ISSN 1743-5080 (print) June 2008 Issue 33

• SQEAC: new method evaluatingaccess and coverage

• Surveying nomads in Mali• Integrated surveillance in

Malawi• Piloting LQAS in Somaliland• Blanket BP5 in North Dafur• Animal husbandry and

agricultural projects in Uganda

1

This issue of Field Exchange is publishedin the ‘after-glow’ of the high profile andlong-awaited Lancet series on ‘Maternaland Child Undernutrition.’ The series of

five papers are summarised in this issue withvarious comments posted in the letters section.In essence, the Lancet series sets out the scaleof nutrition problems globally, their human andeconomic impact, evidence for interventionsthat have (or have not) made some inroads intothe problem, culminating in an analysis of whatneeds to change if the global burden of under-nutrition is to be addressed. Nutrition in emer-gencies gets very little attention in the series(mainly in paper 5). This may well be appropri-ate given the scale of endemic undernutrition innon-emergency contexts. However, the limiteddiscussion and analysis of nutrition in humani-tarian crises is piece-meal, poorly developed orsubstantiated and contributes little to currentdebate and thinking. A major opportunity hasbeen missed.

Emergency issues raised in the seriesinclude politicisation of food aid, remittances,role of private sector funding, early warningand needs assessment, assessment ofprogramme coverage and the lack of evidencebase for what does and does not work. Manycritically important and challenging areas donot get a mention. To name a few, the cashversus food debate, effective means foraddressing moderate malnutrition at populationlevel, scaling up livelihoods interventions, theinstitutional ‘blind spot’ around internallydisplaced populations (IDPs), collaborativeefforts on infant and young child feeding inemergencies, the need for operationally feasi-ble methodologies for determining programmeimpact and lack of accountability of donors. Itmay well be that the authors of paper 5 had toprioritise topics given an extremely limitedword count with which to work – if so, it wouldhave been useful to alert the reader to thisconstraint. Notwithstanding this piecemealapproach to topic coverage, the content of thesections dealing with emergencies also lackshistorical perspective, coherence and mostimportantly, creative thinking around solutions.In effect, the series tends to regurgitate what isalready out there in the published literaturerather than substantially move the subjectforward.

In a comprehensive nutrition series such asthis, some form of historical analysis is critical.This is partly because institutional memory isvery weak in our sector (reflecting factors likethe lack of career development structures andpoor coordination between a multiplicity ofstakeholders) and also because the past hasthe ability to bring into sharp relief what isabsolutely critical to progress in our sector. Itcan show us what has worked and changed forthe better and why, as well as where we havebecome stuck or even gone backwards and theexplanation for this.

Perhaps the most unforgivable weakness inthe series is the failure to advance thinking.Paper 5 (to its credit) discusses the lack ofevidence base and information on costs foremergency interventions. As emergency nutri-tion interventions are very much the ‘bread andbutter’ of this publication, we welcome thisfocus. Indeed this is something that the ENNhave been arguing for years and have repeat-edly (perhaps too often) flagged in editorials.The Lancet series has also highlighted theabsence of an agency or body with responsibil-ity for taking an overview of the effectivenessand cost-effectiveness of different types ofintervention as a reason why the status quoprevails. Again, the ENN has argued this verysame point over a number of years.Unfortunately, these ideas are not developed orelaborated in the series. How useful it wouldhave been if the authors had discussed poten-tial institutional locations and processes forsuch a body. The case for this could have been

From the EditorContents

Any contributions, ideas or topics for futureissues of Field Exchange? Contact the editorial team on email: office@ennonline.net

strengthened by showing historically whatmight have been achieved had such an institu-tional arrangement existed previously. Giventhe profile and considerable public relations onthis series, we feel that an important advocacyopportunity for change has been missed.

The same criticisms cannot, however, belevelled at the interventions and pilot studiesdescribed in field articles in this issue of FieldExchange. All, without exception, entail new ormodified approaches to programming introducedto address specific problems.

An article by Hanna Mattinen working forAction Contre la Faim describes how blanketdistribution of BP5 biscuits to children underfive succeeded in bringing down global acutemalnutrition rates dramatically in internallydisplaced persons (IDP) camps in Darfur,following reduction of amounts of Corn SoyaBlend in the general ration. Although costly, theapproach was seen as a temporary measure toaddress a short-term problem. Tom Oguta,Grainne Moloney and Louise Masese from FAOsFood Security Analysis Unit write about a pilotstudy comparing Lot Quality AssuranceSampling (LQAS) with 30 by 30 cluster nutritionsurveys in IDP camps in Hargeisa, Somaliland.Although similar results were found for preva-lence of malnutrition, morbidity and healthprogramme coverage, there were significantdiscrepancies for household level data such asdietary diversity, access to water, etc.Furthermore, confidence intervals with LQASwere far wider. The authors conclude thatalthough LQAS is up to 60% cheaper and takesmany less person days than a traditional nutri-tional survey, its most appropriate role is prob-ably where monitoring of hot spots is needed orwhen there is limited access to an area.

An article by Elena Rivero, Núria Salse andEric Zapatero of Action Against Hungerdescribes how an integrated food and nutritionsystem has been established in Malawi andused effectively for decision making. Theauthors highlight the advantages of a longitudi-nal surveillance system over a system of regu-lar cross-sectional surveys and also demon-strate how nutrition and food security informa-tion, which is specific to the Malawi context, canbe integrated into one information system.Finally, an article by a regular contributor toField Exchange, Mark Myatt, describes thedevelopment and application (in Ethiopia) of anew method to assess nutrition programmecoverage. The method, which has the easilyremembered acronym SQUEAC (semi-quantita-tive evaluation of access and coverage), hasbeen developed by Valid International in collab-oration with Concern Worldwide, World Vision,and UNICEF. It is less resource intensive thanCSAS (see Field Exchange issue 271) and makesuse of routine data collection as part ofprogramming and small scale surveys.

All these field articles demonstrate a degreeof adaptive programming and pragmatism onthe part of implementing agencies. Theapproaches used are not ‘one-size fits all’approaches but more ‘why not try this in suchand such a context’. It is always encouragingand inspiring to see such innovation on the‘front line’.

We hope you enjoy this issue of Field Exchangeand find something in here that helps you inyour work.

Jeremy ShohamEditor

Field Articles2 SQUEAC: Low resource method to evaluate access and coverage of programmes14 Methodology for a Nutritional Survey among the nomadic population of northern Mali21 Integrated Nutrition and Food Security Surveillance in Malawi 26 Piloting LQAS in Somaliland

30 Blanket BP5 distribution to under fives in North Darfur34 Animal husbandry and agriculture efforts toward programme sustainability

Research6 Somali KAP Study on Infant and Young Child Feeding and Health Seeking Practices7 Fortified maize meal improves vitamin A and iron status in refugees8 Mortality in the DRC8 Moderating the impact of climate change

9 Links between needs assessment and decisions in food crisis responses10 Review of CMAM in Ethiopia, Malawi and Niger11 Summary of Lancet Series on Maternal and Child Undernutrition

News15 Integrated Food Security Phase Classification (IPC) Website Launched16 Regional IFE Workshop held in Indonesia16 Trigger Indicators and Early Warning and Response Systems 16 FANTA Review of Essential Nutrition Actions in Ethiopia16 New MSF website on field research16 Evaluation of TALC’s CD-ROM ‘Community Nutrition’17 New publication on Household Economy Approach17 Regional Consultation on Nutrition and HIV in South East Asia17 Webpage from the World Bank on the Food Price Crisis18 US and Canadian responses to global food crisis18 CE-DAT website launched18 Breastfeeding: Practice and Policy Certificate Course18 Updated Module on IFE available in English and French19 Nutrition in Emergencies short course19 Unified United Nations response to the global food price challenge19 Update for ‘Caring for Severely Malnourished Children’ book19 Public Health in Complex Emergencies training20 COMPAS course in project management/evaluation

Views20 The Effects of Global Warming on Food Security

Letters24 SAM inadequately addressed in the Lancet Undernutrition Series 25 Challenging conclusions and analysis of the Lancet Undernutrition Series

Evaluation28 SC UK’s experiences with Central Emergency Response Funds (CERF)29 Peer review of WFP evaluation process

Agency Profile33 Samaritan’s Purse

37 People in Aid

1 Field Exchange 27 (March 2006). The challenge ofapplying CSAS in DRC. Jason Stobbs, AAH- USA. P28-30. Postscript by Mark Myatt.

2

Field Article

SQUEAC: Low resource methodto evaluate access and coverageof programmes

By Mark Myatt

Mark Myatt is a consultant epidemiologist and senior researchfellow at the Division of Ophthalmology, Institute ofOphthalmology, University College London. His areas of expertiseinclude infectious disease, nutrition, and survey design.

The author would like to acknowledge the contributions of VALID InternationalLtd, Concern Worldwide, World Vision, and UNICEF to this work.

Centric Systematic Area Sampling(CSAS) was developed to estimatecoverage of selective feedingprogrammes. It provides an overall

estimate and a spatial distribution map ofprogramme coverage, and a ranked list ofprogramme-specific barriers to service accessand uptake. As CSAS is resource intensive, ittends to be used in programme evaluationrather than in planning. This means CSASfindings arrive too late in the programmecycle to institute effective remedial action. Inaddition, the community-managed model ofservice delivery is now being adopted indevelopmental and post-emergency settings,which tend to suffer from considerableresource-scarcity, compared to emergencyresponse programmes implemented by non-governmental organisations (NGOs). Thereexists, therefore, a need for a low-resourcemethod capable of evaluating programmecoverage and identifying barriers to serviceaccess and uptake. This article describes someaspects of such a method1 currently beingdeveloped by VALID International, in collabo-ration with Concern Worldwide, WorldVision, and UNICEF.

Outline of the proposed method forevaluating access and coverageThe method is called SQUEAC (Semi-Quantitative Evaluation of Access andCoverage) and uses a two-stage screening testmodel:

STAGE 1: Identify areas of probable low andhigh coverage, and reasons for coverage fail-ure using routine programme data, alreadyavailable data, quantitative data that may becollected with little additional work, and anec-dotal data.

STAGE 2: Confirm the location of areas ofhigh and low coverage and the reasons forcoverage failure identified in Stage 1 usingsmall-area surveys.

Data sources and their methods ofanalysis (STAGE 1)Routine programme dataExperiences with Community TherapeuticCare (CTC) programmes in a variety of emer-gency settings show that programmes withreasonable coverage show a distinctive patternin the plot of admissions over time. As can beseen in Figure 1, the number of admissionsincreases rapidly. These may then fall awayslightly before stabilising and finally droppingaway, as the emergency abates and theprogramme is scaled down and approachesclosure. Major deviations from this pattern inthe absence of evidence of (e .g.) mass migra-tion or significant improvements in the health,nutrition, and food-security situation of theprogramme's target population indicates apotential problem with a programme's recruit-ment procedures. For example, Figure 2 shows

Figure 1: Pattern of admissions over timeover an entire programme cycle for anemergency-response CTC programme

Figure 2: Admissions over time in a CTCprogramme with initially poor communitymobilisation

Figure 3: Pattern of CTC admissions overtime with seasonal calendars of humandiseases associated with acute undernutritionin children and household food availability

1 A more detailed report on the SQUEAC method can befound at http://www.brixtonhealth.com/squeaclq.html

M Myatt, Ethiopia, 2007

Group discussion led by local staff

Num

ber

of

new

adm

issi

ons

Num

ber

of

new

adm

issi

ons

Num

ber

of

new

adm

issi

ons

Hum

andis

ease

sFo

od

eate

n

Time

Time

Time

Fever

ARI

Diarrhoea

FeverFever

ARI

Diarrhoea

Community mobilisationactivities started

3

DNA referrals are more likely than defaultersto be current cases. This means that high DNArates are associated with low programme cover-age. DNA rates can be calculated by monitoringreferrals. Mapping of DNA cases can provideinformation about problems of proximity toservices and other barriers to service access anduptake that may also be spatially distributed(e.g. ethnic or religious groups). Defaulting andDNA rates may also be analysed (classified)using the Lot Quality Assurance Sampling(LQAS) technique presented later in this article.The SPHERE minimum standard for defaultingrates is 15% (maximum). This standard may alsobe used for DNA rates.

Anecdotal dataThree methods of collecting anecdotal data froma variety of sources are used in SQUEAC assess-ments. These are:

1. Informal group discussions with:• Carers of children attending OTP sites.• Relatively homogenous groups of key- informants (e.g. community leaders and religious leaders) and lay-informants (e.g. mothers and fathers).• Programme staff.

2. Semi-structured interviews with key- informants such as:• Programme staff.• Clinic staff.• Community-based informants such as schoolteachers, traditional healers, and traditional birth attendants.• Carers of defaulting and DNA cases.

3. Simple structured interviews, undertaken as part of routine programme monitoring and during small-area surveys, with:• Carers of defaulting and DNA cases.• Carers of non-covered cases found by small-area surveys.

Other methods of collecting anecdotal data (e.g.formal focus groups and more structured and in-depth interviews) may also prove useful in somecontexts. The collection of anecdotal data shouldconcentrate on discovering reasons for both non-attendance and defaulting.

Validating and analysing anecdotal dataIt is important that the collected anecdotal datais validated. In practice, this means that data iscollected from as many different sources aspossible. Data sources are then cross-checkedagainst each other. This process is known astriangulation. The data collected from routineprogramme data and anecdotal data, whencombined, provide information that can beconsidered as a set of hypotheses that can betested. The SQUEAC method uses small-areasurveys to confirm or deny these hypotheses.Hypotheses about coverage should be statedbefore undertaking small-area surveys.Hypotheses about coverage will usually take theform of identifying areas where the combineddata suggest that coverage is likely to be satisfac-tory and areas where the combined data suggestthat coverage is likely to be unsatisfactory.

Figure 8, for example, shows an area of prob-able low coverage identified by mapping homelocations, analysis of outreach activities,defaulter follow-up, and anecdotal data. Thehypothesis about coverage in this area was thatcoverage was below the SPHERE minimumstandard for coverage of therapeutic feedingprogrammes in rural settings of 50% due to:

a plot of admissions over time in an emergency-response CTC programme that had neglected toundertake effective community mobilisationand outreach activities. Admissions initiallyincreased rapidly and then fell away rapidly.Such a pattern is indicative of a programme withlimited spatial coverage, relying on self-referrals.

The pattern of admissions in a developmen-tal setting is likely to be more complicated and,once the programme has been established,should vary with the prevalence of acute under-nutrition. Making sense of the plot of admis-sions over time in such settings requires infor-mation about the probable prevalence of acuteundernutrition. This can be determined usingseasonal calendars of human diseases associatedwith acute undernutrition in children (e.g. diar-rhoea, fever, and acute respiratory infection(ARI)) and food availability. Figure 3 shows aplot of admissions over time with seasonalcalendars of human diseases and food availabil-ity. Deviation from the expected pattern indi-cates a potential problem with a programme’srecruitment procedures.

Using admissions over time ignores the prob-lem of defaulters. Defaulters should be in aprogramme and are, by definition, coverage fail-ures. Figure 4 shows a standard programmeindicator graph from a CTC programme. Thisgraph shows an increasing defaulting rate. Thiswas due to the programme having too fewOutpatient Therapeutic (OTP) sites. More caseswere found and admitted as the programme'soutreach activities were expanded. However,more of these cases defaulted after the initialvisit because beneficiaries and carers had towalk too far to access services.

The home location of the beneficiary isusually recorded on the beneficiary record card.Mapping the home locations of beneficiariesattending each OTP site is a simple way of defin-ing the actual (rather than the intended) catch-ment area of each OTP site. Figure 5, for exam-ple, shows the home location of each beneficiary,attending an OTP site, that was admitted to theprogramme in the previous two months. Thisplot suggests that the programme has limitedspatial coverage with coverage restricted toareas close to OTP sites or along the major roadsleading to OTP sites.

Mapping is also a useful way of assessingoutreach activities. It is also useful to map thehome locations of defaulting cases. Figure 6, forexample, shows the home locations of benefici-aries who defaulted in the previous two months.Most defaulting cases come from villages farfrom the OTP site suggesting that lack of prox-imity to services (either to the OTP site or tooutreach and support services) is a leadingcause of defaulting. It may also be useful torecord and map DNA (did not attend) referrals(i.e. probable current cases that did not attend theprogramme despite having been referred to theprogramme) that you find by referral monitoring(see below). Following-up of defaulting andDNA cases (i.e. with home visits) should also beundertaken in order to identify reasons fordefaulting and non-attendance.

All of the mapping work outlined in this arti-cle can be performed with a paper map of usefulscale, clear acetate sheets, adhesive maskingtape, Post-it™ notes, and marker pens. Figure 7,for example, shows a coverage assessmentworker mapping the home locations of benefici-aries attending an OTP site.

Figure 4: Standard CTC/TFC programme indicator graph

Figure 5: Home locations of OTPbeneficiaries

Figure 6: Home locations of OTP benefici-aries who defaulted in the previous twomonths

Figure 7: A coverage assessment workermapping the home locations of OTPbeneficiaries

Figure 8: Area of probable low coverageidentified by mapping of home locations(shown), analysis of outreach activities,defaulter follow-up, and anecdotal data

Data courtesy of Concern Worldwide

Photograph courtesy of Concern Worldwide

Photograph courtesy of Concern Worldwide

Field ArticlePe

rcen

tage

of

exits

Months since start of programme

Intended catchment

Major road

Towns andvillages

OTP siteHomeLocations

Intended catchment

Major road

Towns andvillages

OTP siteDefaultingcases

The SPHERE minimum standard for coverage oftherapeutic feeding programmes in rural settings is50%. The following rule-of-thumb formula may beused to calculate a value of d appropriate for classi-fying coverage as being above or below a standardof 50% for any sample size (n):

The [ and ] symbols mean that you should rounddown the number between the [ and ] symbols tothe nearest whole number. For example:

With a sample size (n) of 11, for example, anappropriate value for d would be:

For standards other than 50%, the following rule-of-thumb formula may be used to calculate a suitablevalue for d for any coverage proportion (p) and anysample size (n):

For example, with a sample size (n) of 11 and acoverage proportion (p) of 70% (i.e. the SPHEREminimum standard for coverage of therapeutic feed-ing programmes in urban settings) an appropriatevalue for d would be:

An alternative to using the simple rule-of-thumbformulae presented here is to use LQAS samplingplan calculation software. Consideration of the clas-sification errors associated with candidate LQASsampling plans should be informed by the two-stagescreening test model used by SQUEAC2.

Figure 9 shows the data collected in the small-areasurvey of the area shown in Figure 8. The surveyfound 12 cases and 3 of these cases were in theprogramme. The appropriate value of d for a samplesize (n) of 12 and a coverage standard of 50% is:

Since 3 is not greater than 6, the coverage in thesurveyed area is classified as being below 50%.

The LQAS technique may also be used to classifydefaulting and DNA rates. For example, using thedata presented in Figure 10 and a standard for DNArates of 15% (maximum):

In this example there are 7 DNA cases from 15 referrals. Since 7 is greater than 2, the DNA rate forreferrals from this particular community basedvolunteer (CBV) is classified as being unsatisfactory(i.e. above 15%).

4

• A mismatch between the programme's defi- nition of malnutrition (i.e. anthropometric criteria and problems of food security) and the community's definition of malnutrition (i.e. as a consequence of illness, particularly diarrhoea with fever).• Patchy coverage of outreach services partic- ularly with regard to the ongoing follow-up of children with marginal anthropometric status.• Distance to OTP sites and other opportunity costs.

A small-area survey was undertaken in thisarea to confirm this hypothesis. The findings ofthe small-area survey confirmed, in generalterms, the hypothesis under test. It also identi-fied a problem with the application of case-defi-nitions leading to some cases being admitted tothe wrong programme.

Small-area surveys (STAGE 2)SQUEAC small-area surveys use the same in-community sampling and data-collection meth-ods as CSAS surveys. Cases are found using anactive and adaptive case-finding method (seeBox 1). When a case is found, the carer is askedwhether the child is already in the programme.A short questionnaire is administered if themalnourished child is not already in theprogramme.

Severe acute undernutrition is a relativelyrare phenomenon. This means that the samplesize (i.e. the number of cases found) in small-area surveys will usually be too small to esti-mate coverage with reasonable precision (i.e. asa percentage with a narrow 95% confidenceinterval). It is possible, however, to classifycoverage (i.e. as being above or below a thresh-old value) with small sample sizes using a tech-nique known as Lot Quality Assurance Sampling(LQAS).

Analysis of data using the LQAS techniqueinvolves examining the number of cases found(n) and the number of covered cases found. Ifthe number of covered cases found exceeds athreshold value (d) then coverage is classified asbeing satisfactory. If the number of coveredcases found does not exceed this thresholdvalue (d) then coverage is classified as beingunsatisfactory. The value of d depends on thenumber of cases found (n) and the standardagainst which coverage is being evaluated.

Box 1: Active and adaptive case-finding

The within-community case-finding method usedin both SQUEAC small-area surveys and CSASsurveys is active and adaptive:

ACTIVE: The method actively searches for casesrather than just expecting cases to be found in asample.

ADAPTIVE: The method uses information foundduring case-finding to inform and improve thesearch for cases.

Active and adaptive case-finding is sometimescalled snowball sampling.

Experience with CSAS surveys indicates that thefollowing method provides a useful starting point:Ask community health workers, traditional birthattendants, traditional healers or other keyinformants to take you to see “children who aresick, thin, or have swollen legs or feet” and thenask mothers and neighbours of confirmed casesto help you find more cases.

The basic case-finding question (i.e. “children whoare sick, thin, or have swollen legs or feet”)should be adapted to reflect community defini-tions of malnutrition.

It is important that the case-finding method thatyou use finds all, or nearly all, cases in thesampled communities.

Figure 9: Data from the small-areasurvey of the area shown in Figure 8

Figure 10: Example analysis of referrals from a Community BasedVolunteer (CBV)

Figure 11: Villages selected using thequadrat method

Figure 12: Villages selected usingstratified random sampling

Figure 13: An LQAS sampling plancalculator

Data courtesy of Concern Worldwide

AC2V YR4R

Referralnumber

Truecase

Date ofadmission

1 Yes 4/6/072 Yes 11/6/073456 Yes 23/7/07789 No 14/8/071011 Yes 17/8/0712 Yes 17/8/0713 Yes 17/8/071415 Yes 1/10/07

Red: DNA cases Orange: Inappropriate referral Green: Attending cases

Field Article

2 For more information on this, visit:http://www.brixtonhealth.com/squeaclq.html

Intended catchment

Major road

Towns andvillages

Defaultingcases

Intended catchment

Cliniccatchment

Major road

Towns andvillages

Samplinglocations

5

SQUEAC use-studiesThe first SQUEAC use-study examined the abil-ity of the LQAS method to classify coveragecorrectly and was undertaken by VALIDInternational and World Vision in Ethiopia inMarch 2007.

Six small-area surveys were undertaken:• Six OTP sites were selected at random.• Three villages (kebeles) were selected at random from each OTP site’s catchment area.• Five localities (gotts) where then selected at random from each of the selected villages.• Each locality was sampled using active and adaptive case-finding.

The true coverage for each OTP site was esti-mated using data from a CSAS survey under-taken at the same time as the small-areasurveys. The results from these small-areasurveys are shown in Table 1. The LQASmethod correctly classified coverage in each ofthe six OTP catchment areas.

The second SQUEAC use-study was under-taken in a Concern Worldwide CTCprogramme in the Democratic Republic ofCongo in November 2007. This concentrated onthe use and availability of routine programmedata and anecdotal data to identify areas witheither low attendance rates or high defaultingand DNA rates. It also looked at the ability oflocal programme staff to collect and analysedata from these sources and to plan, undertake,and analyse data from small-area surveys.Local staff proved capable of using routineprogramme data to identify probable areas oflow coverage and reasons for non-attendanceand defaulting when these data were availableand presented using simple graphs, tables, andmaps. They had no difficulties mapping routineprogramme data. Local staff also had no diffi-culty undertaking small-area surveys andanalysing survey data using the LQAS tech-nique. However, the collection and analysis of

routine programme data needed to beimproved and standardised. Furthermore,collection and analysis of anecdotal data bylocal programme staff proved problematic,particularly with informal group discussions.This situation may improve with careful selec-tion and training of local staff. Further use-studies will concentrate on resolving this issue.

The SQUEAC method is suited to beingused within a clinical audit framework3. Formore information, visit:http://www.brixtonhealth.com/squeaclq.html

Estimating and classifying ‘headline’ oroverall programme coverageUsing LQAS techniques to derive a ‘headline’coverage classification requires:

A first stage sampling method: This is thesampling method that is used to select thevillages to be sampled. CSAS assessments usethe centric systematic area sampling or quadratmethod to select villages to be sampled. If onlya ‘headline’ coverage classification is required,then a similar method could be used to selectvillages to be sampled. The number of quadratsdrawn on the map can be smaller than wouldbe used for a CSAS assessment (this is the sameas using larger quadrats). The villages to besampled would then be selected by their prox-imity to the centre of each quadrat as is done ina standard CSAS survey (Figure 11). Such anapproach would be appropriate for classifyingcoverage over a wide area, such as a healthdistrict. In developmental settings it may bedesirable to classify coverage over a wide areaand also for individual clinic catchment areas. Insuch situations, a stratified (i.e. by clinic catch-ment area) sample could be taken with villagesselected at random from a complete list ofvillages within each catchment area (Figure 12).

A within-community sampling method: Thiswill usually be an active and adaptive case-finding method (see Box 1) or a house-to-housecensus sampling method.

A LQAS sampling plan: This provides a targetsample size (n) which, together with prevalenceand population estimates, is used to decide thenumber of villages to sample (see below). Thetarget sample size (n) will usually be larger thanis required for the small-area surveys used inthe SQUEAC method. Suitable sampling planscan be selected using an LQAS sampling plancalculator. Figure 13 shows an LQAS samplingplan calculator being used to select a samplingplan. This LQAS sampling plan calculator isavailable, free of charge, from: http://www.validinternational.org/pages/sub.cfm?id=1780

A sample size calculation: This is the targetsample size (n) from the LQAS sampling planwhich, together with estimates of the preva-lence of severe acute undernutrition in thesurvey area and population data, is used to

calculate the number of villages that will needto be sampled:

For example:

Once these decisions and calculations havebeen made, sampling locations can be identifiedand the survey undertaken. A standard ques-tionnaire should be applied to carers of non-covered cases found by the survey.

It is unlikely that a survey will return thetarget sample size (n) exactly. So, a value for theLQAS classification threshold value (d) for theachieved sample size should be calculatedusing the rule-of-thumb formulae presentedearlier in this article. For example:

LQAS sampling plan calculation softwarecould also be used (Figure 13).

Coverage is classified using the same tech-nique as is used for SQUEAC small-areasurveys. For example:

Analysis of data collected in individual cliniccatchment areas is analysed in the same wayalthough errors may be considerable if samplesizes are very small.

ConclusionsThe early work presented in this articlesuggests that the SQUEAC approach is likely toprove a suitable method for frequent and ongo-ing evaluation of programme coverage andidentification of barriers to service access anduptake. The SQUEAC approach is also capableof providing a similar richness of informationas is provided by the CSAS method.

The LQAS method may be used to provideclassifications of ‘headline’ or overall coverageover wide areas. The low sample size require-ment means that surveys using the LQASmethod are likely to be less resource intensivethan standard CSAS surveys. Such surveyswould, however, be limited in their ability toprovide a detailed map of the spatial distribu-tion of programme coverage.

For further information, contact: Mark Myatt, athttp://www.brixtonhealth.com/squeaclq.html

Table 1: Results from six small-area surveys from the first SQUEAC use-study

OTP Site True Coverage

CasesFound (n)

Covered Cases (c)

Is c > d? ClassifiedCoverage

Bezena Benara > 50% 15 9 7 Yes > 50%

Jore > 50% 5 3 2 Yes > 50%

Yebu < 50% 16 6 8 No < 50%

Funto < 50% 21 7 10 No < 50%

Demboya > 50% 5 3 2 Yes > 50%

Adilo < 50% 9 2 4 No < 50%

Target sample size 39

Median village population(all ages) 600

Prevalence of severe acuteundernutrition 1%

Number of villages tosample

Target sample size 39

Achieved sample size 43

Standard 50%

d

n 43

d 21

Covered cases found 29

Coverage classification Satisfactory (since 29 > 21)

Field Article

3 Clinical audit is a quality improvement and monitoringmethod that seeks to improve service delivery throughsystematic review against specific criteria and standards andthe implementation of change.

Data courtesy of World Vision

Measuring MUAC in asmall area survey

M Myatt, Ethiopia, 2007

6

Research

Since the collapse of the Somali CentralGovernment in 1991, the country hasfaced a series of disasters, both natural(floods and droughts) and man-made

(poor governance, armed conflict, and collapse ofinstitution/infrastructure). These have led todisruption of livelihood systems, poor provisionof basic services and erosion/loss of social carenetwork.

