritisation of commu-nity mobilisation/awareness. The question fororganisations like ACF is not how to providethe same support in a different context, butrather, what kind of support does the newcontext require and how can this be provided.The answers to these questions are likely tofundamentally change nutrition support organ-isations – from their staff profiles to theirstrategic objectives – but in doing so it willmake organisations better prepared to dealwith a rapidly changing sector.

For more information, contact: Saul Guerrero,email: s.guerrero@actionagainsthunger.org.ukand Maureen Gallagher, email: mgallagher@actionagainsthunger.org

cases at the start (A). After a short period, moti-vation would naturally decrease leading to adrop in referrals by volunteers (B). In the mean-time, the critical mass or momentum created bythe rapid and visually clear recovery of SAMchildren would lead to a gradual increase inself-referrals (C). Over time, the number ofcases that seek CMAM services spontaneouslywould overtake those referred by volunteersand can (assuming no significant barriers toaccess) lead to a sustainable and comprehensivemodel for ensuring programme coverage.

Achieving this dynamic would require, firstand foremost, a prioritisation of communitymobilisation activities as a key feature of thesupport provided by organisations like ACF. Inpractical terms, this would have implicationsfor the profile of staff, and resources madeavailable, to integrated CMAM programmes.Nutrition organisations like ACF shouldsupport MoH in the design, planning andimplementation of sensitisation activitiesincluding mass media, traditional communica-tion channels and the use of new technologies.It would also mean the involvement of nutri-tion organisations in the training andcoordination of community outreach activitiesby volunteers. All efforts to increase commu-nity uptake of CMAM services, however, willneed to be accompanied by the introduction ofa service delivery structure capable of absorb-ing the increase in caseload, and capable ofproviding effective and appropriate care(including low waiting times, regular supply ofRUTF, positive staff-beneficiary interface, highcure rates, etc). Experience has shown that it isthis third element – capacity to deliver - thatoften proves problematic, in particular for inte-grated-CMAM programmes dependent on theavailability and quality of existing healthhuman resources and infrastructure.

Exploring alternative models of CMAMservice deliveryThe current model for integrated-CMAMprogrammes relies on the utilisation of healthfacilities for the delivery of treatment services.Based on this model, support organisations likeACF are tasked with identifying facilities capa-ble of mainstreaming CMAM activities as partof their daily and/or weekly activities. The aimis then to introduce health system strengthen-ing initiatives (e.g. staff training) designed toprepare these facilities for the arrival of newlyidentified SAM cases from the community.

This model represents a limited vision ofhealth systems. Health systems can also includeadditional tiers such as Community HealthWorkers (CHWs). For many years, public

Figure 2: Programme admissions by source

Num

ber o

f SA

M C

hild

ren

Adm

itted

Volunteer Referrals Community Self-ReferralsMonths

180

160

140

120

100

80

60

40

20

00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

AB

D

C

21 Amthor R, Cole SM and Manary M (2009). The Use of

Home-Based Therapy with Ready-to-Use Therapeutic Food

to Treat Malnutrition in a Rural Area during a Food Crisis. J

Am Diet Assoc. 2009;109:464-46722 Ibid. p. 46423 Linneman Z et.al. (2007). A large-scale operational study

of home-based therapy with ready-to-use therapeutic food

in childhood malnutrition in Malawi. Maternal and Child

Nutrition (2007, 3 , pp. 206–215)24 Proportion of affected population, based on programme

admission criteria, receiving treatment.

Views

Saul G

uerr

ero

, Eth

iopia

, 2008

50

Field Article

Marie-Morgane Delhoume isan agricultural engineerspecialising in agro-develop-ment of tropical regions. Sheled the 2011 impact study ofACF-USA’s Integrated

Programme for the Eradication of Konzo in theTerritory of Kwango in DRC.

Julie Mayans is an agriculturalengineer, specialising in foodsecurity and rural developmentprogramme management. Shewas the ACF-USA Food Securityand Livelihoods Coordinator

for the ACF-USA DRC mission.

Muriel Calo is the Senior FoodSecurity & Livelihoods Advisorfor ACF-USA who providedtechnical support to the ACF-USA DRC mission.

Camille Guyot-Bender is theTechnical ProgrammesAssistant for ACF-USA.

Special thanks go to the whole of the ACF proj-ect and study team and the communities whoparticipated in the study. Thanks to those whoprovided peer review of the article, namelyNick Radin of ACF, Marie-Sophie Whitney ofACF and Dr. J. Howard Bradbury of theAustralian National University. The fundingsupport of the European Union (EU) FoodFacility Programme is gratefully acknowledged.

From December 2009 to October 2011,Action Against Hunger (ACF-USA)implemented a 22-month long inter-vention in the Bandundu province of

the Democratic Republic of Congo (DRC)addressing several factors underlying thekonzo epidemic affecting the population ofKwango district (see Figure 1). The ‘IntegratedProgramme for the Eradication of Konzo in theTerritory of Kwango in DRC’ project wasfinanced by the European Union (EU) FoodFacility. It aimed to eradicate the diseasethrough a cross-sectoral approach that focusedon nutrition education and training, dietarydiversification, improved water access andagricultural processing. A total of 22,000 house-holds are estimated to have benefited fromthese activities. The project was implementedin 396 villages in the Territory of Kwango.

Prior to implementing the project, ACFconducted a baseline study1 in collaborationwith the Minstry of Health (MoH)’sPRONANUT or Programme National deNutrition in DRC in 113 villages, across 51health areas and 11 health zones. A total 2,388suspected konzo cases were screened and 2,218were confirmed. The average incidence ofkonzo was 1.07%. Among confirmed cases,83% were located in savannas, 1% in hillsideareas and 4% in valleys. Kahemba health zonewas found to harbour the highest number ofconfirmed cases (1,639), and placed among thetop three zones for incidence (2.08%) largelydue to its density of population and associatedrisk factors.

Local beliefs and traditional customs werefound to influence strongly the incidence ofkonzo in the area. Local eating customs thatfavour the male head of household were notedas likely contributors to heightening the expo-sure of women and children to konzo. Therewas a widely held belief that the disease iscaused by black magic, while knowledge of thefood-related origins of the disease was low.

Households rely primarily on agriculturefor food and have limited dietary diversity.Cassava is cultivated as a main crop, withmaize, groundnuts and beans as secondarycrops. Diverse environmental factors, such assoil fertility and soil water retention, affect thequantity and quality of harvests.

Water access is a critical factor in konzo inci-dence, with access limited by both distance toand seasonality of water points. Water cover-age levels are very low, with 5% coverage in theFeshi health zone and 4.3% in Kajiji.2 Due tothese challenges, cultivators in rural areas mostoften prefer to soak the cassava directly onriver banks, in ponds or in swampy areas inorder to avoid carrying heavy quantities of

water back to their homes. In semi-urban areas,people prefer to ret the cassava in their homes(in buckets or barrels) due to the likelihood oftheft if the cassava is left overnight in a publicarea. This practice can be hazardous as thequantity of water available in urban areas isoften insufficient, and this often leads to notchanging the water on a daily basis during theretting process which greatly increases the riskof cyanide intoxication.

Implementation strategyThe programme strategy sought to addressdirectly the range of critical factors related tokonzo disease that were identified in the base-line survey. These included knowledge of andattitudes towards the disease, limited agricul-tural and dietary diversity, low water accessand poor knowledge and practices aroundcassava processing. The strategy also aimed toaddress indirectly the high rates of malnutri-tion seen in konzo cases (25.8% global acutemalnutrition (GAM) prevalence in konzoaffected children less than 18 years old, 69.3%of GAM in konzo affected adults). Projectdesign used a cross-sectoral approach toaddress underlying factors in a holistic andintegrated manner.

Community outreach, mobilisation andeducationACF employed a community outreach andmobilisation approach through the creation ofcommunity cells as a forum for discussion onkonzo and nutrition. These served as launchpads for a broadly based educational campaignon konzo, which also extended to churches,schools, training of local health professionals,community volunteers and leaders, traditionalauthorities, etc.

Information, Education & Communication(IEC) materials on food processing and prepa-ration, nutrition and konzo were developed incollaboration with the PRONANUT, includingposters, brochures, training modules and othermaterial with graphic illustration and support-ing text in Lingala, Kikongo and Frenchlanguages. Posters were distributed for displayin places such as public areas, religious sites,health centres, schools and administrativeoffices. Radio messages incorporating songsand stories were crafted for broadcasting ontwo local radio stations. To complement, ACForganised 154 mass sensitisation sessions inchurches, mosques and schools.

Across the intervention zone, 35 senior MoHstaff, 429 community leaders and authorities,and 1052 community volunteers were trained

1 Field Exchange 41, August 2011. A cross-sectoral

approach to addressing Konzo in DRC. p2-52 Based on national standards of user numbers per water

point

Impact ofcross-sectoralapproach toaddressingkonzo in DRC

By Marie-Morgane Delhoume, Julie Mayans,Muriel Calo and Camille Guyot-Bender

What is konzo?

Konzo is a sudden epidemic spastic paraparesis (paralytic) disease which leads to a permanent paralysisof the affected person’s lower limbs. It is a neurological ailment triggered by sustained dietary exposureto the cyanide present in improperly processed cassava. Konzo itself is not fatal, but its debilitatingeffects heighten the risk of morbidity and mortality from other diseases. Its disabling effects result inpractical, social and economic challenges for individuals and families of individuals living with thelimited physical capacity induced by konzo. The disease usually appears in clusters within households,as exposure comes from food consumed as a family meal.

Overall, vulnerability to konzo is heightened by the combination of low protein intake (associated withlow dietary diversity), poor soil conditions (which favour the cultivation and consumption of bittercassava varieties high in cyanide), and a lack of sufficient water resources for thorough processing.

Processing mill

Marie M

org

ane D

elh

oune,

DRC

Field Article

selected purposively and 40 of 395 interventionvillages selected randomly. In each selectedvillage, six beneficiary households wererandomly selected to participate in householdsurveys (234 in total). Of these, 76% had partic-ipated in community cells and 24% had not.Household surveys were supplemented withinformation from key informants and focusgroups.

FindingsKnowledge and attitudes on konzo andnutritionChanges in knowledge at endline compared tobaseline suggest that community outreach andeducation activities were effective in challeng-ing long held local beliefs on konzo andnutrition. At project baseline, 74% of sampledpopulation thought that the disease had a meta-physical or black magic origin; at endline thisproportion had dropped to 7%. Eight-eight percent of the sampled population correctly notedthe food-related causes of konzo at endline,while 3% indicated a viral cause and 8%reported they did not know the cause.3 Thisfinding represents the strongest indicator ofproject impact.

Participation in a cell was found to be corre-lated with knowledge of the food-related causeof konzo. In addition there was a strong inversecorrelation between both ‘participation’ and‘lack of knowledge’, and ‘participation’ and‘belief in a metaphysical origin’. These findingsreveal the importance of outreach and educa-tion activities delivered both within thecommunity cells and directly by ACF.

Similar results were found regarding knowl-edge, attitudes and practice on preventionstrategies. In particular, messaging encouragedappropriate processing of cassava and inclusionof increased levels of protein in diets throughincorporation of maize flour into fufu prepara-tion and legumes (pulses) in the diet. Atbaseline, households reported utilizing maizeflour in their fufu preparation in only a fewcases where milling services were available,while millet flour was used in the north easternFeshi territory. Knowledge of prevention strate-gies linked to food preparation and dietarydiversity was limited. At endline, a majority(78%) of respondents indicated that a diversi-fied diet + correct cassava processing wouldprevent konzo, while 7% believed that witch-craft was the cause of the problem and 8% did

introduction in Kwango was aligned with localagro-ecological conditions and intended tosupport increased consumption of sulphur-based amino acids contained in leguminousfoods (to counteract high levels of cyanide inthe diet) and to complement consumption oftraditional cyanide-heavy bitter cassava vari-eties with varieties low in cyanide. ACFdelivered technical training on agriculturaltechniques to the 12,500 beneficiary householdsof food crop support. The bulk of the sweetcassava cuttings distributed in the frameworkof the project were produced by local agro-multiplier associations partnered with ACF.ACF also supported the installation of 13village based mills to increase access to maizeand cassava milling services and improve thequality of the flour.

