• MUAC v weight for height debate in Philippines • Food ... - ENN

• MUAC v weight for height debate in Philippines• Food by prescription in Zambia• Access to global RUTF supplies• Managing MAM in Guinea• Blanket SFP efficiency in Sudan• Special focus on coverage assessment

- SLEAC in Sierra Leone- SQUEAC in Eastern Sudan, Mali & Mauritania

ISSN 1743-5080 (print) January 2012 Issue 42

1

From the EditorContents

Field Articles

3 MUAC versus weight-for-height debate in the Philippines

30 Examining the integration of Food by Prescription into HIV care and treatment in Zambia

46 Increasing Access to Ready-to-use Therapeutic Foods (RUTF)

59 Improving blanket supplementary feeding programme (BSFP) efficiencyin Sudan

63 Multi-pronged approach to the management of moderate acute malnutrition in Guinea

Focus on coverage assessment

33 Foreword

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

37 Causal analysis and the SQUEAC toolbox

39 Using SLEAC as a wide-area survey method

45 News

• Register now for coverage assessment workshop in Oxford• Coverage assessment forum launched on en-net• Technical Reference for SQUEAC and SLEAC Methods, 2012

Research

6 Qualitative review of an alternative treatment of SAM in Myanmar

9 Dangerous delay in responding to Horn of Africa early warnings of drought

11 Community case management of severe acute malnutrition in southern Bangladesh

12 Mortality risk factors in severely-malnourished children hospitalised with diarrhoea

13 Voices from the field: Optimising performance for humanitarian workers

15 Exclusive breastfeeding promotion by peer counsellors in sub-Saharan Africa

15 Effects of a conditional cash transfer programme on child nutrition in Brazil

16 MUAC and weight-for-height in identifying high risk children

17 Civil-military coordination during humanitarian health action

18 Operational research in low-income countries: what, why, and how?

20 Effects of agricultural and nutrition education projects on child health inMalawi

21 Study of causes of persistent acute malnutrition in north Darfur

22 Effects of performance payments to health workers in Rwanda

23 ‘Zap’ it to me: short-term impacts of a mobile cash transfer programme

24 Revisiting the concept of growth monitoring and its possible role in community-based programmes

26 Political economy of adaptation through crop diversification in Malawi

27 Practical challenges of evaluating BSFP in northern Kenya

News

48 UNHCR Technical Workshop on the Operational Guidance on the use of Special Nutritional Products

49 infoasaid: communication in emergencies

50 Experiences of the Nutrition in Emergencies Regional Training Initiative

53 Minimum Reporting Package (MRP) on Supplementary Feeding Programmes

54 Improving patient assessment: The ‘MOYO’ Weight-for-Height Chart

54 E-learning course on Social Safety Nets

55 En-net update

55 Attractive scholarship for EDAMUS Masters programme

55 Government of Sudan CMAM Training Course on Inpatient Management of Severe Acute Malnutrition: Training Materials (2011)

56 UNHCR standardised nutrition survey guidelines and training

57 Putting nutrition products in their place: ACF position paper

58 Conference on government experiences of CMAM scale up, Ethiopia, 2011

Evaluation

61 Evaluation of Concern’s response to the Haiti Earthquake

63 Review of Integrated Food Security Programme in Malawi

Agency Profile

32 Centres for Disease Control and Prevention (CDC), International Emergency and Refugee Health Branch (IEHRB)

68 Obituaries

This issue of Field Exchange gives extended coverage to abriefing paper just released by Oxfam and SC UK on the2011 response to the Horn of Africa crisis. This paper arguesthat the response was late and led to the unnecessary

deaths of between 50,000 to 100,000 people, at least half of whomwere children under 5 years. According to the authors, there wassufficient early warning to trigger a response as early as November2010 but the main response only unravelled in July 2011, followingdeclaration of famine and concerted media coverage. The paperidentifies the usual litany of reasons for this failure of response, i.e.only responding when media attention is overwhelming, politicallyinfluenced decision-making, time-lags between early warning andappeals, making appeals on the basis of capacity to deliver andaccess rather than need, inability to act on risk and forecasts and thedivide between development and emergency programming andfunding. These reasons are familiar to most of us and were largelyapplied to analyses of previous failures of response going back asfar as the Sahelian famine of 1984. However there is a coherenceand clarity in this paper, particularly in the way it trains its focus onthe need for future response to be based more upon risk reductionand the institutional structural change needed to support such anapproach.

The ENN fully supports the recommendations in this briefingpaper and believes that this important document can provide apowerful advocacy tool for change. Some of the issues around theemergency/development divide raise uncomfortable questions forthe ENN itself, which we will reflect upon in due course. There ishowever one important element of the analysis which we feel is notadequately addressed in the report. This relates to the relationshipbetween early warning and donor response and the recommenda-tion for use of earlier triggers and risk analysis. Theserecommendations are hardly new and have been made repeatedlyover the past 25 years. The piece of the jigsaw that is still missing isthe lack understanding and transparency about ‘how donors makedecisions whether to respond’. The failure of donor response overmany years in certain high profile emergencies suggests that thereare complex political and institutional processes that hinder timelyand effective response, including the type of risk taking advocatedin the Oxfam/SC UK briefing paper. The nutrition community,perhaps not unsurprisingly, continues to focus on ‘technical’ solu-tions, yet until we have a better understanding of the constraintsfaced by donors and their ‘room for manoeuvre’ to effect change,our technical solutions will have little impact on response. We there-fore strongly support any advocacy efforts that encourage donorsto systematically analyse their decision-making processes duringemerging crises and to make such findings publically available.

And now to the rest of this Field Exchange edition. Field articlesin this issue of Field Exchange (no 42) can largely be divided intothose related to the treatment of severe acute malnutrition (SAM)and those related to treatment and prevention of moderate acutemalnutrition (MAM). Three of the SAM related articles describe theexperience of conducting different types of coverage surveys forcommunity based management of acute malnutrition (CMAM)programmes and feature in a special section of this issue on cover-age assessment. An article by Ernest Guevarra, Saul Guerrero, andMark Myatt describes the use of the SLEAC method to assessnational level coverage of CMAM in Sierra Leone. The advantage ofthe approach is that relatively small sample sizes are required tomake accurate and reliable classifications of coverage and to iden-tify barriers to programme access. Assessments can therefore becompleted relatively quickly. The authors conclude that the SLEACmethod should be the method of choice when evaluating coverageof CMAM programmes at regional or national level. An article byJose Luis Alvarez Moran, Brian Mac Domhnaill and Saul Guerrero atAction Contre la Faim (ACF) describes the experience of conductingremote SQUEAC investigations in Mali and Mauritania where certainareas are difficult to reach by external investigators. The approachdoes require greater reliance on field teams, as well as strengthen-ing or modifying certain SQUEAC processes, e.g. separating the datacollection and analysis processes, using new technologies andaddressing supervision and motivation issues proactively. A thirdarticle on coverage assessment describes the use of the SQUEACmethod to undertake a causal analysis of SAM in rural areas of east-ern Sudan. The data collected were sufficient to identify risk factorsand risk markers (i.e. diarrhoea, fever, early introduction of fluidsother than breastmilk) that were associated with SAM. The authorssuggest that it is possible to use the SQUEAC toolbox to collectcausal data using staff trained as SQUEAC supervisors and trainers,although data analysis may require staff with a stronger back-ground in data-analysis.