The Food and AgricultureOrganisation’s (FAO) Food SecurityAnalysis Unit (FSAU) has been conduct-ing nutrition assessments in Somalia forthe last eight years. Together with pastsurveys conducted across South andCentral Somalia since the1980s, thesehave consistently recorded high levels ofacute malnutrition and under five mortal-ity rates, even in years with improved foodproduction. Median rates of global acutemalnutrition in 2007, based on 36 surveys,were 15.2% global acute malnutrition(GAM) and 2.7% severe acute malnutrition(SAM), illustrating the severity of the crisis.Potential contributing factors1 include lack of safewater (only 15% of the population have access),lack of sanitation facilities (25% of the populationhave access), lack of a functional health systemfor basic services (measles coverage estimated at35%) and chronic food insecurity.

Humanitarian agencies’ efforts to addressthese services deficiencies are hampered byrecurrent insecurity. Available data on care prac-tices of young children indicate an alarming situ-ation with regard to breastfeeding, complemen-tary feeding and treatment of childhood illness.Based on the hypothesis that the role of childcare, feeding and health practices are a signifi-cant contributing factor to chronically high ratesof acute malnutrition, the FSAU commissioned aKnowledge Attitudes and Practices (KAP) studyin September 20072 to learn more and come upwith recommendations to address identifiedproblems. The study assessed the key livelihoodsystems3 of Somalia in seven major regions. Datacollection methods were both quantitative andqualitative including literature review, focusgroup discussions, rapid profiles, key informantinterviews and case studies. Respondents

included elderly, pregnant women and womenof child bearing age, men, traditional birth atten-dants, traditional healers, community healthworkers, health professionals, commu-nity/religious leaders, women groups and inter-national/local NGOs.

Not surprisingly, the study found that preg-nant/lactating women do not have access to

formal nutrition education andmainly rely on social networksincluding grandmothers, tradi-tional healers and sheikhs foradvice on child feeding andhealth seeking behaviours.There is no special diet forpregnant women. However,food intake is reduced in the3rd trimester to control thesize of the baby and therebyease its delivery. Some of thefoods prohibited (though in

negligible levels) include honey, lamb meat andghee. The traditional post-natal care for 40-daysafter delivery is routinely implemented in all butthe pure pastoralists’ households. This encour-ages mothers to eat nourishing food and to focuson breastfeeding in the initial weeks after deliv-ery, however mother often cite lack of resourcesto purchase the necessary foods.

Breastfeeding is mainly influenced by socialenvironment, especially by maternal grandmoth-ers and other elderly women in the community.Although breastfeeding is acceptable to all,infants are first introduced to sugary water afterbirth around the 3rd or 4th day after delivery.This is to avoid the colostrum (first milk) that isgenerally discarded by the mother as consideredheavy, thick, coarse, dirty, toxic, and harmful tochildren’s health. Exclusive breastfeeding is notpracticed, as breastmilk alone is considered inad-equate and water essential to cool the baby andquench their thirst. On a positive note, the agreedand acceptable duration of breastfeeding is 24months, based on the Koran,Sural Al-Baqrah, Juz2:233 and is consistent with recommended stan-dards. However, this is rarely implemented fortwo main reasons – very close birth spacing, sinceonce a women discovers she is pregnant she stopsbreastfeeding, and an increasing trend of mothersworking outside the home, resulting in earlyseparation from the infant.

The study also found that there is a lack of orinappropriate knowledge of proper complemen-tary feeding practices across all livelihoodgroups. Early introduction of complementaryfoods (from birth to three months) is common,mainly as cow’s/goat’s milk (involving nospecial preparation or storage), tea and porridge.The poorer households report limited access tomilk, which they replace with tea/porridge after3 months. Lack of dietary variety is typicalamong the riverine and, to a lesser extent, theagropastoral community, where children aremainly fed a cereal-based diet. Tea is often givento children as a snack before meals, whichreduces stomach capacity and appetite and inter-feres with nutrient absorption. During illness,special diets for quick recovery include avoidinganimal protein-based foods, which are believedto aggravate illnesses. Knowledge on the use ofOral Rehydration Salts (ORS) for diarrhoea wasgood, with local adaptations using lemon orwater melon juice.

Access to safe water is of major concern, withmost households relying on unprotected sources.Water treatment at household level rarelyhappens. Unsafe water and childhood diarrhoeaare the strongest associated factor with child-hood malnutrition in Somalia. Sanitation/water-related diseases including diarrhoea, acute respi-ratory infection (ARI), intestinal parasitic infesta-tions and skin/eye infections are the mainreported illnesses. Access to formal health serv-ices is limited or unavailable, especially amongthe rural population. Responses to illnesses tendto follow a generalised pattern of: Prayer Traditional home health practiceTraditional healer Buy medicine GetSheikh to pray Health facility.

This clearly shows the lack of knowledge of,or confidence in, more modern health care prac-tices. Many of the traditional practices are harm-ful, such as burning of the feet for nutritionaloedema, and often result in the child becomingsicker before they are finally taken for conven-tional treatment.

In conclusion, the study revealed glaringevidences that child feeding and care practices arebelow the acceptable standards. Much of this canbe attributed to poor knowledge, with poor prac-tices exacerbated by the collapse of the basic infra-structure resulting from the war in the early 90’s.The study recommends participatory dissemina-tion of the study findings, training, advocacy andactive involvement of key community changeagents (religious/traditional and professionals)and caregivers in generating solutions for thechallenges highlighted. Parallel campaigns andinterventions are needed that focus on basichealth care, hygiene, sanitation, safe water andchild feeding practices. The importance of Islamin imparting positive messages relating to childcare and hygiene is very evident, highlighting aclear entry point for community sensitisation andfuture behaviour change.

For further information or a copy of the fullreport, please contact Grainne Moloney, P.O.Box 1230- 00621, Nairobi, Kenya.Tel: 254-20-3745734/1299; Fax: 254-20-3740598; email: grainne.moloney@fsau.or.ke orJoseph.waweru@fsau.or.ke

Somali KAP Study onInfant and Young

Child Feeding andHealth Seeking

PracticesBy Waweru Joseph Mwaura and Grainne Moloney

Joseph Waweru has worked as Nutrition Project Officer with the FAO Food SecurityAnalysis Unit (FSAU) Nutrition Surveillance Project in Somalia for the last twoyears. He has coordinated and supervised several nutrition assessments and train-ings and reviewed/piloted technical guidelines and methodologies. Previously heworked as a Research Assistant at Kenyatta University.

Grainne Moloney is the Nutrition Project Manager of the FAO/FSAU NutritionSurveillance Project, where she has worked for the last 18 months. Previously,Grainne worked with UNICEF in Darfur, also on nutritional surveillance, and withOxfam and Action Against Hunger UK on nutrition and food security in manydifferent contexts.

The authors would like to acknowledge the great effort of Professor Wambui Kogi-Makau and MsRose Opiyo for leading the study. The authors also express sincere thanks to the FSAU nutritionteam. The contribution of UNICEF, World Vision and Gedo Health Consortium in the study isacknowledged. The study was financed by the Swedish International Development Agency.

1 Source: FSAU nutrition surveys database (n=105) from2000 to 20072 Somali Knowledge Attitudes and Practices (KAP) study.Infant and Young Child Feeding and Health Seeking Practices.FAO/FSAU. December 2007. Available at:http://www.fsausomali.org/rss_article_id_245.php3 Agro-pastorals, Pastoral, Riverine, Urban and CoastalCommunities.

A breastfeedingmother fromBerbera

FSAU

, Som

alia, 2007

7

Research

7

Fortified maize meal improves vitamin A andiron status in refugees Summary of published field trial1

Nangweshi refugee camp wasopened in 2000 in response to theinflux of refugees fleeing theAngolan civil war. It is located in

Western Province of Zambia, about 180 kmfrom the border with Angola. The populationfigure for June 2003 was recorded by the UnitedNations High Commissioner for Refugees(UNHCR) as 26,061, with 8,404 households. Theplanned ration supplied by the World FoodProgramme (WFP) comprised maize meal,pulses, vegetable oil and salt. An initial analysisof the food aid basket was performed to assessthe likely prevalence of micronutrient deficien-cies. This showed that the ration was deficientin a number of micronutrients including vita-min A, iron (Fe), riboflavin and vitamin C.High-dose vitamin A capsules were distributedperiodically to children in the camp andFe/folate tablets were available for pregnantwomen via the antenatal care programme.

Using custom designed mobile milling andfortification equipment, maize meal fortifica-tion was field tested in the camp at the begin-ning of 2003. The equipment was shipped to thecamp where it was installed and operated at acentral location. The production staff weremembers of the refugee population, whileprogramme management was undertaken byCare International. A key aspect of this inter-vention was that maize meal fortified in a coun-try of origin may only have a limited shelf-lifeof a few weeks, due to the rapid onset of fat

rancidity in tropical temperatures and humid-ity. This recently published paper describes theoutcome of the field trial.

The objective of the trial was to assesschanges in Fe and vitamin A status of the popu-lation of Nangweshi refugee camp, associatedwith the introduction of maize meal fortifica-tion. Maize grain was milled and fortified incustom-designed mills installed at a centrallocation in the camp. A daily ration of 400g perperson was distributed twice monthly tohouseholds as part of the routine food aidration. The trial was a pre- and post-interven-tion study using a longitudinal cohort. It wasnot possible to include a control group owing tothe geographical layout of the camp and logis-tics, as well as operational and ethical consider-ations. Two hundred and twelve adolescents(10-19 years), 157 children (6-59 months) and118 women (20-49 years) were selected atrandom by household survey in July 2003 andfollowed up after 12 months.

The study found that during the interven-tion period, mean haemoglobin (Hb) increasedin children (0.87 g/dl; P <0.001) and adoles-cents (0.24 g/dl; P= 0.043) but did not increasein women. Anaemia decreased in children by23.4% (P<0.001) but there was no significantchange in adolescents or women. In adoles-cents, serum retinol increased by 0.16 µmol/l(P<0.001) and vitamin A deficiency decreasedby 26.1% (P<0.001).

1 Seal. A et al (2007). Maize meal fortification is associatedwith improved vitamin A and iron status in adolescents andreduced childhood anaemia in a food aid-dependent refugeepopulation. Public Health and Nutrition, pp 1-9

The authors concluded that there was a clearassociation between the introduction of fortifiedmaize meal and an improvement in the Fe andvitamin A status of a food aid-dependentrefugee population. Improvements in the statusof adolescents and a reduction in anaemia inchildren are important public health gains.Furthermore, although no information wascollected on other micronutrient deficiencies, itis likely that the study population was deficientin a number of other micronutrients. In particu-lar, vitamin B deficiency is frequently found inmaize consuming populations and health staffin the camp reported isolated clinical cases ofpellagra during the intervention period. Niacinand other B vitamins were included in themaize meal fortificant and it is reasonable toassume that this will have led to improvementsin the micronutrient status of the population.

The authors also suggest that this type offood aid fortification is an effective means ofreducing childhood anaemia and improvingadolescent Fe and vitamin A status in highlyvulnerable, aid-dependent populations,although its impact on anaemia in womenremains unclear. The data suggest that locallevel, centralised milling and fortification ofstaple foods should be promoted and theirnutritional impact and cost-effectiveness inves-tigated in other operational contexts.

Premixing at therefugee camp site

Blood samplingduring the house-hold survey (left).Weighing out thepremix.

Milling on site atthe refugee camp

A S

eal, Z

ambia

, 2004

A S

eal, Z

ambia

, 2004

A S

eal, Z

ambia

, 2004

A S

eal, Z

ambia

, 2004

Research

8

Moderating the impactof climate change

Adaptations to guard against the impact on food security ofclimate change are reviewed in a recently published article. Theauthors start with the assertion that relatively inexpensivechanges, such as shifting planting dates or switching to anotherexisting crop variety, may moderate negative impacts. However,the biggest benefits will probably result from more costly meas-ures, including the development of new crop varieties and expan-sion of irrigation. Such measures will require substantial invest-ments by farmers, governments, scientists and developmentorganisations. Prioritisation of investment needs, such as throughthe identification of ‘climate risk hot spots’, is therefore a criticalissue.

The authors consider three components to be essential to anyprioritisation approach: selection of a time scale over whichimpacts are most relevant to investment decisions, a clear defini-tion of criteria used for prioritisation, and an ability to evaluatethese criteria across a suite of crops and regions. The authors focuson food security impacts by 2030, a time period most relevant tolarge agricultural investments, which typically take 15-30 years torealise full returns.

Twelve major food insecure regions are identified. Each of thesecomprises groups of countries with broadly similar diets and agri-cultural systems and that contain a notable share of the world’smalnourished individuals, as estimated by the Food andAgricultural Organisation (FAO). Statistical models for evaluatingclimate change impacts across a suite of crops and regions werethen developed using data on historical crop harvests, monthlytemperatures and precipitation and maps of crop locations.

Based on the various projections, the authors identified a smallsubset of crops that met different prioritisation criteria. Becauseattitude to risk differs and the impact projections for some cropsare more uncertain than for others, various institutions mightderive different priorities. For example, one set of institutionsmight wish to focus on cases where negative impacts are mostlikely to occur, to maximise investment benefits. By this criterion,South Asia wheat, Southeast Asia rice, and Southern Africa maizeappear as the most important crops in need of adaptation invest-ments. Other institutions may want investments to target cropsand regions for which some models predict very negativeoutcomes. This would mean a different subset of crops withseveral South Asian crops, Sahel sorghum and South Africanmaize (again) appearing as the most in need of attention.

Either of these risk attitudes can be applied with an explicitregional focus. For a sub-saharan African institution interested ininvesting where negative impacts are most likely to occur (wheremedian impact projections are substantially negative, or wheremost climate models agree that negative impacts are likely tooccur), priority investments would include Southern Africa maize,wheat and sugarcane, Western Africa yams and ground nut, andSahel wheat.

Despite the many assumptions and uncertainties associatedwith the crop and climate models used, the analysis points tomany cases where food security is clearly threatened by climatechange in the relatively near-term. The importance of adaptationin South Asia and Southern Africa appears particularly robust.The results also highlight several regions (e.g. Central Africa)where climate-yield relationships are poorly captured by currentdata sets, and therefore future work is needed to inform adapta-tion efforts.

The authors conclude that impacts will probably vary substan-tially within individual regions according to differences inbiophysical resources, management and other factors. Therefore,the broad-scale analysis that they present only identifies majorareas of concern and that further studies at finer spatial scales areneeded to resolve local hot spots within regions.

Summary of published research1

1 Lobell. D (2008). Prioritising climate change adaptation needs for food securityin 2030. SCIENCE, vol 319, 1st February 2008, pp 607-610 Available athttp://www.sciencemag.org

Summary of published research1

Since 2000, the International Rescue Committee (IRC) has documented thehumanitarian impact of war and conflict in the Democratic Republic of theCongo (DRC) through a series of five mortality surveys. The first four stud-ies, conducted 2000 - 2004, estimated that 3.9 million people had died since

1998, arguably making DRC the world’s deadliest crisis since World War II. Lessthan 10% of all deaths were due to violence, with most attributed to easilypreventable and treatable conditions such as malaria, diarrhoea, pneumonia andmalnutrition. The fifth and latest survey covered the period from January 2006-April 2007. It used a three-stage cluster sampling technique to survey 14,000households in 35 health zones across all 11 provinces, resulting in widergeographic coverage than any of the previous IRC surveys. The key findings andconclusions were as follows:

More than four years after the signing of a formal peace agreement, the DRC’snational crude mortality rate (CMR) of 2.2 deaths per 1,000 per month is 57%higher than the average rate for sub-Saharan Africa and is unchanged since theprevious IRC survey in 2004. As with previous surveys, mortality rates are signif-icantly higher in the volatile eastern provinces than in the west of the country. Inaddition, mortality rates have risen significantly in the centre of the DRC. Basedon the results of the five IRC studies, an estimated 5.4 million excess deaths haveoccurred between August 1998 and April 2007. An estimated 2.1 million of thosedeaths have occurred since the formal end of war in 2002.

For the period covered by the survey, the only region to record a significantreduction in mortality since 2004 was that encompassing the five easternprovinces. This improvement coincided with a reduction in the risk of violentdeath, as well as a more robust United Nations (UN) peacekeeping effort byMONUC2, the international force in the DRC. Nevertheless, the CMR in thisregion is still 2.6 deaths per 1,000 per month, a rate that is 85% higher than thesub-Saharan average.

Only 0.4% of all deaths across the DRC were attributed directly to violence. Aswith previous IRC studies in the Congo, the majority of deaths were due to infec-tious diseases, malnutrition and neonatal- and pregnancy-related conditions.Increased rates of disease are probably related to the social and economic distur-bances caused by conflict, including disruption of health services, poor food secu-rity, deterioration of infrastructure, and population displacement. Children, whoare particularly susceptible to these easily preventable and treatable conditions,accounted for 47% of deaths, even though they constituted only 19% of the totalpopulation.

The persistent elevation of mortality more than four years after the official endof the 1998-2002 war provides further evidence that recovery from conflict cantake many years, especially when superimposed on decades of political andsocioeconomic decline. These data are consistent with those from other conflict-impacted states. Sustained and measurable improvements in key indicators, suchas mortality, will require committed national and international engagement formany years.

Mortalityin the DRC

1 Dr Coghlan. B et al (2007). Mortality in the DRC: an ongoing crisis. International RescueCommittee and Burnet Institute. Available from http://www.theirc.org/resources/2007/2006-7_congo-mortalitysurvey.pdf2 Mission des Nations Unies en République Démocratique du Congo.

IRC mortalitysurvey team

in transit

IRC surveyteam

travelling by boat

Surveyor questionsa woman during

the DRC mortalitysurvey

IRC,

DRC

IRC, DRC IRC, DRC

9

Rese

arch Links between needs

assessment and decisionsin food crisis responses

Summary of published review1

1 Darcy. J, Anderson. S and Majid. N (2007). A review of thelinks between needs assessment and decision-making inresponse to food crises. Study undertaken for the World FoodProgramme SENAC project. Humanitarian Policy Group, May2007. Available athttp://documents.wfp.org/stellent/groups/public/documents/ena/wfp128556.pdf

Arecently completed study underthe WFP StrengtheningEmergency Needs AssessmentCapacity (SENAC) project

explores the links and disconnects betweenneeds assessment and decision-making, inWFP and its partners, in response to foodcrises. The study is based on four in-depthcase studies conducted in the latter half of2006 (Sudan – principally the Darfur crisis,Pakistan – the 2005 Kashmir earthquake,Somalia and southern Africa – principallyMalawi). Each study involved travel to theregions in question and interviews with keyactors in WFP and other bodies. In addition,a number of other ‘reference’ cases werereviewed through documentation and inter-views. The report suggests that the functionof needs assessment in relation to decision-making is three-fold: to inform internal deci-sions about response throughout the life of aprogramme, to influence others’ responsedecisions, and to justify response decisionsand appeals for funds.

Informing internal decisionsThe study found that in most of the casesreviewed, WFPs own initial decisions aboutresponse were under-pinned by adequateinformation and analysis from assessments,whether conducted by WFP itself or througha collaborative process. Although WFPassessment practice has, in some respects,embraced a wider food security perspective,it is often still geared around one set ofresponse questions, i.e. how much food aid isrequired and by whom? The rationale for theproposed food aid strategy is not alwaysclear from the analysis of context in assess-ment reports and is rarely articulated againsta wider range of potential response options.Furthermore, progress in informing initialprogramming decisions is not yet matchedby an ability to make informed decisionsthroughout the life of a programme. Thestudy recommends that WFP adopt an infor-mation strategy for all major responses andthat this be budgeted explicitly. Overall, thestudy team concluded that there is a relativeunder-investment in the information andevidence base to support response decisions,particularly in monitoring and re-assess-ment. This is particularly evident inprotracted crisis response through ProtractedRelief and Recovery Operations (PRROs).

Different information requirements wereidentified in relation to four types of crisis:rapid onset, slow onset, chronic insecu-rity/displacement and transition/recovery.What is good assessment practice dependson the context, nature of the crisis, and time-frame for decision-making. The rapid onsetcases considered showed the need to agreesimple methods for determining initialresource requirements, clearly articulatedworking assumptions, and the necessity ofre-checking those assumptions as situationsdevelop. The slow onset cases showed theimportance of agreed triggers for action,based on leading risk indicators or definedthresholds, for effective prevention. Theconflict and displacement cases had all theseplus other requirements, including ways of

assessing unmet need in currently inaccessibleareas, ways of understanding the links betweenfood insecurity and exposure to violence, andmore robust methods for calculating the needsof dispersed as well as camp populations. Thetransitional contexts showed the need to investmore in mechanisms (including surveillance) todetermine when a programme should changecourse or wind up.

According to the study, some of the data andanalysis currently produced are simply not rele-vant to the needs of decision-makers, or are notpresented in ways that show their relevance.Some important types of information are oftennot available – such as people’s relative depend-ence on food aid or other assistance, and howthis may change over time and space. On theother hand, a number of good new tools (includ-ing market analysis) were found to be in use,even though the results did not always appearto inform response decisions. The study found apreponderance of quantitative over qualitativemethods of analysis and suggests that a betterbalance needs to be found between them, partic-ularly in livelihood related assessment.

The study team felt that more use could bemade of external (local and international) as wellas in-house expertise in conducting situationalanalysis. This could include sociological andanthropological perspectives in addition to moretraditional food security approaches. Goodneeds assessment – particularly in conflict-related situations – is often dependent on thequality of political and social analysis, as muchas on anthropometric or economic analysis.

Another finding was that analysis from theexisting information and analysis mechanisms– early warning, Vulnerability Analysis andMapping (VAM), Emergency NeedsAssessment (ENA), food security monitoring,etc., is not well integrated. In particular, therelationship between VAM analysis and ENA ininforming crisis response decisions is oftenunclear and demands further attention.

Too little attention is given to feedback ofinformation from the operational level, and theneed to build in better ‘feedback loops’ is essen-tial to more responsive programming. At times,the pressure to implement an agreedprogramme according to plan – especiallywhere complex logistical processes have beenestablished – appears to militate against adap-tive programming as needs change or as analy-sis is refined in the light of local realities. Influencing external decisionsThe link with external decisions was relativelyweak. In some cases, decisions – particularlydonor funding decisions – clearly precede anyformal needs analysis. Many are based onprojections of future need, particularly in thecase of protracted crises, although the basis forthese projections is not always clear.

Donor representatives often claimed thatWFP does not help them prioritise betweencontexts, pointing to the need for a commonreference standard and more explicit WFPjudgements on relative priorities.

Recent efforts to strengthen needs assess-ment in WFP have had a significant effect in

building credibility. However, trust in WFPassessment reports is clearly still an issue.Donors expressed varying degrees of scepti-cism, and some felt that there was a tensionbetween the credibility of WFP’s assessmentsand the messages it put out through the media.Regarding the latter, there was a perceivedtendency to talk up the scale or severity of asituation and WFP’s own role, which was felt tobe at odds with objective needs analysis. Thesecredibility barriers appear to be overcome whena robust but constructive relationship existsbetween donor representatives and WFP coun-try office staff, such that donors can ‘interro-gate’ WFP’s findings locally or be directlyinvolved in the assessment process.

Justifying decisions Moves by WFP towards greater transparency inthe assessment process – notably in the practiceof publishing assessments reports on the WFPwebsite – have gone a considerable way toproviding stronger justification for responsedecisions, as well as enhancing the influence ofthe assessments themselves. From the assess-ment reports reviewed for the study, it is appar-ent that there is a need to distinguish situationalanalysis from response option analysis moreclearly – but also to make the links betweenthem more explicit. Assessments that are heav-ily geared towards a particular organisation’sresponse options have limited potential forinforming and influencing others’ responsedecisions, and provide a relatively weak plat-form for justifying the organisation’s ownresponse decisions. Demonstrating the linksbetween situational analysis and responseoptions is essential.

There appears to be little incentive (and somedisincentive) for WFP country programmes tore-assess situations or to monitor change andimpact, particularly if this is likely to indicate ascaled down programme. More generally, thereappears to be little demand for information andanalysis once an operation has commenced,except when a decision to continue or to exit hasto be justified.

The diversity of donor practice in decision-making was found to be one of the singlebiggest variables in the study. Greater harmon-isation of donor decision-making is a necessarycondition for more timely and appropriate allo-cation of funds. The tendency to allocate fundsat the time of greatest media coverage can leadto delayed response (in slow onset crises), front-loaded funding (in rapid-onset) and under-funding in protracted or low profile cases.

Overall, the study team concluded that WFPhas a significant opportunity to take a lead inestablishing good assessment practice acrossthe sector. This involves a combination ofrigour, adaptability to context, effective collabo-ration and good communication – providingtimely information to decision-makers (internaland external) in a form they can use.

Workers wade into theSea to collect WFP foodat the southern Somaliport of Merka

WFP

/Pet

er S

mer

don,

Som

alia

, 2007

10

Review of CMAMin Ethiopia,Malawi and Niger

Over the long term, commitment by donors to develop and maintain capacities is needed, along with planning for future emergencies and for transition of services post-emergency. ii) Access to CMAM services should be assured in priority districts following initial start-up in learning sites and gradual scale-up. Both inpatient and outpatient care needs to be made available by linking with a community-based network of formal and informal healthcare and community systems. iii) Access to CMAM supplies should be ensured during and after emergencies. While beyond the means of most develop- ing country budgets, it is critical that CMAM supplies of essential drugs and therapeutic foods be secured by MoHs. Long-term donor commitment to provide supplies is necessary.iv) Quality of CMAM services can be assured through adherence to national CMAM guidelines, support to and supervision of CMAM services, and harmonised monitor- ing and evaluation tools that are linked to the national health information system.v) CMAM competencies can be strengthened through integrating pre-and in-service training for CMAM into national curricula for all levels of health care providers (community health workers, nurses, and physicians). Training should be augmented through practical learning experiences at CMAM learning sites, post-training on site mentoring support and supervision, and regular experience sharing at meetings and other fora.

The importance of a health systems approach –during service introduction, expansion or tran-sition from emergency to development contexts– is especially important to ensure that CMAMis integrated into the national health system,while not supplanting other essential services.Accordingly, tools to improve assessment,design, monitoring and evaluation of the intro-duction and scale up of CMAM services areneeded.

In countries with a high burden of acutemalnutrition, national health policies will haveto carefully assess the need for and situateCMAM within other essential health care andnutrition services. National health policies willhave to identify context-specific strategies onhow best to address high levels of acute malnu-trition. CMAM services may be prioritised incertain highly vulnerable areas with a chronicburden of acute malnutrition, in emergency-prone areas, and/or as a nationwide service.

Global-level recommendations include:• Advocate and attract technical and financial

support from donors to integrate CMAM into national health policies and strategic plans.• Support country initiatives to invest in the key elements identified for successful inte- gration and sustainability of CMAM.• Invest in developing tools for improving assessment, design, monitoring and evalua- tion of CMAM.• Update donor project proposal guidelines for CMAM programming and include guid- ance based on the key elements for improved integration identified in the review.• Broaden the base of NGOs providing CMAM support and include grantees that are in best position to address key elements for integration and scaling up of CMAM through their expertise in strengthening health systems.

The review also identified the need to docu-ment in greater detail successful integrationexperiences, specifically examining the processand context of integrated service introductionand scale-up. In addition, evidence of inte-grated CMAM services and comparisons ofthese experiences to those of other integratedservices, such as Integrated Management ofChildhood Illness, is also important to docu-ment. Finally, the various approaches andstrategies employed by NGOs and MoHs glob-ally for CMAM service provision shouldcontinue to be documented and shared throughvarious channels, including workshops, data-bases and publications.

USAID requested FANTA to conduct areview of the integration of commu-nity-based management of acutemalnutrition (CMAM) into the

national health systems of Ethiopia, Malawiand Niger. The review is located within acontext of growing recognition of the impor-tance of employing more sustainableapproaches through existing health systeminfrastructure, to ensure services continue asemergencies subside and organisations, andtheir resource flows, diminish.

The objectives of the CMAM country reviewwere to:• Assess integration of CMAM into national health systems.• Document challenges of integration and lessons learned and identify factors influencing integration.• Provide recommendations for improved integration and guidance for Office for Disaster Assistance (OFDA) proposal guidelines and partner selection.

To meet these objectives, FANTA organisedvisits to Ethiopia, Malawi and Niger betweenApril and June 2007. The review involved docu-ment reviews, field visits with direct observa-tion of CMAM services, semi-structured inter-views with key informants at national, regional,district and community levels, and discussionswith health system staff, community healthworkers, community volunteers, beneficiariesand non-beneficiaries.

The review found a number of similaritiesand differences between the three countryexperiences. Similarities included start-up andscale-up of CMAM services during crises, weakhealth systems with poor access and low cover-age of services, dependence on donor supportfor supplies, and the presence of numerousstakeholders with fragmented referral andtreatment networks. Differences in integrationrevolved around the extent of Ministry ofHealth (MoH), UNICEF and international non-governmental organisations (INGO) leadershipand coordination, and on varying strategies fortransferring responsibility for CMAM to MoHs.The review identified five key elements forsuccessful integration that should be consid-ered by MoHs, NGOs and donors that aredesigning, implementing or coordinatingCMAM programmes:

i) The enabling environment for CMAM demonstrates the importance of MoH technical leadership and coordination. A support unit to the MoH for technical guidance on CMAM is helpful for capacity development at national policy and district implementation levels. National guidelines serve as an important policy tool and lead to better harmonisation of CMAM services.