In order to increase water access, ACF imple-mented a variety of hydraulic constructions:public retting tanks to process cassava, bore-holes, springs, rainwater harvesting systemsand piped distribution networks. To encouragethe participation of the community, the tanksestablished were given a supervisor whoassigned a management team. Although themanagement committee owned the tanks,anyone in the area could use them in exchangefor a small fee.

Impact studyThe impact study was conducted in six of theeleven health zones targeted by the projectalong two main axes - the western axis (Kenge,Boko, Popokabaka, Wamba Luad) and the east-ern axis (Kahemba and Kajiji) (see Figure 1).

A stratified sampling approach was used,with six of eleven intervention health zones

3 Respondents could list more than one cause.

Figure 1: ACF study area in Kwango district, Bandundu province, DRC

51

Sensitisation beingcarried out on konzo ina mosque in Kahemba

Julie M

ayans,

DRC

Cooking demonstration

Marie M

org

ane D

elh

oune,

DRC

western axis eastern axis

on the tools and in turn, used their knowledgeand skills to pass the message more broadlyacross the population. Each community volun-teer presided over a community cell or servedas secretary to the president. Skilled in commu-nity mobilisation and training, this individualwould facilitate community dialogues on aweekly basis and ensure regular reporting tothe local health centre. Each cell had use of anoffice space and comprised one committee pres-ident, one secretary and two advisors. Thecreation of 647 community cells across 395villages offered a setting for demonstration,discussion and exchange. These cells gatheredmembers from a local neighbourhood or villagehamlet and numbered roughly 45 memberseach.

Cooking demonstrations were organised atthe level of each cell to complement discussionsaround balanced diets and promotion ofkitchen gardens. ACF organized 1,808 demon-strations and volunteers organized another2,600 demonstrations for cell members, averag-ing four demonstrations per cell. Improved fufurecipes based on mixed cassava and maize flourwere introduced in the cooking demonstra-tions. ACF also extended practical support tothe groups through provision of tanks and millsfor cassava retting and processing.

Agricultural and hydraulic infrastructuresupportAs part of crop and diet diversification activi-ties, ACF introduced cultivation of improvedvarieties of two food crops over two successiveagricultural seasons, niébé (cowpea) varietiesVita 7 and Muyaya, and sweet cassava varietiesTME119, Mwuazi, Nsasi, Disanka and Butamu.Cultivated in other parts of the DRC, their

200%180%160%140%120%100%

80%60%40%20%

0%

Janu

ary

Febr

uary

Mar

ch

April

May

June July

Augu

st

Sept

embe

r

Octo

ber

Nove

mbe

r

Dece

mbe

r

No. of cases 2010 No. of cases 2011

conducted at the same time of year, whichwould have ensured the greatest comparability,however neither one was carried out in thepost-harvest period when differences in foodavailability are significant. The positive trend indiversity and volume of household food stocksmay be attributed to project impact, in particu-lar the IEC activities around balanced diets andfood processing and preparation, as well asexternal factors such as climate, crop diseaseand seasonal fluctuations.

Cassava retting techniques and wateraccessKnowledge of community leaders and memberhouseholds of community cells regardingcassava retting and drying techniques wereassessed before and after training. On therecommended length of time to ret and drycassava, the share of community leaderscorrectly reporting optimal length (4 days)increased from 60% at baseline to 99% atendline. Member households showed a similarlevel of knowledge at endline but statedconstraints around access to processing sitesand water quality in applying the practice.

At endline, a majority of households (92%)indicated they were processing cassava inrivers or ponds, with a minority using cassavaretting tanks (4%) or containers at home (4%).At baseline, utilization of home retting tech-niques – that rely on prolonged use of the samewater, saturated in acid and less effective incyanide detoxification – was relatively commonin urbanized sites (9% in Kahemba, 10% inPopokabaka, 5% in Kenge). At project end, itwas noted that these practices have been largelyabandoned, partly due to ACF’s implementa-tion of peri-urban water points.

At the end of the project, householdsreported soaking cassava an average of 3.4days, a significant increase from the average of2 days noted at baseline across the interventionarea. Constraints to optimal practice includethe risk of theft of tubers at open river andpond sites, as well as dietary and income pres-

52

Field Article

not know. At endline, 47% of the populationreported preparing fufu with mixed cassava/maize flour. Nearly half (45%) of respondentsprepare their fufu exclusively with cassava flourwhile 8% combine cassava and millet flours.

Lack of access to maize milling servicesexplains 77% of surveyed cases of non-incorpo-ration of maize flour into fufu, which areconcentrated on the western axis of the inter-vention zone (Kenge, Boko, Popokabaka andWamba Luadi). Just 3% of cases justified theexclusive use of cassava flour on the basis offood habits, indicating that the messagesaround fufu preparation were well appropri-ated, but cannot be translated into practicelargely due to practical constraints.

Food stocks and dietary diversificationNew varieties of niébé were largely acceptedacross the intervention zone and integrated intothe diet, notably on the eastern axis (Kahemba,Kajiji) where populations were unfamiliar withniébé. Sweet cassava was readily integrated intoboth east and west Kwango, with results show-ing a general increase in the intercropping ofboth bitter and sweet varieties, as well asincreased cultivation of sweet varieties on theirown. In Kahemba, the bitter cassava varietyMwambo is widely cultivated and consumed tothe exclusion of other varieties, whereas bothbitter and sweet varieties are cultivated andconsumed along the western axis. However,sweet varieties were well accepted in Kahembaas they offer shorter processing times and areimmediately consumable.

Food stocks at baseline (May 2010) andendline (August 2011) were assessed (seeFigure 2). The results reveal a notable improve-ment in both the overall stocks and thediversity of food items held by households,including pulses. The surveys were not

80%70%60%50%40%30%20%10%

0%

Cass

ava

Mai

ze

Grou

ndnu

t

Bean

Niéb

é

Soyb

ean

Igna

me

Swee

t pot

ato

Bana

na

Squa

sh

Tom

ato

Frui

ts

Onio

n

Mea

t

Egg

Fish

Suga

r

Oil

Salt

Baseline Endline

Figure 2: Household food stocks by food item in % of households

Figure 3: Comparison of number of konzo cases inKahemba health zone in 2010 and 2011

sures. Impacts on cassava retting practice fromACF’s establishment of communal retting tanksmaintained and watched over by communitygroups are not yet known as the infrastructureswere in process of installation at the time ofsurvey.

Konzo incidenceA surveillance system for screening and identi-fication of konzo cases in Kahemba health zonewas established by the local health structure in2009, with annual caseload an estimated 1,300individuals in 2009. MoH educational activitiesand ACF integrated activities on konzo werelaunched in early to mid 2010, with a markeddecrease in cases (fewer than 200) recorded thatyear. A further reduction in caseload between2010 and 2011 was noted during the criticalmonths of June, July and August (dry season)with 47 new cases recorded in 2011(see Figure3). This represents an 84% reduction in inci-dence between 2010 and 2011. The greatestreduction in new cases was observed amongthe under 5 years age group. Note that theobserved reduction in konzo incidence cannotbe attributed solely to project activities asnumerous external factors are likely to influ-ence this outcome.

The results on reduced incidence are corrob-orated by ACF analytical findings of urine andcassava flour sample cyanide content, takenfrom 100 randomly selected beneficiary house-holds at project endline (see Figures 4a and 4b).A 50% reduction in flour samples presentingmedium to high cyanide levels (20 to 40 ppm)was observed compared with baseline, as wellas a 16% reduction in thiocyanate levels in urinesamples (>300µmol). These reductions translateinto a slightly lower risk of developing konzo.

Observed reductions in cyanide content ofbaseline and endline samples are similarlyattributed to numerous external factors such asseasonality, migration, agricultural production,health condition, diet composition, water avail-ability, as well as project impact.

70605040302010

0Rate < 20 ppm 20 < Rate < 40 ppm Rate > 40 ppm

Flours 2010 Flours 2011

Figure 4a: Flour test sample results 2010 and 2011 by% surveyed households

70605040302010

0Rate<100 µmol/l 100 µmol/l<Rate

<300 µmol/l Rate>300 µmol/l

Urine 2010 Urine 2011

Figure 4b: Urine test sample results 2010 and 2011 by% surveyed households

Water source inKigwangala

Julie M

ayans,

DRC

53

Field Article

Conclusions and recommendationsACF’s multi-tiered community outreachand education strategy proved effective inthe diffusion of information on a largescale. The community cell approachallowed for a deep, sustained and broadbased appropriation of messages and activ-ities around nutrition education and konzothat would not have been possible if onlytraditional IEC methods and materials hadbeen utilised. Placing community membersin leadership positions to carry out sensiti-sation allowed local taboos to be effectivelymitigated through open discussion. Thisapproach also permitted the affected popu-lation to control the educational process,encouraging better appropriation ofmessages, knowledge transfer and behav-iour change. Impacts achieved through thecommunity outreach and educationapproach were reinforced by improvedaccess to water, agricultural processinginfrastructure and opportunities to diver-sify diets.

Based on these findings, ACF-USA issuedthe following key recommendations:• Continued promotion of messaging by

community cells and MoH staff • Continued epidemiologic surveillance

of incidence of konzo cases by MoH in collaboration with local partners with a focus on high concentration areas

• Expansion of access to village-based agro-processing infrastructure and associated hydraulic infrastructure including cereal mills and cassava retting tanks

• Socio-economic support of konzo victims and their families through targeted support to income generating activities

• Further study on traditional diets, food habits and beliefs, including the contribution of wild foods and game, for improved strategies around dietary diversification

For more information, contact: Muriel Calo,email: mcalo@actionagainsthunger.org

Cassava rettingtank in Feshi

Julie M

ayans,

DRC

By Abukar Yusuf Nur – Nutrition Analyst, Ahono Busili – Nutrition Team Manager, Elijah Odundo – Nutrition Data Analyst, Joseph Waweru – Nutrition Analyst, Louise Masese – Mwirigi-Nutrition Analyst,Mohamed Borle – Nutrition Analyst and Tom Oguta – Senior Nutrition Analyst

The authors constitute the FSNAU nutrition situa-tion analysis team.

2011 famine inSouth Somalia:the role of theearly warninginformationsystem

The authors would like to acknowledge thefollowing people for valuable insights and edits:Grainne Moloney, Anne Bush, Tamara Nanitashvili,Jackline Adero and the FSNAU nutrition analystswho collected the data during this very difficulttime. Special thanks to Oleg Biluka, Leisel Talleyand Curtis Blanton from CDC for their invaluableinput and technical guidance on each and everynutrition survey conducted during the famineperiod. They assisted tremendously in assessingthe data quality of all the surveys.

Finally, thanks to FEWS NET Nairobi andWashington colleagues, UNICEF and WFP Somaliacolleagues, Nick Haan, Dan Maxwell, Peter Salama,Mike Golden, Epicentre and Francesco Checchi fortheir technical support and collaboration duringthe famine.

In July and August 2011, the UnitedNations declared famine inSouthern Somalia based on thefood security and nutrition analy-

sis undertaken jointly byUNFAO/FSNAU (Food Security andNutrition Analysis Unit) and FEWS-NET. The analysis indicatedconvergence of key evidence on famineoutlined in the Global Integrated PhaseClassification of food insecurity (IPC)1.This was not a sudden onset crisis andthis article describes the role and keyfindings of the early warning systemleading up to that declaration.

OverviewSomalia has been without an effectivecentral government since PresidentSiad Barre was overthrown in 1991following a protracted civil war byopposing clans. The opposing clansunfortunately failed to agree on areplacement and plunged the countryinto lawlessness and clan warfare, withdisintegration of the country along clanlines and smaller political regions.