We would like to dedicate this issue of Field

Exchange to Kari Noel Egge and Mr Abdikarim

Hashi Kadiye, whose untimely and premature

deaths saddened the whole humanitarian commu-

nity. We have included short tributes to both of

them in a new Field Exchange obituary section for

humanitarian workers, which we sincerely hope

remains unused in subsequent issues.

1 Simplified LQAS (Lot Quality Assurance Sampling)

Evaluation of Access and Coverage2 Semi-quantitative Evaluation of Access and Coverage

2

components of the programme, as were certain keymessages derived from a KAP (Knowledge, Attitudesand Practices) survey. All pregnant and lactatingwomen and children under five years of age weretargeted with the BSFP, providing approximately 500kcals/day. The pilot found that 68% of malnourishedchildren enrolled in the programme recovered withinfour weeks, although the cost of the programme wasat least twice as high per child as the targeted SFPimplemented by WFP. The authors concluded thatwhen food availability and quality is enhancedthrough the provision of small quantities of highlyfortified food combined with the intensive engage-ment of the community around harmful feedingpractices, the impact of food aid is significantlyincreased. The size of the programme allowed inten-sive monitoring by the Ministry of Health and WFP.The challenge lies in taking the pilot to scale.

A final field article written by Catholic ReliefServices (CRS) in Zambia considers a Food byPrescription (FBP) programme that targets moder-ately and severely malnourished HIV positive adultswith medicalised doses of nutrition supplementsthrough the home based care programme. Individualsachets of the supplement are distributed to reducehousehold sharing and strengthen the understand-ing that the food is a ‘medicine’. Data from the pilotprogramme showed that there was an increase inclient Body Mass Index (BMI) between admission anddischarge. Among adult clients, the average BMI onadmission was 17.6 and the average BMI ondischarge was 20.5. The overall average increase inBMI pre-FBP to post-FBP was 2.9. Most clientsrequired three to six months of nutrition rehabilita-tion to qualify for discharge.

The research section of this issue also covers awide range of subjects. There are two articles on cashtransfer programmes. One is a summary of a study ofthe national ‘Bolsa Familia’ programme (BFP) in Brazil,which is the world’s largest conditional cash transferprogramme. It reaches 5,564 municipalities in the 27states of Brazil and about 11 million families (25% ofthe Brazilian population).Once a family enrols, it mustcomply with certain health and education conditionsto remain in the programme. The study found thatchildren from families exposed to the BFP were 26%more likely to have normal height-for-age than thosefrom non-exposed families; this difference alsoapplied to weight-for-age but not weight-for-height.Another study looks at a cash transfer programme inNiger using mobile phone technology and foundthat in comparison to physical cash transferprogrammes, there was a significantly reduced costto programme recipients, as well as reduced imple-menting agency’s variable costs associated withdistributing cash. There is also a fascinating article onthe political economy of crop diversification policiesand the policy process at government level inMalawi. The article explains how the processes ofdiscussing, negotiating, approving and implement-ing policies are as important as the scientific contentof the policies themselves. The experience with cropdiversification shows that dominant stakeholdersalmost always have their way and that in Malawi,

Moving on, an article by Bernardette Cichondescribes ACF nutrition survey findings in thePhilippines, where three consecutive surveys foundlarge discrepancies in the prevalence of SAM usingeither a weight-for-height cut off below – 3 z-scores ora mid upper arm circumference (MUAC) less than 115mm. The authors conclude that as long as the risk ofmortality in children with a weight-for-height of lessthan -3 z-scores but a MUAC greater than 115mm isn’tproperly understood, all children classified asmalnourished according to both indicators shouldreceive treatment (in this instance, admission to aCMAM programme). The authors recognise that usingtwo indicators complicates programming. Whilst notthe case in this example, this may also have significantprogramme capacity implications. Interestingly, aresearch summary in this issues research sectionbased on an old data set from Senegal examines therisk of dying of children having either a low MUAC or alow weight-for-height (z score) or a combination ofboth in the absence of treatment. Analysis found thatMUAC has a better ability than weight-for-height (zscore) to assess risk of dying. Furthermore, using bothindicators together did not improve the identificationof high risk children.

Also related to CMAM programming, a field articleby Jan Komrska of UNICEF’s Supply Division inCopenhagen describes how UNICEF has been keep-ing pace with the increased demand for Ready to UseTherapeutic Food (RUTF) as CMAM programming hasbeen scaled up rapidly across many countries. Thearticle describes various strategies employed, includ-ing increasing the number of global suppliers inEurope, identifying local producers in Africa and Asia,improving forecasting of demand and the pre-posi-tioning of stocks. Staying with SAM management, aresearch summary of work by Tufts describes thesuccess of a pilot CMAM programme in an Upzila inBangladesh where community health workers wereresponsible for and supported in diagnosing andtreating SAM children. Programme indicators likerecovery and mortality exceeded SPHERE standardsand an extraordinary coverage of 89% was recorded.

A further three field articles deal with the preven-tion and treatment of MAM in different contexts. Anarticle by Naomi Cosgrove and colleagues workingfor ACF in Myanmar describes how reduced dailyrations of RUTF (one sachet instead of two or threeused for SAM treatment) were used to treat success-fully uncomplicated cases of MAM managed withinthe CMAM programme. Fuelled by imported supplyconstraints and a rising caseload, ACF modified thetreatment protocol and introduced a second phaseof treatment, once the child had improved from asevere to a moderate (i.e. MAM) stage of malnutri-tion. A key conclusion from the article was that theprogramme success was partly due to the high qual-ity ration but also the attention given to programmedesign, including well trained staff and goodcommunity mobilisation. Another article written byDr Jean-Pierre Papart and Dr Abimbola Lagunju ofFondation Tdh covers MAM management and theimportant role of quality staff in service delivery inGuinea. Community and facility level screening andprovision of RUTF (for SAM) coupled with cookingdemonstrations, counselling, some food supply (forMAM) and defaulter follow up were undertaken bygovernment staff. Urban community health workersplayed a key role in service delivery. The authorsconcluded that government health facilities candeliver good results with the appropriate technical,material and equipment support. On the job trainingand supervision, feedback on performance and regu-lar higher level supervision of nutrition activitiesplayed an important role in service quality. A WFPpilot blanket supplementary feeding programme(BSFP) in Kassala State, Sudan is the subject of a thirdfield article on MAM, written by Pushpa Acharya andEric Kenefick of WFP. This article details how intensivecommunity engagement and sensitisation were key

implementation of crop diversification has beenconstrained by a dominant narrative that equatesfood security with maize production.

Other research of note in this issue include apsychological study to understand how humanitar-ian workers remain effective in challengingenvironments, a study to assess the effect of perform-ance based payment of health care providers on theuse and quality of child and maternal care service inhealthcare facilities in Rwanda and a position paperto guide country-level health clusters on how toapply IASC (Inter-Agency Standing Committee) civilmilitary coordination principles to humanitarianhealth operations given the “shrinking of humanitar-ian space” in many complex crises.