Summary of research1

1 Review of CMAM in Post-Emergency Context: Synthesis ofLessons in Integration of CMAM into National Health Systems.Available to download at:http://www.fantaproject.org/publications/ethiopia_2007.shtml

A child attending a CMAMprogramme in Malawi

Waiting area in CMAMprogramme in Maradi, Niger

Research

H D

econin

ck,

FAN

TA,

Nig

er,

2007

H D

econin

ck,

FAN

TA,

Nig

er,

2007

11

Research

The first Lancet paper1 presents new analysesto estimate the effects of the risks related tomeasures of malnutrition, as well as tosuboptimal breastfeeding practices, onmortality and disease. The authors estimatethat stunting, severe wasting, and intrauter-ine growth restriction together were respon-sible for 2.2 million deaths and 21% ofdisability-adjusted life years (DALYs) forchildren younger than 5 years.

Deficiencies of vitamin A and zinc wereestimated to be responsible for 0.6 millionand 0.4 million deaths, respectively, and acombined 9% of global childhood DALYs.Iron and iodine deficiencies resulted in fewchild deaths and combined were responsiblefor about 0.2% of global childhood DALYs.Iron deficiency as a risk factor for maternal

growth standards) were related to adultoutcomes (height, schooling, income or assets,offspring birth weight, BMI, glucose concen-trations, blood pressure). Systematic review ofstudies from low and middle income countriesfor these outcomes and for indicators relatedto blood lipids, cardiovascular disease, lungand immune function, cancers, osteoporosis,and mental illness was undertaken. It wasfound that undernutrition was strongly associ-ated, both in the review of published work andin new analyses, with shorter adult height, lessschooling, reduced economic productivity andfor women, lower offspring birth weight.Associations with adult disease indicatorswere not so clear-cut. Increased size at birthand in childhood were positively associatedwith adult BMI and to a lesser extent withblood pressure values, but not with bloodglucose concentrations.

In the new analysis and in published work,lower birth weight and undernutrition inchildhood were risk factors for high glucoseconcentrations, blood pressure and harmfullipid profiles, once adult BMI and height wereadjusted for. This suggests that rapid postnatalweight gain, especially after infancy, is linkedto these conditions.

The review of published works indicatedthat there is insufficient information aboutlong term changes in immune function, bloodlipids or osteoporosis indicators. Birth weightis positively associated with lung function andwith the incidence of some cancers, andundernutrition could be associated withmental illness. The reviewers also noted that height-for-age at 2 years was the best predictor of human

The second paper1 in the Lancet series dealswith the potential long-term implications ofundernutrition. The paper reviews the associ-ation between maternal and child undernutri-tion and human capital and risk of adultdiseases in low-income and middle incomecountries. Data are analysed from five long-standing prospective cohort studies fromBrazil, Guatemala, India, the Philippines andSouth Africa.

The analysis found that indices of maternaland child undernutrition (maternal height,birth weight, intrauterine growth restriction,and weight, height and body mass index(BMI) at 2 years according to the new WHO

Below are short summaries of the recently launchedLancet series of papers on Maternal and ChildUndernutrition1. This high profile series focuses on thedisease burden attributable to undernutrition and theinterventions aimed at addressing the problem bystrengthening household food availability and use,maternal and child care and control of infectiousdiseases. Although this series mainly deals with nutri-

tional issues in stable contexts, it is nonetheless a crit-ically important (as well as much criticised) body ofwork that provides a comprehensive update and snap-shot on the state of nutrition globally. Furthermore,while the ENN and Field Exchange are mandated tofocus on emergency nutrition, it is recognised both bythe editorial team (and almost certainly by our readers)that the distinction between emergencies and stable

contexts is often ambiguous and at times quite arbi-trary. Clearly, many of the findings in the LANCETseries have relevance to emergency contexts. A numberof our readers approached us to provide comment onthe series, considering the ENN and Field Exchange asgood fora for this type of exchange and these areincluded in the Letters section (Ed).

mortality added 115,000 deaths and 0.4% ofglobal total DALYs. Suboptimal breastfeed-ing was estimated to be responsible for 1.4million child deaths and 44 million DALYs(10% of DALYs in children under five years).

In an analysis that accounted for co-expo-sure of these nutrition-related factors, theywere together responsible for about 35% ofchild deaths and 11% of total global diseaseburden. The high mortality and diseaseburden resulting from these nutrition-relatedfactors make a compelling case for the urgentimplementation of interventions to reducetheir occurrence or ameliorate their conse-quences.

Key related research needs identified by theauthors include;

• Prevalence of deficiencies of vitamin A, zinc, iron and iodine in sub-national populations. • Consequences of nutritional deficiencies for mortality from HIV/AIDS, malaria, and other important infectious diseases.• Consequences of nutritional deficiencies for immune competence, brain develop- ment, cognitive ability and other possible effects.• Overlap of micronutrients and their joint effects on mortality and morbidity.• Development of international foetal and newborn growth standards.

capital and that undernutrition is associatedwith lower human capital. The authorsconcluded that damage suffered in early lifeleads to permanent impairment, and mightalso affect future generations. Its preventionwill probably bring about important health,education and economic benefits. Also,chronic diseases are especially common inundernourished children who experiencerapid weight gain after infancy.

A number of areas for future research werehighlighted:• Association between rapid weight and length gain at different age intervals in infancy and childhood with human capital and outcomes related to chronic disease, to define the age after which rapid growth should be avoided.• Long term effects of weight gain in late childhood stratified in previously stunted and non-stunted children, and for children with and without intrauterine growth restriction.• Long term effects of micronutrient deficien- cies in childhood.• Association between undernutrition and longterm changes in immune function blood lipids, osteoporosis, and mental illness.• Improved quantification of the economic effect of undernutrition on adult produc- tivity.• Interactions between genes and environ- mental factors in long term outcomes.

Global and regional exposures and health consequences (Paper 1)

Consequences for adult health and human capital (Paper 2)

1 Black et al, 2008. Maternal and Child Undernutrition 1.Maternal and child undernutrition: global and regionalexposures and health consequences. Published OnlineJanuary 17, 2008. DOI:10.1016/S0140-6736(07)61690-0

1 http://www.globalnutritionseries.org/

1 Victoria et al (2008). Maternal and Child Undernutrition 2.Maternal and child undernutrition: consequences for adulthealth and human capital. Published OnlineJanuary 17,2008DOI:10.1016/S0140-6736(07)61692-4

Mother withher 2 hour old

baby, weighing1.7kg, in

Tillaberg, Niger

WFP

/Ste

phan

ie S

avar

iaud,

Nig

er,

2005

Summary of Lancet Series on Maternal and Child Undernutrition

1212

Research

In the third paper of the Lancet series1, theauthors review interventions that affectmaternal and child undernutrition and nutri-tion related outcomes. These interventionsincluded promotion of breastfeeding, strate-gies to promote complementary feeding,with or without provision of food supple-ments, micronutrient interventions, generalsupportive strategies to improve family andcommunity nutrition and reduction ofdisease burden (promotion of handwashingand strategies to reduce the burden ofmalaria in pregnancy). The authors showed that although strate-gies for breastfeeding promotion have a largeeffect on survival and are deemed an essen-tial intervention for infants younger than 5.9months across all populations, the effect ofbreastfeeding strategies on stunting wassmall. In populations with sufficient food,education about complementary feedingincreased height-for-age Z score by 0.25 (95%CI 0.01-0.49). However, provision of foodsupplements (with or without education) inpopulations with insufficient food increasedthe height-for-age Z score by 0.41 (0.05-0.76).The authors consider complementary feed-ing counselling and support strategies infood insecure populations could substan-tially reduce the burden of stunting andrelated disease. Management of severe acute malnutritionaccording to WHO guidelines reduced thecase-fatality rate by 55% (risk ratio 0.45, 0.32-0.62). Recent studies suggest that newercommodities, such as ready-to-use therapeu-tic foods, can be used to manage severe acutemalnutrition in community settings.

Paper 4 in the Lancet series1 reports on anassessment of actions addressing undernutri-tion in the countries with the highest burdenof undernutrition, drawing on systematicreviews and best-practice reports. Seven keychallenges for addressing undernutrition atnational level are defined and reported on:getting nutrition on the list of priorities andkeeping it there, doing the right things, notdoing the wrong things, acting at scale,reaching those in need, data-based decision-making and building strategic and opera-tional capacity. The authors argue that interventions withproven effectiveness that are selected bycountries should be rapidly implemented atscale. The period from pregnancy to 24months of age is a crucial window of oppor-tunity for reducing undernutrition and itsadverse effects. Programme efforts, as well asmonitoring and assessment, should focus onthis segment of the continuum of care.Nutrition resources should not be used tosupport actions unlikely to be effective in thecontext of country or local realities. Nutritionresources should not be used to supportactions that have not been proven to have adirect effect on undernutrition, such asstand-alone growth monitoring or schoolfeeding programmes.

In addition to health and nutrition inter-ventions, economic and social policiesaddressing poverty, trade and agriculturethat have been associated with rapidimprovements in nutritional status should beimplemented. There is a reservoir of impor-tant experience and expertise in individualcountries about how to build commitment,develop and monitor nutrition programmes,move toward acting at scale, reform orphase-out ineffective programmes, and otherchallenges. This resource needs to beformalised, shared and used as the basis forsetting priorities in problem-solving researchfor nutrition.

A number of research priorities to supportnational nutrition actions are suggested.• Research on strengthening leadership and strategic capacity for advancing national nutritional agendas and actions. Positive experiences in Madagascar, Senegal, Thailand, Chile, Costa Rica and other countries have shown that leadership and strategic capacity are key ingredients for advancing the national nutrition agenda and action. Among other roles, these capacities are crucial for leveraging commitment and resources from govern- ment, international partners and the

Effective micronutrient interventions forpregnant women included supplementationwith iron folate (which increased haemoglo-bin at term by 12g/L, 2.93-21.07) andmicronutrients (which reduced the risk oflow birth-weight at term by 16% (relative risk0.84, 0.74-0.95). Recommended micronutrientinterventions for children included strategiesfor supplementation of vitamin A (in theneonatal period and late infancy), preventivezinc supplements, iron supplements for chil-dren in areas where malaria is not endemicand universal promotion of iodised salt. The authors used a cohort model to assessthe potential effect of these interventions onmothers and children in the 36 countries thathave 90% of children with stunted lineargrowth. The model showed that existinginterventions that were designed to improvenutrition and prevent related disease couldreduce stunting at 36 months by 36%, mortal-ity between birth and 36 months by about25%, and disability-adjusted life years associ-ated with stunting, severe wasting, intrauter-ine growth restriction and micronutrientdeficiencies by about 25%. To eliminatestunting in the longer term, these interven-tions should be supplemented by improve-ments in the underlying determinants ofunder-nutrition, such as poverty, poor educa-tion, disease burden, and lack of women’sempowerment. The authors identified a number of evidencegaps:i) The effectiveness and cost-effectiveness ofnutritional interventions in national healthsystems need urgent assessment. Both singleand packaged interventions that affectgeneral nutrient micronutrient intake in

women and children should be assessed fortheir effect on stunting rates and weight gain. ii) Few nutritional interventions for mothershave assessed a wide range of outcomes atsufficient scale. In particular, those withmultiple micronutrients and calcium need tobe assessed with long-term tracking of effecton maternal and child health.iii) Few studies of large-scale interventionsfor promotion of breastfeeding have assessedtheir effects on feeding patterns and growthoutcomes beyond infancy. With the newWHO growth standards, such studies areneeded to assess the effectiveness of strate-gies for breastfeeding promotion and appro-priate complementary feeding for growthand morbidity in various age-groups. iv) There is a need for large-scale studies toverify the irreversibility of stunting in chil-dren aged 36-49 months and older. v) Whether the adverse effects associatedwith stunting, e.g. cognitive impairment orrisk of infection disease, can be amelioratedor reversed.vi) Although the efficacy of preventive zincsupplementation is proven, studies of theeffectiveness of various zinc delivery strate-gies (fortification, supplementation, and bio-fortification) are urgently needed.vii) Since community-based preventive andtreatment strategies for severe acute malnu-trition have been the subject of only a fewstudies, robust experiments in this areashould be prioritised.

private sector. Research is needed to docu- ment the capacities, strategies and tactics present in successful countries, to guide international investments, and to facilitate the exchange of experience between developing countries learning in this important area.• Large-scale effectiveness assessments that can expand the evidence base for strate- gies, and tactics to achieve high, sustained and equitable coverage with proven inter ventions to address undernutrition are also needed.• Development and assessment of valid indicators and methodologies that can be used at national level and below to provide rapid feedback on progress in generating political commitment, strate- gies and operational capacities, coverage and effect. • Links between nutritional status and broader initiatives, such as food for work and micro-credit initiatives, need to be substantiated and used as the basis for assessing their effect on nutrition outcomes.

Interventions for maternal and child undernutrition and survival (Paper 3)

Effective action at national level (Paper 4)

1 Bhutta et al (2008). Maternal and Child Undernutrition 3.What works? Interventions for maternal and child under-nutrition and survival. Published Online January 17,2008DOI:10.1016/S0140-6736(07)61693-6

1 Bryce et al (2008). Maternal and Child Undernutrition 4.Maternal and child undernutrition: effective action atnational level. Published Online. January 17,2008DOI:10.1016/S0140-6736(07)61694-8

Summary of Lancet Series on Maternal and Child Undernutrition

13

The last paper in the Lancet series (Paper 5)1

opens with the statement that many transna-tional organisations work to support efforts toeliminate maternal and child undernutritionin high-burden countries. Financial, intellec-tual and personal linkages bind these organi-sations loosely together as components of aninternational nutrition system. The paperthen goes on to argue that such a systemshould deliver in four functional areas: stew-ardship, mobilisation of financial resources,direct provision of nutrition services at timesof natural disaster or conflict and human andinstitutional resource strengthening.

The authors review quantitative and quali-tative data from various sources to assess theperformance of the system in each of theseareas, and find substantial shortcomings.Fragmentation, lack of an evidence base forprioritised action, institutional inertia andfailure to join up with promising develop-ments in parallel sectors are recurrent themes.Many of these weaknesses can be attributedto systemic problems affecting most organisa-tions working in the field and these areanalysed using a problem tree approach. Theauthors also make recommendations to over-come some of the most important problems.They propose five priority actions for thedevelopment of a new international architec-ture - a new global governance structure, amore effective United Nations (UN), fewerparallel organisations but fewer mandategaps, more investment in capacity strengthen-ing in high-burden countries, and researchleadership in areas that matter.

The authors also make a number of inter-esting observations and points about emer-gency nutrition work, including:

Emergency food aid went principally tojust six countries between 2000 and 2004 -Ethiopia, Sudan, Afghanistan, Angola, Iraqand North Korea. They claim that this clearlyshows a politicised distribution and that foodaid allocations at the macro level have tradi-tionally served primarily domestic agricul-tural interests and foreign policy objectives.They also emphasise that information oncoverage of nutrition services in emergenciesis difficult to obtain. The dynamic nature ofemergencies, and their resulting disruptedand mobile populations, makes estimatingcoverage challenging, and some have ques-tioned the ethics of doing even appliedresearch in these environments.

The authors also suggest that assessmentsand reviews of nutrition actions in emergen-cies have largely focussed on the effect ofvarious feeding programmes on nutritionoutcomes, such as growth and micronutrientsstatus. What is often lacking is a clear analysisof the cost-effectiveness of different interven-tions. This would enable recommendationson the optimum rations composition, target-ing and exit criteria, and the appropriate mixof complementary activities to improvehealth and nutrition outcomes.

They state that one key challenge is theabsence of an agency with responsibility fortaking an overview of the effectiveness andcost effectiveness of different types of inter-vention. Nutrition in emergencies is a mix of

multiple agencies, agendas, protocols, andmethods. The general lack of coordinationand leadership has allowed the institutionalstatus quo to prevail. Thus, agencies that havebuilt up expertise and mandates aroundcertain types of intervention (or interventiondesign) will continue to practice these inter-ventions in emergencies without seriousexamination or challenge.

Several groups – including the UN Inter-Agency Standing Committee NutritionCluster, the Sphere Project, SMART,ReliefWeb, and the Emergency NutritionNetwork - are providing guidance on bestpractice in emergency settings. Building onand consolidating these experiences willgenerate a minimum set of operational stan-dards and a source of much needed documen-tation.

On research, the authors state there is alack of rigorous programme evaluation dataon which to build strong evidence-basedguidance for nutritional nutritionprogrammes. Market forces and institutionalbarriers in international research centres haveconspired to distort the research agenda awayfrom solution-oriented analysis of the prob-lems that contribute to the greatest burden ofdisease and lost human potential. Improvingthe quality and relevance of nutrition researchis a crucial part of strengthening the interna-tional nutrition system.

The authors single out the following researchpriorities:• Research into the accountability and responses of governments to their nutrition relevant commitments under international conventions. • Rigorous analysis of the linkages between nutrition outcomes and global change processes such as climate change, trade liberalisation, etc.• Research into the quantity and effectiveness of international aid for improved nutrition.• Evaluation and prioritisation of interven- tions to ensure timely and relevant emer gency responses. This should include improved problem analysis, more efficient sampling methods for estimation of preva- lence of severe acute malnutrition and targeting of interventions, and cost-effec- tiveness studies of different food commodi- ties and therapeutic foods. • Research on the gap between current and required manpower, training capacity and training resources, at national, regional and global levels.• Research on best practices for designing and delivering the pre-service training, continuing education, and knowledge management system that practitioners need to address undernutrition effectively.• Meaningful self-assessment and peer- assessment of the effectiveness of individ- ual organisations involved in designing and delivering goods, services, and ideas relevant to the elimination of maternal and child undernutrition.

Effective international action against undernutrition (Paper 5)

Research

1 Morris et al (2008). Effective international action againstundernutrition: why has it proven so difficult and what canbe done to accelerate progress? Published Online. January17, 2008. DOI: 10.1016/S0140-6736(07)61695-X

A WFP shipment of 14,000 MT of rice arrivingat the Iraqi port of Umm Qasr in 2003.

WFP

/Anto

nia

Par

adel

a, I

raq,

2003

WFP

/Ger

ald B

ourk

e, D

PR K

ore

a, 2

005

North Korean dock workers baggingdonated wheat – part of a 39,500 tonneshipment in 2003.

14

Field Article

Acción contra el Hambre España (ACF-E) has been working in Gao andKidal, two regions in northern Mali,since 1996. The population comprises

largely farmers and herders in the valley area ofthe Niger River. Those living in the remainingvast expanse of the two regions are nomadic ortranshumant. Both regions have similar annualrainfall ranging from 100 millimeters (mm) in thenorth to 350mm in the south. The nomadic popu-lation tends to practice an opportunistic type oflivestock farming. They are continuously on themove following a set route that only varies whenfaced with a bad year that requires them to adaptin some way.

The ecological, geographical and social envi-ronment of this nomadic population does notallow for the application of standard methods formonitoring and assessing their vulnerability.Particular challenges for monitoring are that it isa large area of land with a low populationdensity, it is hard to assess available resources atthe end of the rainy season and the movement ofthe population and herd. These factors also makeit very difficult to assess the nutritional situation.On previous occasions, nutritional assessmentswere mainly conducted on sedentary nomadicpopulations1 in accessible camps2. However,these survey findings could not be extrapolatedto the whole population of the region - in somecases, more than half of the population livesoutside camps, is on the move, and has a verydifferent set of problems to the sedentary resi-dents. The challenge, therefore, is how torandomly select a representative sample of thewhole population in the region.

Survey optionsThe census carried out in Mauritania in 1965 wasthe first large scale survey of a primarilynomadic population. For this census, the ‘frac-tion’3 was taken as the survey unit and lists of allpeople belonging to a fraction and its geographic

area were drawn up. Heads of each fraction werealso recorded. The problem with this methodol-ogy was that it required a lot of preparation timeand not all of the tents belonging to the differentfractions could be identified4.

Exhaustive sampling methodologies had beencarried out previously in Mali, using wateringpoints and sedentary areas as starting points.However, given the wide geographic dispersionof the population and the difficulty in establish-ing the exact location of the camps, this methodand ensuing results were also flawed.

Five different group sampling methods fornomadic populations have been put forward byGraft-Johnson (1979) based on a revision of themethods proposed by the United Nations (1977)5 :

The group assembly method: The difficulty withthis approach lies in trying to get all of the popu-lation to gather at a given point.

Camp method: In order to employ this methodthere has to be good population census data orlists of those in the camps with exact locationsknown.

Social structure methodology: This is only viableif the head of each fraction can draw up a list ofthe exact composition and whereabouts of eachcamp. There is a risk of omitting or double count-ing subjects, and finding the subjects can also beproblematic.

Numbering by areas: As with other methodolo-gies, there is an inherent problem with coverage,as it is not possible to cover an entire area (evenif small). There is also the risk that subjects couldmove from one area to another.

Watering point method: This method can only beapplied if the subjects of the survey are the samepeople who go to the watering holes.

Given the options, in November 2005, AcciónContra el Hambre designed a pilot study to

assess nutritional status in the study area. Thiswas based on the area numbering method and onthe cluster sampling approach advocated bySMART (Standardised Monitoring andAssessment of Relief and Transition6). In theSMART methodology, if the settlements are smallthere is the option to have more clusters, withfewer children per cluster, to ensure there will bea sufficient number of children at each site.

Pilot study methodologyThe study used weight/height indicatorsexpressed as z-scores7 and oedema. The numberof clusters was calculated taking account of thenumber of children that a team could test perday. In a nutritional survey pilot test, carried outin the Anderamboukan district (Gao region), 14children were selected per cluster. The samplesize was calculated on the basis of the populationin each area (as indicated by official statistics),the expected prevalence, the precision and thehomogeneity of the sample (design effect).

Population estimates were reached inconjunction with the local authorities, given thatthere are no official figures per town, only per

Methodology for a Nutritional Survey among thenomadic population of northern Mali By Eva Vicent and Núria Salse

Eva Vicent has a background in nursing studiesin the University of Valencia and is currentlyworking for Action Against Hunger as NutritionCoordinator in Swaziland. She has been involvedin health and nutrition programmes since 2000,working in several countries including Angola,Guinea Bissau, Indonesia, Malawi and Mali.

Núria Salse is the Health and Nutrition Advisor in Acción Contra elHambre. Previously she spent several years working on nutrition andmedical programmes in Angola, Guinea Conakry, Niger and Argentina.

The authors acknowledge Thierry Métais who contributed to develop-ing the methodology, and Waltraud Novak and Mathias Grossiord whoundertook the field work and contributed to improving the method.

Table 1: Distribution of clusters, population size and corresponding grid

Geographical unit

Populationsize

Assigned cluster

Andremboukan 3000 1,2,3,4

Tinagorof 2000 5,6,7

Alladi 1400 8,9

Tamalate 2000 10,11,12

Inchinanan 1500 13,14

Tagalalte 1500 15,16,17

Goussou 2000 18,19

Inkalafane 1500 20,21,22

Hanouzigoren 1300 23,24

Nomade 7780 25,26,27,28,29,30,31,32,33,34,35

Figure 1: Grid distribution over Anderamboukan neighbourhood

1 Save the Children UK. Nutrition Assessment in Gola OdaWoreda, East Haraghe, Ethiopia. March 26 April 6, 20012 Susanne Jaspars and Hisham Khogali. Oxfam Food andNutrition Group. Oxfam's Approach to Nutrition surveys inemergencies, May 20013 Fraction: Administrative unit used in Mauritania as amethod of social organisation within a tribe.4 Jaques Brenez. L’óbservation démographique des milieuxnomades. L'enquête de Mauritanie ». Population. Année1971, Volume 26, Number 4 p 721-736 5 William D. Kaslbeek, University of North Carolina, Anne R.Cross, Westinghouse Health Systems. Problems in samplingnomad populations.6 http://www.smartindicators.org7 These were calculated using the NCHS (National CentreHealth Statistics) tables and the recent WHO 2005 standardtables. NCHS tables were used to compare with previoussurveys and for programme planning (as NCHS were used asa criteria of admission). The low acute rate prevalence foundin this population did not allow for detecting major differ-ences between the NCHS references and WHO standards.

Figure 2: Choosing acamp or settlement ineach cluster

15

fraction or family. An estimate of the popula-tions of the different sedentary areas andnomadic camps was made assuming thenomadic area had a homogenously distributedpopulation throughout the survey area andbased on discussions with different members ofthe community. The Emergency NeedsAssessment computer programme was used toanalyse the clusters (example in Table 1).

In the nomadic area, each cluster is repre-sented by a square8. Clusters were only retainedif the central point was inside the survey area.Figure 1 shows the grid distribution inAnderamboukan for the 11 selected clustersthat correspond to the nomadic area.

In order to select the children for the survey,the central point (centroid) of each cluster wastaken as the starting point. This point was iden-tified by using the ‘MapInfo’ programme andthen transferring it to the Global PositioningSystem (GPS). On arrival at this point, thesurvey team followed the Epi method, whichconsists of throwing a pen into the air andfollowing the direction it is pointing in when itfalls to ground until arriving at a camp orvillages. If no village was arrived at by the timethe limit of the cluster had been reached or if thedirection the pen was pointing in proved to beinaccessible, the team would return to thecentral point and throw the pen again. Once theteam had reached the first settlement it thenmoved on to the next, and so on until the clusterwas fully covered (See Figure 2).

If the centre of the cluster proved to be in aninaccessible place (for example, a lake), thestarting point became the closest accessibleplace to the western edge of the central point.If, having set off in eight different directions, thecluster was still not complete, the survey teamproceeded to the cluster directly to the right ofthe one they were working in.

If, following the direction indicated by thepen, a town or settlement was reached that hadalready been tested, It was skipped by stopping3 km before the town or settlement and restart-ing 3 kilometres from the other side of thesettlement or village in the same direction.

DiscussionIn previous years, Acción contra el Hambre hadelected to only test sedentary areas or to use anexhaustive methodology to assess nutritionalstatus of the population in the region. This time,by applying different methods, we haveobtained data on both the sedentary andnomadic populations. Also, by predeterminingthe sample and including the most isolatedareas, as opposed to using an exhaustiveapproach, we have obtained a highly represen-tative sample.

At the same time, the reduction in thenumber of subjects per cluster has allowed us tocomplete the surveys in each cluster. On aver-age, only one cluster per survey has had to becomplemented and therefore completed withchildren from a subsequent cluster.

This method also ensures the random selec-tion of subjects, although two limitations didhave to be taken into account. First, the ‘bordereffect’. Selecting only those clusters that havetheir geographical centre inside the district andtesting them in full means that sometimes thearea surveyed does not coincide exactly withthe existing administrative boundaries.Secondly, sampling in nomadic areas does nottake into account the fact that the population

distribution is not homogenous and that someareas are more densely populated than others.

The use of GPS technology during thesurvey did not pose any difficulties as it is verysimple to use and was included in the teamtraining from the start. Sometimes, however,geographical circumstances and difficult terrainmade finding the central points, and the teams’work in general, more difficult. Following the first pilot test, it was decidedto carry out exploratory trips before the surveysin order to find the central points and identifyany possible geographical difficulties. Thismeant that the teams had received preciseinstructions on how to get to the central pointsand were able to modify the methodologyaccording to the terrain, if necessary. Thisexploratory work made the teams’ work consid-erably easier although it also increased thepreparation time and cost of the survey.

Conclusions and recommendationsUsing a group methodology for nutritionalsurveys in pastoral areas where the populationis scattered presents problems in terms of locat-ing and measuring certain groups of the popu-lation.The method of selecting groups that wasproposed for the pastoral areas and described inthis article facilitates the principle of randomselection of children. Reducing the number ofchildren per group also ensures that all therequired measurements for each group werecompleted. The methodology used cancontribute towards assessing the nutritionalsituation of the pastoral population of Mali.

If further nutritional surveys are to be carriedout using this methodology, certain problemsencountered should be addressed as follows:• Reduce further the number of subjects per cluster in order to reduce the daily work load.• Plan the time and resources required for identifying nomadic clusters, as this will greatly facilitate implementation of the survey.• Make necessary improvements in epidemio- logy statistics to the methodology, in order to reduce possible bias in the selection of the target population and to increase the repre sentativeness of the sample.

Acción contra el Hambre recommends ongoingresearch into appropriate assessment methodsfor pastoral areas. For example, recent studiespublished by Mark Myatt9 indicate that ifWeight/Height indicator and Mid Upper ArmCircumference (MUAC) are used in surveysthey lead to different findings in terms of preva-lence of malnutrition as a result of differences inbody shape. Future research on pastoral popu-lations should include the use of MUAC togauge the prevalence of acute malnutrition.

With this in mind, Acción contra el Hambrewill be busy in 2008 improving and consolidat-ing pastoral survey methods through newresearch involving its international network ofagencies. A new publication on this subject willbe coming out in the second quarter of 2008.

For further information, contact: Núria Salse,email: nsalse@achesp.org tel: 00 34 91 391 53 00

8 The sampling area was represented as a grid as per theCentric Systematic area sampling (CSAS) method 9 Report concerning the analysis of data collected for theMUAC/weight-for-height/body-shape research study. MarkMyatt, Institute of Ophthalmology, University CollegeLondon. 30th May 2007.

News

The Integrated Food Security PhaseClassification (IPC) is a standardised tool thataims at providing a ‘common currency’ forclassifying food security. Using a commonscale, which is comparable across countries,will make it easier for donors, agencies andgovernments to identify priorities for inter-vention before they become catastrophic.