The three major political zones areSomaliland, Puntland and South andCentral Somalia. Somaliland, in thenorthwest, has remained relativelystable since its unilateral declaration ofindependence in 1991 with functioninginstitutions and a peaceful transfer ofpower via democratic elections in 2010.Puntland, in the northeastern zone ofSomalia has, since mid-1998, beenreferred to as the Puntland State ofSomalia and has an estimated popula-tion of two million people whichinclude 140,000 internally displacedpeople (IDPs) from parts of South andCentral zone. Puntland is, in general,politically and socially stable but withsporadic incidents of insecurity. Therest of the country is what is currentlyreferred to as South and CentralSomalia and is currently governed bythe Transition Federal Governmentwith the administrative capital inMogadishu. South and Central zones ofSomalia are the epicentre of the currentanarchy in Somalia, which hasimpacted negatively on lives, liveli-hoods and access to humanitariansupport.

The FSNAU in collaboration withFEWSNET, UNICEF, WFP and other

1 See FSNAU website: www.fsnau.org

A sibling takes care of anewborn in Mogadishuamongst IDPs (July 2011)

Moham

ed M

oalim

/FSN

AU

, Som

alia,

July

2011

FSNAU pressreleases

Field Article

54

Field Article

partners provide evidence-based analysis of Somali food,nutrition and livelihood security, to enable both short-termemergency responses, and long-term strategic planning infood security and nutrition well-being. It was first establishedas the food security assessment unit (FSAU) in 1994 post1992/93 famine, evolving into the FSNAU in 2000 with theintegration of food security and nutrition projects. FSNAUworks to develop the capacity of other agencies (both govern-mental and non-governmental) to collect evidence-basedinformation and focus more on the overall analysis of the foodsecurity situation. FSNAU analysis also contributes to policyand strategy development.

Deterioration in food security and nutrition situation(2010/11)In Gu (April-June) 2010, the food security situation improvedin most of rural Somalia, leading to a reduction in number ofpeople facing acute food insecurity (see Box 1 for livelihoodssystem overview). This was attributable to an exceptional Guseasonal performance across most of the agricultural liveli-hoods, as well as improved livestock production in thecountry. The number of the urban population facing acutefood insecurity also significantly decreased during this perioddue to reduced inflation, increased wages and overallimproved food production in the country. However, duringthe following season (Deyr ‘10/11), there were clear signs of aworsening food security situation as flagged in the FSNAUand FEWSNET early warning system, in most livelihoods ofSomalia2. This was the result of unusually below average rain-fall caused by the La Niña meteorological phenomenon. Theimpact of this dismal seasonal performance was demon-strated in failed crops in most of the southern crop-producingregions, and considerable water and pasture shortages inmost of the key pastoral areas of the country – reaching only19% of average (Figure 2).

These developments resulted in significantly reduced foodproduction (cereals, milk and meat), which subsequently wasreflected in increased food prices and a rising number of thepopulation facing acute food insecurity in early 2011.Specifically, in Deyr 2010/11 a total of 2.4 million people facedacute food security, representing a 20% increase from Gu 2010.In the South, the rural/urban population in crisis increased by64% from Gu 2010, reaching 855,000 people. The La Ninaevent continued further through Gu 2011 season.

Restricted humanitarian assistance and substantialconstraints to food access (Table 1) were observed mostly inthe south. The food security situation was further aggravatedby the food price inflation, limited labour opportunities andoverstretched social support. However, the camel rearingpastoralists were less affected in comparison to the riverineand agropastoral communities (cattle rearing) as they optedto out-migrate to areas with slightly nearer average condi-tions. Nonetheless, nationwide high livestock deaths, as wellas increased pastoral destitution, were reported during thisperiod, especially for cattle and sheep.

Nutrition surveillance in SomaliaA wide range of indicators are derived from the nutritionsurveillance system in Somalia: acute malnutrition, deathrates, proportions at risk based on mid upper arm circumfer-ence, and nutrition trends at health facilities. In any particularseason, the severity in level of findings by indicator rangesfrom Acceptable, Alert, Serious, Critical, Very Critical toExtremely Critical mostly based on internationally recognizedthresholds. Convergence of evidence on the severity level ofthe findings is the basis for categorizing the nutrition situa-tion in a given phase. Where representative nutrition surveysare conducted, global acute malnutrition (GAM) is the coreoutcome reference indicator.

An integrated analysis of the nutrition situation3 in south-ern Somalia from the Gu (April-June) in 2007 depicts varied

The rural livelihood systems of Somalia (pastoralism, agropastoralism, farming), revolvearound the seasonal rainfall pattern. There are four seasons in South Somalia: the hotand dry Jilaal (January-March), and Hagaa (July-September), and the cool and wet Gu(April-June), and Deyr (October-December).

The Gu and Deyr rainy seasons provide pastoralists with water and pasture for livestockproduction and sustenance, while enabling crop establishments and with it, increasedlabour and income opportunities in the agro-pastoral and riverine livelihood zones simul-taneously. Whereas pastoralists reside in their regular settlements during the wetseasons, they out-migrate towards the rivers and other water points for water andpasture during the dry seasons. Agropastoralists in the south are more inclined to cropthan focus on livestock production and, together with the riverine population group(pure farmers), are generally sedentary. Urban livelihoods rely heavily on the rural areasfor food (milk, cereal, fruits and vegetables), accessed through purchase. Changes in rain-fall pattern therefore adversely affect food access across all population groups,significantly impacting on health and nutrition outcomes.

Box 1 : Livelihood systems in Somalia

25,000

20,000

15,000

10,000

0

Maize Sorghum PWA 5 year average

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year

Figure 2: Post-Gu plus off season cereal production compared with the Post War Average (PWA)(1995-2011)

Livelihoodzone

Proportionpoorhouse-holds

Sources offood (poorhouseholds)

Relative importance offood source (% ofminimum foodbasket in thebaseline year)

Change in2010/11compared tobaseline year

Likelyimpact onAug/Sept2011 foodaccess

BakoolAgropastoral

40% Own cropproduction

25% 88% decline incrop production

Poor house-holds onlyable toaccess 40-50% of foodneeded forsurvival

Purchasesfunded bywage labour

25% 57% decline inTerms of Trade(ToT) Wage/Sorghum

Purchasesfunded bylivestock sales

25% 76% decline ToTin Goat/Sorghum

Other 25% Net decline likely

BayAgropastoralHighPotential

35% Purchasesfunded bywage labour

60% 80% decline incrop production

Poor house-holds onlyable toaccess 40-50% of foodneeded forsurvival

Purchasesfunded bywage labour

10% 82% decline inToT Wage/Sorghum

Other 30% Net decline likely

LowerShabelleRiverine

38% Own cropproduction

75% 64% decline incrop production

Poor house-holds onlyable toaccess 40-50% of foodneeded forsurvival

Purchasesfunded bycrop sales

15%

Purchasesfunded bywage labour

10% 46% decline inToT Wage/ Maize

Table 1: Evidence of substantial constraints to food access by livelihood group, August 2011

2 FSNAU Press Release April 27, 2011: Somalia: Drought Impact Intensifies as

Rains Delay3 The nutrition situation classification is based on ‘The Nutrition Classification

Framework’ developed by FSNAU in consultation with key partners invoved in

nutrition activities in Somalia and in the Region.

IDPs migrate to Mogadishu(Shobelle region) in July 2011

Som

alia Y

oung S

old

iers

Agency

(SO

YD

A),

Som

alia,

July

2011

55

4 Press Release July 20, 21011: Expanding Famine

Across Southern Somalia5 Press Release September 5, 2011: Famine

Spreads into Bay region; 750,000 people face

Imminent starvation6 Press Release February 03, 2012: Famine over yet

31% of the population remain in Crisis

Box 2: Chronology of events (early warnings and alerts) leading to 2011 famine declaration

August 2010:FSNAU warned that though Somalia received above aver-age rains, the gains made could easily be reversed giventhat the outlook for the next rainy season was poor. FEWSNET also issued an alert on La Niña and Food Security inEast Africa. They projected that the impacts of a La Niñaevent from August 2010 could include significantFebruary 2011 crop harvest deficits in south-easternKenya, Somalia, and northern Tanzania, depending on theseverity of the La Niña event. Additionally, reduced range-land resources (water and pasture) in key pastoral areas inthe Horn of Africa between October 2010 and March 2011and possible reduction in 2011 long rains agriculturalproduction were predicted.

September 28, 2010:FEWS NET issues La Nina Brief, indicating the likelihoodthat it would continue into early 2011 with significantfood security implications.

November 2, 2010:FEWS NET East Africa indicates that pre-emptive liveli-hood support could mitigate likely La Niña impacts inthe eastern Horn.

November 26, 2010:FSNAU issues a press release on the early impact of poorrains seen in Somalia.

December 16, 2010:FSNAU preliminary analysis on the early indication of theoutcome of the Deyr season performance indicates thatthat large scale severe crisis would sustain in most partsof southern Somalia given prevailing restrictions onhumanitarian interventions.

January 26, 2011:FSNAU issued a news release on the crop failure, severewater crisis for both human and livestock, following fail-ure of the short Deyr rains, heightening fears ofdeepening humanitarian crisis in coming months.

March 15, 2011:The FEWS NET led inter-agency publication issued analert that the existent crisis following the October-December, 2011 drought was likely to worsen, based onbelow-average March to May rains forecast in theEastern Horn.

March 21, 2011:FEWS NET/FSNAU issue an alert: extreme food insecuritylikely due to drought and lack of humanitarian response.

April 2011:FSNAU issued a press release indicating that the countrycould slide into an even deeper crisis due to the combi-nation of drought, skyrocketing food prices and constant

population displacement from ongoing conflict. Theprevailing drought had already displaced some 50,000Somalis within the country, according to UN estimates.FSNAU, through the Nutrition Update publication,further issued an alert that the nutrition situation in theSouth was Very Critical. Further deterioration in nutritionstatus is expounded in July 8th, 2011 edition.

May 6, 2011:FEWS NET East Africa alert: Poor performance of Aprilrains brings major food security concerns in the EasternHorn

June 7, 2011:FEWS NET led multi-Agency East Africa Alert warns thathouseholds in the pastoral and marginal cropping areasface moderate to extreme levels of food insecurity due tothe drought, deteriorating purchasing power, and in someareas, limits on the delivery of humanitarian assistance.

June 20, 2011:FSNAU issues a press release as well as a Quarterly Brieffocusing on Post Gu Season Early Warning. They high-light the deepening crisis in Somalia in the 2nd half of2011 with food prices hitting a new record high, follow-ing the two consecutive poor rainy seasons.

July 20, 2011: The United Nations declared famine intwo areas of southern Somalia (Bakool agropastorallivelihood zones and all areas of Lower Shabelle) basedon joint FSNAU FEWS NET analysis4.

August 3, 2011:The agropastoral areas of Balad and adale districts ofMiddle Shabelle, the Afgoye corridor IDP settlement, andthe Mogadishu IDP community, were declared in faminewith an additional 50,000 people in the cropping areas ofGedo, Juba and pastoral areas of Bakool facing famine-level food deficits.

September 5, 2011:A joint FSNAU FEWS NET release on September 2011reported that Bay region was facing famine conditionsbased on new nutrition survey data . In addition, theJuly/August Post-Gu seasonal assessment analysisreported that poor households in this region facedmassive food deficits due to a combination of poor cropproduction and deteriorating purchasing power.

February 3, 2012:The UN declared an end to famine conditions in Somaliabut warned that the crisis in the Horn of Africa was notover and continued efforts were required to restore thefood security in the region.6

The reference documents are accessible atwww.fsnau.org and www.fews.net

Deyr ‘06/07

Deyr ‘09/10 Deyr ‘10/11

Deyr ‘11/12

Gu ‘07

Gu ‘09 Gu ‘10 Gu ‘11

Gu ‘08Deyr ‘07/08 Deyr ‘08/09

Source: FSNAU website. Note: To see the detail, download from http://www.fsnau.org/downloads/Prgrogression_of_Estimated_Nutrition_Situation_Deyr_06_10_to_Gu_11.pdf

Figure 3: Progression of estimated nutrition situation Deyr’ 2006/07 – Gu 2011

Field Article

nutrition phases ranging from Serious inJuba and Shabelle regions, to Very Criticalin Gedo, Bay and Bakool regions. (Referto progression maps in Figure 3). Themaps show the variation and progres-sion of the nutrition situation in thedifferent regions of South Somalia(circled) between Gu 07 and Gu 11.