Our opening 2012 issue is a reminder of theongoing innovations and inevitable compromisesthat are sometimes necessary in humanitarianprogramming. You can have the best designed inter-vention, but national and sub-national capacity todeliver on the ground and supply chain difficultiescan make or break a quality programme. The articlesfeatured also reflect some of the ‘grey’ areas in theemergency nutrition sector. For example, whilst SAMmanagement has rapidly developed and improvedthrough CMAM programming over the past 10 years,there is a limited evidence base for optimal program-ming for the prevention and management of MAM.The jury is still out on how best to deal with high andfluctuating levels of MAM and field practice remainsa bit of a ‘free for all’. A news piece by ACF, laying outthe agency’s position on the role of products in thetreatment and prevention of global acute malnutri-tion, sets out certain boundaries whilst highlightingsome of the more contentious areas. Our prediction isthat the role of products in the prevention and treat-ment of MAM will be one of the big issues in thecoming year. And we hope the SCUK/Oxfam paperhas just generated another - the 2011 Horn of Africacrisis highlights once again a clear failure of responseat the highest level of the international aid commu-nity. Is it not incumbent upon donor governments toanalyse why that occurred and to work with the restof us to address the reasons for this failure as urgentlyas possible?

We hope you enjoy this diverse issue of FieldExchange and wish all our readers a healthy, happyand productive 2012.

Jeremy Shoham, EditorMarie McGrath, Sub-editor

Allis

on S

helley/C

once

rn,

Port

-au-P

rince

Haïti, 2

010

Figure 1: Prevalence of SAM according to weight-for-height and MUAC

Figure 2: Prevalence of GAM according to weight-for-height and MUAC

3

MUAC versus weight-for-heightdebate in the PhilippinesBy Bernardette Cichon

Bernardette is a Public HealthNutritionist who at the time of thework described in this article,worked with Action Contre la Faim(ACF). She is currently working withMédecins sans Frontières.

The author would like to thank ACF Spain, inparticular Elisa Dominguez (Senior NutritionAdvisor) and Amador Gomez (TechnicalAdvisor), as well as the ACF Philippinestechnical team in the mission and all thesurveyors involved in data collection.

In the 39th issue of Field Exchange, anarticle was published by Jennifer Carterand Joel Conkle1 that discussed thedifferences in prevalence of acute

malnutrition in Cambodia determined usingeither weight-for-height or mid upper armcircumference (MUAC) measurements. Inthis article, we share similar observationsfrom the Philippines that challenge the imple-mentation of a community-basedsurveillance system for early detection andreferral of acutely malnourished children totherapeutic feeding programmes. The article

discusses possible causes of the differencesbetween MUAC and weight-for-height andcalls for more research into this topic.

MUAC measurements have been usedworldwide for identification, referral andadmission of severely malnourished childrenaged 6-59 months to nutrition programmes.It is a good indicator of both muscle massand mortality risk. MUAC measurementsenable community volunteers with minimaltraining to carry out emergency needs assess-ments and active case finding. It is a quick,easy and cheap alternative to weight-for-

height measurements and has recently beenendorsed by the World Health Organisation(WHO).2,3

However, in some populations, largedifferences have been observed betweenprevalence of acute malnutrition measuredby weight-for-height versus MUAC. In linewith this, three surveys carried out by ActionContre la Faim (ACF) in the Philippines founda big discrepancy between the two. This raisesa number of questions, such as what causessuch a big difference between MUAC andweight-for-height based prevalences in somepopulations and not in others, what are theprogrammatic implications, and which is thebetter indicator for measuring acute malnutri-tion in the Philippines?

Relationship between MUAC andweight-for-height in Filipino populationsIn 1984, Johnson et al first reported in aPhilippines study that arm circumferencemeasurements resulted in a lower prevalenceof malnutrition as compared to weight-for-height and weight-for-age. They concludedthat arm circumference measurements were

1 Jennifer Carter and Joel Conkle (2010). CMAM in

Cambodia – indicators of acute malnutrition for screen-

ing. Field Exchange, Issue No 39, September 2010.

p47. http://fex.ennonline.net/39/cmam.aspx2 UN Standing Committee on Nutrition. Task Force on

Assessment, Monitoring and Evaluation. Fact Sheets on

Food and Nutrition Security Indicators/Measures: Mid-

Upper Arm Circumference (MUAC).3 Myatt M., Khara T. and S. Collins. A review of methods

to detect severely malnourished children in the

community for their admission into community based

therapeutic care programmes. 2005.

*Location: 1= Joint emergency nutrition and food security assessment of the conflict-affected persons in Central

Mindanao (UNICEF/WFP, January- March 2009), 2= Follow-up emergency nutrition assessment in the conflict affected

communities in Central Mindanao (Save the Children, March/April 2010), 3=Anthropometric and mortality survey in the

municipality of President Roxas, Central Mindanao (ACF, October/November 2010), 4= Anthropometric and mortality

survey in the municipality of Arakan, Central Mindanao (ACF, November 2010) , 5= Anthropometric and mortality survey

in the municipality of Kapatagan, Central Mindanao (ACF, November/December 2010).

** The MUAC cut-off used for SAM in the UNICEF/WFP survey was 110mm not 115mm.

Weight-for-height<-3 z-scores

MUAC <115mm

Am

ount

(%)

Location*

1** 2 3 4 5

2.2

0.3

1.1

0.4

2

0.2

0.9

0.3

1

0

Weight-for-height<-2 z-scores

MUAC <125mm

Am

ount

(%)

Location*

1** 2 3 4 5

9.8

6.37.8

14.8

10.3

1.3

5.9

1.5

6.9

1.1

Fie

ld A

rtic

leMUAC measurement of achild in the Municipalityof Arakan

B C

ichon,

Philip

pin

es,

2010

Municipality Height groups (cm) W/H <-2 z-scores MUAC <125mm W/H <-3 z-scores MUAC< 115mm

President Roxas ≥ 24 months≥ 24 months

12.39.3

3.20

2.71.6

0.40

Arakan ≥ 24 months≥ 24 months

5.45.7

50

01.3

10

Kapatagan ≥ 24 months≥ 24 months

11.84.6

3.30.4

1.50.7

00.2

4

PresidentRoxas

Arakan Kapatagan

% of children with a W/H ofless than -2 z-scores thathave a MUAC of less than125mm

7.8% 15% 14%

% of children with a W/H ofless than -3 z-scores thathave a MUAC of less than115mm

17.6% 0% 0%

% of children with a MUAC ofless than 125mm that alsohave W/H z-score of less than-2

77.6% 60% 89%

% of children with a MUAC ofless than 115mm that alsohave W/H z-score of less than-3

100% 0% -

not adequate for detecting undernutrition inFilipino children4. Even though standards andcut-offs used then differ from those used today,five surveys carried out in the southernPhilippine island of Mindanao betweenJanuary 2009 and December 2010 have foundsimilar results. Prevalence of severe acutemalnutrition (SAM) was much lower accordingto MUAC than weight-for-height in all fivesurveys (see Figure 1) and global acute malnu-trition (GAM) in four out of five surveys (seeFigure 2).

Not only did the percentage of acute malnu-trition differ between the two indicators, butthe children selected were not always the same.While a large proportion of children identifiedas acutely malnourished according to MUACwere also classified as malnourished accordingto weight-for-height, only a small proportion ofchildren classified as malnourished accordingto weight-for-height were also classified asmalnourished according to MUAC (see Table 1).

Programmatic implications and dilemmasThe three surveys carried out by ACF betweenOctober and December 2010 were baselinesurveys for a new four year integrated foodsecurity, nutrition, water and sanitationprogramme funded by the SpanishCooperation (AECID). One of the activities ofthis programme was to enable local health serv-ices to provide SAM treatment for those whoneed it. Active case finding was planned as apriority activity to be carried out by community

nutrition workers in the villages using MUAC.However, the large difference between MUACand weight-for-height means that this approachhas to be reviewed. If MUAC is used for activecase finding, only a small proportion of chil-dren classified as malnourished according toweight-for-height would be identified andreferred to the programme (see Table 1).