A new website has been set up by the sevenagencies and international NGOs supportingthe IPC (see Global Partners below). Contentincludes IPC maps and food security classifi-cations for many countries, information abouttraining workshops, publications and newsarticles, and updates to the IPC technicalmanual. An e-learning course and relatedtraining materials and interactive features,such as forums, are also being planned.

Originally developed in Somalia by theFood Security Analysis Unit (FSAU), IPCmaps are now being produced in Somalia andKenya - where the IPC has been officiallyadopted by the national government. IPCrelated activities have been substantial inSouthern Sudan and Burundi, while start-upactivities are taking place in the DemocraticRepublic of Congo, Uganda and Tanzania. IPCrelated training, awareness raising activitiesand pilot tests have also taken place inSouthern Africa and Asia.

The IPC Global Partners include the Foodand Agriculture Organisation (FAO), TheWorld Food Programme (WFP), CareInternational, the European Commission JointResearch Centre (EC JRC), the Famine EarlyWarning System Network, Oxfam GreatBritain, and Save the Children UK and US. TheEuropean Union, the Dutch government, theCanadian International Development Agency(CIDA) and the U.K.’s Department forInternational Development (DFID) are provid-ing funding.

Integrated Food Security Phase Classification(IPC) website: http://www.ipcinfo.org

Integrated FoodSecurity PhaseClassification (IPC)Website Launched

Field Article

16

News

USAID's Office of Food for Peace (FFP) and theFood and Nutrition Technical Assistance(FANTA) Project have released ‘Trigger Indicatorsand Early Warning and Response Systems inMulti-Year Title II Assistance Programmes’,No. 5 in FFP's Occasional Paper Series.

In geographic areas and populations chroni-cally vulnerable to food insecurity, early warn-ing and response (EWR) systems can providenecessary information to help modifyprogramme interventions and increaseresources in response to shocks. Trigger indica-tors (TIs) are one such EWR mechanism.

The paper has been published to allow TitleII Cooperating Sponsors a greater degree of

The ENN, as co-ordinator of the IFE CoreGroup1, co-facilitated a regional strategy work-shop on infant and young child feeding inemergencies on 10-13th March 2008 in Bali,Indonesia. The theme of the meeting was‘Infant and Young Child Feeding inEmergencies: Regional Experiences andChallenges in Achieving Optimal EarlyResponse’. This workshop was organised underthe auspices of the Inter Agency StandingCommittee (IASC) Nutrition Cluster that is ledby UNICEF. Within the Nutrition cluster, themain organiser was the ENN, representing theIFE Core Group, in cooperation with UNICEFEast Asia and Pacific Regional Office, UNICEFSouth Asia Regional Office, UNICEF IndonesiaOffice, and the Ministry of Health, Indonesia.

The workshop was funded by the IASCNutrition Cluster and a contribution from

1 An interagency collaboration concerned with policy guid-ance and capacity building on IFE. Current members areUNICEF, UNHCR, WFP, WHO, IBFAN-GIFA, CARE USA, ACFand ENN. Associate members are IFRC and SC UK. See athttp://www.ennonline.net/ife

Regional IFEWorkshop held in Indonesia

IBFAN-GIFA to support attendance of keydelegates who were not self-funding.

108 participants from a mixture of keygovernment, UN, donor, international andlocal NGO representatives from 16 countriesattended the workshop. A press release wasissued by UNICEF following the workshop, inresponse to the pledge for action agreed by the participants (see at http://www.unicef.org/eapro/media_7973.html).

The proceedings will be available from theENN in late June 2008 and will be summarisedin the next issue of Field Exchange.

flexibility in responding to emerging crises andshocks in their areas of operation, without therisk of potentially undermining advancesbeing achieved by development interventions.While written specifically for recipients of TitleII assistance, it has relevance for others work-ing in the food security context.

Support for the development of the guidewas provided by USAID's Bureau forDemocracy, Conflict and HumanitarianAssistance's Office of Food for Peace andBureau for Global Health's Office of Health,Infectious Disease and Nutrition.

The guide can be downloaded from FANTA'swebsite at http://www.fantaproject.org

Trigger Indicators and Early Warningand Response Systems

The Food and Nutrition Technical Assistance(FANTA) Project has announced the release ofthe Review of Incorporation of EssentialNutrition Actions into Public HealthProgrammes in Ethiopia Report (January 2008).

The Essential Nutrition Actions (ENA) pack-age is an approach to expand the coverage ofseven affordable and evidence-based actions toimprove the nutritional status of women andchildren, especially those under two years ofage. The review, requested by USAID/Ethiopia, examined a number of facilitatingand inhibiting factors to ENA integration in thecontext of Ethiopia’s health system. It foundthat the approach has been incorporated intothe Ethiopia Federal Ministry of Health system

and multilateral and NGO programming.However, improved training and other stepsare necessary to further institutionalise theapproach.

Support for the synthesis report wasprovided by USAID's Bureau for GlobalHealth's Office of Health, Infectious Diseaseand Nutrition and the USAID Mission inEthiopia.

The report can be downloaded from FANTA'swebsite, http://www.fantaproject.org orcontact: Information Associate, FANTA Project,1825 Connecticut Ave., NW Washington, DC20009, USA. Tel: +1 (202) 884-8000 fax: +1 (202) 884-8834 email: fanta@aed.org

FANTA Review of Essential NutritionActions in Ethiopia

MSF has launched a new website that archivesall its scientific articles published in peer-reviewed journals. The goal is to make MSFresearch experience available to health careworkers, policy makers and researchers indeveloping countries without the usual costsassociated with retrieving research articles.

The online and searchable content comprises:• Published research and commentary for PubMed articles• Conference abstracts• MSF's Ethics Review Board• Programme descriptions that give lessons learned from the field• Collaborating partners.

The articles are available free, with full text,and in an easily searchable format. No loginis required.

Users are invited to send comments aboutthe website, published articles or othersubjects via a Feedback section where allsuitable comments will be posted.

Visit online at www.fieldresearch.msf.org

TALC (Teaching-aids At Low Cost)is undertaking amini-evaluation oftheir CD-ROM on‘ C o m m u n i t yNutrition’. ThisCD contains documents andother materials relevant to health workers inlow-income countries. In order to collect datafrom intended users, TALC will send a freecopy of the CD and a one-page evaluationform to tutors of nutrition courses in low-income countries. Up to 10 free copies of theCD will be sent to all those who fullycomplete and return the form.

If you are a tutor, or know of any tutors,who can participate in this evaluation, sendnames, postal addresses and email addressesto TALC, email: info@talcuk.org or to MarlouBijlsma, email: marloubijlsma @yahoo.co.ukor Ann Burgess, email: annpatriciaburgess@yahoo.co.uk

Three free copies of the CD are available toanyone who sends in advance UK£1 or US$2to cover postage costs. For more details, email: info@talcuk.org

The Community Nutrition CD can also beordered online at www.talcuk.org/cd-roms.htm

New MSFwebsite on fieldresearch

Evaluation ofTALC’s CD-ROM‘CommunityNutrition’

Webpage from the World Bankon the Food Price Crisis

17

News

Save the Children UK (SCUK) and FEG Consulting(formerly known as theFood Economy Group(FEG)) have just broughtout a new publication‘The Household EconomyApproach: A guide forprogramme planners and policymakers’. It is part of a toolkit for the southern AfricaRegional Hunger and Vulnerability Programme(RHVP), aimed at strengthening the capacity of stafffrom government and national and international non-governmental organisations (NGOs) to undertakevulnerability assessment and analysis in southern Africa.

To date, guidance on the Household EconomyApproach (HEA) has been provided by the manual ‘TheHousehold Economy Approach – a resource manual forpractitioners’ (Save the Children, 2000), and byresources and training materials produced since then bythe FEG and SC UK. The toolkit aims to bring togetherand consolidate this material and provide up-to-dateguidance on the approach, its use in the field and itsapplication for particular purposes.

The toolkit comprises three elements:i) The Practitioner’s Guide to the HEA1, which is a ‘how to’ guide for those participating in field work.ii) The Household Economy Approach, which is a guide for programme planners and policy-makers, targeted primarily at those who are involved in using assessment results to inform decisions on response and to assist in programme planning. It aims to help policy-makers and programme plan- ners understand the methodology, interpret results and engage critically in the process of translating results into programme and policy recommendations.iii) The HEA training guide, targeted at those facilitating HEA trainings.

The new guide begins with an overview of the analyti-cal framework (Chapter 2), outlining essential steps inHEA and why these are necessary. Chapter 3 describesthe application to which HEA has been put over the past15 years. The guide then gives an overview of how HEAis carried out – how information has been collected inthe field to date and the tools that assist in HEA analysis(Chapter 4) and considers why such field methods areused and whether the information and analysis is reli-able (Chapter 5). The linkages between HEA and otherapproaches and areas of inquiry are described inChapter 6, which also outlines how HEA can be appliedto a number of issues of relevance in the analysis ofvulnerability, poverty and chronic food insecurity. Someof the criticisms that have been made of HEA arediscussed in Chapter 7. Finally, a number of the prod-ucts that can be generated from HEA analysis areoutlined in Chapter 8.

The HEA Guide for programme planners and policy-makers is available to purchase from SC UK, price £14.99sterling plus postage and packing (rates vary accordingto number of copies ordered and destination). Fororders, contact: Save the Children Publications, c/oNBN International, Estover Rd, Plymouth PL6 7PY, UK.Tel: 0044 (0)1752 202301 Fax: 0044 (0)1752 202333 e-mail: orders@nbninternational.comhttp://www.savethechildren.org.uk

1 Download at http://www.savethechildren.org.uk/en/54_4200.htm

A regional consultation on nutrition andHIV in South East Asia was held inBangkok, 8-11 October 2007.Represenativesof countries from the region, together withscientists, researchers, programmers, deci-sion makers, United Nations (UN) agen-cies, non-governmental organisations(NGO)s, civil society groups, caregivers,people living with HIV/AIDS groups,donors and bilaterals, gathered to reviewevidence, operational challenges andlessons. Recommendations for action werereleased in a participants’ statement,summarised here.

Nearly 4 million living with HIV/AIDSlive in South-East Asia, a region where 79%of the world's malnourished children resideand where several countries report over40% stunting. With its large and rapidlygrowing populations, widespread malnu-trition and burden of infectious and chronicdisease, the region is particularly vulnera-ble to the HIV epidemic.

Nutrition must be incorporated into allaspects of HIV prevention, treatment andcare programmes as a high priority - thechallenge is to integrate nutrition into thecontinuum of care. 'Nutrition care andsupport' in this context should be definedbroadly to include not only the provision offood and livelihood security, but also clini-cal care guidelines to include nutrition,counselling, capacity building to supportintegration of nutrition into care, and oper-ational research.

Food and nutrition support is a criticalcomponent of a comprehensive response toHIV. Benefits include reduced susceptibil-ity to infections, improved medicationcompliance and quality of life, preventionof mother-to-child transmission of HIV, andimproved overall HIV-free survival ofexposed infants through appropriate infantfeeding practices.

Participants called for stronger commit-ment from policy makers and donors tointegrate nutrition and HIV/AIDS intoexisting policies and make adequateresources available. A call for immediateactions included:

• Advocacy for greater awareness amongst policymakers and donors of

New publication onHouseholdEconomy Approach

As concerns grow over the emerging food crisis across the globe, the World Bank has createda page to inform the general public about their coordinated efforts to combat this growingproblem. In brief, the World Bank's ‘New Deal’ on global food policy includes creating safetynets such as school feeding, food for work, and conditional cash transfers. Additionalelements of this policy include doubling agricultural lending to Africa over the coming yearand also providing immediate aid to the country of Haiti to feed poor children and othervulnerable groups. The ‘Understanding the Crisis’ section provides brief synopses of thesituation in different parts of the world, including Africa and Asia. The site also contains acalendar of events and links to papers from the World Bank on related topics.

Visit http://www.worldbank.org/html/extdr/foodprices/

Regional Consultation on Nutritionand HIV in South East Asia

the critical link between Nutrition and HIV and the responsibility to incorpo- rate nutrition and HIV considerations into existing national food, nutrition and HIV policies and plans.• Improve food security and livelihoods of families and communities affected by HIV. • Review and update existing policies, programmes, plans of action and guide lines to reflect the nutritional require- ments of people living with HIV/AIDS. • Promote and support optimal infant feeding practices for all children, including those infected with HIV, and meet the nutritional needs of HIV-posi- tive pregnant and lactating women. • Involve adults and children living with HIV in the design and provision of nutritional support interventions and actively pursue gender equity and elim- ination of stigma as obstacles to food security and access to health services. • Urgently and rapidly collect country- specific HIV and related nutritional surveillance data. • Continue building the evidence base through bio-medical, socio-cultural and operational research.• Ensure multi-sectoral coordination and adequate resource allocation. • Call for actions and commitments at the country level and urge country teams (Ministries/UN/NGOs) to draw up action plans that address these recomm- endations and regularly follow up and evaluate progress, preferably every two years. • Make the highest level representation by UN agencies through global and regional forums such as the South Asian Association for Regional Cooperation (SAARC), the Association of Southeast Asian Nations (ASEAN), and the World Health Assembly, and through specifically organised meetings and workshops.

A report on the meeting, along with tech-nical background papers, the participants’statement and related links are at:http://www.who.int/nutrition/topics/hiv_regional_consultation_bangkok/en/index.html

News

18

Module 2 on Infant Feeding in Emergencies(IFE) was first produced by the IFE CoreGroup1 and collaborators in December 2004(version 1). It targets health and nutritionworkers working in emergency situations. Anupdated version (v1.1, December 2007) is nowavailable from the Emergency NutritionNetwork (ENN) in English and, for the firsttime, in French. The new version includesupdated key references, a revised section oninfant feeding and HIV to reflect the latestWHO guidance, and incorporates the currentversion of the Operational Guidance on IFE(v2.1, Feb 2007).

Module 2 comprises four parts – a coremanual, additional material (includingsections on artificial feeding and managementof acute malnutrition in infants under sixmonths), annexes and slides content. These areavailable to download separately or as acomplete document.

The update and translation into French wasfunded by the UNICEF-led Inter-AgencyStanding Committee (IASC) Nutrition Clusteras part of a package of IFE activities imple-mented by the ENN and the IFE Core Groupin 2007.

Module 2, v1.1 is available online in bothlanguages at http://www.ennonline.net/ife/view.aspx?resid=4 or follow the link from theENN homepage at http://www.ennonline.net

Print copies in French (funded by theWHO) and in English are available from theENN (postage and packaging charges apply,negotiable for recipients in developing coun-tries). To save on distribution costs, regionalsources of print copies in French are beingestablished, including in East and West Africaand in the US. Module 2 is also included on a CD of IFEresources developed for a regional workshopin Bali (see news piece) and CD copies areavailable from the ENN, at a cost of £2 each(plus P&P).

For requests, contact the ENN, tel: +44 (0)1865 324996/249745 or email: office@ennonline.net

1 The IFE Core Group is an interagency collaborationconcerned with the development of policy guidance andcapacity building on IFE. Current members are UNICEF,UNHCR, WFP, WHO, IBFAN-GIFA, CARE USA, Action Contrela Faim and ENN. Associate members are SC UK and IFRC.See at http://www.ennonline.net/ife

Updated Moduleon IFE availablein English andFrench

In response to the global food crisis, a pressrelease on 1 May 20081 detailed a call byPresident Bush to the US Congress to providean additional $770million to support food aidand development programmes. This is in addi-tion to $200 million in emergency food aidrecently made available via the Bill EmersonHumanitarian Trust (at the AgricultureDepartment). This extra funding would gotowards existing emergency food aidprogrammes and agricultural developmentprogrammes in developing countries.

Acknowledging that the US must change theway food assistance is delivered, there was alsoa call to approve a proposal submitted toCongress in 2008, to purchase up to 25 percentof food assistance directly from farmers in thedeveloping world. The need for the US admin-istration to work with others was also high-lighted, including securing food aid commit-ment from other G-8 countries and workingtowards the conclusion of a successful DohaRound agreement, to reduce and eliminatetariffs, other barriers, and market-distorting

1 http://www.whitehouse.gov/news/releases/2008/05/20080501-22.html2 Johannesburg, 5 May 2008 (IRIN)

US and Canadian responsesto global food crisis

subsidies for agricultural goods. The US presi-dent also urged that countries lift restrictions onagricultural exports and remove barriers toadvanced crops develop through biotechnology.

Canada is the latest major donor country to‘untie’ its food aid, by removing restrictions onwhere the food can be purchased2. It will nolonger insist on sending domestically grownfood but will instead provide aid agencies withcash, giving them the flexibility to source cheaperfood in the region or beneficiary country.

This leaves the United States of America, theworld largest donor of food, as the only devel-oped country with tied food aid. Canada alsoannounced a donation of US$230 million tofood aid programmes, joining the UnitedKingdom, the European Union and Japan, whohave also pledged significant amounts to dealwith the current crisis.

Dates: 9 – 27 June 2008

This three-week masters-level certificate held atthe Centre for International Development(CIHD) in London, UK is the only advanced-level international training course on breast-feeding and related topics available worldwide.It is held annually in collaboration with WHOand UNICEF.

The course has been revised to provide anup-to-date scientific, technical and practicaltraining on all aspects of breastfeeding, includingpolicy and programme implementation. It istargeted at doctors and other health or alliedprofessionals involved in national or local infantfeeding programmes. Teams from the samecountry or facility are particularly welcome. Thecourse takes an international perspective and isrelevant for participants from all types ofeconomies. The course is limited to 30 partici-pants and is conducted in English.

The full three week Certificate course is intwo parts, which can be taken together (recom-mended) or separately:

Part 1: Clinical Management and Public Health(9-20th June) provides a comprehensive in-depth study of breastfeeding, including physi-ology, practical skills and counselling andhospital/community visits to meet mothers andbabies.

Part 2: Addressing Challenges to OptimalPractice and Implementation (23-27th June)provides a series of themed study days oninfant and young child feeding, includingComplementary Feeding, HIV and Infant andYoung Child Feeding in Emergencies (IFE).

The IFE study day (Wed 25th June) is beingcarried out in collaboration with the EmergencyNutrition Network (ENN) and will be open toexternal candidates working in the emergenciessector.

Further details and a registration form areavailable atwww.ich.ucl.ac.uk/education/short_courses/courses/2S24 or contact the Course Director, Sandra Lang,email: sandra.lang1@virgin.net

Breastfeeding: Practice and PolicyCertificate Course

The new version of the Complex Emergency Database (CE-DAT)website has been launched. It offers an interface that allows usersto access CE-DAT data through a map portal and a timeline tool.The CE-DAT Complex Emergency Monitor also provides anoverview of the nature of mortality in complex emergencies.

Access the website at http://www.cedat.be

The latest issue of the CE-DAT Scene newsletter is available at:http://www1.cedat.be/Documents/Newsletter/newsletter_apr08.pdf

CE-DAT website launched

19

News

On 29 April 2008, a press release1 detailed a unifiedUnited Nations (UN) response to the recentdramatic escalation in food prices that is unfoldingas a worldwide crisis. This followed a meeting ofthe Executive Heads of the United Nationsspecialised agencies, Funds and Programmes andBretton Woods institutions2, in Bern on 28 and 29April 2008. Under the chairmanship of the UNSecretary-General, a common strategy was agreed insupport of developing country governance toconfront the global food crisis.

To meet immediate emergency requirements, theChief Executives Board (CEB)3 called upon the inter-national community to fund urgently and fully theemergency requirements of US$755 million for theWorld Food Programme (WFP), deliver on itspledges and provide maximum flexibility to targetthe most urgent needs.

The CEB also asked for action to be taken tocounteract the effects of escalating energy, fertilizerand input prices by providing developing countryfarmers with the support required to ensure thenext harvest. To address this:• The Food and Agriculture Organisation (FAO) Emergency Initiative on soaring food prices has called for US$ 1.7 billion in funding to provide low income food deficit countries with seeds and inputs to boost production.• The International Fund for Agricultural Development (IFAD) is making available US$200 million to poor farmers in the most affected countries to boost food production by providing essential inputs.• The World Bank is exploring a rapid financing facility for grant support to especially fragile, poor countries and quicker, more flexible financing for others.

To address short to medium term challenges, the UNsystem will co-operate to provide for development ofemergency safety nets and social protection of themost vulnerable and rapid employment and incomegeneration programmes.

At the country level, UN Resident andHumanitarian Coordinators, Heads of the WorldBank missions, and the UN country teams will

urgently meet with humanitarian agencies in affectedcountries to draw up support strategies for nationalgovernments and vulnerable populations and seekinternational support for their implementation.

The CEB has called upon countries where exportrestrictions on food have reduced supplies andcontributed to price hikes, urgently to reconsiderthose policies.

In the medium to long term, the UN system willbring together its technical and analytical capabili-ties to fill research and knowledge gaps in order tosupport governments with the best information foragricultural decision-making to boost productionand productivity. The CEB called for a rapid conclu-sion of the Doha Development Round to result inscaling down trade distorting subsidies that havedamaged developing countries production capacity.

To address the long term challenges, the CEBurged that structural and policy issues, includingchallenges posed by climate change to productivesystems that have substantially contributed to thiscrisis, be urgently addressed. Further research mustbe undertaken on the impact of diversion of foodcrops to bio-fuel production and all subsidies tofood-based bio-fuels should be reviewed.

The specific needs of Africa as the most affectedregion should be addressed.

The World Bank, International Monetary Fund(IMF), IFAD and Regional Development Banks andrelevant agencies of the UN system will collaborateto develop a long-term strategy. In order to create aprioritised plan of action and coordinate its imple-mentation, a Task Force on the Global Food Crisis isbeing established immediately under the leadershipof the Secretary-General bringing together theHeads of the United Nations specialised agencies,Funds and Programmes, Bretton Woods institutionsand relevant parts of the UN Secretariat.

1 http://www.reliefweb.int/rw/RWB.NSF/db900SID/EGUA-7E6MKV?OpenDocument2 The Bretton Woods Institutions are the World Bank and theInternational Monetary Fund (IMF). They were set up at a meet-ing of 43 countries in Bretton Woods, New Hampshire, USA inJuly 1944.3 http://www.unsystemceb.org/

A short 4-page update for the book ‘Caring forSeverely Malnourished Children’ has beenprepared to reflect some important developments inthe management of severely malnourished childrensince it was first published in 2003. In particular,the update includes community-based care andthe use of ready-to-use therapeutic foods andenriched family foods, and antibiotic treatment ofchildren with HIV.

The update is being inserted into all copies of thebook sold by Teaching-aids At Low Cost (TALC), which costs £4.00(£5.35 with an accompanying CD) plus postage and is available fromTALC, P.O. Box 49, St Albans, Herts, AL1 5TX, UK. Fax: +44 1727 846852, email: info@talcuk.org website: http://www.talcuk.org

The update is also available in e-format from TALC, email: info@talcuk.org, or the authors, email: Ann.Hill@lshtm.ac.uk orannpatriciaburgess@yahoo.co.uk.

The Centre for Public Health Nutrition atthe University of Westminster, London isrunning its short course, Nutrition inEmergencies, again in London, UK (1-5September 2008).

The five-day course provides partici-pants with technical knowledge and up-to-date information regarding the deliv-ery of nutritional support to those affectedby humanitarian emergencies. The topicscovered on the course include food secu-rity, malnutrition, supplementary andtherapeutic feeding programmes, includ-ing Community Therapeutic Care, andother interventions.

Participants will explore types ofmalnutrition, their direct and underlyingcauses, how they are assessed, andcommon nutritional interventions.Participants are drawn from a wide rangeof backgrounds, although the course isparticularly aimed at nutritionists,doctors, nurses and programme managersworking in developing countries.

The fee for the course is £650. There aretwo full scholarships available – to qual-ify, applicants should be from a develop-ing country, employed in the sector in anarea prone to disaster. More scholarshipdetails are online at the following linkhttp://www.wmin.ac.uk/scholarships/page-9319 (scroll down the page to‘Other University of Westminster schol-arships’ section).

For a course brochure, go tohttp://www.wmin.ac.uk/sih/page-979 or contact Mark Armstrong on tel: +44 (0)20 7911 5883 or email: armstrm@westminster.ac.uk

The next Public Health in Complex Emergencies training will take placein Bangkok, August 11-23 2008 and in Uganda, November 3-15 2008.

This course is designed for NGO, INGO, UN and Ministry of Healthstaff working in humanitarian assistance programmes providing healthand health-related services. Sectors covered include nutrition, epidemiol-ogy, reproductive health, psychosocial issues and coordination.

For more information on courses and for details regarding online applica-tions, check the PHCE website: http://phcetraining.org

Contacts for specific trainings are:Bangkok: Asian Disaster Preparedness Centre, Thailand. Tel: (66-2) 516 5900 ext 351, Fax: (66-2) 524 5360 524 5350, email: janette@adpc.net Website: http://www.adpc.net (click Trainings and Workshops)

Uganda: Makerere University School of Public Health Tel: (256-41) 543872, (256-41) 263158/9 Fax: (256-41) 531807, email: P_nalubega@yahoo.co.uk Website: http://www.musph.ac.ug (click Upcoming Events)

Update for ‘Caring forSeverely MalnourishedChildren’ book

Public Health in ComplexEmergencies training

Unified United Nations response to theglobal food price challenge

Nutrition inEmergenciesshort course

20

Views

The Quality COMPAS method andDynamic COMPAS software areproject and information manage-ment tools for humanitarian proj-ects. They will be the subject of ashort training course to be held atGroupe URD’s head office inPlaisians (Provence, France) from23rd to 27th June 2008. The coursewill be conducted in English. Thedeadline for applications is 13thJune 2008.

The course has been designed fornational and international aidworkers involved in projectmanagement activities such asneeds assessment, design, monitor-ing, self-evaluation and evaluation.Drawing its content from theCOMPAS method, the course willcover subjects such as conducting asituation analysis which goesfurther than a simple needs analy-sis, designing a project beyond thelogical framework, defining objec-tives and indicators in keeping withall the quality criteria, developingand implementing a monitoringsystem, understanding the differ-ence between monitoring and eval-uation, etc. The course is organised arounda case study that gives participantspractical experience of qualitymanagement and using theCOMPAS method and the DynamicCOMPAS software. To downloadthe software and/or for more infor-mation on COMPAS, visit:http://www.compasqualite.org/en/index/index.php

Good value accommodation andfood is available to participants.For more information and torequest a registration form, contact:Pierre Brunet, Training UnitCoordinator, Groupe URD, LaFontaine des Marins, 26170Plaisians, FranceTel/Fax: 00 33 (0)4 75 28 29 35 email: formations @urd.orghttp://www.urd.org

With so many reports on climate changecoming out from different quarters, itcan be hard to know which are basedon reliable information. Those of us

who work in the food and nutrition sector do havereal cause for concern, as it is highly probable thatthere will be a general increase in food insecurity inmany of our most vulnerable populations over thecoming years. Additionally, it is very likely that therewill be an increased frequency of natural disasterscaused by extreme weather, including flooding, heatwaves and drought.

Sourcing objective informationThe Intergovernmental Panel on Climate Change(IPCC) was set up in 1988 to review objectively andconsolidate the latest scientific literature on thesubject. In 2007, they released their FourthAssessment Report (AR4)1 based on 29,000 observa-tional data series from 75 studies around the world(though data from developing countries are seri-ously lacking). The AR4 reflects that global warmingis a reality, the average air temperature has warmedby 0.74o C (0.56-0.9 o C) over the past 100 years. Thisis an average global figure and doesn’t truly reflectlocalised temperature variations. It is also feared thatthe rate of global warming is increasing - eleven ofthe twelve years to 2006 rank among the twelvewarmest years on record since 1850. Land regions arewarming faster than the oceans and the northern lati-tudes are warming faster than southern ones withthe average arctic temperature increasing at twice theglobal rate. There are also increased sea levels(global average 1.8, 1.3-2.3 mm per year 1961-2003).The rate of increase of the sea level accelerated in thelast decade of the century but it is not yet clear if thiswas due to decadal variation or not. The world’sthird largest natural disaster of recent years, (afterthe tsunami of 2004, Pakistan earthquake of 2005)was, in fact, a heatwave that killed over 64,000people2 in southern Europe in July/Aug of 2003.

Observed changes in the climate‘Warming of the climate is unequivocal as is nowevidenced from observations of increases in global averageair and ocean temperatures, widespread melting of thesnow and ice caps, and rising global average sea level ‘(IPCC, 2007).

Warm air is able to hold more water vapour thancooler air and this means that warm air is less likelyto provide rainfall until it cools. Cooling, however,results in very heavy, localised rainfall coming fromthe water laden skies. Simplistically, this means hotregions become hotter and drier whilst cooler onesbecome warmer, wetter and with a higher potentialfor flooding. The rate of global warming is thoughtto be increasing at a rate of 0.2 degrees centigrade/decade.

The AR4 observes that trends in precipitationhave changed, with eastern parts of North and SouthAmerica, northern Europe and northern and centralAsia having significantly higher rainfall in contrastto lower rainfall in the Sahel, Mediterranean, south-ern Africa and parts of southern Asia. Globally, thearea affected by drought has almost certainlyincreased since the 1970s. There has also been an increase in the frequencyand/or intensity of extreme weather events over thepast 50 years. Hot days and nights have becomemore frequent and cold days and nights lesscommon. It is likely that heat waves are morefrequent over most land areas, that the frequency ofheavy precipitation has increased over most areas,and that the incidence of extremely high sea level hasincreased at a wide range of sites worldwide.