The main areas of concern highlightedin the FSNAU/FEWSNET early warningsystem from August 2010 were the rain-fed agropastoral areas of southern andcentral Somalia where crop productionwas largely impacted by the poor rain-fall, cattle-breeding pastoralcommunities in South-East Pastoral, alllivelihoods of Hiran and Bakool regions,the entire Coastal Deeh livelihood, theAddun Pastoral, Nugal Valley and SoolPlateau. Most of these areas had sufferedfrom consecutive seasons of below aver-age rainfall. Therefore, the capacity towithstand the looming crisis was verylimited.

The maps highlight the deterioration ofthe nutrition situation to Very Critical in allsouthern regions by Gu 11. The deteriora-tion was mainly attributed to lack ofaccess to food due to rapid increases infood prices, massive death of livestock,and limited access to relief food due to thewithdrawal of humanitarian agencies forsecurity reasons. The situation was exac-erbated by disease outbreaks such ascholera, acute watery diarrhoea andmeasles in southern Somalia.

In South Somalia, the nutrition situa-tion remained worrying from Gu 2007up until Gu 2011 when the levels of acutemalnutrition reached the peak, with

56

Field Article

7 The Somalia Nutrition cluster approach of estimating the cases of malnourished

children is such that the total cases of acute malnutrition, including those severely

malnourished state, is obtained by applying the GAM and SAM prevalence rates to

the under-five population for each of the assessed population groups

GAM as high as 55% in Bay region. This increase was the result ofthe severe food insecurity described earlier that followed the poorrainfall performance in the area and shrinking humanitarian spacesince 2009, following heightened insecurity. In 2009 for example,international organisations operating in southern Somalia werecompelled by local militia to vacate, leaving the communities withminimal access to humanitarian support. In addition, diseaseoutbreaks, such as acute watery diarrhea/cholera and measles in anenvironment with extremely low access to health services aggra-vated the situation. At this point, the population’s general resilienceto shocks that sustained the nutrition situation at Serious phase insome seasons seemed to have collapsed.

Early warnings and alerts from the FSNAU/FEWSNET earlywarning system were issued through 2010 and 2011) leading up todeclaration of famine in July 2011. These are listed in Box 2.

Evidence of famineThe movement restrictions in southern Somalia meant that imple-mentation of nutrition surveys was remotely managed and mostwere implemented successfully. In July 2011, FSNAU conducted 18representative nutrition (SMART two-stage cluster) surveys inSouthern Somalia, covering the rural livelihoods, Mogadishu, andAfgoye IDPs. Sixteen repeat SMART surveys were conducted inAugust 2011to monitor the situation. There were data limitationsfor Bakool and Hiran regions identified by CDC in October 2011.However this was not a significant constraint for early warningpurposes given the available food access analysis and the nutritionand mortality data from surrounding areas.

The July 2011 surveys indicated extreme levels of acute malnu-trition across the south, with GAM rates exceeding the IPC faminethreshold of 30% in all livelihood zones except for the riverinezones of Middle Shabelle and Hiran Regions. In some regions,prevalence of acute malnutrition exceeded 50% (Figure 4). Findingsfor the retrospective crude death rates (CDR) confirmed famine inthese areas according to IPC thresholds, especially for Bay andBakool agro-pastoralists, Lower Shabelle riverine and agro-pastoralists and Afgoye and Mogdishu IDPs where they exceed theIPC famine and WHO emergency threshold of 2/10,000/day(Figure 5). Based on the August 2011 survey findings, the CDR wasat 2.11/10,000/day among Bakool agro-pastoralists. For LowerShabelle, in the riverine and agro-pastoralist livelihood zones, theJuly survey findings indicated CDR of 6.12 and 4.29/10,000/dayrespectively. In the other population groups, the rates were lower,ranging between 1 and 2/10,000/day, indicating excessive deathsand a doubling or trebling of the FSNAU median rate of0.7/10,000/day in Southern Somalia.

The July 2011 findings indicated significant deterioration in thenutrition situation across all population groups in the south fromthe preceding six months. FSNAU and FEWSNET analysisconfirmed that the core IPC famine outcomes of food security,nutrition and mortality indicators had been met. Therefore, in Julyand August 2011, the UN declared famine inMogadishu andAfgoye IDPs, Bay and Bakool Agro-pastoral, Lower Shabelle andAdale and Aden Yabal districts in Middle Shabele (see Figure 6Nutrition Situation Map).

Crisis population and responseAccording to the FSNAU and FEWS NET August 2011, post Gu2011, a total of 4 million people were in crisis nationwide, of which3 million were in the south. The worst affected areas were inSouthern regions where most people lacked food access, with750,000 experiencing famine level outcomes.

At national level, approximately 450,000 were estimated7 to beacutely malnourished children, translating to 30% of the 1.5 millionSomali children. Of these, 190,000, or 13% of the 1.5 million Somalichildren were severely malnourished. The south was worst hit, andhost to 336,000 (or 74%) of all the acutely malnourished children,160, 800 of whom (84% of the national estimate) were severelymalnourished.

In response to the crisis, the Nutrition Cluster in Somalia under-took scale up of nutrition services both through static and mobile

Table 2: Nutrition survey results, July 2011

Survey GAM SAM Crude Mortality Rate

Bay agro-pastoralists 58.3% (52.1-64.2) 22.1% (18.2-26.5) 2.15/10,000/day

Mogadishu IDPs 45.6% (40.5-50.8) 23% (19.2-27.2) 4.02/10,000/day

Afgoye IDPs 46% (40.8-51.3) 24.7% (20.2-29.8) 5.68/10,000/day

90

60

30

0

Jul Jul Jul Jul Jul Jul Jul Jul Jul Jul Jul Aug Jul Aug Jul Aug Jul Aug Jul Aug Jul Aug Jul Jul

Gedo

Agr

opas

t

Gedo

Rive

rine

Hira

n Ag

ropa

st

Hira

n Pa

stor

al

Hira

n Ri

verin

e

L.Sha

belle

Agr

opas

t

L.Sha

belle

Rive

rine

M.S

habe

lle A

grop

ast

M.S

habe

lle R

iverin

e

Bako

ol P

asto

ral

Bako

ol A

grop

ast

Juba

Pas

tora

l

Juba

Agr

opas

t

Juba

Rive

rine

Gedo

Pas

tora

l

Bay A

grop

ast

Mog

adish

u ID

Ps

Afgo

ye ID

Ps

Figure 4: Prevalences of GAM and SAM, Southern Somalia, July – August 2011

GAM SAM

22201816141210

86420

L. Shabelle

Dea

th ra

te

M. Shabelle

IDPS-South

BayBakool

M&L JubaM&L Juba Gedo Hiran

Rive

rine

Agro

past

oral

Rive

rine

Agro

past

oral

Afgo

ye

Mog

adish

u

Agro

past

oral

-July

Agro

past

oral

-Aug

Agro

past

oral

Past

oral

-July

Past

oral

-Aug

Rive

rine-

July

Rive

rine-

Aug

Agro

past

oral

-July

Agro

past

ora-

Aug

Past

oral

-July

Past

oral

-Aug

Rive

rine

Agro

past

oral

Past

oral

Rive

rine

Agro

past

oral

Past

oral

Figure 5: Crude and under 5 death rates, per 10,000 per day, June-August 2011

CDR U5DR

Figure 6: Summary of the nutrition and food security situations in August 2011

A case of kwashirokorin Juba riverine

Ibra

him

Moham

ed,

FSN

AU

, Som

alia,

July

2011

Nutrition situation map based on the FSNAU Nutrition Situation Classification Framework and FoodSecurity situation map based on the IPC.

57

8 FSNAU Nutrition Update, May-July 2011. Cross

border nutrition.9 A Dangerous Delay. The cost of late response to

early warnings in the 2011 drought in the Horn

of Africa. Oxfam, Save the Children UK. Joint

agency briefing paper. 18 January 2012.

Available from:

http://policy-practice.oxfam.org.uk

Interview by Marie McGrath, ENN

Atrip to Washington in early October,gave the opportunity for the ENN tointerview Bread for the World andBread for the World Institute. We first

came across Bread for the World when attendinga SUN meeting in June 2010 in Washington thatthey co-hosted with Concern Worldwide. Theywere obviously in the thick of the political andpolicy setting aspects of nutrition and we werecurious to learn more about what they wereabout.

I interviewed Asma Lateef, Director, and ScottBleggi, Senior Policy Analyst for Hunger andNutrition, at the Bread for the World Institute attheir light-filled office overlooking Capitol Hill,the seat of the US government. Scott remarked aswe shook hands, that while the view is a ‘sellingpoint’ for any prospective staff, it reminds themdaily of the focus and target of their work: highlevel influence on US government policy.

Asma joined Bread in 2000 as a policy analyston the government relations team; a role involv-ing direct lobbying at “the hill”. She left for a briefperiod and returned in 2007 as Director of Breadfor the World’s research and education affiliate,Bread for the World Institute. She “came to nutri-tion late”, her primary training was as aneconomist. Scott has worked in Washington formore than 30 years, originally as a US ForeignService officer (Agriculture) and then as a consult-ant in trade and development. He joined theInstitute 2 years ago to work on mother and childnutrition and linkages with health and agriculturein US development policies and programmes.

Bread is best understood by hearing the storyof how it began. Asma described how Bread wasestablished in 1974 by a pastor, Art Simon, in NewYork City whose parish in the Lower Eastside ofManhattan was then populated with hungry andpoor people. He realised he wouldn’t be able tomeet the needs of his community by “doleing outsoup”, which led him to question what is the roleof the church in addressing the systemic issueshunger? He and a handful of others gatheredregularly to study what they could do solvehunger and care for those whom Jesus calls “theleast of these”. From this small start in TrinityLower Eastside Lutheran, he moved toWashington DC to be closer to Congress and tofurther build Bread for the World. Art was presi-

Ag

en

cy

Pro

file

NET Post Deyr 2011/12 seasonal assess-ment results indicated that famineconditions defined by IPC no longerexisted in Southern Somalia in February3rd 2012, though nearly a third of thecountry’s population remained in crisis,unable to fully meet the essential foodand non-food needs. With these results,the end of famine was declared by FAOin February 2012.

The information presented in thispaper illustrates the availability of suffi-cient early warning situational analysisto trigger an appropriate response in atimely fashion. A SCUK/Oxfam jointbriefing paper on the 2011 Horn ofAfrica crisis concluded that the scale ofdeath and suffering, and the financialcost, could have been reduced if earlywarning systems had triggered anearlier, bigger response9. Securing thenecessary resources to respondremained a challenge until after thefamine declaration that brought signifi-cant assistance. However, even wherethere was a response, insecurity limitedhumanitarian access.

The well being of populations inSomalia remain extremely fragile due toa combination of multiple natural,economic and political factors,compounded by insufficient measures inplace to mitigate loss of life and liveli-hoods. Close monitoring of the foodsecurity and nutrition situations is not inand of itself sufficient to protect thischronically vulnerable Somali popula-tion without simultaneous commitmentto intervene early where early warninginformation indicates such a necessity.

For more information, contact FSNAU:email:info@fsnau.org

Field Article

A mother and child in Mogadishu (IDPs)

Moham

ed M

oalim

/FSN

AU

, Som

alia,

July

2011

Laura

Eliza

beth

Pohl/

Bre

ad f

or

the W

orld,

US,

2012nutrition centres:

• Treatment of severe acute malnutri-tion (SAM) scaled up from 388 outpatient therapeutic programmes (OTP) reaching 7,500 children per month in early 2011, to 443 OTPreaching 9,000 per month in August 2011.

• Treatment of moderate acute malnu-trition MAM) scaled up from 512 Supplementary Feeding Programmes(SFPs) reaching 18,000 children per month in early 2011, to 641 SFP facilities reaching 25,000 per month in August 2011.