In addition, the caseload would be so lowthat it is difficult to justify an intervention. Inthe municipality of President Roxas, for exam-ple, survey results showed 10% GAM and 2%SAM according to weight-for-height, a preva-lence high enough (in alert levels) to justify theneed for community based management ofacute malnutrition (CMAM). Prevalenceaccording to MUAC, however, was much lower(see Figures 1 and 2: 1.3% GAM and 0.2% SAMprevalence, respectively). If MUAC turned outto be a better indicator of malnutrition in thispopulation, the type of intervention neededwould differ. The decision to intervene wasbased on prevalence according to weight-for-height, but is this correct? It seems that beforewe can answer this question, more research isneeded to understand the weight-for-height/MUAC relationship and to determinewhich indicator is more appropriate in thePhilippine context.

Why is there such a big discrepancybetween MUAC and weight-for-height?A study carried out by Myatt et al in 2007 thatincluded data from 560 surveys from 31 coun-tries showed that while a similar prevalence ofacute malnutrition (GAM and SAM) accordingto MUAC and weight-for-height was found forthe whole data set, there were differences in theMUAC/weight-for-height relationship betweenand even within countries5. While in somepopulations the prevalence of acute malnutri-tion was the same according to both indicators,in others MUAC led to a higher prevalence thanweight-for height or vice versa. Similar to theresults from the Philippine surveys above,prevalence of acute malnutrition according toMUAC was found to be lower in parts ofEthiopia and Kenya, Sudan, Chad, the Indiansubcontinent and the Hispanic populations6.Equivalent problems have also been observedin Cambodia in the national survey carried outin 20087. Not only did the prevalence differaccording to the two indicators, but the childrenselected were not always the same8.

A number of different factors have beenassociated with the MUAC/weight-for-heightrelationship such as body shape, age andmortality (see below). Although it is possiblethat inaccuracies in the measurements are inpart responsible for the large differences found,the fact that five surveys carried out by differ-ent organisations found the same problemmeans that this issue deserves further attention.

The multi-centre growth reference study(WHO 2006) confirmed that ethnicity and envi-ronment influence growth of infants andchildren, but according to the WHO thesedifferences were not “large enough to invali-date the general use of the WHO growthstandards population as a standard in all popu-lations”9 . Contrary to this, studies have shownthat body shape does seem to influence weight-for-height in some populations. In Ethiopia, forexample, prevalence of acute malnutrition inpastoralist populations as measured by weight-for-height was found to be much higher thanaccording to MUAC: 20% versus 7%. In agrar-ian populations in the same country, bothindicators led to similar estimates. Studies haveshown that in these populations, body shapewas associated with the MUAC/weight-for-height relationship. Children from pastoralistpopulations have longer legs and shortertrunks and thus lower sitting to standing heightratio (SSR) than agrarian populations. Whereasweight-for-height was strongly influenced bybody shape, the effect on MUAC was verysmall10. It is possible that body shape also playsa role in the Philippines and other Asian popu-lations but no data were found to support thissupposition.

It has been shown that children identified asacutely malnourished using MUAC are, onaverage, younger than those selected byweight-for-height11. As a result, it has beenshown that the MUAC/weight-for-height rela-tionship is better in children under 24 monthsand height-adjusted MUAC cut-offs have beensuggested12. While malnourished children werefound in all age and height groups according toweight-for-height in the Philippine surveys, nochildren with a MUAC of less than 125mmwere found in the taller children (see Table 2)and in two of the three surveys no children over24 months had a MUAC of less than 125mm(see Table 3).

MUAC has been shown to be a better predic-tor of mortality than weight-for-height13. In thePhilippine surveys, both mortality rates andprevalence of acute malnutrition according toMUAC were relatively low whereas prevalenceaccording to weight-for-height was much

Table 1: The relationship between weight-for-heightand MUAC in three municipalities in CentralMindanao (October-December, 2010)

Municipality Height groups (cm) W/H <-2 z-scores MUAC <125mm W/H <-3 z-scores MUAC< 115mm

President Roxas ≥65 and ≥75≥75 and ≥90≥ 90

16.198.3

6.40.20

4.51.71.1

1.300

Arakan ≥65 and ≥75≥75 and ≥90≥ 90

9.455

6.20.60

00.63

1.500

Kapatagan ≥65 and ≥75≥75 and ≥90≥ 90

13.454.7

4.80.50

1.80.80.8

00.30

Table 2: Prevalence of acute malnutrition according to weight-for-height and MUAC by height group in threemunicipalities in Central Mindanao (October-December 2010)

4 Johnson MD, Yamanaka WK, Formation CS. A Comparison

of Anthropometric Methods for Assessing Nutritional Status

of Preschool Children. The Philippines Study. Journal of

Tropical Pediatrics 1984;30:96-104.5 Myatt M, Duffield Isabella. A review of survey data

collected between September 1992 and October 2006.

22nd October 2007.6 Myatt and Duffield, 2007. See footnote 5.7 Carter J and Conkle J. CMAM in Cambodia-indicators of

acute malnutrition for screening. Field Exchange Issue 39,

September 2010.8 WHO, UNICEF, 2009. WHO child growth standards and

the identification of severe acute malnutrition in infant and

children. A joint statement by the World Health

Organisation and the United Nations Children’s Fund.9 WHO, UNICEF, 2009. See footnote 8.10 Myatt M, Duffield A, Pasteur F. The effect of body shape on

weight-for-height and mid-upper arm circumference based

case definitions of acute malnutrition in Ethiopian children.

Ann Hum Biol. 2009 Jan-Feb;36(1):5-20.11 See footnotes 5 and 8.

Table 3: Prevalence of acute malnutrition according to weight-for-height and MUAC by age group in three munic-ipalities in Central Mindanao (October-December 2010)

Field Article

Field Article

5

higher. It is possible that morbidity plays a rolein this relationship.

Malnutrition can be caused by lack of foodand/or infection. While in both cases childrenlose weight, it is possible that effect on bodycomposition and thus arm circumferencediffers. In malnourished children, once glyco-gen stores are used, fat becomes the main fuelfor energy production and protein breakdownis reduced. During infection, protein turnover isincreased due to an increased need for proteinsinvolved in the body’s response to infection. Ithas been shown that while protein breakdownand synthesis are reduced in malnourishedchildren, it increases if the malnourished chil-dren have an underlying infection14. Muscle is agood source of protein and MUAC is directlyrelated to arm muscle mass. It is thereforepossible the increased protein breakdown ininfected malnourished children leads to a lowerMUAC than uninfected malnourished children.This might also explain why MUAC is a betterpredictor of mortality than weight-for-height.Furthermore, it is possible that in locationswhere infection is generally high, for exampledue to an unhealthy environment or highprevalence of HIV/AIDS, MUAC leads tohigher estimates in malnutrition than in othercontexts.

MUAC, weight-for-height, or both?The WHO recommends that where differencesbetween these two indicators exist, both shouldcontinue to be used. Some programmes haveused a two stage approach where MUAC hasbeen used for referral and weight-for-height foradmission to programmes15. This has led to theproblem of rejected referrals and therefore someorganisations are now using both as admissioncriteria. While such an approach is possible inpopulations where prevalence of acute malnu-trition is higher according to MUAC, in thePhilippines this would not be possible. UsingMUAC for identification and referral at thecommunity level would mean that, asmentioned above, a very large proportion ofchildren that are malnourished according toweight-for-height would be missed and thecaseload would be so low that it would be diffi-cult to justify the need for a programme. On theother hand, using only weight-for-height foractive case finding, referral and admission iscostly, takes more time and requires muchbetter trained staff. One possibility might be toincrease the MUAC cut-offs to a value thatcorresponds better to weight-for-height.