What does this mean for food security?Naturally, these climate changes have direct effectson agricultural production. It is anticipated that formoderate global average temperature increases (esti-mated between 1-3oC), there will be an overallincrease in global food production. Additionaltemperature increases, however, would cause anoverall fall in food production. So who are likely to be the winners as the climatechanges and who are likely to lose out? In verygeneral terms, some regions at lower latitudes willbecome hotter and drier with a shortened growingseason. Small scale and subsistence farmers will be atparticular risk. The AR4 has also confidentlypredicted that by 2020, rain fed agricultural produc-tion will fall by 50% in many African countries. Anumber of arid and semi-arid areas may simply fallout of agricultural production. In contrast, otherregions, in higher latitudes and including parts ofEurope and the western USA, will become warmerand wetter with an extended growing season. Thisprovides the potential of an increased level ofproduction, though producers may need to adaptand change their agricultural techniques and thetypes of crops grown. The ability of a country torespond to this may well depend on its preparednessand wealth.

Future changes in the climateIf predictions based on the current levels of globalwarming are realised, between 75 and 250 millionpeople will be exposed to increased water stress dueto climate change. By 2030, production from agricul-ture and forestry is projected to decline over much ofsouthern and eastern Australia due to increaseddrought and fire. By 2050, freshwater availabilitywill be decreased in Central, South, East and S EAsia. Coastal areas will be at increased risk of flood-ing from the sea or the river megadeltas.

Overall, this means that we can expect malnutri-tion levels to increase in some of the worlds mostvulnerable populations. Additionally, we need toanticipate more droughts, heat-waves and floods.This article has focused on the impact of climatechange upon food security through agriculturalproduction effects alone. However, the impact ofclimate change on other sectors like human healththrough changes in infectious disease vectors willalso impact food security but are difficult to quantify.

To contact the author, email:K.Godden@westminster.ac.uk.

Additional information, including the entire IPCCreport series, can be found at http://www.ipcc.ch

The Effects ofGlobal Warmingon Food SecurityBy Kate Godden

Kate is a Senior Lecturer inthe Centre for Public HealthNutrition at the Universityof Westminster where sheteaches food security,programme planning andemergency nutrition. Shehas over 10 years experi-ence within the emergen-cies sector.

1 IPCC 2007. Climate Change 2007. Synthesis Report. A Report ofthe Intergovernmental Panel on Climate Change. WorldMeteorological Organisation (WMO) and the United NationsEnvironment Programme (UNEP). Available at http://www.ipcc.ch 2 Centre for Research on the Epidemiology of Disasters. (CRED),http://www.cred.be

Unified United Nations response to theglobal food price challenge

COMPAS course in project management/evaluation

21

Field Article

The Ministry of Health and Ministry ofAgriculture of Malawi, in partnershipwith Action Against Hunger (ACF), havebeen running the Malawi Integrated

Nutrition and Food Security Surveillance(MINFSS) system since May 2003. The system wasinitially piloted in six districts, but currentlycovers the whole country. The MINFSS systemaims to provide information on nutrition trends ofchildren under five years, as well as the householdfood security situation in Malawi. Any significantdecline in the nutrition and food security situationdetected by the system should lead to a detailedinvestigation, using standard nutrition surveyand food security assessment methodologies.These would estimate the actual prevalence ofacute malnutrition or the level of food security.

The MINFSS system monitors nutritional statusof a sample of 9,100 children (350 per district) fromfive growth monitoring clinics (GMC) in each ofthe 26 districts in Malawi. These children arerandomly selected from a population of childrenattending the GMCs and therefore include healthy,malnourished, and sick children. Out of the 9,100children, 1,300 children (50 per district) areselected to gather household food security infor-mation. These same children are followed over theperiod of a year.

NUTRITION Information Collection andAnalysisIndicatorsAt each of the GMC sessions, the following infor-mation is collected from the 70 selected children:• Child Identification • Child age in months• Sex• Height (using a height board to the nearest mm)• Weight (using a Salter scale to the nearest 0.1 kg)• Mid-upper-arm circumference (MUAC) (using MUAC tape to the nearest mm)• The occurrence of acute diarrhoea in the past 2 weeks • Presence or absence of bilateral oedema

MethodologyFive health centres are selected within eachdistrict to ensure that all livelihood zones arecovered. The centres are called sentinel sites. Ineach sentinel site, a number of GMC sessions are

conducted with caretakers of children under 5years of age who regularly attend these clinics. Inorder to ensure all children attending a GMC havean equal chance of being selected, a randomsampling is carried out at all the GMC sessions ina particular month.

The MINFSS system aims to create an informa-tion channel from rural Malawi Health Centres andAgriculture Extension Development Officers(AEDOs) through to their respective ministries inLilongwe where data are entered and analysed.Results are reported to national and internationaldecision/policy makers. Data are sent to Lilongweat the end of each month and the 10th of the follow-ing month is the cut-off point for these data beingreceived at Ministry offices. These data are enteredinto the Nutrition and Food Security databases.

AnalysisThe anthropometric data are analysed withANALYNUT (see box) to produce anthropometricindices, using the National Centre for HeathSurveillance (NCHS) tables of reference. Theanalysis uses repeat measurements by pairing thesame child from the previous month in order tomake comparisons over time.

The weight for height z-score (WHZ)1, as anindicator of acute malnutrition, is of primaryinterest. The proportion of children with GAM(global acute malnutrition) and SAM (severe acutemalnutrition) are calculated. Data on trends inheight for age z-score (HAZ)2 are used as an indi-cator of chronic malnutrition or stunting. Thisinformation is disseminated once a year but can bemade available upon request. Other nutritionindicators, such as the proportion of children withlow mid upper arm circumference (MUAC), diar-rhoea and oedema are also calculated using theANALYNUT programme.

FOOD SECURITY Information Collectionand AnalysisIndicatorsOnce the sampling has been carried out, a baselinesurvey is conducted to gather household food

Integrated Nutrition and Food SecuritySurveillance in Malawi

By Elena Rivero, Núria Salseand Eric Zapatero

Elena Rivero is currently working for ActionAgainst Hunger as Surveillance Advisor inthe Malawi Integrated Nutrition and FoodSecurity Surveillance System (MINFSS).Since 2000, she has been involved inhumanitarian aid and nutritional develop-ment programmes in several countries,including Algeria, Colombia, CentralAmerica, Sudan (Darfur) and Lebanon.

Núria Salse is the Health and NutritionOfficer in Acción Contra el Hambre.Previously she spent several years workingon nutrition and medical programmes inAngola, Guinea Conakry, Niger andArgentina.

Eric Zapatero is the Food Security Officer inAcción Contra el Hambre. Previously hespent several years working on food secu-rity in Burundi, Tanzania, Ethiopia and theDemocratic Republic of the Congo.

The authors would like to acknowledge NeilFisher who designed, implemented and hasprovided continued support to developingthe systems methodology. They would alsolike to acknowledge Phil Mckinny, SilkePitsh, Brad Palmer, George Mvula , MajaMunk, Mercy Chikoko, María Bernandez,Jacob Asens, Rebecca Brown, Marta Valdésand Peter Hailey who contributed to launch-ing and improving the surveillance system.They would especially like to acknowledgethe work of the Ministry of Health and theMinistry of Agriculture of Malawi in imple-menting the methodology.

This article describes a surveillance system developed and implemented by AAH and partners in Malawi, that monitors nutrition and food security (through a foodstress index) in a sample of children under 5 years at sentinel sites. The development of user friendly software programmes for data analysis is facilitating the inte-gration of the system into government structures.

ANALYNUT can be used to calculate monthly trends by extracting the relevant information from themonthly data files and calculating the mean and standard deviation for each sentinel site that areentered onto a worksheet. As a second stage, ANALYNUT carries out an adjustment for missing sites. Theprogramme estimates a value for every child included at least once in the run of months for which theanalysis is set up. This is done for every missing site in a district. The programme essentially calculatesa best estimate of the value for each missing cell in a reiterative fashion using a missing data procedure.The programme provides a best estimate for the district based on the available data of previous months.For those districts with missing sites, the best estimates are only calculated for the mean, as confidenceinterval calculations need information on the number of children.

ANALYNUT is an ideal data analysis package from the point of view of capacity building of the Ministryof Health, since the programme is easy to use and can be run by individuals that have limited data analy-sis experience. The programme is purely Excel based (i.e. data entry, cleaning, flagging of biologicallyimplausible indices, analysis, and output is all carried out in an excel programme). The results arecomparable with other anthropometric data analysis packages such as EPI NUT contained within the Epi-Info programme (http://www.cdc.gov/epiinfo/) and Emergency Nutrition Assessment (ENA) Software(http://www.nutrisurvey.de/ena/ena.html).

More information on the ANALYNUT programme is available from the authors (see contacts details at theend of this article).

1 WHZ = (actual weight – reference weight)/reference standarddeviation2 HAZ = (actual height – reference height)/reference standarddeviation

22

Field Article

security related information. The main topicscovered are:• Structure of the household• Assets ownership• Land, crops and cultivation practices• Cash income and income sources• Loans• Food consumption and preferences• Sickness and health• Water and sanitation

A shorter questionnaire is used on a monthlybasis to monitor changes occurring in the house-hold. It includes:• Changes in the household (births, deaths, movements…)• Cash income and income sources• Food availability• Food consumption patterns • Shocks

MethodologyA sample of 10 children is selected from the list ofthe 70 children enrolled in nutrition follow-up.The selection is random using a sampling inter-val procedure. This child’s household is thenassessed via questionnaire each month toprovide information on the main food securityindicators. This sampling method allows fortracking changes over time and takes account ofthe need to combine nutritional and food secu-rity indicators in the same locality.

AnalysisThe analysis of the food security data isconducted using two programmes developed byACF for this project. These programmes weredeveloped in Excel using Visual Basic in order tomake it more accessible for users.

Surveyprogramv10 is used for:• Simple accumulation of coded categories• Mean, standard deviation, median and quartiles• Accumulation of coded categories with more than one per cell• Mean, median, etc with zero values ignored• Accumulation of string variables

It is used for cleaning the data and to calculatethe Food Stress Indexes. It can also be used tocross tabulate sets of data and calculate meansand medians, etc.

Monthstrendsv5 is the latest version of aprogramme that is used to calculate monthlytrends. It extracts the relevant information fromthe monthly data files, and assembles the meanfor each sentinel site onto a worksheet. In asecond stage, it carries out an adjustment formissing sites. Through the analysis of the data itis possible to follow several food security indica-tors each month in each district. It is also possibleto compare the data between districts, regionsand years.

Food Stress IndexThe nine indicators from the monthly question-naire which showed a discernible trend over thehunger period were combined into a ‘food stressindex’. This provided a useful summary of foodinsecurity for the 2004-05 hungry season andallowed comparison with 2003-04.

Food Stress Index mark 1(FSI1)Nine of the variables which showed significantdifferences between months were combined intoa food stress index. The percentage of house-holds reporting each indicator is shown in Table1. The mean of the nine indicators is the foodstress index mark 1 (FSI1). A value of zero wouldimply that all households:• had more than 50kg of cereal in store

• had root crops available• had a cash income greater than MK1000 per month• ate three meals per day• ate cereals every day• ate groundnut or other legume every day• did not have to undertake any form of food rationing.

At the other extreme, a value approaching 100would indicate that few households experiencedsuch conditions.

However a number of problems becameapparent with the FSI1 indicator. The FSI1 was tooheavily based towards maize whereas in someareas, other staple foods have a very importantrole, e.g. cassava in Nkhata Bay. Following discus-sion with the Malawi Vulnerability AssessmentCommittee (MVAC), it became apparent that newindicators needed to be included in the index. Forexample, ganyu, the system of casual pieceworkwhich sustains the most vulnerable non-food-self-sufficient households through the hungry season.Also, the emphasis placed on meal frequenciesand enforced rationing (reduced portions and‘whole days without a staple food’) was felt to betoo high.

Food Stress Index mark 2 (FSI2)Given the limitations of FSI1, a new index wasdeveloped, the ‘food stress index 2’ (FSI2). Eightindicators were combined into the FSI2. Adjustedweightings reduced the importance of foodconsumption indicators relative to the availabil-ity of starch food (cereals and root crops) andcash earnings. FSI2 is better able to deal withdistricts where cassava is one of the staple crops.It also includes information on ganyu, which wasnot available from the questionnaire in usebefore October 2004. The following eight indicators were proposed forthe new food stress index (FSI2):

1. The proportion of households that have very low supplies of starch staple food: less than 20 kg of maize, other cereal or dry cassava and no cassava or sweet potato ready for harvest (weighted 1.0).2. The proportion that have a potential shortage in the longer term: less than 50kg of maize, other cereal or dry cassava and no cassava or sweet potato ready for harvest in the next two months (weighted 1.0).3. Households with income less than MK1000 per month (weighted 1.0).4. Households having difficulty finding ganyu (weighted 1.0).5. 100 x (3 minus meal frequency) (weighted 0.33).6. Households who have not eaten groundnut or legume on the previous day (weighted 1.0).

7. Households reporting that they did not have enough food at some time in the month (weighted 0.33).8. Households going whole days without a staple food (weighted 0.33).

The first two indicators are closely connected butincluding them both emphasises the importanceof immediately available food to the index. Themeal frequency indicator operates on theassumption that most households would eatthree meals a day if they had adequate food.Occasionally households report 4 meals per daybut this is treated as 3 meals. Indicator 6 onlegume consumption is indicative of food qualityas well as changes in food access. Having exper-imented with various weights for the differentindicators it was decided to downgrade theimportance of the three indicators of foodconsumption (meal frequency, not enough foodand whole days without staple) so that the threetogether carry the same weight as any one of theother indicators.

Each month, a Bulletin is issued with theresults of the data analysis for both the Nutritionand the Food Security information. All reportsand bulletins are shared with the ministries andany other interested institution. Results are alsosent to the district representatives of theministries and presented at Nutrition and FoodSecurity meetings.

Challenges and ConstraintsPossible biases of the methodologyDefaulting is an issue that arises and may be asource of bias because the same children arefollowed over time. If a large number of chil-dren default, the sample may cease to berandom. However, for the purposes of trackingtrends over time, the system needs a broadlyrepresentative sample rather than a randomone. The statistical theory and methods areessentially those of time series rather than ofsurveys. All that is necessary is that the sampleshould be representative of the population.

Another potential source of bias is the natureof clinic attendees. For example, if there is atendency for especially sick children or childrenof sick mothers to attend the GMC, sick chil-dren will be over-sampled. However, a counter-balancing factor may be that that more healthconscious caregivers attend the GMC.

Another potential bias is introduced bysampling children at the health centres and notthe outreach clinics, so that more remote settle-ments are under-represented resulting in wealth-ier families being over-sampled. However thisdoes not appear to be the case. At the start ofimplementation, both static and outreach clinics

Table 1: Percentage of households reporting each indicator for Food Security Index mark 1 (2003/04)

Nov Dec Jan Feb Mar Apr

Less than 51kg in store 46 60 66 77 80 49

100-access to cassava/sweet potato 55 66 50 46 39 39

Income<MK1000 56 74 68 73 69 59

100*(3-meal frequency) 77 66 89 105 104 90

100-cereals taken 7 8 15 9 10 8

100-legumes taken 58 61 61 66 63 44

Reduced amount/meal 59 58 63 65 65 49

Reduced meals/day 53 59 62 60 61 47

Entire days without staple 23 28 24 22 18 12

Mean (FSI1) 48 53 55 58 57 44

Field Article

23

were used in each district for the follow up of thenutrition and food security situation. Following anevaluation of the pilot, it was concluded that therewas no significant difference between the popula-tion sampled from the outreach posts and thosesampled from static GMCs (for both Nutrition andfood Security).

Interview fatigueFamily’s tire of being asked the same food securityquestions each month without any kind of ‘reward’In addition, it is sometimes difficult for AEDOs tocomplete the questionnaires, resulting in low ques-tionnaire submissions rates.

Data returnDespite our best efforts, the dataset is at timesincomplete, as completed questionnaires fromspecific sites in specific months are not alwaysreceived in Lilongwe. There are two problems thataffect data return. First, there are not enough chil-dren to be measured/monitored. Secondly, not alldata are used due to poor quality. Insufficientnumbers of datasets received every month deter-mines that, currently, comparisons between regionsare more useful and give a more accurate picture ofthe situation than district by district comparisons.

DiscussionThe surveillance system implemented in Malawihas several important practical and statistical differ-ences from a survey based system. The same chil-dren are measured over months as a time serieswhereas surveys use a point time data analysismethodology. Surveillance is effectively a movingpicture while a survey is a snapshot of a movie. Thesample size for surveillance is different from asurvey. The pairing of the data (following the samechildren each subsequent month), requires a smallersample size to produce good statistical power andprecision in demonstrating significant changes overtime. Currently the nutrition surveillance system atits maximum, providing a sample size of 350-paireddata per district.

A priority aim of this surveillance system was tounderstand the linkage between nutrition and foodsecurity. Thus, a sub-sample of those households

with children monitored under the nutrition surveil-lance component was chosen for food security assess-ment monitoring. It was hoped that this would clar-ify the impact of the food security situation at house-hold level on the nutritional status of the children.

The Surveillance System is proving an importantsource of information for all stakeholders and keydecision makers. This was demonstrated during thefood crisis of 2005 when the Integrated Nutritionand Food Security System provided informationwhich allowed timely intervention when the situa-tion appeared to be deteriorating.

The surveillance system is also providing usefuland pertinent information to the MVAC3 in Malawi.It provides secondary data on food security andnutrition indicators which are used by the MVACfor the annual assessment report. This informationis provided at harvest time when the MVAC isanalysing the situation in Malawi. The system is alsoused by the MVAC to track the situation over timeand monitor the predictions made.

Strategies are currently in place to ensure asmooth and sustainable integration of the systeminto governmental structures. Comprehensivehandover to the Ministry of Health and Ministry ofAgriculture is in progress and will hopefully lead tofull Government ownership.

ConclusionsMINFSS applies statistical tests that are based onanalysis of variance of WHZ from repeat measure-ments on the same children. This pairing of the samechildren improves the precision of the analysis indemonstrating changes over time, reduces the needfor large samples and is more capable of showing achange in acute malnutrition over time. The foodstress index allows comparison of food securitybetween months (is it improving or faltering?),between years and between geographic units(regions, districts or livelihood zones). It does notpredict the situation that will emerge later in theyear nor does it provide an exact description of thesituation in each district. However, after some yearsof running the system, it should be possible topredict the scale of deterioration in a ‘hungry’season as there will be data over a number of yearson which to draw. For the sake of continuity of theinformation and in order to establish valid trendsover time, it is recommended that the same house-holds are selected for one year and a new set ofhouseholds selected thereafter on an annual basis.

For further information, contact: Núria Salse, email:nsalse@achesp.org and Eric Zapatero, email: ezap-atero@achesp.org tel: 00 34 91 391 53 00

Table 2: Components of the Food Stress Index mark 2

Weighting Oct Nov Dec Jan Feb Mar Apr

Starch not available now 1 6.0 12.6 15.0 21.0 20.1 14.5 6.7

Starch not available soon 1 10.2 18.0 16.3 24.1 19.4 14.5 9.6

Income < MK1000 1 52.7 55.7 56.4 59.3 57.8 62.4 56.9

Difficult to get ganyu 1 27.8 34.2 43.0 39.6 32.4 35.0 33.0

3 minus meal frequency 0.33 80.2 79.0 82.9 81.6 77.9 74.1 56.9

No legume previous day 1 59.0 65.8 73.3 78.3 66.1 56.5 53.0

Not enough food 0.33 54.8 59.3 61.9 67.0 61.1 56.0 41.7

Whole days without staple 0.33 4.2 4.7 11.0 9.5 11.2 6.0 4.5

Weighted average 33.7 39.0 42.6 45.8 41.0 38.0 32.3

FSI2 34.9 39.3 43.6 47.0 42.4 39.0 33.1

3 MVAC (the Malawi Vulnerability Assessment Committee) is aconsortium committee of government, NGO and UN agencies thatis chaired by the Ministry of Economic Planning andDevelopment. MVAC members have contributed to livelihoodzoning of Malawi. Livelihood zones feature baselines profiles thatare monitored and re-assessed yearly.

SAM inadequately addressed inthe Lancet Undernutrition Series

Dear Editor,

In 2003, The Lancet captured andfocused attention on saving children’slives with the publication of a five-partseries on child survival. In 36 succinctpages, this series described the“where” and “why” children aredying and highlighted the contribut-ing role of undernutrition to approxi-mately 50% of all deaths in childrenunder 5 years of age1. With the recentMaternal and Child Undernutritionseries, The Lancet continues itscommitment to shine the spotlight onnutrition as “a desperately neglectedaspect of maternal, newborn, andchild health.” The toll exacted bychildhood undernutrition is felt everyday by Médecins Sans Frontières(MSF) field workers in more than 60countries around the world.Development of nutrition programmesis an increasingly important focus ofour work. MSF has gone from treating100,000 malnourished children duringthe period from 2000-2005, to reaching150,000 such children in 2006 andagain in 2007.

MSF welcomes this series andcommends the efforts to emphasise thevital importance of nutrition forvulnerable populations, particularlychildren under 2 years of age. Theseries also highlights the glaringunder-allocation of resources to thisarea. However, we think the seriesmisses the mark in some importantareas; underestimation of diseaseburden of severe acute malnutritionand failure to acknowledge thetremendous potential of new treat-ment strategies.

Underestimation of the diseaseburden of severe acute malnutritionand the associated mortalityIn paper 1 of the series, the burden ofdisease is estimated, by prevalence, tobe 19 million (3.5% of 555,729,000)children suffering from severe acutemalnutrition (SAM), and 178 million(32% of same population) stunted chil-dren2. As SAM is a transitory condi-tion, with known seasonal variationsin many parts of the world, a cross-sectional survey ‘snapshot’ at any onepoint in time will certainly lead to anunderestimation of the caseload. Inaddition, no mention is made in paper1 (or any of the other papers in theseries) of kwashiorkor, a commonform of SAM in central and easternAfrica, further underestimating theburden of SAM. According to the series, SAM isidentified as a significant cause ofdeath, responsible for 449,160 deathsper year among children in develop-ing countries. This estimate is substan-tially lower than the 1 million SAM-attributable deaths reported in the UNJoint Statement on Community-basedManagement of Severe Acute

Letters

Measuring MUAC in a GMC session

E Rvivero, ACFS, Malawi, 2007

24

Malnutrition3 and again raises questions aboutthe validity of calculating mortality fromprevalence surveys.

The risk of death associated with SAM isreported as 9.4 times higher than for a non-wasted child, yet the global case fatality ratecalculated using the authors’ data is only 2% -this with only a minute fraction of the 19million children receiving treatment for SAM.In 2007, MSF’s nutrition project in Maradi,Niger, treated 22,250 children for SAM usingWHO standards for admission criteria. Thisprogramme reported an 83% recovery rate,while case fatality and defaulter rates were2.7% and 8.8%, respectively (G. Harczi,personal communication). Thus, even witheffective and proven treatments, case-fatalityrates are not as low as suggested in paper 1.

Despite the increased risk of death associ-ated with SAM, however, the undernutritionseries focuses its attention on stunting, andinterventional strategies targeting the 36 coun-tries that account for 90% of this condition.Countries such as Turkey are included in thislist, while Chad, Somalia, Central AfricanRepublic, Sierra Leone, Liberia, and Haiti arenot. This type of analysis prioritises the needsof the less undernourished over those mostimmediately at risk of death.

Failure to acknowledge the major impact ofCommunity-based Treatment of SAMTreatment of SAM is not a ‘key message’ inpaper 3, despite its huge potential to savelives. Relying solely on data from randomisedcontrolled trials (RCTs), the authors limit theirrecommendation to treat SAM in facility-basedprogrammes instead of endorsing community-based management. This fails to accuratelyreflect current reality of practice and does notacknowledge community-based care strategiesusing ready-to-use therapeutic foods (RUF) asan enormous advance in the ability to treatthis lethal condition and its potential to reachvast numbers of children.

In paper 3, the authors give several reasonsfor not widely recommending community-based care. Firstly, their intent is to discussintervention effects in national populations,not in “special circumstances of crisis.” Thisdifferentiation ignores the fact that, althoughinitially developed and implemented for moreeffective response in emergency relief4, outpa-tient management of SAM is gaining favourwith national governments. Ethiopia, Malawi,Sri Lanka, and Niger are institutionalisingcommunity-based SAM management proto-cols, with many more countries following theirlead. The vast majority of children affected bySAM do not live in areas of conflict or “specialcircumstances,” but rather live in families thatare fundamentally food insecure.

Secondly, the authors of paper 3 cite thelack of robust, randomised studies as a reasonfor not endorsing community-based care. Inweb table 5, the authors present their searchcriteria, which led to the selection of RCTs,including those with historical controls. Theirsearch methods returned 19 articles addressingthe use of RUF in community management ofacute malnutrition, of which only 5 wereconsidered suitable for analysis. Of the 14studies excluded, 5 were observational studies,one of which enrolled 2,131 severely malnour-ished children5. However, large-scale observa-tional studies can provide equivalent or better

evidence, especially when randomised studiesare small and inadequately control forconfounding variables6.

Thirdly, the authors of paper 3 focus onreduction of case fatality rate as the basis formaking their recommendations. Thus, facility-based care according to WHO protocols isrecommended for implementation (Table 1 ofpaper 3) because it is deemed effective atreducing the case fatality rate for children withSAM, compared with those not treated accord-ing to this protocol. Yet none of the fivepublished studies cited and used for thispooled analysis (Ahmed, Ashworth, Deen,Falbo, Wilkinson) for supporting improvedfacility-based care used randomisation. Thesestudies used historical controls from the samefacility or non-random controls from caseseries. Furthermore, no mention is made of thecoverage or defaulter rates associated withfacility-based treatment. In MSF’s experiencein Niger, exclusive facility-based treatment hasbeen associated with substantially higherdefault rates compared with combined orstrictly outpatient care: 28% vs 16.8% and5.6%, respectively7. Thus, actual mortalityfrom stand-alone facility-based programmes islikely to be underestimated.

Compared with resource-intensive, facility-based inpatient management, community-based care is a practical, large-scale interven-tion, as evidenced by ~60,000 children treatedin Niger in 20068, and >26,000 treated inMalawi, Ethiopia, and Sudan in 2001-20069.Coverage of >70% with community-basedtreatment, versus <10% in facility-basedprogrammes, suggests community care has amuch greater impact at the population level10,with improved outcomes11. Furthermore, it isimportant to remember that community-basedtreatment includes WHO protocol facility-based care; it is an effective strategy because itoffers inpatient care for complicated SAMwhile achieving high patient coverage throughan outpatient network, and minimisingdefaulter rates by decreasing opportunity costsfor mothers.

While the authors of the undernutritionseries are to be commended for their rigor inestablishing a solid evidence base for makingrecommendations, the fundamental questionremains: what rules of evidence should beapplied? Of the 10 original papers cited asreferences for the UN Joint Statement onCommunity-based Management of SevereAcute Malnutrition, four were expresslyexcluded from consideration in the SystematicReview of Management of Childhood SevereMalnutrition for the undernutrition series.Among these, there is a controlled trial withsystematic allocation, a clinical trial, a retro-spective study and a review (paper 3, webtable5). It may be time to re-examine the rules ofwhat constitutes valid evidence.

The three studies from the UN JointStatement that are included in the paper 3SAM review are RCTs. Although RCTs havebecome the gold standard for clinical decision-making, the same standard is not sufficient forpublic health decision-making where the path-ways from intervention to impact are multipleand complex12 Evidence-based public healthdecisions must therefore rely on a variety ofdata types, not just from RCTs, ranging fromhighly controlled efficacy trials, to observa-

tional studies with control or comparatorgroups, to the reporting of results obtainedfrom large-scale programmes in differingcontexts.

Lastly, many of the interventions recom-mended for the 36 countries in Table 1 ofpaper 3, e.g. hand washing and behaviourchange communication, are not supported bystudies showing an effect on nutritional statusor mortality. Different standards thus seem tobe used to recommend some interventions andnot others.

Moving forwardThe Lancet Undernutrition series has takenimportant steps forward in acknowledgingand beginning to define the problems withnutrition for women and children under 2. Butsome of the recommended interventions–exclusive breastfeeding, complementary feed-ing practices and hand washing - rely onbehaviour change, presuming that mothers caneasily change their busy, overburdened lives.Furthermore, Vitamin A and zinc supplementsaddress only part of the nutritional deficits.

As the Lancet series points out, funding iswoefully inadequate. If the nutrition commu-nity is serious about addressing the crisis ofmaternal and child undernutrition, it must notbe limited by inexpensive solutions that inter-vene on the margins. Funding must increaseand programming developed to ensure thatenergy-dense, nutrient rich foods or supple-ments get into the mouths of young children.Scaling up community-based treatment ofSAM is a major step in the right direction, butit is not sufficient. A child should not have todeteriorate to the point of severe wasting to“qualify” for nutritious food.