Faced with severe food deficits, thou-sands of people fled from the southernregions of Somalia to neighbouringcountries of Kenya and Ethiopia to seekassistance. According to UNHCR, anaverage of 10,000 new Somali refugeeswas arriving in Kenya’s Dadaab campsper month (at least 1,300 per day fromJune) while 5,000-6,000 per monthpeople were arriving at Dolow Adocamp in Ethiopia. As of 26 June 2011, atotal of 60,200 Somalis were registeredin Kenya – more than a 100 per centincrease as compared to the same timein 2010. The refugees arriving in thesecountries were in dire situation withGAM rates amongst new arrivals inrefugee camps in Ethiopia and Kenyaranging from 30-40%. Three in five chil-dren arriving in refugee camps inEthiopia from southern Somalia weremalnourished8. In refugee camps inKenya, more deaths were recordedamong Somali children in the therapeu-tic feeding centres in the first quarter of2011 than in all of 2010.

Following the declaration of faminein parts of Sothern Somalia, substantialhumanitarian assistance was providedto the affected population. At the sametime, the subsequent Deyr 2011 rainfallperformance was above average result-ing in average harvests which mitigatedthe extreme food deficits and reducedmortality. The FSNAU-FAO and FEWS-

58

Name: Bread for the World and Bread for the World Institute

Address: 425 3rd Street SW, Suite 1200, Washington, DC 20024

Phone: (202) 639-9400

Email: institute@bread.org

Website: http://www.bread.org

Year founded: 1974

Director: Asma Lateef, director, Bread for the World Institute

No. of staff in Bread

for the World (HQ):

80 (10 at the Institute, 70 at Bread for the World)

donations. The balance comes from founda-tions. The type of individual donor varieshugely – from one individual who “empties hispockets, puts it in an envelope and sends it tous,” to monthly donors, and larger contribu-tions from various Christian denominations.Foundations that support the work of theInstitute and Bread include the Gates andHewlett Foundations.

Asma and Scott reflected how the 2008Lancet series was a critical launch-pad forBread’s focus on nutrition. Recognising thatyou cannot fight hunger without addressingnutrition and food quality had a massiveimpact on their advocacy work and marked thepoint that Bread “jumped into action” on nutri-tion. This coincided with the 2008 global foodprice crisis, and by happy coincidence, anumber of other initiatives, such as Bread’sown mother and child nutrition projects, the

Scaling up Nutrition (SUN) Movement and theUSAID’s FTF launch in 2010. Bread has beenactively involved in developing the SUN frame-work, in the preparation for the 2010 UNGeneral Assembly Meetings and the launch ofthe simultaneous ‘1,000 Days Call to Action’.Bread and Concern Worldwide co-hosted thecivil society SUN event in June 2011. Asmaemphasised the importance of this meeting as itbrought together civil society partners, givingmany of the country representatives their firstexposure to SUN -“they suddenly had a voicein this global policy initiative.” Bread wereheartened to learn just a few months ago thatone of the country delegates that attended isnow the government representative of SUN inGuatemala, and reports directly to the country’snewly-elected president.

Agency Profile

(Left to right): Kay DeBlance, Rebecca Walker, Aaron Marez andDavid Ramos of Texas walk through the Russell Senate Office

Building on their way to a meeting in Sen. Kay Hutchison's office(R-TX). They visited the office as part of Bread for the World's

Lobby Day in Washington, D.C., on Tuesday, June 12, 2012.

dent until 1991, when David Beckmann tookover. Bread began with the values that embod-ied its work – in the early days, staff salarieswere on a needs basis, meaning that the mostsenior staff were not paid the highest. Thischanged, Asma explained, as they had to adopta more mainstream “competitive” approach toattract new staff. However the fundamentalconcern of addressing the systemic causes ofpoverty, hunger and undernutrition remains atthe heart of Bread’s work.

Scott explained the difference between Breadand the Institute – which largely relates to fund-ing and enabling tax breaks for individualdonations. The Institute began in 1978 as theBread for the World Education Fund, whichenabled donors to get a tax deduction for theircontributions. These funds cannot be used forlobbying (Americans donating to Bread do notget a tax break as the funds can be used forgrassroots and direct lobby-ing). Essentially the Institute isthe research and educationaffiliate of Bread, providingcritical analytical thinking thatunderpins and supportsBread’s advocacy andcampaigning work. Goodexamples are the analyses thathave influenced policy aroundmaternal and child nutritionand Feed the Future (FTF),USAID’s global hunger andfood security initiative.Outputs take the form of socialmedia blogs, briefing notes,with summary key recommen-dations providing a digestedread and ‘need to know’ forpolicy-makers.

There are approximately80 – 100 staff at Bread and theInstitute. A lot of the staff members undertakeeducation outreach. They are organised asdepartments such as church relations, commu-nications, government relations, organising andgrassroots capacity building. Bread has morethan 90,000 influential members and around5,000 local congregations, enabling an outreachof about one million people for its advocacywork. All members and partners are US based,enabling Bread to have a presence in all thecongressional districts of the country. Oneexception is the Canadian Food Grains Bankthat supports the annual hunger report; theirengagement reflects that the Institute’s policywork, particularly on global hunger, is highlyrelevant to Canadians as well.

In terms of funding, around 60% of theirresources come from members/individual

Bread and the Institute are not direct serviceproviders. They work in advocacy. Scottdescribed how the Institute “occupies a spacethat has one foot in understanding technicalnutrition issues and one foot in influencingpolicy”. But both Scott and Asma were clear tostate that they are not nutritionists, andcertainly not technical nutrition experts. Theyhave great respect for the nutrition expertiseout there. Their job is to “distill” and “unpack”what nutrition is, how and where it is relevantat a political and policy level, and to holdgovernment players to account. For example,the Lancet Maternal and Child Nutrition series,FTF and SUN have helped crystallise thinkingon nutrition for children under 2 years of ageand how FTF (focused on agriculture and liveli-hoods) is integrating nutrition acrossdevelopment sectors. Bread is also work onimproving foreign aid effectiveness. Bread’spresident David Beckmann (who is the 2010

World Food Prize laureate) is co-chair of the ModernisingForeign Assistance Network,informing country plans and“pushing the US to be betterdonors, with more transparencyand more M&E”.

There have been markeddevelopments, they feel, overthe last few years, in terms of USgovernment awareness andengagement on nutrition. At theG8 in Washington, US PresidentBarack Obama addressed nutri-tion, and his Secretary of State,Hillary Clinton, has become ahuge advocate for improvingnutrition, especially for womenand children. During Clinton’strip to Tanzania, she spokeabout nutrition and in all likeli-hood influenced the

development of that country’s national nutri-tion plan. The Administrator of USAID, Dr.Rajiv Shah, is talking about nutrition and is amember of the SUN lead group. Not only thatnutrition is being spoken about, but the “depthof the conversation” has increased while therehas been a progression from “how do we makelinkages” to “look at what we are doing”. Froman economic point of view, nutrition investmentmakes huge sense and “pays off”. Scottdescribed early research in Guatemala thatshowed how malnutrition affects the GDP of anentire nation. Having donors build an evidencebase showing the successes of early nutritioninterventions helps Bread and other advocacygroups make the case for sustaining fundinglevels in the US Congress for poverty-focuseddevelopment assistance.

The morning plenary session forMonday's "1,000 Days to Scale

Up Nutrition for Mothers andChildren: Building Political

Commitment" at the NationalGathering on June 13, 2011.

Speaking is Maria Otero, U.S.Under Secretary of State for

Democracy and Global Affairs.

Jim

Stipe,

US,

2011

Bread is concerned not to lose momentumafter these initial gains. ENN’s visit toWashington coincided with the eve of thesecond series of debates between the US presi-dential candidates, some three weeks before theelections. For the first time, Asma reflected,they have had a US Secretary of State who has“talked about development as a main pillar ofher job”. Indeed, Scott added, the “centre ofgravity for nutrition policy has been at the statedepartment rather than at USAID”. Both agreedthey will be hard pushed to get anotherSecretary of State with such an interest andengagement. So in the coming months, a “hugere-education” will probably be required of newkey players, as the new secretary takes officeand newly-elected members of Congress cometo Washington.

So, I asked, how exactlydoes the Bread network oper-ate? Asma described howBread mobilises its networkand local congregations, toengage their members ofCongress on domestic andglobal hunger and povertyissues, especially around anannual legislative campaign.Bread organisers work withgrassroots activists, trainingthem on advocacy andeducating them on key issues.Activities include buildingrelationships with local news-papers, with Congressionalrepresentatives throughoutthe calendar year. Breaddefines key issues and clearactions, providing a kit, educational materialsand website to support this. Once a year,hundreds of Bread’s members come toWashington DC. One good example of suchaction and impact was around budget cutsproposed by the Republican presidential candi-date’s running mate, Paul Ryan, who hadsuggested deep cuts to programmes vital tohungry and poor people. Bread and its partnerscalled on Congress to create a “circle of protec-tion around these key programmes”. So far, nomajor cuts have happened and external consult-ants have attributed this in large part to thework of Bread and its faith partners.

A key aspect of their work is building thecapacity of the faith community to “carry thetorch on nutrition” through ‘grassroots’campaigns. For the first time, Bread engagedpartners in the form of US women’s groupsamongst their members of various Christiandenominations on the issue of maternal andchild nutrition. This led to the launch of theWomen of Faith for 1,000 Days Movement. Thisin turn led to the ‘1,000 conversations about1,000 Days’ initiative. The campaign was builtaround women committing to having “1,000conversations” about the 1,000 days – from talk-ing to their member of congress to coffeemornings to chat with a neighbour. What struckAsma and Scott in this process is that thewomen proved very powerful. “The fact thatthe root causes of malnutrition – access to food,knowledge, education – are the same whereveryou are in the world hit a chord with USwomen and enabled these conversations”.Another example of action on the domesticfront has been the Women, Infant and ChildProgramme targeting children.

elections looming. “We can’t afford to let politi-cal winds change our course”.

A key area of work they are engaged with atthe moment is how to define nutrition sensitivedevelopment. Bread have “called on thecommunity” to contribute to early drafts onpapers, and will be producing a paper on nutri-tion sensitive programming that will reflectstrong inputs from UN agencies (FAO),research, academia and non-governmentalorganisations. Sometimes they build consensusin the process of collaborative work but that isnot their primary goal. They see their role morein terms of providing a platform to ask thequestions, from which consensus may emerge.They are ever mindful of their role - their job is

not to write the strategy, but toask what the strategy is andhow it can be improved?

Another key issue, Asmafeels, is how we talk about anduse the term ‘hunger’. Asmadescribed being struck by apresentation by the head ofBRAC in Bangladesh, whodescribed how we no longertolerated famine and starvation,yet seemed to tolerate stuntingat an outrageous scale. Theconnections between poverty,nutrition and health are becom-ing clearer and enabled by theconversation around the 1,000days window. When it comes toemergencies, their perspectiveand that of their members islong term. When emergencieslike Haiti and Japan happened,

they didn’t see a peak in donations, nor adecline. Bread understands the need for invest-ment in systemic change. Emergencies are partof the bigger picture.

Bread and the Institute are constantly look-ing ahead – they produce an annual HungerReport where they forecast and analyse keyissues in the coming year. Another imminentpiece of work is the 2013 Hunger Report, duefor release the week of the US Thanksgivingholiday, looks at the post 2015 agenda aroundthe Millennium Development Goals. Some oftheir work also combines domestic and globalinterests. For example, they have worked morerecently on US immigration, prompting discus-sions on the root causes (hunger, poverty) thatdrive people to come to the US, as well as consid-ering the conditions they live in, in the US.

Asma concluded by reflecting how “thepower is in the people – in government, theyknow about Bread because of the letters theyreceive from Bread’s members”. Action atgrassroots level really impacts on higher levelpolicy making.

As Scott left for his high level governmentmeeting, I was struck by Bread’s capacity toengage at two extreme levels – the ‘person onthe street’ through their grassroots work andthe ‘man in the White House’, through theirhigh level policy work. It seems that the visionof their founder holds as true today as it didmore than 35 years ago, amongst this refresh-ingly curious and determined organisation.Making sense of nutrition - and making nutri-tion make sense – to the powers that be is nosmall task. Long may they continue.......