However, Myatt warns us not to automati-cally accept weight-for-height as the goldstandard because it can overestimate the preva-lence of malnutrition in some populations, asshown in pastoralist populations in Ethiopia(see above). So the question that we face in thePhilippines is not how best to find childrenwith a low weight-for-height in the communitybut what is the better descriptor of nutritionalstatus.

We must remember that both MUAC andweight-for-height are anthropometric measure-ments and are not direct indicators ofnutritional status. Weight-for-height comparesweight to height and while it is very likely thata child with a weight-for-height of less than -3z-scores is malnourished in some form oranother, a child with a z-score of greater than -2is not necessarily well nourished. Weight-for-

height cannot tell us anything about nutrientstores and body composition, the latter ofwhich does influence weight. Muscle, for exam-ple, weighs more than fat. Body Mass Index(BMI), another anthropometric indicator thatcompares weight to height, has been known toclassify healthy athletes with a low percentageof body fat as obese. Contrary to weight-for-height, MUAC is directly related to musclemass16 and therefore gives us some idea aboutbody composition. In addition it is a betterpredictor of mortality than weight-for-heightand is independent of body shape. Admittingchildren based on MUAC might therefore focuslimited resources to those most at risk of death.However, it is well known that with the 2006WHO standards, children are selected earlierwhich increases case load but improves chancesof survival. Another question that thereforeneeds to be answered is whether in thePhilippines (where many more children areclassified as severely malnourished accordingto weight-for-height than MUAC) waiting untilchildren have a MUAC of less than 115mmwould reduce their chances of survival.

ConclusionsIn the immediate term, as long as we do notfully understand the risk of mortality in chil-dren with a weight-for-height of less than -3z-scores but a MUAC greater than 115mm, allchildren classified as malnourished accordingto both indicators should receive treatment inthe Philippines. The easiest option to find asmany malnourished children in the communityas possible would be to increase the MUAC cut-off used for referral. This cut-off would need tobe carefully chosen to avoid the problem ofrejected referrals. Alternatively, weight andheight measurements could be used in addition

12 Van Herp M, Leurquin B, Verwulgen A, Delchevalerie P. Can

height-adjusted cut-offs improve MUAC’s utility as an

assessment tool? Field Exchange Issue 30, April 2007.13 Myatt et al, 2005. See footnote 3.14 J.C. Waterlow. Metabolic changes In: Protein Energy

Malnutrition. Smith-Gordon and Company ltd. 2006.15 Myatt et al, 2005. See footnote 3.16 Myatt and Duffield, 2007. See footnote 5.

to MUAC, which would ensure that childrenclassified as malnourished according to bothindicators would be found. Since it will not bepossible for community volunteers to carryheavy height boards, measuring tapes or heightsticks have been suggested as alternatives. Theproblem is, of course, that height measurementswould not be as accurate as with the heightboards, active case finding would be more timeconsuming and it is unclear how many of thebarangay nutrition scholars (community volun-teers) are able to read weight-for-height tables.

Before a more conclusive recommendationcan be made in the Philippines, the weight-for-height/MUAC relationship in this populationneeds to be better understood. This will requireresearch into the influence of body shape andmorbidity on the weight-for-height/MUACrelationship. The appropriateness of the use ofWHO standards in this population should beinvestigated, as well as the risk of death in chil-dren with a weight-for-height of less than -3z-scores but a MUAC of greater than 115mm. IfWHO standards turn out to be appropriate andweight-for-height a better indicator thanMUAC, MUAC cut-offs need to be revised forthis population.

For more information, contact BernardetteCichon, email: [email protected] or ElisaDominguez, email: [email protected]

Length measurement of achild in the Municipality ofArakan, Mindanao

B C

ichon,

Philip

pin

es,

2010

Admission criteria in Therapeutic Nutrition Programmes in 2009Age 0-59 months

and WHZ <-3 (WHO) and/or MUAC <110 mm with a length of >65 cmand/or oedema grade 1 or 2

and no medical complicationsand appetite

Willing caregiver

Treatment phase 1 (usual protocol):RUTF given according to body weight

Systematic medical treatmentHealth education given to caregiver

Treatment phase 2: Alternative protocol: If MUAC > 110 mm and WHZ ≥-3 + 200g(<65cm) or +300g (>65cm) and nomedical complications, no oedema andappetite

1 sachet RUTF (92g)/ child/dayTreatment phase 2 (usual protocol)

Treatment Phase 2: Usual protocol: If oedema grade 1 & 2, if age <6months

RUTF given according to bodyweight Treatment phase 2 (usual protocol)

Discharge criteria15% weight gain after 2 consecutive weighings

MUAC > 110mm and WHZ ≥ -2

6

Research

Qualitative reviewof an alternativetreatment of SAM inMyanmar By Naomi Cosgrove, Jane Earland, PhilipJames, Aurélie Rozet, Mathias Grossiordand Cecile Salpeteur

Naomi Cosgrove has over 13 years of experience inthe food industry. in addition to experience in thehumanitarian sector, including a local Mental Health& Development NGO in Sri Lanka and as a FoodSecurity, Livelihood & Hygiene Advisor in NorthernArgentina. This research project was completed aspart of her Masters in Human Nutrition.

Jane Earland is a registered Dietitian and PublicHealth Nutritionist and works in nutrition andresearch at the Universities of Liverpool andSheffield. Her background includes nutrition educa-tion and training in Papua New Guinea for 11 years,Field Director for Save the Children and short termwork in Indonesia, Malaysia and Papua New Guinea.

Aurélie Rozet is a nurse trained in nutrition and hasbeen working with ACF since 2006 in Asia in partic-ular. She was a Nutrition Programme Manager inMyanmar at the time of the programme evaluatedin this article and now supports theACF FranceNutrition team in Paris.

Mathias Grossiord is a Public Health Nutritionist(MSc) and was a Nutrition Programme Manager inMyanmar at the time of the programme evaluatedin this article. He is now Nutrition ProgrammeManager for ACF in India.

Phil James was a Masters student with LSHTM in2010 analysing the performances of the alternativetreatment of SAM in Myanmar and is nowEmergency Nutrition Coordinator for ACF UK. He ispreparing a scientific article with ACF on this MScthesis.

Cecile Salpeteur is a public health nutritionist andis ACF HQ Operational Nutrition ResearchFacilitator and HIV focal point. She has six years ACFfield experience in implementing a wide range ofnutrition and food security programmes.

in some instances, better than when theStandard Protocol was used in 2008.However there were limitations to compar-ing these two data periods, mainly due to thedifferent standards being used (NCHS in2008 versus WHO in 2009). Nevertheless,results of the Alternative Treatment greatlyexceeded the international Sphere Standards.

Paediatricians and scientists who devel-oped the normal treatment protocol for SAMwere aware that that the quantity of RUTFsachets given to children was rounded up tothe higher figure and maintained throughoutthe treatment until complete recovery, inorder to simplify implementation by healthpersonnel. However, the nutritional needs ofthe child for catch up growth are expected todecrease as his/her nutritional statusimproves. Thus the quantities administeredin the latter stages of treatment are not fullyjustified from a nutritional perspective6.