Susan Shepherd, MDNutrition AdvisorAccess Campaign for Essential Medicines

1 Black, RE et al. Where and why are 10 million childrendying every year ? Lancet 2003; 361: 2226-2234.2 Black, RE et al. Maternal and Child Undernutrition : Globaland regional exposures and health consequences. Lancet20083 WHO, WFP, UNSCN, UNICEF. Community-based manage-ment of severe acute malnutrition. Geneva: World HealthOrganisation, 2007.http://www.who.int/nutrition/topics/statement_commbased_malnutrition/en/index.html (accessed on May 12, 2008).4 Collins S, Sadler K. Outpatient care for severely malnour-ished children in emergency relief programmes: a retrospec-tive cohort study. Lancet; 2002. 360: 1824-30.5 Linneman Z, Matilsky D, Ndekha M, A large-scale opera-tional study of home-based therapy with ready-to-use thera-peutic food in childhood malnutrition in Malawi. Matern ChildNutr. 2007 Jul;3(3):206-215.6 Shrier I, Boivin J-F, Steele RJ, et al. Should meta-analysesof interventions include observational studies in addition torandomized controlled trials? A critical examination of under-lying principles. Am J Epidemiol 2007; 166: 1203–9.7 Gaboulaud V, Dan-Bouzoua N, Brasher C, Fedida G,Gergonne B, Brown V. Could nutritional rehabilitation at homecomplement or replace centre-based therapeutic feedingprogrammes for severe malnutrition? J Trop Pediatr. 2007;53: 49-51.8 Defourny I, Seroux G, Abdelkader I, Harczi G. Managementof moderate acute malnutrition with RUTF in Niger. FieldExchange 2007; 31: 2–4.9 Collins S, Sadler K, Dent N, et al. Key issues in the successof community-based management of severe malnutrition.Food Nutr Bull 2006; 27 (suppl 3): S49–82.10 Collins S, Dent N, Binns P, Bahwere P, Sadler K, Hallam A.Management of severe acute malnutrition in children. Lancet2006; 368: 1992-2000.11 Ciliberto MA, Sandige H, Ndekha MJ, et al. Comparison ofhome-based therapy with ready-to-use therapeutic food withstandard therapy in the treatment of malnourished Malawianchildren: a controlled, clinical effectiveness trial. Am J ClinNutr 2005; 81: 864–70.12 Victora CG, Habicht, JP and Bryce J. Evidence-Based PublicHealth: Moving beyond randomized trials. Am J Public Health;March 2004; 94, 3. 400-405.

25

Challenging conclusions and analysis ofthe Lancet Undernutrition Series

Dear Editor,

We welcome the opportunity provided by theLancet to highlight the “desperately neglected”subject of nutrition through its series on maternaland child undernutrition. We appreciate thesubstantial scope of the series as well as thescientific rigour applied in producing it. We findthe analysis in many instances to be boththoughtful and perceptive. However, we takeissue with some of the major conclusions of theseries and the fact that certain areas of analysishave been over-simplified and poorly developed.

There are four key conclusions in the serieswhich we would like to briefly comment upon:• There is a lack of evidence about effective programming.• There are a small group of ‘single’ interven- tions that are effective in reducing undernu- trition. • There are a small number of countries who have successfully reduced undernutrition. • There is a lack of leadership and ‘fragmen- tation’ in the international nutrition system.

Lack of evidence baseThe authors cite a limited evidence base forprogramme effectiveness but, at the same time,throw out a great deal of data because of failureto meet criteria for admissible evidence. Thevast bulk of the acceptable evidence is for singleinterventions, which address the immediatecauses of malnutrition. Single interventions areeasier to evaluate and demonstrate impact onmalnutrition levels than multi-sectoral interven-tions. Randomised control methodologies maybe applied to the former, but are very difficult,in practical terms, to apply to the latter.

We feel that there has been a lack of cautionfrom authors in highlighting which interven-tions are ‘proven’ to be effective. The impressiongiven is that unless there is evidence of effec-tiveness using strict scientific criteria, an inter-vention should not be implemented. For exam-ple, it is stated in paper 4, page 1, “Nutritionresources should not be used to support actionsthat have not been proven to have a direct effecton undernutrition....” This would effectivelymean that all emergency food aid programming(with the exception of foods used in therapeuticcare programmes) should cease as there is nofirm evidence base for its effectiveness (Duffieldet al 2004 ).

What we would like to have seen addressedis the question why there haven’t been moreevaluation studies which meet the scientificcriteria (randomised control trials) set by theLancet? The obstacles to such evaluation needto be understood, in particular the need todevelop more rigorous and operationally feasi-ble methods to evaluate the impact of multi-faceted approaches. Scientific evaluation ofimpact needs to be planned at the outset ofprogrammes. Donors need to be aware and will-ing to provide the resources needed to fundrigorous scientific evaluation. Currently, donorfunding mechanisms for interventions tend tofavour agencies with a ‘track record’ for deliver-ing outputs rather than outcomes (for examplethe delivery of food rather than the impact ofthe food on undernutrition). The focus onoutputs rather than outcomes at donor levelreflects administrative pressures and the lack ofnutritional expertise within donor agencies.

We would have liked the Lancet to highlightthe lack of epidemiological expertise, withindonor and implementing agencies, to developand implement appropriate methodologies forimpact evaluation.

Another key factor which allows this ‘statusquo’ to prevail is the lack of institutionalaccountability for programme outcomes. Thereis no single agency with a mandate or responsi-bility to ensure that nutrition interventionimpact is assessed and, therefore, no overviewwithin the system as to which programmes ordesign features deliver the desired impact.Consequently it is left to entrepreneurial indi-viduals or agencies to amass such evidence andultimately push for better designed interven-tions. Such an ad hoc situation is unsupport-able. While the Lancet series highlights thisdeficiency (paper 5), there is no institutionalanalysis of how to remedy this situation. Animportant opportunity for advocacy has there-fore been missed.

Small group of effective interventionsThe Lancet identifies 14 interventions of whichthree (breastfeeding counselling, vitamin Asupplementation and zinc fortification) have thegreatest proven benefits in terms of reducingstunting and micronutrient deficiencies. Most ofthese 14 interventions are ‘single’ interventions(eight on micronutrient supplementation orfortification, three on promotion of good infantand young child feeding through counsellingand behaviour change, one on treatment ofsevere malnutrition, one on intervention toreduce tobacco consumption and one on hand-washing and hygiene intervention).

The reviewers in paper 3 report that theyhave “analysed large-scale nutrition programmes, toderive estimates of population effect, achievablecoverage levels and sustainability” but it is unclearwhat conclusions have been drawn about thesuccess or otherwise of multi-sectoral nutritionprogrammes. Only one large-scale nutritionprogramme met the criteria for prospectiveevaluation (Progressa in Mexico) and theauthors conclude that “given the paucity of effec-tiveness data, strengthening of monitoring andrigorous assessment of large-scale nutritionprogrammes are imperative”.

Furthermore, certain interventions wereexcluded from the analysis (education, untar-geted economic strategies or those for povertyalleviation, unconditional cash transfers andmicrocredit programmes etc.) Overall, the impli-cation is that the short route of single interven-tions is the way to reduce undernutrition. Thereis no discussion about the added benefit of link-ing interventions together (for example growthmonitoring alone cannot impact on nutritionalstatus but linked with vaccination, vitamin A,iron/folate supplementation, breastfeedingcounselling, de-worming, may well have animpact). Neither is there discussion about link-ing short-term with longer-term strategies,though implicit throughout the series.

We were disappointed by the failure of analy-sis. We would at least like to have seen an analy-sis of funding decisions in relation to nutritionprogramming. For example, has funding beenbiased in favour of short term, single interven-tions, such as micronutrient supplementation,and not on longterm investment in governmentsand longer-term multi-sectoral programmes?Small group of countries whereundernutrition has been reducedA number of countries are reported to have

successfully reduced undernutrition. These areCosta Rica, Cuba, Sri Lanka, Thailand, Mexicoand, more recently, China. However, the analy-sis as to how these improvements occurred islimited and is largely descriptive. Crucially, thelink between the 14 nutrition interventionsidentified in paper 3 and the country levelimprovements is not made explicit. For exam-ple, did improvements in nutritional status inChina arise mainly due to micronutrient supple-mentation and improved rates of breastfeedingor are more complex factors or longer termfactors at play and if so, what are these?

Furthermore, there is no analysis at a countrylevel of the ‘how’ in relation to nutritionprogramming rather than the ‘what’. We wouldhave liked the ‘how’ question to have beenaddressed in this series. How have some coun-tries been able to reduce undernutrition whilestill remaining poor, while in other countries theopposite seems to have occurred? What social,political and funding mechanisms were neededand how much capacity development wasnecessary to achieve these reductions?

Lack of leadership and fragmentationThe lack of responsibility for, and leadership in,international nutrition is recognised and the‘fragmentation’ in the sector identified. Butthere is no analysis of why or recommendationsfor what can be practically done.

What is lacking in this section is an historicalanalysis of past initiatives to improve coordina-tion and coherence in the sector and why thesehave petered out over time. For example, thereis no analysis of why post-InternationalConference on Nutrition initiatives evaporated.What de-railed the process?

Analysis of the nutrition ‘architecture’ forresponding to emergencies is especially weak.Generalisations about the politicisation of foodaid are poorly supported by the evidencepresented, while the issues highlighted are aregurgitation of the recent literature and addlittle insight into ways forward. The authorscould have made far better use of historicalanalysis and positive examples to indicate howto make progress where this is needed.

In conclusion, we were disappointed by thesuperficial analysis of the Lancet series in anumber of important areas especially in light ofthe efforts put in to the quantitative analysis.We feel that there has been a lack of answers tocritical questions relating to the lack of anevidence base, the ‘how’ of nutrition program-ming, funding decisions and accountability.

What most concerns us is what a seniorgovernment figure from a developing countrywould make of the conclusions. Will s/heconclude that national nutrition policy shouldbe re-written to focus only on micronutrientsupplementation/fortification and promotion ofgood infant and young child feeding practices?What other policy and strategy insights has theLancet series given us to support improvementsin nutrition in what is acknowledged to be this“desperately neglected” area which claimsmillions of young lives every year?

Regards

Carmel Dolan and Fiona Watson,NutritionWorks

1 Duffield A et al (2004). Review of the Published Literaturefor the Impact and Cost-effectiveness of Six NutritionRelated Emergency Interventions. Report prepared by theENN; December.

Letters

26

Field Article

This article presents a pilot study1 conducted in Somaliland by the UN Foodand Agricultural Organisation Food Security Analysis Unit, (FAO/FSAU).It uses the Lot Quality Assurance Sampling (LQAS) method to assess thenutritional situation and compares the results to a 30x30 cluster surveyconducted simultaneously in the same sample population.

The North West Region of Somalia, whose borders follow thoseof the former British Somaliland Protectorate, declared itselfindependent (Republic of Somaliland) in May 1991 (see Map1). However, the declaration of independence is not recog-

nised in other parts of Somalia or internationally. The region has,nonetheless, managed to avoid the protracted conflict and violencethat has afflicted much of southern Somalia. Hargeisa town is thebiggest urban centre in the region and is the capital town of Somaliland.Hargeisa is the concentration for public administration private sector,and hosts large numbers of persons from the international aid commu-nity. Due to its dynamic market and labour opportunities, it is also thedestination for many refugees, returnees and internally displacedpersons (IDPs). Hargeisa currently hosts more than 80,000 IDPs.

Since 1997, large numbers of UNHCR’s ‘official’ returnees toSomaliland from camps in Ethiopia have selected Hargeisa as theirchosen destination. Several returnee settlement areas have sprung upon the outskirts of the city and have grown considerably sinceUNHCR began its voluntary repatriation programmes in 1997. Manyreturnees have been settled in permanent settlement areas, namelySheikh Nur, Mohammed Mooge, Aw Aden and Ayaha. Most othersremain in temporary settlements in three different areas (State House,Stadium, A and Daami) and many more are unaccounted for, scatteredin different sites within the municipality. Another poor settlementcalled Sheedaha is also emerging in the northern part of Hargeisatown, mainly inhabited by the urban poor who are unable to pay forrents for better housing in town. These comprise pastoral dropoutswho lost their animals in previous droughts and are seeking alterna-tive livelihoods, and those who have currently been displaced fromthe renewed conflicts in Mogadishu.

Nutritional statusThe returnees and IDPs have been identified as nutritionally vulnera-ble over the past few years, reflected in a series of nutrition surveysthat have found rates of global acute malnutrition (GAM) over 10%.These high rates result from inconsistent and limited income generat-ing opportunities and lack of basic services such as sanitation,protected water and health care. Routine assessments are conductedon this population to monitor the nutrition situation and informappropriate response. In September 2007, the Food and AgriculturalOrganisation (FAO) Food Security Analysis Unit (FSAU) with UNICEF,Ministry of Health and partners, conducted one 30x30 cluster nutritionsurvey and two exhaustive surveys in the three most concentrated andprotracted IDP populations in Somaliland - Hargeisa, Berbera andBurao. The opportunity was also taken to trial an alternative samplingmethodology, Lot Quality Assurance Sampling (LQAS), to assess itssensitivity in determining the nutrition situation.

LQAS and prevalence of malnutritionLQAS is a method of sampling derived from the manufacturing indus-try and, over the last two decades, has been also applied to the healthsector (see Box). More recently, adaptation of LQAS principles hasbeen explored in relation to estimating the prevalence of GAM. Thesampling method traditionally used to assess the prevalence of acutemalnutrition in emergencies is a 30x30 cluster survey. This methodprovides statistically reliable results if implemented correctly but witha sample size requirement of 900, it can be time-consuming and expen-sive to carry out. FANTA (Food and Nutrition Technical Association2)has been exploring the use of LQAS as a rapid and cost-effective alter-native for assessment of the prevalence of acute malnutrition. A studyby FANTA, Catholic Relief Services (CRS), and Ohio State University(OSU)3, field-tested the use of the LQAS designs in an emergencysetting in Ethiopia4. The study concluded that LQAS designs providestatistically appropriate alternatives to the more time-consuming30x30 cluster survey, though the variance is larger, resulting in widerconfidence interval round the point prevalence, and additional fieldtesting was recommended.

By Tom Oguta,Grainne Moloneyand Louise Masese

Tom Oguta has been working with FAO/FSAU in the NutritionSurveillance Project in Somalia as a Nutrition Project Officer for the lasttwo years. Previously, he worked as a Research Officer at KenyaIndustrial Research & Development Institute and has a specific interestin sampling approaches in nutrition assessments.

Grainne Moloney is the Nutrition Project Manager of the FAO/FSAUNutrition Surveillance Project, where she has worked for the last 18months. Prior to FSAU, Grainne worked with UNICEF in Darfur, also onnutritional surveillance, and previously with Oxfam and Action AgainstHunger UK in many different countries and contexts.

Louise Masese works with FSAU as a Nutrition Project Officer, planningand coordinating nutrition assessments. Prior to FSAU, Louise workedwith a consultancy firm as a survey specialist, with WFP’s KenyaEmergency Drought and Flood Operation and with UNICEF Kenya.

The authors would like to express sincere thanks to the FSAU nutritionteam in the implementation and write up of this report, to the partnersUNICEF and the Ministry of Health Somaliland, to the local authorities ofSomaliland for all the logistical support in the implementation of thestudy and to the community of the IDP camps, who participated in thisstudy.

Piloting LQASin Somaliland

1 The results of this study have been published in the FAO/FSAU monthly NutritionUpdate, August 2007.2 More information on LQAS is available on the FANTA website: www.fantaproject.org3 Deitchler et al (2007). A Field Test of Three LQAS Designs to Assess the Prevalence ofAcute Malnutrition. International Journal of Epidemiology. May 2007.4 Field testing LQAS to assess acute malnutrition prevalence. Summary of publishedresearch. Field Exchange 31. September 2007, p4

Measuring weight andheight of a child.

FSAU

, Som

alia

, 2007

27

FSAU PilotIn September 2007, FSAU undertook to pilotand field-test the LQAS methodology amongthe IDPs in Hargeisa, in order to compare thefindings with a 30x30 assessment and exploreits potential application in the nutritionsurveillance system in Somalia.

There are two sampling options for the esti-mation of the prevalence of acute malnutritionusing LQAS; a 33x6 cluster design, where 33clusters are selected with 6 children per cluster,and a 67x3 cluster designs, where 67 clustersare selected each with 3 children. Althoughboth methods provide similar precision in rela-tion to estimation of malnutrition, the latter hasshown a higher precision for household leveldata. However, as 67x3 clusters requires moretravel time, FSAU opted to trial the 33x6approach.

MethodologyA 33x6 cross-sectional LQAS assessment wasconducted alongside a standard 30x30 clusterassessment, amongst the protractedlydisplaced populations concentrated in sevensites5 in Hargeisa town of Somaliland. Bothstudies used the same sample frame and wereconducted by two separate teams concurrently.

For both studies, a two-stage clustersampling methodology was used to select theclusters and the households. In the case of theLQAS, 33 clusters were drawn from thesampling frame, while 30 clusters were selectedfor the 30x30 cluster surveys. The clusters wererandomly selected using the Nutrisurvey6 soft-ware. The recommended SMART (StandardisedMonitoring and Assessment of Relief andTransitions) method7 was used for the secondstage sampling of households and children.The same tools were also used for both studies,with quantitative data collected through a stan-dard household questionnaire for nutritionassessment. This included data on childanthropometry, morbidity, vitamin A supple-mentation, measles and polio immunisationcoverage, dietary diversity, and access to waterand sanitation. Qualitative data were collectedthrough focus group discussions and keyinformant interviews by an interagency teamcomprised of assessment supervisors and coor-dinators, to provide further understanding ofpossible factors influencing nutritional status.

For the LQAS study, as all eligible childrenin a sampled household were assessed, this

resulted in a total of 204 children. For the 30x30surveys, a total of 905 children were sampled.Two households overlapped between the twostudies. As only 198 children were required forthe LQAS analysis, the six extra children wererandomly eliminated at the analysis stage usinga table of random numbers. For both studies,household and child data were entered,processed (including cleaning) and analysedusing EPI68 and Nutrisurvey software.

Thresholds and Decision Rule (DR) for LQASLQAS is a hypothesis test to determine whetheran outcome is ≥ or < a defined threshold. LQASuses two thresholds, an upper and a lowerthreshold, to define the alpha and beta (α andβ) errors (tolerable statistical error)9. Therefore,to design an LQAS sampling plan, the thresh-old of interest for an indicator (e.g. GAM preva-lence), and tolerable statistical error (α and β)are defined in advance. The upper threshold isthe threshold at which the area is at risk (e.g.≥10% for Hargeisa IDPs). The alpha error is theprobability of incorrectly classifying an area asnot being at risk when the true GAM preva-lence is ≥the threshold of interest. For thepurposes of this pilot study, the lower thresh-old is the GAM prevalence at which an areawould not be considered a priority for an inter-vention (e.g. ≤ 5%) and the beta error is theprobability of incorrectly classifying an area asbeing at risk when the GAM prevalence is < thethreshold of interest.

The null hypothesis assumes the GAM preva-lence is ≥ the upper threshold. To classify GAMprevalence as ≥ or < the upper threshold, thenumber of children with GAM is counted andthen compared against a Decision Rule (DR)determined using binomial probabilities10.GAM prevalence is judged as ≥ the upperthreshold if the number of children with GAMin the sample is > than the DR. Similarly, GAMprevalence is judged as < the upper threshold if

Lot Quality Assurance Sampling (LQAS)

LQAS is a method of sampling derived in the1920s for assessing quality of lots (or batches) ofproducts in the manufacturing industry. By the1980s, the LQAS sampling concepts were recog-nised as having universal applications and theapproach is now being used all over the world toassess coverage of maternal and child health,family planning and HIV/AIDS programmes, qual-ity of health workers performance and diseaseprevalence. It is based on the principle thatinspection of a small, representative sample of a‘lot’ will, with high probability, allow for thevalid rejection of the entire lot, should thenumber of defective goods in that sample exceeda predetermined allowable number.

A global review (covering a total of 805 LQASsurveys) of the use of LQAS surveys to assessaspects of health care including service delivery,health behaviour and disease burden was carriedout by the World Hleath Organisation (WHO) in2006. LQAS surveys were found to be a practicalfield method increasingly applied in the assess-ment of preventative and curative health serv-ices and for measuring variation in behaviourchange when collected recurrently at multiplepoints in time. Most LQAS surveys have beenused to assess risk factors for HIV/AIDS and sexu-ally transmitted infections, although substantialnumbers have also been conducted to assessimmunisation coverage, growth and nutrition,and post-disaster health status of communities.

5 Ayaha, Aw Aden, Sheikh Nur, Daami, Mohamed Mooge,Stadium, and State House.6 http://www.nutrisurvey.de/7 The method involves a modification of the standardExpanded Programme on Immunisation (EPI) method toreduce the bias in sampling the households in the centre ofthe settlement8 http://www.cdc.gov/epiinfo/Epi6/ei6.htm9 Megan D et al (2007). A field test of three LQAS designs toassess the prevalence of acute malnutrition. InternationalJournal of Epidemiology. May 2007, pp 2-5.10 Valadez J, Weiss B, Leburg C and Davis R. (2003).Assessing Community Health Programs: Using LQAS forbaseline surveys and regular monitoring. London: Teaching-aids at low cost.

Field Article

WHZ: Weight for height z score HAZ: Height for age z score WAZ: Weight for age z score a

Measuring MUAC

FSAU

, Som

alia

, 2007

Table 1: Decision Rules for 33X6 LQASdesign (n=198) for various GAM thresholds

Decision Rule –Number of childrenidentified as acutelymalnourished in a33x6 LQAS study

Outcome

≤ 13 GAM is < 10%

> 13 but < 23 GAM is ≥ 10%

> 23 but < 33 GAM is ≥ 15%

> 33 GAM is ≥ 20%

Table 2: Comparative Analysis of findings using 30x30 standard cluster and 33x6 LQAS methodologies

Indicator Prevalence 95% CI CI Width Standard Error Design Effect

30x30 33x6 30x30 33x6 30x30 33x6 30x30 33x6 30x30 33x6

Global acute malnutrition (GAM) (WHZ) 10.3 9.6 8.4 – 12.2 6.1 – 13.1 ± 1.9 ± 3.5 0.06 0.08 0.91 0.72

Severe acute malnutrition (SAM) (WHZ) 1.1 1.0 0.5 – 1.7 0 – 2.4 ±0.6 ±1.4 0.003 0.007 0.70 0.98

Stunting (HAZ) 19.2 19.2 16.3 – 22.2 12.7 – 25.7 ± 3.0 ± 6.5 0.09 0.10 1.30 1.42

Underweight (WAZ) 18.0 20.7 14.8 – 21.3 15.0 – 26.4 ± 3.3 ± 5.7 0.06 0.08 1.68 1.02

Reported diarrhoea 15.3 13.6 11.7 – 19.9 7.9 – 19.4 ± 1.8 ± 5.8 2.07 2.94 2.91 1.46

Reported acute respiratory tract infection 15.4 19.2 11.3 – 19.4 11.9 – 26.5 ± 2.3 ± 6.3 2.08 3.71 3.00 1.74

Reported febrile illness 3.0 2.5 1.2 – 4.7 0.5 – 4.6 ± 1.4 ± 2.1 0.90 1.06 2.53 0.90

Suspected measles 2.2 1.6 0.82 – 3.64 0.0 – 3.9 ± 1.3 ± 2.3 0.72 1.17 2.03 1.65

Measles immunisation 58.3 57.7 52.3 – 64.3 43.6 – 71.8 ± 3.4 ± 14.1 3.08 7.20 3.33 4.01

Vitamin A supplementation 60.9 43.4 53.5 – 68.2 33.4 – 53.5 ± 3.4 ± 10.1 3.74 5.13 5.33 2.12

Polio immunisation 89.4 93.9 86.4 – 92.4 88.7 – 99.2 ± 2.2 ± 5.3 1.53 2.70 2.23 2.53

Dietary diversity (>3 food groups) 80.9 90.6 76.5 – 85.3 84.1 – 97.0 ± 3.8 ± 6.5 2.23 3.30 1.63 1.35

Access to safe water 84.7 97.2 75.9 – 93.5 94.2 – 100 ± 8.8 ± 3.0 4.50 1.55 7.82 0.94

Access to latrine 68.4 74.5 58.2 – 78.6 65.1 – 82.5 ± 10.2 ± 9.4 5.18 4.73 6.25 1.25

28

11 Only 13 acutely malnourished children are required tohypothesize GAM levels of ≥10%.

In May 2007, Save the Children UK (SCUK) surveyed all of its fieldprogrammes operating in countrieswhere Central Emergency Response

Funds (CERF) had been granted over theyear. This was a follow up to a publishedposition paper on CERF in January 2007,which argued that unless non-governmentalorganisations (NGOs) achieve direct accessto CERF funding, the CERF would neverachieve its full potential. The January paperhighlighted several concerns, including thespeed of forward disbursements, lack oftransparency, problems in United Nations(UN) /NGO relations in the field and poorcommunication.

Of the 37 countries that had receivedCERF funding in 2007 at the time of thesurvey, SC UK was operational in seventeen.Of those 17, SC UK had successfully appliedfor grants in two countries – Mozambiqueand Liberia. The Somalia response team alsobenefited from CERF, through CERF fundedUN cargo flights that assisted SC UK reliefprogramming in Hiran region.

Currently, each UN agency requires usingits own sub-agreement form with NGOs.Also, overhead costs still need to be negoti-ated on a case-by-case basis that takes time.Experience in Mozambique revealed thatCERF did not provide the critical jump-startfunding for either UN agencies or NGOs inthe wake of the flooding in February 2007.According to the Inter-Agency real-timeevaluation of the response to the February2007 floods and cyclone in Mozambique,there would have been considerable delaysin assistance had UN agencies and NGOsnot had their own seed money to begin oper-ating right away. The report also states that“the most difficult area of the CERF was theapplication for funds for NGOs via theCERF”.

According to the SC UK survey it is prob-lematic that UN agencies do not adequatelykeep track of when and how much CERFfunding is passed through to NGOs.Without these data, it is impossible to gaugethe impact of CERF funding on beneficiaries.To its credit, the CERF Secretariat has intro-duced new reporting requirements whichshould improve the transparency. The UNand the International Organisation forMigration (IOM) are now required to submitan annual report, including lessons learnedon accessing funds from the CERF andanalysing its impact. TheHumanitarian/Resident Coordinator(HC/RC) annual reports are meant to docu-ment forward disbursements to NGOs.

Another finding from the survey is thatthe relationship between the UN agenciesand NGOs in the field varies greatly. Mutualsuspicion and a lack of communication canprevail in certain contexts, and a lack oftransparency feeds this problem. SC UKfound that there was confusion (and scepti-cism) amongst Country Directors about the

UN agencies’ motivations for applying forCERF funding. There was a perception thatthey use the CERF to ‘top up’ their ownbudget shortfalls, which is not the intendeduse of CERF. In two countries, SC UKCountry Directors made clear how frus-trated they were at being excluded from thedecision-making processes. In both cases,they were assured they would be consultednext time.

Several of the SC UK programmes weresimply unaware that CERF funding hadbeen granted to the countries in which theywork. This points to not only a lack ofcommunication but also to the state of UN-NGO relations.

SC UK, Oxfam and representatives fromUN agencies attended a CERF Training ofTrainers in April 2007, which helped financestaff understand better the mechanisms,limitations and opportunities with the CERF.Trainings from the Humanitarian LiaisonSupport Unit in the UN Office for theCoordination of Humanitarian Affairs(OCHA) have also helped the wider NGOcommunity – particularly field based staff –know more about the CERF.

SC UK makes a number of recommendationsregarding the CERF:• The CERF Secretariat should establish pre-approval procedures or standardised Letters of Understanding (LoU) between the CERF Secretariat and NGOs with proven competency. Alternatively, a stan- dardised global LoU would also be useful between any UN agencies and NGOs with proven capability.• The CERF Secretariat and UN agencies must continue to improve the transpar- ency of CERF funding, including regular public reporting of the speed and impact of CERF funding for humanitarian programmes on the ground. There should be clear guidelines for timing, minimum percentage of funds to NGOs and percentage of overheads. • The CERF Secretariat should pursue a project tracking system, to track funds all the way from CERF to beneficiaries. • The CERF Advisory Group and UN member states must hold UN agencies accountable to higher levels of transpar- ency and speed for disbursal of funding from the CERF, in order to improve this important mechanism and better demon- strate impact for beneficiaries on the ground. A real-time evaluation of CERF’s speed and transparency should be piloted.• The Inter-Agency Standing Committee (IASC) Working Group and CERF Secretariat should navigate a way to allow NGOs with established capabilities direct access to CERF funding, in order to improve the timeliness and effectiveness of humanitarian response.

the number of children with GAM is ≤ than the DR.This is illustrated in Table 1 based on a 33x6 LQASstudy. Here, if more than 13 children but less than24 children are identified as acutely malnourished,then the global acute malnutrition rate can be esti-mated to be 10-15%.

Results and DiscussionOverall, the 33x6 LQAS design produced more orless similar results compared to the conventional30x30 design for the child data (malnutrition,morbidity and health programmes coverage) butless correlation for the household data (householddietary diversity, access to water and access tosanitation facility). However, as expected giventhe smaller sample size of the LQAS design (198children), the standard error is larger, and conse-quently the confidence intervals are wider for theLQAS results. The design effects were generallylower for the LQAS design (See Table 2).