59

Another role for Bread members is educationof the general public around how the US can“afford’ overseas programmes in tighteconomic times, emphasising for example, howlittle the US actually spends on overseasprogrammes, and how it is a good investment,serving national security. As well as researchand education, the Institute is involved in non-legislative advocacy, e.g. policy work. Areal-time example was Scott having to dash offas the ENN interview ended, to a meeting at theUS Office of Management and Budget (OMB)on budget planning and investment on foodsecurity and nutrition for fiscal year 2014.

The Christian ethos is deeply engrained inBread. There is no requirement to be a Christian

to work at Bread, however many staff –Christian and non-Christian alike - find itenables them to “come and live out their faith,working on social justice issues”. They alsoreach out to other religious groups with sharedvalues, in fact, the ‘Alliance to End Hunger’was created in 2001 to facilitate just this.Recognising that Christian organsations werenot going to be able to “do it alone”, Breadcreated this second affiliate to engage otherfaith groups, corporations, and universities andprovides management and staff services to theAlliance. Tony Hall, a former Congressmember, is Executive Director,

Bread doesn’t purport to have all theanswers, far from it. They are curious and thequestions they struggle with are addressed togovernment in order to find a solution. Theirquestions are not rhetorical or conceptual butare grounded in the practicalities – “DoesUSAID have a strategy? Do we [US govern-ment] need to organise ourselves differently torealise what we want do? Do we need to recruitnew people for the expertise we need todeliver?” And when they identify needs andgaps, Scott explained, “we go and make thecase on the Hill’ with elected officials. They askquestions on accountability, Asma added –“Who is the nutrition focal point in the USgovernment? We are asking all these SUNcountries to identify a focal point, where isours? Who is accountable for how we deliveron nutrition? Who is responsible for coordina-tion between US State Dept and USAID? Whoseneck is on the line if we fail to deliver?” Theseare critical questions as they enter into a transi-tion period in the US government with

Agency Profile

Anna Lartey, associate professor andformer head of department, nutrition

and food science, University of Ghana,presisdent-elect of the internationalUnion of Nutritional Sciences. She is

speaking during the panel entitled"1,000 Days to Scale Up Nutrition:

Building a Movement."

Jim

Stipe,

US,

2011

Adèle Fox is currently based inConcern Worldwide Burundi office asMaternal, Newborn and Child HealthTrainee. Adèle completed a Mastersin Public Health from the London

School of Hygiene & Tropical Medicine in 2011. She hasprior experience with Save the Children UK, working asPolicy & Research Coordinator and with Oxfam GB, asSystems & Resources Team Leader.

The author would like to acknowledge the involvementof Alive & Thrive for providing support and materials forthis project. She would also like to thank Peter Gottertand Tigiste Abate of Alive & Thrive and Gashaw Abera,Gwyneth Cotes, Suzanne Fuhrman, and Pankaj Kumarof Concern Worldwide for their assistance in providinginformation and reviewing this article.

By Adèle Fox

Integrating Infantand Young ChildFeeding and theProductive SafetyNet Programmein Ethiopia

Concern Worldwide has worked inEthiopia since 1973, primarily inthe areas of health and nutrition,livelihoods, and HIV and AIDS.

Two years ago, Concern documented thepoor nutritional situation in Dessie ZuriaWoreda and the multiple obstacleshampering previous efforts to improve thesituation1. Concern concluded that a multi-sectoral approach to improve infant andyoung child feeding (IYCF) practices andto increase access to food was among theresponses needed. In 2010, the ‘IYCF intoProductive Safety Net Programme (PSNP)’project (IYCF-PSNP) was launched as amulti-sectoral pilot project aimed at reduc-ing malnutrition in Dessie Zuria. Theproject targets poor households enrolled inthe existing PSNP, as well as the generalpopulation, and addresses both the directand root causes of malnutrition. The proj-ect aimed to develop an effective,sustainable and scalable model to improveIYCF practices in the most vulnerablehouseholds. The final results have beenimpressive, with large improvements inIYCF practices and a positive responsefrom the communities and stakeholdersinvolved in the project.

BackgroundDessie Zuria Woreda (district) is situated inthe Amhara Region of Ethiopia, to thenorth of Addis Ababa. It is made up of 31kebeles (clusters of villages), half of whichare largely dependent on a typically unreli-able and short rainy season. Dessie Zuriahas been listed by the Ethiopian govern-ment as chronically food insecure for thepast 11 years, and almost 40% of its popu-lation are eligible to participate in theGovernment’s PSNP (see Box 1).

Over half of all children in Dessie Zuria(54%) are chronically malnourished andthe woreda suffers from persistently highrates of acute malnutrition. Annual nutri-tion surveys show that between 2000 and2011 global acute malnutrition rates haveonly once, in 2004, dropped below 10%2.

Concern Worldwide has supported theEthiopian Ministry of Health to providenutrition services for children in theworeda since 2000, through Community-based Management of Acute Malnutrition(CMAM). However, the persistently highrates of acute and chronic malnutritionshowed that management of malnutritionalone was insufficient, and that a preven-tive approach to undernutrition wasneeded.

Early assessments identified food inse-curity, high rates of illness, and poor IYCF

practices as contributing factors leadingto high rates of malnutrition. A baselinesurvey conducted by Concern inNovember 2010 found low rates of exclu-sive breastfeeding (31%) and poorcomplementary feeding practices3. Only13.3% of children received meals with therecommended dietary diversity, and veryfew children ate foods rich in iron (4%)and Vitamin A (13.7%). The sample sizefor the baseline survey was 694 childrenaged 0-23 months.

Project descriptionConcern Ethiopia started the two-yearproject in October 2010 with funding fromAlive & Thrive, a Gates Foundation initia-tive. The project aimed to improve IYCFpractices among the most vulnerablehouseholds in Dessie Zuria Woreda inorder to reduce malnutrition, as well as totest the effectiveness of this model forscaling up and replicating elsewhere.

Concern Worldwide aims to assistpeople living in extreme poverty to makelasting improvements in their lives. Inline with this goal, Concern’s pro-grammes target the poorest and mostvulnerable households in the areas whereConcern is at work. In Ethiopia, thegovernment’s PSNP programme providessupport and assistance to the poorestmembers of each community. This projectwas designed to link with the PSNP sothat nutrition support would be targetedto those most in need of assistance. ThePSNP presented an entry point to helpfamilies improve their IYCF practices, asit targets the poorest households, who aremore likely to have undernourished chil-dren. The PSNP also provides manypotential contact points to deliver behav-iour change communication (BCC)messages.

The project was implemented througha range of strategies, but the primaryfocus was on fostering social and behav-ioural change. At the start of the project, abaseline survey was conducted and form-ative research was carried out. The projectused a Trials of Improved Practices (TIPs)approach to identify key barriers toimproving IYCF practices, as well as toidentify and pre-test a set of simple, real-istic actions that mothers could take toimprove their child’s nutrition.

1 Style S. History of nutritional status and Concern’s

response in Dessie Zuria Woreda, Ethiopia, Field

Exchange, Feb 2011, Issue 40.2 GAM based on z-scores, using NCHS 1977 reference

as per Ethiopian National Guidelines.3 Standard WHO indicators were used in baseline and

endline surveys of IYCF practices.

For the past decade, the Government of Ethiopia has been implementing a federal FoodSecurity Programme (FSP) which aims to ensure food security for five million chronically foodinsecure people and for 10 million more who are negatively affected by food shortages duringdrought years. Started in 2005, one of the three pillars of the FSP is the Productive Safety NetProgramme (PSNP) whose objectives are to reduce household vulnerability, improve house-hold and community resilience to shocks and break the cycle of dependency on food aid.

It operates through payment for labour intensive public works and direct support throughcash or food transfers for those unable to participate in public works. Beneficiary householdsare intended to acquire sufficient assets to graduate from the PSNP, at which point theyreceive assistance for one further year.

Box 1: What is the Productive Safety Net Programme (PSNP)?

A PSNP beneficaryand Model Mother

School directors preparing special porridge with egg andvegetables, supervised by Hawla Eshetu , VCHW for TebasitKebele. They are wearing t-shirts with various IYCF messages

Adele

Fox,

Conce

rn W

orldw

ide,

Eth

iopia

, 2012

Adele

Fox,

Conce

rn W

orldw

ide,

Eth

iopia

, 2012

Field Article

60

Field Article

Building the capacity of multi-sectoralactors at woreda, kebele and communitylevel to deliver effective IYCF messages andencourage behaviour changeConcern used a cascading approach to train keydecision-makers and community members onmalnutrition, the PSNP, optimal IYCF practicesand methods for promoting behaviour change(see Figure 1). At the woreda level, a task forceof key officials from across different sectors wasformed and trained. The woreda officials thentrained the kebele-level Food Security TaskForce members, including Health ExtensionWorkers (HEWs), Development Agents (DAs),kebele administration staff, school directors,girl’s clubs in schools, Women’s Affairsmembers and religious leaders. HEWs thentrained Voluntary Community Health Workers(VCHWs) to work directly with mothers’support groups, which included one VCHWand six female members of the community.Through the mother’s groups, the VCHWpromoted key IYCF messages to the community

Supporting the promotion of effective IYCFmessages at key contact points, includingthose linked to the PSNPBCC was a central aspect of the IYCF–PSNPproject. Findings from formative research wereused to help design programme activities andmessages, and the project utilised many differ-ent contact points for disseminating messages.Health workers carried out education sessionsat PSNP pay days, public work days, childhealth days, screenings for growth promotion,and targeted supplementary feeding progr-amme sites. Local radio and drama were alsoused to convey messages to the community.

Concern made use of existing resources asmuch as possible. One key tool was the ChildNutrition Card developed by the Alive &Thrive project, which lists ‘7 Excellent Actions’(See Figure 2). This tool emphasised sevensimple actions that mothers and fathers couldcarry out to improve their child’s nutrition.Families were given the poster-sized card,which lists the seven actions with a tick boxnext to each one for the mothers to complete.Mothers were encouraged to place them ontheir walls and show them off proudly to theirneighbours. The messages in the nutrition cardwere reinforced through mother’s groups,school clubs, and community outreach. Oncemothers completed the seven actions, theyattained the title of “Model Mother” and grad-uation ceremonies were held to celebrate theirsuccess. Another aspect of BCC focused onequipping community members with practicalskills to improve child feeding practices. Healthworkers and mothers’ support group membersconducted house visits to provide one-on-onecounselling on breastfeeding and preparationof complementary foods. Cooking demonstra-tions conducted at health facilities andcommunity events allowed people to learnthrough observation and participation. Recipeswere developed by Concern, based on foodavailability in the area and vegetablespromoted through seed distribution.

Enabling communities to improve theirIYCF practicesFormative research at the start of the projectfound that there were a number of barriers toadopting recommended IYCF practices,beyond limited knowledge and skills. Giventhe high levels of food insecurity in the woreda,

the project found it was necessary to supportfamilies and communities with practical toolsand resources to enable them to carry out therecommended feeding practices.

To improve food production, DevelopmentAgents disseminated seeds and tools to thepoorest households, provided advice andsupport on how to diversify their crops, andpromoted home gardens as well as IYCFmessages. To improve the health of women andchildren, VCHWs referred pregnant mothers tohealth institutions to encourage iron supple-mentation during pregnancy as well as vitaminA during the postnatal period. Protected springswere also constructed by Concern to improvewater, hygiene and sanitation practices.

Engaging the entire community to fostersocial changeThe IYCF – PSNP programme looked beyondthe traditional target population of mothers andyoung children, harnessing the power of abroad range of influential groups to changeoverall social norms around IYCF. For example,BCC targeted fathers to encourage them toprovide healthy foods for their children’smeals. Religious leaders were trained on keyIYCF messages and were involved in local taskforces.