Rationale for proposed study As the key drivers behind this successfulprogramme outcome were not fully under-stood, ACF wanted to carry out furtherstudy.

Aims & objectivesThe overall aim of this study was to identifythe factors that contributed to the success ofthe Alternative Treatment. The objectives were: 1. To identify all factors associated with

success of the intervention, as well as areas for improvement in the Myanmarprogramme using the Alternative Treatment.

2. To develop a feasibility grid system for identifying another country with these optimum conditions for further testing ofthe Alternative Treatment.

MethodsSee Figure 2 for a summary of the eight stagestudy design used in this research projectand described in more detail here.

1. A literature review was conducted toidentify new innovative tools and methodsfor qualitatively evaluating programmes andthe factors that may influence the success offeeding programmes. This was used toinform the development of the interviewguide and analysis7.

2. ACF documents and reviews werecollected and reviewed that included capital-

In 2009, Action Contre la Faim (ACF)treated an estimated 18,000 childrenunder five years for severe acute malnu-

trition (SAM) in Maungdaw and ButhidaungTownships and Sittwe, Rakhine state, west-ern Myanmar. This followed a change ofprotocol in January 2009 where the identifi-cation of malnourished children wasswitched from being based on NationalCentre for Health Statistics (NCHS) stan-dards to the 2006 WHO International ChildGrowth Standards (ICGS)1,2. As a result ofthis change, the number of children fallinginto the category of severe malnutritionincreased dramatically (a multiplicationfactor of 5.6) so that there was an increasedamount of product required to treat thesechildren. In addition, in April 2009, therewere complications with the import of theready to use therapeutic food (RUTF)(Plumpy’Nut©). This meant that there wasinsufficient stock to cover the case load ofSAM affected children until the end of theyear and ACF had to identify a solution tothe problem3.

Modified treatment protocolACF decided to modify the treatment proto-col and introduced a second phase oftreatment, once the child had improved froma severe to a moderate (MAM) stage ofmalnutrition (see Figure 1 for existing andmodified treatment protocols). Eligible chil-dren for this ‘Alternative Treatment’ wereuncomplicated MAM cases, without oedema,above six months of age and with increasingweight. The intake of RUTF for this secondphase was reduced from two or three sachetsper child per day (depending on body weightas defined in the usual protocol) to only onesachet and hence a reduction in kilocalories.This reduction ranged from 116% (< one yearold) to 62% (> four years old) of the child’sdaily energy needs, based on an averagerequirement of an adequately nourishedchild within that age category4. As a result,ACF staff advised caregivers to make up theenergy requirements of the child with foodavailable at home5.

This Alternative Treatment was imple-mented from July 2009 to January 2010 andthe data were collected, analysed andcompared to the same period the year before.Despite the reduced ration using theAlternative Treatment, the performance ofthe programme was found to be as good and

1 NCHS. (2011). National Centre for

Health Statistics. Retrieved 28.04.11,

20112 WHO. (2011). World Health

Organisation. Retrieved 28.04.11, 20113 ACF (2009). ECHO Report - Integrated

Approach to malnutrition through

nutrition, health and care practices.4 FAO, WHO, UNU Human energy

requirements: Report of a Joint

FAO/WHO/UNU Expert Consultation.

FAO: Rome, 20045 James, P. (2010). Evaluation of an

Alternative Protocol for the Treatment

of Severe Acute Malnutrition, imple-

mented by ACF Myanmar from July

2009 to January 2010 Masters in

Nutrition, London School of Hygiene

and Tropical Medicine. Also see

Footnote 3.6 Golden, M. (2011). RUTF Sell -

Prevention and Treatment of Severe

Acute Malnutrition Forum Area.

http://www.en-net.org.uk/question/

362.aspx also Footnote 4.

Figure 1: Summary of the admission and discharge criteria for theOutpatient Therapeutic Programme and the two phased treat-ment for the Alternative Treatment

Children attending astabilisation centre

N C

osgro

ve/A

CF,

Myanm

ar

7

isation reports from the relevant field ProgrammeManagers during 2009 and ACF Country OperationalStrategy Reports (2009 and 2011) from ACF HQ Paris.Notes were taken from these documents to inform the ques-tions for the interview guide. This information was alsoused to validate the data gathered from the interviews at alater stage, i.e. a triangulation approach8.

3. Using information gathered from steps one and two, theinterview guide with key questions was developed as a toolfor conducting the interviews. These questions needed tobe open ended to ensure accurate, non-biased answers.Probes and follow up questions were added where neces-sary, to ensure the question was fully answered. Seven key

areas were identified as areas to explorein the interview. Questions were devel-oped for each and were incorporated intothe interview guide. These areas were:• General introduction and context of

Myanmar.• Opinions and descriptions of the

Nutrition Programme• Management Style• Training and Capacity Building• Community Involvement• Other ACF programmes• Other international non-governmental

organisations (INGOs) working in the area.

4. In-depth interviews were conducted,lasting up to an hour, with key inform-ants across a range of professionaldisciplines to obtain an accurate andbalanced perspective of the programme,processes and relationships. Where possi-ble, the interviews were held face-to-face- this was not always possible due tologistics and therefore several were donethrough Skype. Key people interviewedincluded the Nutrition Advisor in ACFParis HQ, regional and local NutritionManagers, the local Human ResourcesManager for the programme, the Head ofBase (logistics and administration) andlocal Programme Managers in other ACFprogrammes.

5. A Pattern, Theme and ContentAnalysis method was used to analyse thedata collected. The analysis consisted ofidentifying core consistencies and mean-ings from the material and interviews.Patterns and themes were searched foracross all information provided (bothinterview transcripts and notes from therelevant documents) and re-occurringwords and texts were identified, includ-ing their frequency and the context inwhich they were used9. Data werepresented in the form of quotations withsufficient context to ensure that theycould be interpreted.

6. The early findings were discussed in aworkshop which included two highly

valued individuals who had a great dealof field experience. The five key ques-tions which were debated in theworkshop were:• How solid, coherent and consistent is

the evidence in support of the findings?

• To what extent and in what ways do the findings increase and deepen understanding of the situation/ success of the Alternative Treatment for SAM?

• How do these qualitative factors complement the quantitative outcomes and help to explain the success of the programme?

• To what extent are the findings consistent with other knowledge?

• To what extent are the findings useful for use in other programmes globally?

The initial findings were also sent to twoof the interviewees to get reactions andadditional comments.

7. A one page success factor matrix wasdeveloped and discussed in the work-shop. The aim of this matrix was tovisually represent the findings and corre-late them with the quantitative outcomes.

8. A simple, one page feasibility gridwas also developed and discussed in theworkshop. This involved creating a series

7 Literature review: Dersham, L. (2011). Design,

Monitoring and Evaluation - Save the Children.

Retrieved 05.05.11.

Draper, A. and J. A. Swift (2011). Qualitative

research in nutrition and dietetics: data collection

issues. J Hum Nutr Diet 24(1): 3-12.

Green, J. and N. Thorogood (2009). Analysing

Qualitative Data. Qualitative methods for Health

Research. D. Silverman, SAGE publications Ltd:

195-228.

Patton, M. Q. (2002). Qualitative Research &

Evaluation Methods, Sage Publications Ltd. .

Pilnick, A. and J. A. Swift (2011). Qualitative

research in nutrition and dietetics: assessing quality.