Analyses of the findings from the two assess-ment designs provide similar estimates of acutemalnutrition. An acute malnutrition rate (WHZ<-2 and/or oedema) of 9.6% (CI: 6.1 – 13.1) wasreported using the LQAS (33x6) design.Comparable results (confidence intervals overlap-ping) were reported from the conventional 30x30design with acute malnutrition rates (WHZ<-2and/or oedema) of 10.3% (CI: 8.4 – 12.2). Forhypothesis testing against a threshold of below orabove GAM of 10%, 19 children were found to beacutely malnourished, indicating levels above 10%11

and <15% according to the DR shown in Table 1.

ConclusionsThe LQAS approach required fourteen less persondays and cost approximately 60% less ($5,600compared to $13,700) than the 30x30 approach. Afurther benefit was to the staff who reported beingless tired and more motivated.

Given these findings, there would appear to bea role for LQAS in the nutrition surveillancesystem for Somalia – especially for filling informa-tion gaps during the seasonal assessments and forareas with limited accessibility. In addition, giventhe chronically high rates of acute malnutritionreported in Somalia, this method could be used toidentify hot spots and help prioritise interven-tions where the nutrition situation has deterio-rated significantly. There is a considerable amountof baseline nutritional information in Somaliawhich would allow for this type of comparison.

However, for interpretation purposes, moreemphasis will be needed on adapting the decisionrule approach to provide a range estimate ratherthan an absolute prevalence estimate, given thesmall sample and the wide confidence intervals.This may, in turn, limit the use for monitoring thesituation over time except when a significantchange in the nutrition situation has occurred.

FSAU is planning to conduct further studiesusing the LQAS approach in less secure and acces-sible areas in Southern Somalia in 2008. As withthis study, parallel standard surveys will beconducted to allow for comparability of theresults. The results of these studies will bereported in the monthly Nutrition Updates.

For more information, please contact: Tom Joseph Oguta or Grainne Moloney, P.O. Box1230- 00621, Nairobi, Kenya. Tel: 254-20-3745734/1299 or 0722392499 Fax: 254-20-3740598 email: tom.oguta@fsau.or.keor grainne.moloney@fsau.or.ke

SC UK’s experiences with CentralEmergency Response Funds (CERF)Summary of evaluation report1

1 SC UK (2007). Save the Children’s Experience withCERF in 2007. May 2007. Available fromhttp://www.scuk.org

Evaluation

29

their role in decentralised evaluation. Theirevaluation function involves public critique ofprogramming while they must also ultimatelymaintain collegial day-to-day relationshipswith Country Offices.

CredibilityAlthough WFP has an evaluation policy, it is alayered series of documents that detracts fromclarity and applicability. Evaluation policy isnot sufficiently used to guide practice.Evaluators and Regional Bureaux have beenunclear regarding what is expected in terms ofquality, due to a lack of specification withinOEDE itself and concerns that headquartersexpectations do not take into account resourceand time constraints in the field. On the whole,OEDE evaluation function is impartial with theviews of all stakeholders often sought.However, there appears to be an unevenemphasis on stakeholders who are more acces-sible and articulate, with beneficiary views, inparticular, under represented. The process ofpreparing for evaluations, management, andadvising and supporting teams in the field ishandled in a highly professional manner byOEDE. Terms of reference are generally of goodquality, but they are, at times, too standardisedand over ambitious. The quality of evaluationsis mixed but should improve with the plannedestablishment of new quality standards byOEDE. A failure to take into account the costimplications of recommendations, togetherwith factors related to the nature of prioritysetting in WFP, has damaged credibility ofevaluations among some WFP staff.

UtilityEvaluation is insufficiently integrated intomany of the processes by which WFP sets,monitors and analyses policies. Evaluation isprimarily focused on outputs, as opposed tooutcomes and impact, which reflects thedemands of many stakeholders. In a narrowsense of contributing to an understanding ofhow to ‘do things right’, evaluation makes anotable contribution to programme design andmanagement. In a wider perspective of learn-ing about ‘doing the right thing,’ performanceis not so good. There have been some effortswithin evaluations to present evidence that canstimulate greater reflection within WFP - overthe changing role of food aid, for example.However, the corporate view of evaluation hastended to focus primarily on its utility formaking modest adjustments to existingapproaches. New OEDE plans, to tie evalua-tion closely to logical frameworks, mayenhance utility through a focus on outcomes;but lack of prevailing understanding and use oflogical frameworks within WFP will make itdifficult. At decentralised levels there is a closelink to utility, since there is a direct desire to useevaluation to inform and justify newprogrammes and phases.

Evaluation makes an inadequate contribu-tion to overall knowledge building within WFPand virtually none among partners. Access toreports and findings through website, debrief-ings, etc, is acceptable, but promotion of the

Arecently commissioned peer reviewof the evaluation process and func-tion within the World FoodProgramme (WFP) has been

completed and posted on the WFP website2.The review assessed the central evaluationfunction, i.e. the Office of Evaluation (OEDE) ofWFP, as its starting point, but also includedanalysis of decentralised evaluation in WFP.The review was based upon preparatory deskstudy, field visits to WFP regional bureaux andcountry offices, a meta-evaluation of twelveOEDE evaluations, a web-based survey of theviews of WFP staff (87 responses) and peerpanel interviews with selected stakeholders.

The panel concluded that the independenceof WFP evaluations was adequate in compari-son to similar organisations, however the cred-ibility of products was uneven. While theprocess of evaluations was somewhat morecredible, they were also problematic. The crite-ria of utility of evaluations were partially metwith regard to contributing to programming,but structures and mechanisms to promoteutility were weak in most other respects. Inaddition, OEDE is a strong unit with commit-ted, well-trained and highly motivated staff.Over the past seven years, OEDE has investedmuch effort on improving evaluation.However, evaluation is of more variable qual-ity at the level of Regional Bureaux andCountry Offices, with levels of motivation andinvested resources dependent upon the interestand priorities of the offices concerned.

IndependenceEvaluation resources are currently safeguardedwhile OEDE is outside of line managementwhilst, at the same time, sufficiently integratedinto WFP leadership structures to facilitateimpact. However, accountability for the imple-mentation of recommendations is unclear.Some OEDE staff are concerned that theircareers may be affected by their evaluationrole, which could lead to inappropriate riskaverse behaviour in their management of sensi-tive evaluations. There are also insufficientsafeguards to prevent partiality and conflicts ofinterest amongst external evaluators. The roleof Regional Bureaux in both oversight andadvisory support to Country Offices has prob-lematic implications for the independence of

use of evaluation products is not sufficientlyproactive.

RecommendationsOEDE should develop an evaluation policy asa transparent vehicle for promoting greatercommunication among internal and externalstakeholders, regarding the aims and intendedutility of evaluations.

OEDE should develop an ‘accountabilitymap’ of key WFP stakeholders, both internaland external, to help in clarifying roles andresponsibilities.

OEDE should look for ways of promoting,and providing incentives for staff to adoptmore participatory approaches to evaluation.Engagement with partners at country, regionalor global levels is primarily a responsibility ofother parts of WFP.

After an evaluation has been submitted tothe Executive Director, OEDE should not beinvolved with drafting or compilation ofresponses from different parts of the organisa-tion. The management response mechanismsshould include rules about the timeframe forthe response and procedures for follow up, aswell as for reporting to the Executive Board. Asimilar system for management responseshould be used for decentralised evaluations.Management response and follow up mecha-nisms should be transparent with relevantdocuments easily accessible for WFP and part-ners and routinely posted in electronic form.

The capacity of OEDE staff should be main-tained over time to stimulate interest in theevaluation field and encourage professionalism.

The learning element of evaluations shouldbe linked to a larger organisational knowledgemanagement strategy. OEDE should continuerecent efforts to systematically harvest lessonsfrom existing evaluations, as well as externalfora such as ALNAP3, the Inter-AgencyStanding Committee (IASC)4 and relevant part-ners.

OEDE should develop a more transparent,rigorous and competitive approach to selectionof team leaders. All evaluation teams shouldinclude at least one evaluation specialist,preferably the team leader.

In order to address concerns that only asmall portion of the overall evaluation budgetis within the direct control of OEDE, WFP’ssenior management should devise ways tosafeguard the funding allocated to evaluationsfor the next biennium. The establishment of acentrally managed fund for both OEDE evalu-ations and decentralised evaluations should beinvestigated.

Peer review of WFP evaluationprocess Summary of report1

1 Baker. J et al (2007). Peer review of the evaluation func-tion at WFP. Stockholm, November 5th, 2007. The report isavailable at http://www.uneval.org/indexAction.cfm?module=Library&action=GetFile&DocumentAttachmentID=2003 oron the WFP website (see footnote 2)2 http://wfp.org/operations/evaluation/doclist.asp?section=5&sub_section=8&EBDoc=WFP/EB.1/2008/7-A&Year=2008&ID=WFP1536523 Active Learning Network for Accountability and Performancein Humanitarian Action, http://www.alnap.org4 http://www.humanitarianinfo.org/iasc/

Evaluation

A motherwaits for WFPfood aiddistributionwith her childin Malawi

WFP

/Mik

ael Bje

rrum

, M

alaw

i, 2

006

30

Field Article

The current conflict in Darfur, Sudan startedin 2003. Since then, large-scale violence hasdecreased but fighting and attacks continueover large parts of the territory despite on-

going peace building efforts. Over 3.5 million peopleare estimated to be affected by the conflict, with morethan 1.8 million people having been displaced fromtheir homes.

Widespread looting and destruction of assets,displacement and restricted movement has had asignificant impact on people’s lives and liveli-hoods (farming, livestock herding, trade andmigration)1. The region, which was formerly self-sufficient in food except in unusually baddrought years, became a major recipient of foodaid. Most livelihood strategies are now restrictedand poorly remunerated, markets and trade areseverely disrupted and coercion and exploitationare deep-rooted2.

Despite food aid and other external assis-tance, the prevalence of global acute malnutri-tion (GAM) in the region is estimated at 16.1%,with the highest rates in North Darfur at 20.5%3.The malnutrition rates are found to fluctuateconsiderably in time and in space and an overallincrease in GAM from 12.9% in 2006 has beennoted in greater Darfur. This article focuses on the situation in AbuShok and As Salaam internally displaced popula-

tion (IDP) camps in North Darfur State, whereAction Contre la Faim (ACF) has implementednutritional activities since 2004. More specifically,it looks at the impact of a blanket distribution ofhigh energy biscuits to under five year olds,following the detection of extremely high acutemalnutrition rates in these camps in June 2007.

Situation in the campsAbu Shok and As Salaam IDP camps are situ-ated in North Darfur State, 4 km north-west ofEl Fasher town. Abu Shok camp was opened inApril 2004 with 42,000 IDPs registered. As themaximum capacity of Abu Shok was reached, anew camp, As Salaam, was set up within sightof Abu Shok in June 2005. The camps were offi-cially closed to new arrivals in November 2005,but sporadic registrations still take place. Thecurrent caseload is approximately 50,000people in both camps.

General food distributions are implementedby the Sudanese Red Crescent in collaborationwith the World Food Programme (WFP),providing the IDPs a full daily ration. WES(Water and Environmental Sanitation)/Government of Sudan, Unicef and Oxfam are incharge of water and sanitation programmes.Primary health care and kindergartens are alsopresent. ACF’s nutritional activities in bothcamps include supplementary feeding

programme (closed in October 2006), a thera-peutic feeding programme, involving a thera-peutic feeding centre and outpatient therapeu-tic programme, active case finding and nutri-tional surveillance5.

Malnutrition rates showed an overall declin-ing trend with some seasonal variations6, untilthe GAM peaked in June 2007 at 30.4%7, asshown in the Figure 1. This GAM rate is signif-icantly higher than the GAM rate found at thesame period during the previous year (22.8%).In addition, over 30% of the children screenedduring the survey were found to be at risk ofmalnutrition (80-85% weight/height percent-age of the median). There was no statisticallysignificant difference between the severe acutemalnutrition (SAM) rates. The extremely highJune 2004 malnutrition rates followed a measlesoutbreak, just after a major population displace-ment at the beginning of 2004.

Causes of malnutritionVarious causes of the increased GAM rates wereidentified by staff in the field and corroboratedthrough a Nutritional Causal Analysisconducted by ACF in the camps in November2007.The following associations suggestingimmediate and underlying causes of malnutri-tion have been identified8: • Malnourished households generally have a

Blanket BP5 distribution tounder fives in North Darfur

By Hanna Mattinen, ACFSince 2005, Hanna Mattinen has been FoodAid Advisor at the Action contre la Faim (ACF)headquarters, focusing on policy and opera-tional issues around food assistance andcash-based interventions. Previously sheworked with ACF as Food Security coordinatorin Guinea, Liberia, Chechnya/ Ingushetia,Nepal and Indonesia.

Many thanks to Philippe Crahay, Hanibal Abiy Worku and Loreto Palmaera (ACF FoodSecurity Coordinators in Darfur), David Mahouy (ACF Food Aid Officer in Darfur), EmilieCrozet (ACF Food Security Expert in Darfur), Sophie Laurence and Aurelie Fournier (ACFNutrition Coordinators in Darfur), Béatrice Mounier (ACF Nutrition SurveillanceCoordinator in Darfur), Dorothy Dickinson (ACF Nutrition Officer in Darfur) and theirteams, as well as to Olivia Freire, ACF Nutrition Advisor in Paris, all of whom have beeninvolved in implementing the project and collecting and analysing the data used in thisarticle. Thanks also to Rebecca Brown and Andrew Mitchell for proof reading.

1 Buchanan-Smith M and Jaspars S (2007). Conflict, campsand coercion: the ongoing livelihoods crisis in Darfur.Disasters, 31 (s1):s57-76. Blackwell Publishers, London. 2 See footnote 1.3 WFP, FAO, Unicef, MoA/GoS, MoH/GoS and CDC (2007).Food Security and Nutrition Assessment of the conflict-affected population in Darfur, Preliminary Results.August/September 2007. 4 Blanket under-5 distribution refers to a distribution thattargets children under 5 years of age (6-59 months) in agiven geographical area.5 For a detailed analysis of ACF nutritional programmes inNorth Darfur, please refer to the evaluation Action Contre laFaim nutritional intervention in North-Darfur 2004-2007carried out by B. Feeney, Valid International in March-April2007.6 The traditional hunger gap period falls roughly betweenMay and September/October each year.7 Action contre la Faim. June, 2007. NutritionalAnthropometric and Retrospective Mortality Survey, Childrenaged 6-59 Months. Abu Shok and As Salaam IPD Camps. ElFasher, North Darfur State. GAM reported as weight/heightbelow – 2 Z-scores and/or oedema; SAM defined asweight/height below – 3 Z-scores and/or oedema.

Figure 1: Evolution of malnutrition rates for children under 5 years (6-59 months) in Abu Shokand As Salaam camps

Bags of BP5packed, ready

to distribute

Dav

id M

ahouy,

ACF

Sudan

.

31

Field Article

from the distribution to promote exclusivebreastfeeding. A preventive approach wasfavoured as opposed to a curative approach dueto the large number of children at risk of malnu-trition. Targeted supplementary feeding washence not an option. SPHERE standards alsostipulate blanket supplementary feeding as aresponse when malnutrition rates are so highthat it may be inefficient to target the moderatelymalnourished and all individuals meetingcertain at risk criteria (e.g. those 6-59 months)may be eligible15. The blanket distribution was designed to becomplementary to the general food distribution,with a ration covering approximately 60% of theenergy needs of a child under five years of age16

until the end of the hunger gap period. Four 15-day distributions were organised betweenAugust and October 2007 targeting all the chil-dren in the camps. The end of the hunger gapwas considered the appropriate moment to endthe complementary distributions because tradi-tionally, food availability increases as theharvests start and food prices decline17. A subse-quent nutritional survey was planned after theend of the blanket distribution to measure theimpact and the potential need for a follow upintervention. The project was funded by theEuropean Commission Humanitarian Office(ECHO) and cost approximately 1.35 millioneuros18.

BP5 biscuits were chosen as the appropriatefood item due to their high energy and micronu-trient content and the fact that they could easilybe used as porridge to feed younger children.Additionally, previous ACF distributions of BP5in Darfur had demonstrated great acceptabilityamong beneficiaries. The biscuits were crushedand pre-packaged by ACF prior to distributions,in order to limit re-sale and to promote the use ofthe product as porridge. ACF also consideredSP450, but the product could not be sourced inadequate quantities in the required timeframe.Plumpy’nut and Supplementary Plumpy werediscarded as options as they were developed forthe treatment of malnutrition, not prevention.CSB-oil-sugar mix was not considered satisfac-tory, given its low acceptability and poorperformance in treating malnutrition in ACFsupplementary feeding programmes in Darfur19. ACF staff registered children under five at theoutset of the project with the help of local leaders(umdas and sheiks). The height of a child wasused as a proxy for their age20. In total, 15,337children were registered and over 150 MT of BP5was distributed during the project period. Large-scale information and awareness-raising sessionswere organised to explain the aim of the projectto the population. Special efforts were made to

As Salaam was reduced in January 2007. Variousreasons were put forward for the decision to cutthe ration – restrictions regarding the use ofgenetically modified maize in the country wereunder discussion, disruption in the availability ofCSB and reports of excessive sales of CSB by thepopulation. The reduction in the CSB rationtranslated into lower overall energy andmicronutrient10 content of the ration, as well asless availability of food items particularly suit-able for feeding children less than five years ofage. Focus group discussions with women imme-diately after the nutritional survey confirmedthat they lacked porridge to feed their childrenafter the ration was cut. However of note, theJune ACF nutritional survey did take placeduring the traditional hunger gap and at thepeak of preparations for the upcoming agricul-tural season when the workload of caretakerswas particularly high. Younger children wereoften left in the care of their older siblings. Inaddition, the survey was conducted during theschool holidays and some mothers mentionedthat during this period, the morning meal forchildren was given less importance11.

Water supply is also an issue of concern in thecamps as it depends substantially on handpumps, with 12 to 15 pumps having run dry oryielding a reduced volume of water over the past3 years12. Assessments show that only about halfof the water is used for drinking and bathingpurposes, whereas the rest is used for brickmaking and other livelihood activities13.However, no particular disease outbreak tookplace before or during the survey.

Blanket under-5 distributionOrganisationACF decided to implement a blanket food distri-bution to children between 6-59 months as thelack of adequate food for children14 was identi-fied as one of the key causes for the dramaticincrease in the prevalence of acute malnutrition.Infants under 6 months of age were excluded

lower income than non-malnourished families. Main income sources in the camp include urban casual work in the neighbouring town of El Fasher and to a lesser extent, rural casual work, as well as petty trading and sales of food items from the general food distribution. The income sources are largely similar for both malnourished and non-malnourished households.• The majority of households with at least one malnourished child were found to have settled in the camp more recently and were more likely not to have a general food distrib- ution card than families without malnour ished members.• Food diversity in families with a malnour- ished child is slightly lower, especially in

terms of the consumption of lentils, meat and Corn-Soya Blend (CSB) porridge and/or snacks.• Children in households with at least one malnourished child are more susceptible to diseases than those in families without malnourished children (level of hygiene was found to be lower).• Mothers with malnourished children have a slightly higher workload than mothers in households without malnourished children.• A lower proportion of mothers in families with malnourished children attend antenatal or postnatal consultations than mothers in families without malnourished children. • Families with malnourished children under five years were more likely to leave their children with their older siblings than those with malnourished children.• Exclusive breastfeeding until 6 months of age is practiced by less than 30% of the households interviewed in all camps9. Water and animal milk are introduced early in the diet of an infant (increasing the risk of diar- rhoea, particularly if the water is not clean).

In addition to these findings, the quantity of CSBin the general food distribution in Abu Shok and

8 Action contre la Faim (2007). Nutritional Causal Analysis. North Darfur Camps. North Darfur State.9 Caregivers were interviewed on breastfeeding practices and an event calendar was used to define the time period, whenneeded. It is acknowledged that this method may induce bias and allows only for a rough estimation.10 CSB is the most richly fortified item in the food ration. 11 ACF closed down its supplementary feeding programmes in Abu Shok and As Salaam camps in October 2006. While it isunlikely that this measure impacted the drastic increase in malnutrition rates, monitoring of admission numbers in suchprogrammes may have provided useful early warning information on the deteriorating nutritional situation. 12 Tearfund (2007). Darfur: Water supply in a vulnerable environment. Darfur, Sudan.13 ACF is currently rehabilitating a dam approximately 2 km from the camps and it is hoped that the water made available willbe used for livelihoods activities. 14 The lack of adequate food was particularly clear for children under 5 years, especially at the age when complementary foodswere being introduced.15 SPHERE (2004). ‘SPHERE. Humanitarian Charter and Minimum Standards in Disaster Response’. http://www.sphereproject.org16 The nutritional requirements of a child under five are estimated at 1,200 kcal/day, varying according to the child’s age. Theration of three BP5 bars a day provides 756 kcal and covers 63% of the daily requirements in energy and micronutrients. 17 The end of the blanket distribution was not conditional on reinstatement of the full CSB ration in the general food distributionbecause there were no clear provisions made for this. 18 This amount excludes the cost of a reserve stock and its transport, which was not distributed within the current project. 19 Caregivers noted dissatisfaction with CSB-oil-sugar mix as one of the reasons for defaulting in ACF supplementary feedingprogrammes in Darfur. 20 According to WHO standards, children measuring under 110cm are considered under 5 years of age.21 Action contre la Faim (2007). ‘Post Distribution Monitoring. Blanket distribution for children under 5 years old. Abu Shok andAs Salaam IDP camps’. El Fasher, North Darfur State.

Women having just received the BP5 ration

Dav

id M

ahouy,

ACF

Sudan

.

32

Field Article

ensure the involvement of umdas and sheiks,who could relay the message to the population,further promoting its acceptance. ACF nutritionscreening teams also organised hygiene educa-tion sessions, awareness raising on breastfeedingpractices and cooking demonstrations, andvisited homes in-between the distributions togive further information and assistance.

ImpactPost Distribution Monitoring, organised aftereach distribution round, showed that the numberof meals for children under 5 years increasedduring the project period and that almost all theBP5 was consumed within the household21. Lessthan 1% of the product was sold or exchanged.Intra-household sharing, as well as using BP5 tofeed infants under 6 months of age, which werecommon during the first round of the distribu-tion, decreased substantially due to additionalefforts put into sensitisation (Figures 2 and 3).Education sessions for women waiting in distri-bution lines and the time given to discuss the useof BP5 were important factors in limiting intra-household sharing.

A nutritional survey organised by ACF afterthe end of the blanket distributions in November2007 showed a radical improvement in the nutri-tional status22. The GAM rate in children 6-59months decreased from 30.4% to 14.3% and theSAM rate decreased from 2.8% to 0.6%. For thefirst time since the beginning of the conflict, themalnutrition rates dropped below the emergencythreshold. The GAM and SAM rates were foundto be significantly lower when compared to thesituation the previous year at the same timeperiod (respectively 22.6% and 2.7% inNovember 2006 – see Figure 1)23.

Caregivers reported during focus groupdiscussions that children were given a first mealof BP5 porridge in the morning and that childrenrarely asked for more food before the fatur(lunch). This is reportedly common when otherfood items (most commonly those provided inthe general food distributions) are used. In termsof satisfaction with the product, motherscompared it favourably to the traditionalporridge, madida, which was used prior to theconflict and was composed of millet, oil, sugar,milk and salt.

While the provision of adequate complemen-tary food during the peak of the hunger gapundoubtedly contributed to improvement of thenutritional situation, seasonal factors alsoimpacted positively. Overall food availabilityincreased with the onset of the harvests and theworkload for caregivers decreased. In addition tothis, sensitisation and awareness raisingcampaigns associated with the blanket distribu-

tion put child feeding and care practices high onthe agenda in the camps. Beneficiary satisfactionwith BP5 biscuits, their acceptance and under-standing of the aims of the project were also keyfactors in the success of the operation.

Conclusions and outstanding questionsThis experience from Darfur shows that timelyblanket distributions, which use appropriateproducts and are accompanied by sensitisationand awareness-raising, can be an effective meas-ure to tackle transitory peaks in malnutrition.Such distributions, however, suffer from twomain drawbacks - the distributions remain costlyand their impact, especially as a stand-aloneresponse, is not sustainable.

Cost is particularly high when foreign-produced, highly sophisticated and expensiveproducts are purchased and airlifted to the proj-ect area. In this case, roughly 35% of the budgetwas used for the purchase of BP5 and 30% forinternational air transport. However, there wereno appropriate local alternatives available at thetime of the crisis. Further research into local forti-fication and/or local production of supplemen-tary foods may provide future solutions that areless costly and more sustainable than the optionsthat are currently available. The organisation of aparallel registration exercise, two-week distribu-tion cycles and thorough monitoring alsoincreased the cost of the intervention, but werecrucial to ensure adequate quality. The role of food aid in saving lives andpreserving livelihoods in the current conflict inDarfur is widely recognised24. At the same time,it is clear that food aid alone, including blanketdistributions, does not provide sustainable longterm solutions for preventing malnutrition, evenif, in the short term, its impact can be life-saving.Food aid can have a sustainable impact on liveli-hoods, food security and ultimately malnutritiononly when the immediate, underlying and basiccauses of malnutrition are understood and tack-led. In the long term, security measures, landtenure and market issues as well as peace build-

Figure 2: Use of BP5 within the households

Figure 3: Sources of information for the distributions

ing at the community and national levels areessential for food aid to have a meaningful effecton livelihoods in Darfur25. Deep-rooted culturalpractices are among other causes of malnutrition.These require careful analysis and the develop-ment of long term approaches to achieve behav-ioural change. The current experience also highlights theneed for food security surveillance to rapidlydetect potential problems. A change in rationcomposition is one of the key early indicators inareas where food aid is the main source of foodand income. Targeted supplementary feedingprogrammes, where these are in place, may alsoplay a role in early warning through the monitor-ing of admission numbers.

Another question arising from this experienceis whether the ration composition in the generalfood distribution is adequate for all populationgroups in terms of macro- and micro-nutrientsand their bioavailability. The GAM ratesincreased to above 30% while the general fooddistributions were ongoing with a theoreticalration providing over 2000 kcal/person/day, at atime when morbidity and mortality rates had notchanged significantly26.

For further information, contact: HannaMattinen, email: hmattinen@actioncontrelafaim.org

22 Action contre la Faim (forthcoming): ‘NutritionalAnthropometric and Retrospective Mortality Survey, Childrenaged 6-59 Months, November 2007. Abu Shok and AsSalaam IPD Camps.’ El Fasher, North Darfur State. 23 Note that the malnutrition rates in November 2006 werefound to be higher than expected due to the delayed rainyseason and high prevalence of disease.24 Gelsdorf K, Walker P and Maxwell D (2007). Editorial: thefuture of WFP programming in Sudan’ Disasters, 31 (s1):s1-8. Blackwell Publishers, London.25 Young H (2007). Looking beyond food aid to livelihoods,protection and partnerships: strategies for WFP in the Darfurstates. Disasters, 31 (s1):s. Blackwell Publishers, London.26 Note that WFP plans to include in the general food distri-butions in North Darfur a supplementary ration including CSBmixed with milk powder for children under five years of ageduring the lean period in 2008.

MUAC screening at adistribution site

Dav

id M

ahouy,

ACF

Sudan

.

33

The ENN recently interviewed Mary LouFisher, Health and Nutrition Advisorfor Samaritan’s Purse, based in NorthCarolina, USA. Mary Lou started her

professional life as a registered nurse. In 1996she received a degree in advanced practice nurs-ing from Johns Hopkins University (JH) andcontinued her career as an adult nurse practi-tioner in the emergency department. Havingalways wanted to serve overseas, Mary Lou tookevery opportunity for short-term medical workwith Samaritan’s Purse while still working forJH who allowed short secondments. Memorableearly experiences in the Balkans includetrekking over the Albanian Alps with Kosovarswishing to return home. While there, she alsoworked in a MASH for Albanian Kosovars and awestern-style emergency medical department inGjakova. Her other short-term assignmentsincluded work in Afghanistan and westernDarfur. In November 2006, she became the full-time International Health and Nutrition Advisorfor Samaritan’s Purse.

Samaritan’s Purse was founded in 1970 by Dr.Bob Pierce who, 20 years earlier, had createdWorld Vision. Those of you who know yourBible (not me) will realise that the nameSamaritan’s Purse comes from the NewTestament parable about someone helping aninjured stranger by the road side whom othershad passed by. Mary Lou described how thestory symbolises the importance of helpingabsolutely anyone in need no matter who thatperson is or how critical you may feel aboutthem. Samaritan’s Purse is committed to helpinganyone in need. As representatives of theChristian faith, staff understand the importanceof following Jesus’ command to love one’sbrother (or sister).

Samaritan’s Purse began as a strictly emer-gency-focused organization, providing relief inthe form of feeding programmes, housing,water, and sanitation. However, over the yearsthe organization has also branched out intodevelopment and will often remain in-country toestablish longer-term infrastructure and servicesafter an emergency intervention.