The Smart and Strong Schools Approach isanother unique concept piloted by the project.Through this activity, children in girl’s after-school clubs are taught about breastfeeding,complementary feeding and hygiene, andencouraged to bring home the lessons theyhave learned. School directors and studentsenthusiastically embraced the project.According to School Director EndrisMohammed, “The girls are very motivated to sharethe messages with their friends and families becausethey understand their importance.”

ResultsDuring the two year project, 978 familiesachieved a ‘model family’ status. More than 850Mother’s Support Groups were trained. Anendline survey4 was conducted in June 2012, ona sample of 807 children aged 0-23m. There arefour main seasons in the woreda: Kiremt/longrains (July-September), Meher/harvest (October-January), Bega /dry summer (February to May)

Concern trainsWoreda level

Task Force

Woreda trains Kebelelevel Food Security

Task Force members

Kebele level Food SecurityTask Force members train

Volunteer CHWs

Volunteer CHWs identify and formvillage level IYCF support groups

HEWs, DAs, VCHWs and support groupschannel messages to wider community

Figure 2: Nutrition Card of ‘7 excellent actions’

4 Hailu, W, Infant and Young Child Feeding Practices in

Dessie Zuria Woreda. End line survey report, Concern

Worldwide, 2012

Figure 1: Cascade of training to multi-sectoral actorsat woreda, kebele and community levels

A PSNP beneficary andModel Mother

Adele

Fox,

Conce

rn W

orldw

ide,

Eth

iopia

, 2012

61

HEW: Health Extension Worker, DA: Development Agent,

VCHW: Voluntary Community Health Worker

Table 3: Changes in complementary FeedingPractices, Dessie Zuria, 2010-2012

Field Article

change significantly despite recipes andmessages promoting the consumption of eggsand meat. The agricultural aspect of the projectdid not have a livestock component, so therewas no support for increasing production ofanimal source foods. This is something thatmay be considered in future projects. AnotherConcern project in Ethiopia has found that eggsare fairly widely available, but they are usuallysold for profit.

Interviews with programme participantsfound that this project has been extremely wellreceived by stakeholders at all levels, fromWoreda officials to Model Mothers, many ofwhom can now list the recommended actionswithout hesitation.

One unexpected outcome of the project hasbeen increased implementation of appropriatematernity leave for pregnant and lactatingwomen. Prior to the project, these womennormally participated in public work days aspart of the PSNP. The project has been effectivein creating an understanding of the importanceof rest for mothers to give them time to recoverand establish breastfeeding. As a result, womenwere excused from doing public works duringpregnancy and for the first 4 months after birth,attending health and nutrition sessions instead(PSNP payments continued during this time).

Lessons learnedA number of factors contributed to the successof the IYCF – PSNP project. The project took amulti-sectorial approach, involving actorsacross a wide range of groups and sectors. Itwent beyond simply BCC, targeting theenabling environment as well as social norms,and involving the community at large. The proj-ect used multiple platforms and approaches todisseminate messages, and used a targeted

approach to behaviour change, basing projectactivities and messages on formative researchand emphasizing simple, do-able actions ratherthan health education messages.

The project engaged with actors from agri-culture, education, women’s affairs, and healthsectors. Each sector worked together towards acommon purpose, leading to increased owner-ship and accountability. In an interview, oneSchool Director described IYCF as now being“everyone’s responsibility.” This approach alsoprovided greater opportunities for engagingwith communities. Cooking demonstrations,school clubs, and agricultural support were allcombined to provide an overall aim of prevent-ing malnutrition among children.

The project worked across levels, creatingstrong links between woreda, kebele andcommunity levels through a cascading style oftraining and through the continued provisionof support and supervision. This multi-levelapproach has had many advantages; for exam-ple, Hawa Hussein, a VCHW in GuguftuKebele explained that if she has a problem, suchas a lack of knowledge on an issue, she cancontact the HEW or other member of the KebeleFood Security Task Force.

The project went beyond traditional commu-nication approaches, using a social andbehavioural change model. Early assessmentsshowed that simply providing messages alonewas unlikely to be effective, given widespreadfood insecurity and other barriers to behaviourchange. This project aimed to influence thewider community and social norms as a whole,as well as address barriers to practising recom-mended IYCF behaviours. For example,policies around women’s maternity leave werestrengthened, agricultural support addressed

and Belg/short rainy season (May to June). Thebaseline study was done in November, which isthe beginning of the harvest and the end linewas in June which is the beginning of thehunger period. Both periods can be consideredas similar with regards to the food security situ-ation and child feeding practices. Comparingbaseline with endline strongly suggests that injust a short period of time, IYCF practices havegreatly improved. The percentage of motherswho reported that they had started breastfeed-ing their child within the first hour after birthrose from 26% to 75%, and the proportion ofchildren less than 6 months old who wereexclusively breastfed rose from 36% to 91% – aremarkable increase (see Table 1). It is thenature of the EBF indicator that the exclusivebreastfeeding rate is likely to be somewhatoverestimated, as it only measures whether achild received breastmilk exclusively in theprevious 24 hours, and includes very younginfants who are more likely to be exclusivelybreastfed. There was a reduction in the percent-age of mothers who reported giving theirchildren water in the first month of life, from48% to 18%, but almost one in five children arestill given water at a very young age.

The end line survey also noted improve-ments in complementary feeding practices,although these were less dramatic (see Table 2).Approximately one-third of children (32.7%)aged 6-23 months consumed a minimumacceptable diet, up from 10% at baseline. Thisincrease is due to improvements in both dietarydiversity and meal frequency. However, themajority of children still received foods fromonly 3 food groups or fewer in the previous day.

Several other indicators showed littlechange. There was a small increase in thepercentage of children aged 6-8 months whohad started complementary foods (see Table 3).The consumption of iron rich foods did not

100%90%80%70%60%50%40%30%20%10%

0%Early

initiation ofbreastfeeding

Consumptionof water

between 0-1months

ContinuedBreastfeeding

at 2 years

Exclusivebreastfeeding

2010 2012

Table 1: Changes in breastfeeding practices, DessieZuria, 2010 to 2012

100%90%80%70%60%50%40%30%20%10%

0%Minimum

dietary diversityMinimum

acceptable dietMinimum

meal diversity

2010 2012

Table 2: Changes in quality of diet among childrenaged 6-23 months, Dessie Zuria, 2010 to 2012

100%90%80%70%60%50%40%30%20%10%

0%Introduction of

solid, semi-solidor soft foods

Consumption ofvitamin A rich

foods

Consumption ofiron rich foods

2010 2012

At Kebele level school directors andgirls' clubs are trained

Schools facilitate question/answer sessions and promote

school gardening activities

Students share key IYCF messagesin their homes and communities

At the end students participate in model house and

family certificate ceremony

Figure 2: Non-traditional community targeting on IYCF

The multi-level Guguftu Kebele team including Zewditu Jemal and AmakelAhmed (HEWs), Hawa Hussein (VCHW), Fato Assen (support group member)and Seada Mohammed and Aminat Seid (Model Mothers), who work togetherto reach all households in the kebele with the essential IYCF actions.

Adele

Fox,

Conce

rn W

orldw

ide,

Eth

iopia

, 2012

62

food insecurity, and support to water andhealth services aimed to prevent commonillnesses.

Conclusions Using a comprehensive, multi-sectorialapproach to social and behaviour changehas resulted in impressive improvements ina number of IYCF practices over a shorttime. The IYCF-PSNP project has beencommended as innovative, effective, andsustainable by a wide range of stakeholderswithin Ethiopia, and interviews withcommunity members showed that mothersand other community members have enthu-siastically embraced the project andmaterials.

The approach has been able to reach alarge number of people who are widelydispersed over challenging terrain.Channelling activities through the PSNPcreates additional contact points andensures targeting of the poorest households.The two year pilot also suggests that despitestrong linkages with the Government and asense of community ownership, longer termsupport will continue to be necessary toaddress underlying factors of malnutritionsuch as food security, health, and hygiene,as well as to reinforce the gains made.

Possible areas for future investigationinclude engaging the private sector forproducing ready to use fortified comple-mentary foods, increased water andsanitation activities within communities,strengthening linkages with livestocksupport to promote consumption of iron-rich foods, and the integration of ConcernIYCF centres with community gardens toalleviate food insecurity.

Concern is currently in the process ofreplicating this project in four other woredasin Ethiopia.

For more information, contact: PankajKumar, Assistant Country Director,Concern Ethiopia, email:pankaj.kumar@concern.net.

Erica Roy Khetran is theCountry Director for HelenKeller International (HKI) inBangladesh. She has lived andworked in Asia since 2004 witha focus on emergency, food

security and nutrition programmes.

Helping homesteadgardeners mitigate theimpact of soil salinity

By Erica Roy Khetran

Homestead food production(HFP) is an effective way tohelp poor families increaseaccess to nutritious food and

new sources of income. HFP enableswomen to access fresh vegetables forthemselves and their children directly,instead of relying on a male familymember to purchase them. Furthermore,proceeds from household gardens areusually controlled by women and thusmore likely to be used for education,healthcare and other activities, whichdirectly benefit women and children.Helen Keller International (HKI) hasimplemented HFP programmesthroughout Bangladesh since the early1990s. As part of the global Project LaserBeam initiative, the Kraft FoodsFoundation is currently supporting HKIto increase women’s asset base and foodsecurity through HFP, improve nutrition,address gender barriers and intra-house-hold communication and strengthenfarming groups.

However, a changing climate requiresthat new practices be integrated intostrategies to promote HFP, particularlyin the vulnerable areas of southernBangladesh that face frequent floods andcyclones. Shymnagar, a sub-district insouth-western Bangladesh bordering theBay of Bengal, is particularly vulnerableto floods and storms. The impact of theseevents on food production can persist foryears. In May 2009, the area was hit bytropical cyclone Aila, a category 1 storm.Heavy rain and wind led to a tidal surgeof up to 10 feet and several rivers brokethrough embankments (already weak-ened by nearby shrimp farms which cutholes through the embankments to floodtheir ponds), causing widespread inlandflooding. In some areas, 2-3 feet of waterremained for up to 20 days. In additionto immediate destruction of homes,

crops and assets and widespread displace-ment, the waterlogging damaged fruit andtimber trees and resulted in increased soilsalinity which, three years later, continuesto inhibit agricultural production.

The soil salinity is worst during dry peri-ods. The spring of 2012 was particularlydry, with no rainfall during the month ofMay, according to the local farmers and theDepartment of Agricultural Extension.With support from the Kraft FoodsFoundation, HKI surveyed the impact ofsoil salinity on household gardeners inShymnagar during this period and rolledout strategies to help families continuevegetable production.

Soil salinity in formerly flood-affected areasA survey, including visits to 144 gardensand three focus group discussions, wasconducted in late May, 2012, by HKI’sProject Laser Beam staff. The purpose of thesurvey was to assess the impact of soilsalinity on vegetable cultivation inShymnagar, identify the most tolerant vari-eties of vegetables, document localpractices to cope with salinity and, ulti-mately, develop recommendations to assistfamilies in areas with high salinity withcontinuing vegetable cultivation. All house-holds had previously been provided witheight varieties of summer vegetable seedsand HFP training by Project Laser Beam.

Of the surveyed household gardens,levels of soil salinity were very high in 38%of gardens and moderate in another 34%. Atotal of 14% of gardens were completelydestroyed by soil salinity, meaning that noseeds could successfully germinate. In themost saline areas, farmers and the localDepartment of Agricultural Extensionreport that, following cyclone Aila, standingflood waters persisted for 15-20 days. Inmoderately saline areas, flood waters stoodfor seven days or less. No gardens were

Improved compost pile

HKI,

Bangla

desh

, 2012

Field Article

The soil salinity survey was led by ProjectLaser Beam team members, Md. HafizurRahman and Moshfikur Rahman-Sr. Thanks toMd. Amun Uddin for his extensive inputs intothis article (including photos) and leadershipin assisting homestead gardeners inShymnagar.

Fie

ld A

rtic

le

63

Beyu Mohammed, a PSNP beneficiary andModel Mother in Tebasit kebele, received

vegetable seeds from Concern and hasgrown carrots, kale, potatoes and beetroot,all of which she uses in the complementaryfood she has learned to feed her child. Shesells any vegetables that are left over and

has been able to buy soap with the money.