J Hum Nutr Diet 24(3): 209-214.8 Green, J. and N. Thorogood (2009). Analysing

Qualitative Data. Qualitative methods for Health

Research. D. Silverman, SAGE publications Ltd:

195-2289 Patton, M. Q. (2002). Qualitative Research &

Evaluation Methods, Sage Publications Ltd.

1. Identify new innovative tools & methods for evaluating Programs

qualitatively

2. Review all relevant existing documents

from ACF Myanmar

3. Use above two inputs to

develop an interviewguide

The analysis can bebroken down into 4further stages: • Inductive Analysis • Classification Framework Development • Descriptive Phase• Interpretative Phase

The interviews were:• Face-to-face andthrough Skype with apurposeful sample ofkey multi-disciplinarypeople, who weredirectly involved withthe programme.

4. Conduct in-depth interviews

5. Pattern, Theme & Content Analysis

6. Discuss findings in a workshop with supervisors

7. Develop aSuccess Factor Matrix

Local Context

CommunityInvolvement

ACF Nutrition

ProgrammeDesign &Management

Caregiver/Mother

BNF/Child

8. Develop a one Page Feasibility Grid for replication globally

• Ideal Timing with absence of natural disaster & ‘normal’HH security

• ACF Well Known with Good Reputation in the region

• Community Sensitisation, Mobilisation & Support• Early Referrals

• Adoption of WHO Standards• Time for Careful Planning• Staff Training & Capacity giving High Quality of Care• Simple, Organised Processes• Consistent, Experienced & Strong Leadership

Local Context

LocalContext

Idealtiming

Communitysensitisation &

mobilisation

Adoption ofWHO

Standards

*90% RecoveryRate

* Quantitative outcomes were measured from admission to discharge. * The remaining % of children: 6.71% had an unknown outcome, 0.07% transferred due to medical complications and 0.91% were non

responders* Children spent a median of 14 days on phase 1 and 21 days on phase 2 of the Alternative Treatment

Levels ImpactQuantitative OutcomesQualitative Inputs

A simple, highly successful, cost effective programme with outcomes that exceed Sphere Standards

0% MortalityRate

*2% DefaultRate

*Av. 42 days inTreatment

4 Month followpp with dry

Rations

Close monitoring& use of BNR

Involvement ofthe mother andhome feeding

The productquantity

Integration ofcare practice

Staff training& capacity

building

Highlyorganisedprocesses

Earlyreferrals

ACF well known withgood reputation

Time forplanning

Consistent, experienced &

strong leadership

CommunityProgramme

Design &Management

Caregiver/Mother

Beneficiary/Child

CommunityInvolvement

ACF Nutrition

ProgrammeDesign &Management

• Work to minimise regional constraints further andmore advocacy to help the discriminated population

• Develop a strategy to improve relationship withgovernment at all levels

• Further integration with other ACF programmes aswell as co-ordination with other INGOs in the area andfocus on a more global, preventative longer termapproach

• Improve Community Ownership further in terms ofaddressing the root causes of malnutrition and educatingthem on the reasons why the programme exists.

• Consider ways to simplify the protocol further for easeof use

• Develop new ways of increasing staff loyalty further and ensure complete empowerment

• Although there was a good level of staffing, there waslimited resource for home visits, which is a key need

• Ensure the focus is on curing the children and not justgood results to report

• Integration of Care Practices giving psycho-social support• Involving the Mother and Home Feeding in the

Recovery Process

• Close Monitoring & Use of BNR Methodology• The Product Quantity was more likely to be eaten by the

child• 4 Month Follow-Up with Dry Rations

BNF: BeneficiaryThese success factors were then placed at the relevant levels in terms of local context,community involvement, the Nutrition Programme, the mother/caregiver and the BNF/child.

(Patton 2002)

Figure 2: Eight Step Study Design

Figure 3: Placing the 14 success factors at the relevant level

Figure 4: Placing the eight areas for improvement at the relevantlevel

Figure 5: The Success Factor Matrix

Research

8

of questions and a scoring system. The aim ofthis grid is to identify another country withoptimum conditions for further testing theAlternative Treatment.

ResultsThe fourteen factors that were consistentlyidentified as key contributors to the success ofthe Alternative Treatment are summarised inFigure 3. Eight areas of improvement were alsoidentified and are summarised in Figure 4.

DiscussionThe discussion below focuses on the keyoutcomes including the Success Matrix and theFeasibility Grid and also raises additional find-ings and considerations from the research.

The Success MatrixThe Success Matrix was developed to build apicture of the data as a whole to aid systematicanalysis and link the qualitative findings to thequantitative outcomes (see Figure 5).

The Feasibility GridThe Feasibility Grid needed to be in a spread-sheet form that could be used for identifyingother countries that also have optimum condi-tions for further testing the AlternativeTreatment. Input and feedback was given by allthose people that would be using it to ensure itwas user-friendly and relevant.

The grid was developed as a series of ques-tions to be answered by Country NutritionCo-ordinators for each country being consid-ered for replicating the Alternative Treatment.There is a simple scoring system. If the answeris yes to a question, 1 point is given and if no, 0points are given. This enables each country tobe ranked, in order to identify the optimumcountry for further testing the alternative proto-col. All questions were given the sameweighting although weighting might be consid-ered in future versions of the grid. The countrythat scores the highest points is the recom-mended country.

The grid could potentially be adapted andused for on-going programmes in the future,including as an annual quality check. A portionof the spreadsheet is shown in Table 1. The fullspreadsheet is available on request from theauthor.

Discussion on what makes a good qualityprogrammeResponses were not consistent to the interviewquestion on what makes a good qualityprogramme. However there were several recur-ring themes, including good programmemanagement, staff training, educating the care-giver on the causes of malnutrition and caringfor the child, and having a preventative strat-egy in place. The review of published papers onprogramme effectiveness indicated that therehas been very little discussion at an interna-tional level on programme quality and theimpact this has on the outcomes. There appearsto be an emphasis on quantitative outcomesand limited understanding of the factors whichcontributed to these outcomes, i.e. the ‘how’and ‘why’ questions, which are typicallyaddressed through qualitative research. Thissuggests a need for a clear definition and set ofguidelines regarding programme quality, aswell as more published studies examining thequalitative aspects of humanitarianprogrammes globally.

Cost saving of using less RUTF versusadditional time and resource needsAlthough there were some cost savings in usingthe Alternative Treatment in terms of product,the research shows that more time and resourceare required by the staff and management toensure successful implementation. Althoughthe amount of additional time and resources toimplement the alternative protocol in theMyanmar programme are believed to be rela-tively low due in part to exceptionalmanagement and highly competent, welltrained staff, this will not always be the casegoing forward in other missions and countries.

Ideal setting in MyanmarThere are a number of reasons as to why thiswas an ideal setting for testing the AlternativeTreatment. These included absence of naturaldisaster during 2009, home feeding being partof the treatment, and existence of well trainedand experienced staff. However it is importantto consider that if any of these factors were tochange, the outcomes could be very different. Itis also worth noting that ACF had completecontrol of the programme and did not dependon the government and local amenities for anyaspect of programming. This will not always bethe case, especially as a key objective for manyINGOs today is to empower local governmentsso that they can implement nutritional services.

Sustainability of the programmeSome of the constraints to sustainability includethe high turnover of local medical staff, the highcost of the RUTF product (over 50% of totalprogramme costs)10 and the need to import prod-uct. Insufficient government involvement in theprogramme also impedes sustainability. Thissuggests the urgent need for strategic review.Despite high levels of community awareness,there appears to be no improvement in overallmalnutrition rates in the intervention area of

Level Questions Yes No Comments

LocalContext

• Has it been confirmed that there are no natural disasters or any other potential/planned risks in the coming 6 months, which could jeopardise household (HH) FoodSecurity?