The types of feeding programmes thatSamaritan’s Purse implements include generalfood distributions and selective feeding, i.e.supplementary and therapeutic feedingprogrammes. After the NATO bombing inKosovo ended, Samaritan’s Purse distributedover 4,000 loaves of bread daily. While inAfghanistan, they fed hundreds of men involvedin reconstruction. Skip back one decade andSamaritan’s Purse distributed food and livestock

in Bosnia and food and medical supplies inMogadishu. More recently, feeding programmeshave been established in Darfur. In line with theshift towards more developmental program-ming, Samaritan’s Purse is now looking atsetting up a Child Survival project in Niger andalso has a grant proposal with USAID and WFPto implement nutrition and health care program-ming, including school feeding, in Guatemala.

Views on nutrition are evolving withinSamaritan’s Purse. Mary Lou believes that therecent Lancet nutrition series is very timely andreally highlights the need to refine programmefocus on children two years old and under tomake sure they get optimum nutrition. This isespecially important given the risks of chronicillnesses associated with overfeeding after theage of two. The series also highlights additionalinterventions for maternal and child undernutri-tion and survival, including international action.

In line with many other humanitarian agen-cies, Samaritan’s Purse is moving away from thecentre-based model for treatment of severemalnutrition to community-based care, e.g.community management of severe malnutrition(CMAM). Another shift is employing more qual-ified staff, like nurses with public health degreesand experience. Samaritan’s Purse has learnedthat “one answer does not work for all situationsand that you need good assessments everywhereyou work to make sure you get context-specificsolutions.” Other lessons include the value ofmaximising the involvement of national staff andthe importance of networking by attending nutri-tion sector meetings.

The ENN asked Mary Lou to talk about herbest and worst programme experiences. Herwork in Hamish Koreb, a village in easternSudan with a high rate of malnutrition, fittedboth bills. The area is populated by the Beja, atribe that had previously been pastoralists.Many of the residents had lost their livestockbecause of conflict and were forced to live asvirtually destitute IDPs. Samaritan’s Pursebecame involved in a number of sectors between2004 and 2006, including food, medical, water,and agriculture programming. The organisationset up its main office across the border inAsmara, Eritrea, and at one time had 45 Eritreanand expatriate staff living in the Hamish Korebcompound. The programme also established anursing school for men and women. Throughthis project, illiterate women, with no previousschooling, were successful in learning diseaseprevention and treatment. In order for the Bejawomen to understand, lessons were verballytranslated from English to Tigrinian to Arabic to

Bedouit with the help of three translators. Thestudents were given oral examinations of thematerial and did surprisingly well. Sadly,following the signing of the peace agreementbetween the SPLA and the SudaneseGovernment, Samaritan’s Purse was asked toleave the area. The Beja appreciated how staffwere respectful of their culture and wanted theorganisation to return. Samaritan’s Purse is nowworking to re-establish this programme. Recentassessments have shown that the situation hasmore or less gone back to ‘square one’ with highrates of undernutrition, especially amongwomen and children.

Samaritan’s Purse gets 95% of its fundingfrom private donors and foundations. It is a faith-based evangelical organisation and a number ofemployees are the children of missionaries whogrew up in underdeveloped countries. Althoughmost of the work is standard emergency-typeprogramming, Samaritan’s Purse is also involvedin building and repairing churches destroyedduring conflict.

Mary Lou feels that Samaritan’s Purse hasmany positive qualities. They always do whatthey say they are going to do even when facedwith danger. She describes the valour of somestaff as “beyond understanding” because theyare willing to work in places that many othersmight not countenance. She also feels that whenSamaritan’s Purse “sets up shop somewhere,”everyone knows about their work. When she isable to run for a bit of recreation while in-country,she often encounters local citizens who recognizeher and know about the organisation’s work.

Samaritan’s Purse is working to more effec-tively document the impact of the work beingdone. The organisation is also looking to expandinto more countries.

It is clear that the work of Samaritan’s Purseis being increasingly recognised by fellow non-governmental organisations, governments, andother agencies as staff participate in interna-tional forums and networks. As a result, they arebecoming more involved in cutting-edgeprogramming approaches and issues. Mary Lousaid that they are using models such as PositiveDeviance (PD) Hearth and CARE Groups forinterventions, while Lot Quality AssuranceSampling (LQAS) may also be a way forward inassessment.

After thanking Mary Lou for the interviewwe agreed that we would look out for each otherthe next time one of us was out running alongthe banks of the Nile. I got the feeling that shemight be hard to keep up with.

Name: Samaritan's Purse

Address: Samaritan's Purse, 801 Bamboo Road,Box 3000, Boone, NC 28607, USA

Email address: info@samaritan.org

Website: www.samaritanspurse.org

Founded: 1970

Director: Franklin Graham, President

No. of staff in Boone(Headquarters):

No. of staff in thefield (international):

389

175

Projected income(2008):

$304 million USD

Agency Profile

Children in a village in Guatemala, pictured by Mary Lou's colleague whilst "scouting" foracute malnutrition (left). Mary Lou with one of the women helping in the SFP in Kenya.

ML

Fish

er/S

amar

itan

's P

urs

e, K

enya

, 2007

ML

Fish

er/S

amar

itan

's P

urs

e, G

uat

emal

a, 2

007

By Jeremy Shoham, ENN

34

Bundibugyo District, situated on the western side of the RwenzoriMountains, is one of the most disadvantaged districts in Uganda. Thedistrict is recovering from Allied Democratic Force (ADF) rebel attacks inthe late 1990s, and is on the border with the Democratic Republic of the

Congo (DRC) so that there is a constant threat of instability. There are no pavedroads and no electricity, making the transport and storage of goods (includingfood) very difficult. Cocoa, the most popular cash crop, increasingly usurps landpreviously used to grow food crops.

A nutrition survey conducted in January 2007 (dry season) indicated that bothacute and chronic malnutrition are problematic in Bundibugyo (global acutemalnutrition (GAM) 3% and stunting 45%1). Ugandan and Congolese childrenwho are either severely malnourished or at-risk for malnutrition present at localhealth centres. However, case presentation was complicated by a recent Ebolaoutbreak (29 November 2007 - 26 February 2008) that decreased patient willing-ness to attend health centres, while eroding health centre capacity to provideauxiliary health services.

World Harvest Mission (WHM) is a Christian non-governmental organisation,with a presence in Bundibugyo since 1985. In 2003, WHM began a partnershipwith the World Food Programme (WFP) and a local government health centre(Nyahuka Health Centre IV), wherein supplemental food rations were distributedto malnourished and at-risk children (motherless, HIV-affected), as well as womenenrolled in the prevention of mother-to-child transmission of HIV (PMTCT) proj-ect. This partnership ended in November 2006, as WFP left Bundibugyo to servein other needy areas.

This article describes WHM’s and the Nyhauka Health Centre’s efforts tocontinue nutrition services originally involving WFP-supported SupplementaryFeeding Programmes (SFPs). The ‘BundiNutrition’ programme focuses onincreasing sustainability through two animal husbandry projects and an agricul-ture project. Increasing sustainability is especially important as providing thera-peutic milk for all children who are malnourished or at-risk is cost-prohibitive,logistically difficult and absorbs much of WHM’s resources. All BundiNutritionprojects have the following goals: • Caring for individual children who are malnourished or at risk for malnutrition• Encouraging caregiver reciprocity• Building local capacity to produce high protein foods • Encouraging food crop cultivation.

Priority groups of children served are paediatric inpatients with severe acutemalnutrition, recently discharged inpatients, moderately malnourished outpa-tients without complications, motherless infants under 1 year, and children aged6 – 18 months whose mothers are HIV-positive and who are attempting to weanthem2. Although the antenatal clinic-based prevalence of HIV infection amongwomen in Bundibugyo is low at 2.9% (Scott Myhre, personal communication), thechildren born to HIV-positive women need special medical attention and nutritionadvice to prevent transmission of the virus from mother to child.

The BundiNutrition ProjectsThe BundiNutrition projects include a chicken project and dairy goat project toprovide animal protein to malnourished and at-risk children, and the ByokuliaBisemeye mu Bantu (‘Good food for people’) project, which promotes cultivationof high-protein food crops through seed distribution and return, and provideslocally ground groundnut paste and soybean flour to malnourished outpatients(see Box 1).

Chicken ProjectThe BundiNutrition programme previously purchased and distributed eggs tochildren. To move toward greater sustainability, and to encourage local farmers toinvest in chickens, a demonstration chicken coop was built in January 2007, tohouse 200 hybrid ‘layer’ chickens. A coccidiosis epidemic (March-April 2007)reduced the original number to 39 hens that began laying in August 2007. Eggs aregiven to HIV-affected children, surrogate breastfeeders (also called wet nurses) ofmotherless infants, and malnourished inpatients on the Nyahuka Health Centrepaediatric ward (in the latter case, to supplement the therapeutic milk onceappetite is regained).

Matiti Dairy Goat ProjectFor motherless infants, surrogate breastfeeding or wet nursing by a HIV-negativewoman in the child’s family is promoted. When a wet nurse is not available, theproblem becomes harder to solve. Dairy cows are scarce, as is processed cow’smilk from other areas of East Africa. While Bundibugyo District is a harsh envi-ronment in which many animals fail to survive, goats are a valued part of the

Animal husbandryand agricultureefforts towardprogrammesustainabilityBy Stephanie Jilcott, Karen Masso, LamechTugume, Scott Myhre and Jennifer Myhre

Stephanie Jilcott, PhD, recently finished an 18-monthterm with World Harvest Mission as a NutritionProgramme Officer in Bundibugyo District. She began theChicken Project and the Byokulia Bisemeye Mu BantuProject.

Karen Masso is a WHM Nutrition Programme Officer. Shebegan the Matiti Dairy Goat Project and has over 10 yearsof experience caring for malnourished children and moth-erless infants in Bundibugyo.

Lamech Tugume is a WHM Agriculture Extension Officer,and has been working with BundiNutrition Projects inBundibugyo for over two years.

Jennifer Myhre, MD, MPH, is a WHM Team Leader inBundibugyo, and has been a volunteer pediatrician forBundibugyo District Health Services for 14 years, duringwhich she has had extensive experience working withmalnourished children, motherless infants, and HIV-affected children. She began the therapeutic feedingprogramme at Nyahuka Health Centre.

We wish to acknowledge the tireless work of PaulineKentaro and Geofrey Batisibwa, both WHM AgricultureExtension Officers We also thank health centre staff suchas Basaliza Costa, Kyamukano Robinah, Babika Kudula,Kabasomi Naumi, Tayali Yonason, Kiyaya Tadeo, BihukoMisaki, Bwambale Expedito, and Ithungu Regina, as wellas volunteers Besemeliya Sipisoza, Biira Annet, MbusaAugustine, and Musinguzi Johnson. This work could nothave been done without cultural insight from BatigwaAngelite. We gratefully acknowledge those who spear-headed Byokulia Bisemeye mu Bantu Project food produc-tion, including Mbusa Florence, Tarunji Rodah, MangadaMperwa, Mugisa Justina, Butulibwabo Antonina, BazaraPeace, Asiimwe Mary, and Mugisa Jane. We are gratefulfor the daily, diligent care shown by Basaija, Jackson, andEngonzi to the goats and chickens.

1 Jilcott SB, Masso KL, Ickes SB, Myhre S, Myhre J (2007). Surviving but not quite thriving:Anthropometric survey of children 6-59 months in a rural western Uganda district. J Am Diet Assoc. 2007Nov;107 (11):1983-8.2 For the most recent recommendations on infant feeding and HIV, visit: HIV and Infant Feeding: Updatebased on the Technical Consultation held on behalf of the Inter-agency Task Team (IATT) on Prevention ofHIV Infection in Pregnant Women, Mothers and their Infants, Geneva, 25–27 October 2006, WHO, 2007.http://www.who.int/child-adolescent-health/New_Publications/NUTRITION/

Lamech (WHM Agriculture ExtensionOfficer) and beneficiary’s father witha Matiti Project goat

S J

ilcott

, W

HM

Mat

iti Fa

rm,

Ugan

da,

2007

Field Article

35

culture and seem to survive well. Thus, theMatiti Dairy Goat Project emerged. Hybriddairy goats from Masaka, Uganda, are selected,purchased and transported to Bundibugyo tobe given to at-risk families. Potential recipientsare identified through health centre staff andvolunteers at Nyahuka Health Centre, and areinvited to dairy goat sensitisation meetingsconducted by a WHM-hired agriculture exten-sion officer. The recipient list is finalised basedon the child’s need, number of trainingsattended, ability to manage dairy goats well, andconstruction of a goat pen. Recipients are askedto return the firstborn female to the project.

The Byokulia Bisemeye mu Bantu (‘Good foodfor people’) ProjectThis Project is located at two satellite healthcentres (Busunga Health Centre II and BusaruHealth Centre III. Children are eligible for the5-week programme if they are below theUgandan Child Health Card weight-for-agegrowth curve, or have a mid-upper armcircumference less than 12cms. Supplementalfood consists of three cups of groundnut pasteand three cups of soybean flour, mixed withdried moringa leaf powder when available.This supplement is produced locally as hand-powered grinders were distributed to women’sgroups (production teams) who volunteer toproduce the supplemental food. (Hand-powered nut shellers donated to WHM by theFull Belly Project were also given to commu-nity groups to make the process of shellinggroundnuts less labour-intensive.) Thesoybeans and groundnuts are purchased byWHM, distributed to production teams, whoroast, grind, and package them, keeping aportion of what is produced. WHM staff trans-

port the finished product from productionteam to the satellite health centres. In Octoberand November of 2007, volunteers and healthcentre staff received nutrition education andtraining, as well as training to identifymalnourished children and distribute the foodsupplement.

In addition, WHM-hired agriculture exten-sion officers to conduct agriculture trainingsfor local farmers. WHM also distributesgroundnut, sesame, and soybean seed to farm-ers in exchange for a portion of the harvest toredistribute to other farmers. Seed is distrib-uted based on farmer preferences. Fields aremonitored prior to seed distribution to ensureproper preparation.

Implementation evaluationChicken and Matiti Dairy Goat ProjectsFrom August – November 2007, 2,433 eggsfrom the Chicken Project were distributed atthe local health centre. Although the project isnot currently cost-effective, there are manybenefits. The chicken coop serves as a demon-stration project, eggs are easy to transport tothe nearby health centre, and they provideanimal protein to malnourished and at-riskchildren. The biggest implementation hurdlethus far was the coccidiosis epidemic.

In 2006, 36 hybrid dairy goats were givenout, and by the end of 2007, 26 were still alive.In April 2007, 72 were distributed, of which 10died and four were stolen. In total, there are 89mature male and female hybrid dairy goats inthe field, 32 exotic goat kids, and 23 other crossbreeds. Fourteen goats were returned to theproject and passed on to other families in 2007and the beginning of 2008. Several goat

Table 1: Seed returned as a percent of seedGroundnut Soybean Simsim

Bubandi Sub-county 361/673 = 53.6% 84/169 = 50.0% 48/70 = 68.6%

Busaru Sub-county 186/712 = 26% 173/413 = 41.9% 70/212 = 33.0%

Table 3: Average weight gain for moderately malnourished children

Health Centre Cycle Number enrolled Average weight gain in g/kg/d

Busunga Pre-ebola, 13 Nov – 11 Dec 21 4.3

Post-ebola, 22 Jan – 19 Feb 21 2.0

Busaru Pre-ebola, 14 Nov – 28 Nov 21 4.3

Post-ebola, 23 Jan – 20 Feb 16 2.0

Table 2: Goat location and milk yield

Location ofgoat

Duration of lactation(months)

Average amountof milk obtained(mls/day)

Hakitara 6 1750

Kagora 6 1000

Kinyante 5 1000

Lamia 5 1000

Busunga 6 1750

Busunga 4 500

Busunga 4 1000

Kisonko 4 1000

Nyankonda 4 1000

Bunyangule 3 1000

Lamia 2 750

Murungitanwa 3 750

Kabatabule 3 750

Tombwe 2 500

Njuule 3 1000

Nyahuka 2 750

Nyahuka 5 1500

Bundibugyo 2 750

Mukidungu 3 1000

Box 1: Project description, beneficiaries and sustainability issues for the BundiNutrition Projects

Project Description Beneficiaries Sustainability/ local capacity building

ChickenProject

Hybrid layer chickensare cared for by WHMstaff and eggs aregiven to malnourishedand at-risk children.

• Children 6 – 18 months of HIV-positive mothers trying to wean• Malnourished inpatients• Motherless infants < 1 year

• Eggs are a local food (overcomes transportation difficulties).• Eggs are a good source of animal protein.• The chicken coop and garden for chicken fodder are a demonstration project for local farmers.• The agriculture extension officer is a resource for local farmers with chickens.• The project could be self-sustaining if more chickens are brought in.

Matiti DairyGoat Project

Hybrid dairy goats arebrought in fromMasaka, Uganda andgiven to caregivers ofneedy children.

• Children 6 – 18 months of HIV-positive mothers trying to wean• Underweight HIV-positive children• Motherless infants < 1 year

• Mating of local female goats with hybrid males should produce greater lactation potential in offspring. • Sensitisation and on-farm trainings for farmers and beneficiaries regarding fodder establishment, record-keeping, and general goat care.• Reciprocal relationship with beneficiaries as they are asked to return the first born female to the project to be given to another needy child.• Participatory monitoring and evaluation.

ByokuliaBisemeyemu BantuProject

Local productionteams volunteer toroast and grindgroundnuts andsoybeans, which aregiven to moderatelymalnourished outpa-tients in two satellitehealth centres.

• Moderately malnourished (underweight) children served as outpatients

• Hand-powered grinders (also for community use) given to women’s groups who volunteer to produce the supplemental food (reducing reliance on outside foods).• Volunteers and health centre staff have been trained and equipped for growth monitoring, giving nutrition education, identifying malnourished children, and distributing food supplement.• Sensitisation seminars and agriculture trainings for local farmers • Reciprocity is encouraged as caregivers are asked to bring moringa leaves to the health centre each week.• Reciprocity and community ownership also increased by groundnut, sesame, and soybean seed distribution, where farmers are asked to return a portion of the harvest to the project be given to others in the community.

A member of Busunga Production Teampreparing dried moringa leaf powder,added to the groundnut paste distributedin satellite feeding programmes

S J

ilcott

, Buban

di Cen

tral

Vill

age,

Ugan

da,

2007

Field Article

36

management seminars were conducted.Implementation problems include frequentdivorces so that the goat ends up in a house-hold without the malnourished child.Additionally, goats are not given to childrenfrom the DRC due to the difficulty of givingquality veterinary care in an unstable region.Finally, some beneficiaries have reported thatgoats have a difficult time conceiving.

Byokulia Bisemeye mu Bantu ProjectTo learn more about how to improve theprogramme, qualitative interviews wereconducted with the chairpersons of twoproduction teams. Reported benefits of beingon the production team included getting ashare of the product, being able to grind with-out paying, learning more about preparinghealthy food for the family, and an increase inlocal groundnut cultivation. Difficultiesinclude that grinding is physically difficult,teams are working without any salary, and onone team there is reported distrust amongmembers due to disputes over profit share.

Although production teams were told toallow the community to use the grinders for asmall fee (100 Uganda Shillings per 2 cups ofgroundnuts), they reported that the commu-nity did not use the hand-powered grinders.When random community members wereasked why they did not use the grinders, theyreported that this was due to lack of awareness,not having groundnuts, the distance from theirhomes to the grinder, belief that the grinderwas for the production team members only,and the belief that the production teamscharged too much money for grinding.Additional implementation issues include thedifficulty of procuring moringa powder duringthe rainy season, as well as systematic qualitycontrol of the groundnut paste and soybeanflour, which would be different in appearancefrom week to week.

Despite implementation hurdles, produc-tion teams have thus far met programmetargets of 63 cups of soybean flour and 63 cupsof groundnut paste per week, in addition to 12cups of groundnut paste as an incentive forhealth workers and volunteers administeringthe programme.

Table 1 shows the amount of seed returnedto the Byokulia Project as a portion of the seeddistributed. The seed return discrepancybetween Busaru and Bubandi sub-counties isattributed to leadership issues, including nottaking initiative to motivate constituents toreturn the seed – this has been especially diffi-cult since some leaders themselves did notreturn seed they received.

Evaluation of impact on patients servedHIV-exposed childrenIn addition to dairy goats, twenty-five HIV-exposed children also receive biweekly foodsupplements (groundnut paste, eggs andbeans), distributed when the mother attendsthe clinic for antiretroviral treatment. Table 2shows the location of the nineteen lactatinggoats in the field, the duration of goat-lactationto date, and the average amount of milkobtained per day from each goat.

Two home visits were conducted to examinethe impact of the animal husbandry on HIV-exposed children. In the first case, the HIV-positive mother received a hybrid goat in theApril 2007 distribution. Before its untimelydeath two weeks prior to this (not replaced),she reported that it had trouble conceiving.This mother weaned her infant at six monthsand he is normal weight and HIV-negative. Bi-weekly she receives eggs and beans as shecomes to the clinic for antiretroviral drugs.

The second child and mother visited weregiven a lactating goat in April 2007, just as thechild turned six months, because her mother(HIV-positive) reported being ready to wean.The family obtained approximately 2 cups ofmilk per day, milking in the morning andevening. The milk was prepared by boiling onehalf volume of water to milk, mixing in ateaspoon of sugar, then pouring it into a plasticcup to cool. The child is now weaned, normalweight, and HIV-negative. The motherreported never giving milk to other children inthe compound. In addition to goat’s milk, thechild received bi-weekly food aid (groundnutpaste, beans, eggs) at clinic visits.

Both mothers reported giving the bi-weeklyfood to other children in addition to the childenrolled in the programme.

Moderately malnourished outpatientsTwo 5-week programme cycles werecompleted at each of the health centres. AtBusunga Health Centre, one child enrolled hada congestive heart defect, died, and wasremoved from analysis. At Busaru HealthCentre (post-ebola outbreak cycle), follow-uptimes varied, from 14 – 28 days. Table 3 showsaverage weight gain in each cycle.

Six home visits were made and caregiverswere asked to demonstrate preparation of thegroundnut paste and soybean flour. Most care-givers began by peeling and boiling bananas,then adding one or two tablespoons of ground-nut paste and soybean flour to the sauce. Manycaregivers then added cabbage, dodo, ormoringa. Overall, the sauces were very dilute.

Mothers did not report feeding the food toother children in the compound.

Future activities • The Matiti Project plans include identifying model farmers (‘early adopters’), i.e. those who are doing well with fodder establish- ment and goat management. These farmers will then be trained to educate other farmers on best practices for goat care and manage- ment. Matiti Project staff also plan to selec- tively breed to establish a line of local dairy goats.• An addition to the chicken coop is being constructed with a plan to bring in a new stock of 200 day-old chicks in May 2008. • Future work includes adding cooking demonstrations (requested by caregivers) to the nutrition education component of each Byokulia Project cycle. • Because several children who enrolled in the first 5-week cycle qualified again for enrolment in the second 5-week cycle, the l ength of the Byokulia Programme was extended from 5 to 10 weeks.• To decrease the number of defaulters, weekly return to the health centre is being heavily promoted, using the idea that the food is like a medicine.• In response to complaints about the quan- tity of food given at outpatient centres (perhaps because of previous emergency relief efforts), caregivers are being educated about the nutritional value of the food received. • The current food supplement given in the Byokulia Project is not used by caregivers as a ready-to-use supplemental food (RUSF). Future work includes working with production teams to standardise production methods so that the food supplement can be marketed to caregivers as RUSF rather than a supplement to sauce. • Production teams are working very diligen- tly as volunteers. Therefore, microfinance initiatives enabling teams to benefit from the work are greatly needed.• Due to low seed return rates, the next seed distribution will include caregivers of malnourished outpatients, volunteer health workers, and those from the previous distribution who returned seed.

For further information, contact: StephanieJilcott, stephjilcott@yahoo.com, Karen Masso,masso@iname.com, and Drs. Scott andJennifer Myhre, drsmyhre@yahoo.com, WorldHarvest Mission, PO Box 1142, Bundibugyo,Uganda, East Africa

Costa (health worker) with the caregiver of a severely malnourishedchild, an inpatient on the paediatric ward

A member of Busunga Production Team demonstrates the grinder

S J

ilcott

,Ugan

da,

2007

S J

ilcott

,Ugan

da,

2007

Field Article

37

People in aid

Graduates of WHM nutrition trainings pose in their 'Byokulia Bisemeyemu Bantu' t-shirts

Participants of the Regional Workshop on Infant and Young Child Feeding in Emergencies, 10-13 March, 2008

Busaru Production Team in WHM programme (see field article), BumateVillage, Uganda.

Geo

frey Batisib

wa, U

gan

da, 2

007

S Jilco

tt, Busaru

Health

Cen

tre, Ugan

da, 2

007

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, donorsand academics. The Network is the result of a shared commitmentto improve knowledge, stimulate learning and provide vitalsupport and encouragement to food and nutrition workersinvolved in emergencies. The ENN officially began operations inNovember 1996 and has widespread support from UN agencies,NGOs, and donor governments. The network aims to improveemergency food and nutrition programme effectiveness 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,FieldExchange, which is devoted primarily to publishing field levelarticles and current research and evaluation findings relevant tothe emergency food and nutrition sector.

The main target audience of the publication are food and nutritionworkers involved in emergencies and those researching this area.The reporting and exchange of field level experiences is central toENN activities.

The Team

Editorial teamDeirdre HandyMarie McGrathJeremy Shoham

Office SupportRupert GillSarah FosterMatt ToddDiane Crocombe

DesignOrna O’Reilly/Big Cheese Design.com

WebsitePhil Wilks

Contributors for this issueTom OgutaGrainne MoloneyLouise Masese Mark MyattAndy SealNúria Salse Kate GoddenEva VicentStephanie JilcottKaren MassoLamech TugumeScott and Jennifer MyhreHanna MattinenElena RiveroEric ZapateroWaweru Joseph MwauraMary Lou FisherSusan ShepherdCarmel DolanFiona Watson

Pictures acknowledgementGrainne Moloney/FSAUMark MyattAndy SealGeofrey BatisibwaStephanie JilcottDavid MahouyFSAUWFP/Gerald BourkeWFP/Stephanie SavariaudWFP/Antonia ParadelaWFP/Peter SmerdonConcern WorldwideInternational Rescue CommitteeWFP/Mikael BjerrumWFP/Fred NoyH Deconinck/ FANTAMary Lou Fisher/Samaritan'sPurseKate Godden

On the coverAn airdrop by an Ilyushincargo plane in Oringi, Sudan.WFP/Fred Noy, Sudan, 2007

The opinions reflected in FieldExchange articles are those ofthe authors and do not neces-sarily reflect those of theiragency (where applicable).

The Emergency Nutrition Network (ENN) is a registered charity in the UK(charity registration no: 1115156) and a company limited by guarantee andnot having a share capital in the UK (company registration no: 4889844)Registered address: 32, Leopold Street, Oxford, OX4 1TW, UKENN Directors/Trustees: Marie McGrath, Jeremy Shoham, Bruce Laurence,Nigel Milway, Victoria Lack, Arabella Duffield

Field Exchangesupported by:

We are sorry to say that Dan George has left theENN, to pursue fulltime study. Dan has been theFinance Assistant with ENN since the office moved toOxford from Dublin in late 2004. He will be greatlymissed, not just for his mastery of Excel spread-sheets but for tales of his latest biking antics. We allwish him the best of luck in his next adventures.

Rupert Gill isENN officemanager andfundraiser,based inOxford.

We are delighted to welcome a new addition to theENN Team, Diane Crocombe. Diane recently joinedthe ENN as part-time Project and Finance SupportOfficer based in Oxford. Diane has extensivefundraising and management experience withOxfam for the past 25 years, both in Oxford andoverseas including Kenya, India and Cambodia.

Jeremy Shoham (Field Exchangetechnical editor) and Marie McGrath(Field Exchange production/ assistant editor) are both ENNdirectors.

Orna O’ Reillydesigns andproduces all ofENN’s publications.

Sarah Foster works part-timewith ENN tosupport mailing ofField Exchange and IFE Resources.

Phil Wilks (www.fruity solutions.com)manages ENN’s website.

Matt Todd is the ENNfinancial manager,overseeing the ENNaccounting systems,budgeting and financial reporting.

38

Many people underestimate the value of their indi-vidual field experiences and how sharing them canbenefit others working in the field. At ENN, we arekeen to broaden the scope of individuals and agen-cies that contribute material for publication and tocontinue to reflect current field activities and expe-riences in emergency nutrition.

Many of the articles you see in Field Exchangebegin as a few lines in an email or an idea sharedwith us. Sometimes they exist as an internal report

that hasn’t been shared outside an agency. Theeditorial team at Field Exchange can support you inwrite-up and help shape your article for publication.

To get started, just drop us a line. Ideally, send us(in less than 500 words) your ideas for an articlefor Field Exchange, and any supporting material,e.g. an agency report. Tell us why you think yourfield article would be of particular interest to FieldExchange readers. If you know of others who youthink should contribute, pass this on – especially to

Invite to submit material to Field Exchangegovernment staff and local NGOs who are under-represented 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 1PX,UK. Tel: +44 (0)1865 324996 Fax: +44 (0)1865 324997

Emergency Nutrition Network (ENN)32, Leopold Street, Oxford, OX4 1TW, UK

Tel: +44 (0)1865 324996Fax: +44 (0)1865 324997Email: office@ennonline.netwww.ennonline.net