Adele Fox, Concern Worldwide, Ethiopia, 2012

64

Field Article

found to be completely without soil salinity. Inthe most affected areas, household ponds, theprimary source of irrigation, were also extremelysaline and thus unsuitable for irrigation.

About half of households were alreadyimplementing practices to cope with soil salin-ity. Among these, 38% were using organiccompost and 34% were planting crops in pitswhich were first leached with water. However,households with the knowledge and means toadopt these practices tended to be among thebetter off; poor households who are morereliant on their gardens for food and incomehad fewer coping mechanisms and were thusmost affected by the salinity.

It is worth noting that the increased soilsalinity since cyclone Aila has affected otheraspects of the food production system inShymnagar. Small farmers report that produc-tion of rice, an important cash crop and sourceof food, has decreased since the storm.Households with access to larger plots of landhave moved from rice cultivation to shrimpcultivation, which provides fewer day-labouropportunities for poor households.

Mitigating impact of soil salinity onhome gardensBased on survey findings, HKI promotedseveral practices to help families whose landswere very affected by salinity, continue HFP.This strategy focused on practices that requiredminimal or no financial investment, recognizedthe limited availability of fresh water to irrigateand flush the soil, and could be implementedprimarily by women.

Key practices included mulching with ricestraw, coconut coir or other locally availableorganic materials to increase water retention ofthe soil. Households were able to obtain thesematerials free of cost, in most cases from theirown family farms. Development of compostpits near homestead gardens and use ofcompost in preparing and managing the soilwas also encouraged. Compost effectivelyincreases organic content and available nutri-ents in the soil, helping to counteract the effectsof salinity. Composting is free of cost, butrequires an ongoing investment of time andspace on the part of families. Lime, purchasedby households from the open market, was alsoused to treat saline soil. The cost of lime wasminimal and easily managed by households.However, given cultural limitations on access tomarkets, women frequently needed to requestthat a male family member purchase the lime.

Leaching requires covering the area withsufficient water to fully infiltrate the soil andstand on the surface before being drained away.This is effective, but for families with limitedaccess to fresh water leaching is impractical forlarge areas. HKI encouraged two practices thatenabled households to increase production byleaching much smaller areas. First, sowing ofseeds in beds, with the soil first leached of saltand then prepared with mulch and compost.Families then transplant seedlings to largerfields after heavy rain, when the soil is lesssaline. Secondly, HKI promoted ‘pit cropping’,in which households sowed crops in small pitswhich were first leached and prepared withcompost. This was most effective whencombined with other practices, including seedbeds and use of mulch and lime.

Finally, HKI’s survey found significantlygreater resilience in salinity among somevegetable crops. These include Indian spinach,sweet gourd, okra and Kangkong- a leafy greenvegetable rich in vitamins A and C as well asiron and calcium. In most cases, promotion ofthese vegetables required distribution of seedsdirectly to households. This is particularly truefor Kangkong, which was among the mostsaline-resistant crops but is relatively new tosouthern Bangladesh. Thus this strategy, whilehighly effective, may not be practical inprogrammes which do not have resources foragricultural inputs.

Container gardeningIn areas most affected by soil salinity, evenleaching and measures to increase soil waterretention and nutrient content will not enablehouseholds to grow vegetable throughout theyear, or will only enable gardeners to produce alimited number of crops (mainly kankong andIndian spinach) in the dry season. Containergardening is an effective strategy to increase thenumber of planting seasons and garden diver-sity in these areas, though ongoing training andtechnical support is required to introduce thisnew practice.

Containers available to households inBangladesh at low- or no-cost include nylonbags (such as those used for animal feed), juteshopping bags, old metal and plastic bucketswith holes in the bottom and used clay pots.Containers should be about 20 litres in size. Avariety of media can be used for planting,including compost, rice hulls, sawdust andchopped wheat straw. ‘Manure tea’ (water inwhich cured manure has been steeped) is onelow-cost method to fertilise these media.

Fully damagedgarden, with thegarden's owner

Partially damaged garden

HKI,

Bangla

desh

, 2012

When fertilised and irrigated regularly, thisstrategy enables households with very limitedaccess to non-saline soil grow vegetablesthroughout the year. HKI’s experience has beenthat homestead gardeners are at first scepticalthat planting in non-soil media will work, andthus demonstration sites are critical to ensuringthe practice is adopted.

ConclusionsThis set of practices was identified to meet thespecific needs of households in southernBangladesh. However, most are applicable tothe growing number of areas in the worldwhere poor households are affected byincreased soil salinity. Most methods are verylow cost and, with the exception of promotingnew varieties, can be achieved with no addi-tional inputs. The ability of households toreplicate all practices was increased by workingthrough Village Model Farms, which offereddemonstration sites and a space for meetingand training to all homestead gardeners in avillage.

Considering effectiveness in mitigating soilsalinity and low cost to beneficiaries in terms oflabour, money and time, mulching and plantingof seed beds are highly replicable strategies thatcan be easily integrated into any HFPprogramme. Digging and treatment of pits forcrop plantation are also very effective and mosthouseholds do not find it to be overly labourintensive. However, introducing this practicerequires a relatively high level of expertise byprogramme staff in order to demonstrate thecorrect method of soil management and plant-ing for various types of crops. Composting is apractice that poor households have found moredifficult to adopt. Properly managed compostpits require allocation of appropriate space(which can be scarce on small homesteads) andongoing effort to maintain. Nonetheless,composting is very effective in mitigating soilsalinity and increasing soil water-holdingcapacity. A key component of healthy gardens,composting also brings other benefits such asdiscouraging the use of chemical fertilizers andincreasing crop productivity in any environ-ment. It is therefore worth developing tools andtechniques to promote composting in areaswhere vegetable cultivation is a priority strat-egy to increase nutrition and income for poorhouseholds.

For more information, contact: Erica RoyKhetran, email: ekhetran@hki.org

Village model farmerdemonstrates addingcompost to pumpkinsplanted in pit crop system

65

People in aid

Participants in UNICEF/GNC/SCUK workshop on IYCF in development and emergencies (see news piece)

Congratulations to Philand Gemma Wilks on thearrival of Matilda on 30thNovember. If the ENNwebsite starts slowingdown, blame the baby......

Josephine Ippe (GNCCoordinator), with herhusband, James, and theirbeautiful new arrival, Joy.

Invite to submit material to Field Exchange

Many people underestimate the value of their individual field experiences andhow sharing them can benefit others working in the field. At ENN, we are keen tobroaden the scope of individuals and agencies that contribute material for publi-cation and to continue to reflect current field activities and experiences inemergency nutrition.

Many of the articles you see in Field Exchange begin as a few lines in an email oran idea shared with us. Sometimes they exist as an internal report that hasn’tbeen shared outside an agency. The editorial team at Field Exchange can supportyou in write-up and help shape your article for publication.

To get started, just drop us a line. Ideally, send us (in less than 500 words) yourideas for an article for Field Exchange, and any supporting material, e.g. an agencyreport. Tell us why you think your field article would be of particular interest toField Exchange readers. If you know of others who you think should contribute,

pass this on – especially to government staff and local NGOs who are underrepre-sented in our coverage.

Send this and your contact details to:Marie McGrath, Sub-editor/Field Exchange, email: marie@ennonline.netMail to: ENN, 32 Leopold Street, Oxford, OX4 1TW, UK. Tel: +44 (0)1 865 324996 Fax: +44 (0)1 865 597669

Visit www.ennonline.net to update your mailing details, to make sure you getyour copy of Field Exchange.

If you are not the named recipient of this Field Exchange copy, keep it or pass iton to someone who you think will use it. We’d appreciate if you could let us knowof the failed delivery by email: office@ennonline.net or by phone/post at theaddress above.

66

Editorial teamJeremy ShohamMarie McGrathDeirdre Handy

DesignOrna O’Reilly/Big CheeseDesign.com

The Emergency Nutrition Network (ENN) is a registered charity in the UK (charity

registration no: 1115156) and a company limited by guarantee and not having a

share capital in the UK (company registration no: 4889844). Registered address: 32,

Leopold Street, Oxford, OX4 1TW, UK. ENN Directors/Trustees: Marie McGrath,

Jeremy Shoham, Bruce Laurence, Nigel Milway, Victoria Lack, Arabella Duffield

Jeremy Shoham (Editor), MarieMcGrath (Sub-editor) andCarmel Dolan are ENNTechnical Directors.

Orna O’ Reillydesigns andproduces all of ENN’spublications.

Phil Wilks (www.fruitysolutions.com)manages ENN’s website.

After 4 years working with the ENN asmailing assistant, Katherine Kaye hasrecently left the ENN. With her valuablehelp, many of you will have receivedField Exchange and our other resourcesover the years. The ENN team warmly

acknowledge Katherine's contribution to the work ofthe ENN and wish her the very best in her new ventures.

Chloe Angood is anutritionist workingpart-time with ENN ona number of projectsand supporting HumanResources.

Thom Banks is theENN's Desk OperationsOfficer and provideslogistical and projectsupport to the ENNteam.

Michele Toler hasjoined the ENN asOperations andFinance Assistant,based at the ENN'soffice in Oxford, UK.

Matt Todd is the ENNfinancial manager, over-seeing the ENNaccounting systems,budgeting and financialreporting.

The opinions reflected in Field Exchangearticles are those of the authors and donot necessarily reflect those of theiragency (where applicable).

The Team

Field Exchangesupported by:

The Emergency Nutrition Network (ENN)

grew out of a series of interagency meetings focusing on food andnutritional aspects of emergencies. The meetings were hosted byUNHCR and attended by a number of UN agencies, NGOs, donors andacademics. The Network is the result of a shared commitment toimprove knowledge, stimulate learning and provide vital support andencouragement to food and nutrition workers involved in emergencies.The ENN officially began operations in November 1996 and haswidespread support from UN agencies, NGOs, and donor governments.The network aims to improve emergency food and nutrition programmeeffectiveness by:

• providing a forum for the exchange of field level experiences• strengthening humanitarian agency institutional memory• keeping field staff up to date with current research and evaluation

findings• helping to identify subjects in the emergency food and

nutrition sector which need more research.

The main output of the ENN is a tri-annual publication, Field Exchange,which is devoted primarily to publishing field level articles and currentresearch and evaluation findings relevant to the emergency food andnutrition sector.

The main target audience of the publication are food and nutritionworkers involved in emergencies and those researching this area. Thereporting and exchange of field level experiences is central to ENNactivities. ENNs five year strategy (2010-2015) is available atwww.ennonline.net

Contributors for this issue

Thanks for the pictures to:

Cover

Office SupportKatherine KayeMatt ToddThom Banks

WebsitePhil Wilks

Kate SadlerFreddy G.HoungbeLaura Elizabeth Pohl/Bread for the WorldJim StipeSaul GuerreroSamuel Hauenstein SwanBen AllenGwyneth CotesErica Roy KhetranCarlos Grijalva-Eternod Martha MwangomeAngelo TomediAli MaclaineRoshan RamlalPascale FritschHelpAge InternationalMélody Tondeur and Sarah Style

A woman in receipt of WFP food assis-tance in Al-Mahweet, Yemen;WFP/Abeer Etefa, Yemen, 2012

Susana MorenoElisa Dominguez Brigitt OlagivelNatasha LelijveldAdèle FoxCamille Guyot benderJulie MayansMuriel CaroMarie Morgane DelhouneGeoff BrouwerClaire BaderSarah StyleMelody TondeurSaul GuerreroMaureen GallagherKate SadlerGwyneth CotesErica Roy KhetranAbukar Yusuf Nur Ahono BusiliElijah OdundoJoseph WaweruLouise Masese-MwirigiMohamed BorleTom OgutaSaleem SumarLaila Naz TajIqbail KermaliHenry MarkAsma LateefScott Bleggi

Emergency Nutrition Network (ENN)32, Leopold Street, Oxford, OX4 1TW, UK

Tel: +44 (0)1 865 324996

Fax: +44 (0)1 865 597669

Email: office@ennonline.net

www.ennonline.net