• Does each HH have food access and its availability guaranteed for 6 months (whetherthrough good agricultural season or through food assistance)?

• Has each HH access to drinkable water in the area of CMAM implementation andguaranteed for 6 months?

• Are health services available and functioning in the area for the next 6 months? • Is the security level of the mission and base =< level 2?• Is ACF established and been active in the area for at least 1 year?• Does ACF have a good working relationship with national & local authorities as well

as humanitarian actors in Health and Nutrition in the area?

CommunityInvolvement

• Have WHO international standards been adopted? If not possible, has considerationbe given to earlier identification of children using NCHS ref. but with a cut-off of.WHZ < -2.5?

• Has the community been sensitised and mobilised in the area in terms of awareness,education and support of the programme?

• Have key members from the community been identified and trained as communitycaregivers for continuous screening of children and ensuring early referrals to thecentres? If not, has this been built into the planning stage of the project?

Table 1: Further testing the Alternative Treatment: A sample of the Feasibility Grid

ACF in Myanmar since the programme began in2003 and children continue to be admitted.

ConclusionsThe aim and objectives of this study weresuccessfully achieved using both review ofdocuments and in-depth interviews. The studyhas shown the importance of combining quali-tative and quantitative research to give acomprehensive picture and meaning to thefigures. This combined learning has led to adeeper understanding of the AlternativeTreatment.

The results from this research project havegiven invaluable insights into the AlternativeTreatment of SAM. The findings confirm that itis not only the high quality of RUTF itself whichis necessary for success, but a large number ofquality considerations some of which may bespecific to the local context in Myanmar. It is acombination of all of these factors that resultedin the quantitative outcomes far exceeding theSphere Standards and it is these factors whichneed to be considered before the AlternativeTreatment is replicated globally.

This research project indicates that thereneeds to be a stronger emphasis on nutritionprogramme design. Key design issues andfactors include careful planning, existence ofwell trained staff and community mobilisation.

The findings have also shown that thecurrent programme is not sustainable inMyanmar and that future strategies mustaddress this challenge.

For more information, contact; Cécile Salpéteur,email: [email protected]

10 James, P. (2010). See footnote 5.

Research

security situation until May or June, andpredicted “localised famine conditions [insouthern Somalia], including significantlyincreased child mortality… if the worst casescenario assumptions are realised”. TheFSNWG also warned that “failure of the Marchto May rains is likely to result in a major crisis”.At this stage, humanitarian actors were advisedto begin large-scale contingency/responseplanning immediately and to implementexpanded multi-sectoral programming. Yet thiscall was not adequately heeded.

The various early actions taken by nationalgovernments affected by the crisis are set out inthe briefing paper, e.g. in Ethiopia, the govern-ment’s Agricultural Task Force, supported bythe Food and Agriculture Organization of theUnited Nations (FAO), developed a roadmapfor interventions in early 2011. Meanwhile inthe regional autonomous state of Puntland, thePresident announced a drought emergency inNovember 2010, and called on the internationalcommunity and aid agencies to providehumanitarian assistance. However, it is theinternational response which the briefing paperfocuses mainly upon.

The donor response at scale was too slow.Figure 1 shows the level of humanitarian fund-ing to Ethiopia, Kenya, and Somalia in thelead-up to the crisis. An increase can be seenafter the first warnings in late 2010 and the UNConsolidated Appeals Process (CAP) appeal(November 2010). But it was only after majormedia coverage in June/July 2011, and after theUN declared a famine in Somalia, that donorsdrastically increased the funds available. Somedonor representatives in the region were alsoaware of and acted on the impending crisismuch earlier. For example, key donors in Kenya– European Commission Humanitarian Aid &Civil Protection (ECHO), UK Department forInternational Development (DFID) and UnitedStates Agency for International Development(USAID) – met in December 2010 to co-ordinatetheir initial response. Indeed, Oxfam accessedECHO funding for work in Turkana, Kenya, inApril 2011, and this was quickly scaled up inJuly. Nevertheless, most donors were not ableto access funding at scale from their headquar-ters until malnutritionrates were at dangerouslevels and media atten-tion broke the story.

In Somalia, a furthercomplicating factor waspresent: the internationalcommunity failed toprioritise growinghumanitarian concernsover political considera-tions, reflected in Oxfamand Save the Childrenstruggling to find fund-ing for work in SouthCentral Somalia.

Figure 1: Humanitarian funding for Ethiopia, Somalia, and Kenya, May 2010 toOctober 2011

9

According to a briefing paper justreleased by Oxfam and Save theChildren UK (SC UK), the 2011 crisis in

the Horn of Africa has been the most severeemergency of its kind this century, with morethan 13 million people still affected andhundreds of thousands at risk of starvation.The authors assert that this crisis, affectingprimarily the drylands of Ethiopia, Somalia andKenya, unfolded despite having beenpredicted. All three countries were almostequally affected by the total failure of theOctober to December 2010 rains and the poorperformance of the March to May 2011 rains,resulting in crop failure and animal deaths. Thesituation was much worse in southern andcentral Somalia, where conflict further impededtraditional drought coping mechanisms, andreduced access for humanitarian agencies. Thebriefing paper is principally concerned withhow the international system responded (ornot) to early warnings of the oncoming crisis,and why it was allowed to spiral into a disaster.

The countries affected by this drought werein very different situations. For example, inEthiopia, there has been considerable effort tobuild resilience through the development of theProductive Safety Net Programme (PSNP),investment in new health posts which enabledhuge increases in access to nutrition responses,investment in pastoral areas through thePastoral Community Development Programme(PCDP) and the promotion of disaster riskmanagement policy and practice. This contrastswith Somalia, where such work has beenlargely absent, due to access restrictions, acomplex environment and the unwillingness ofdonors to invest.

The emergency in the Horn of Africa in 2011was not a sudden-onset crisis. Forecasts of theimpending crisis started in August 2010, aschanging weather conditions linked to the LaNiña phenomenon were confirmed. Thesepredictions became more strident in earlyNovember 2010, when the October toDecember short rains were forecast to be poor.This prediction was accurate, prompting theFood Security and Nutrition Working Groupfor East Africa (FSNWG) to set up a La Niñatask force. In December 2010, it stated that “pre-emptive action is needed to protect livelihoodsand avoid later costly lifesaving emergencyinterventions” and called on the humanitariancommunity (donors, United Nations agencies,non-governmental organisations (NGOs)) “tobe prepared now at country level.”

Multi-agency scenario planning took placein February 2011. A Famine Early WarningSystems Network (FEWSNET) food securityalert dated 15 March 2011 made it clear that thecurrent situation was already alarming andwould deteriorate further if the March to Mayrains were as poor as expected. It stated thateven average rains would lead to a critical food

The UN’s humanitarian appeal in November2010 seriously underestimated the number ofpeople in need of emergency aid. This is partlybecause the timeline of UN appeals is notaligned with the seasons in the Horn of Africa.Assessments were carried out in September,before the failure of the short rains (whichnormally start in October) and did not take intoaccount the future weather predictions. And forSomalia, more recent appeals were based moreon what programming can be achieved (withinthe constraints of access and partners) ratherthan what funding would be required to avertdisaster, thus potentially giving a misleadingpicture of needs within the country. TheConsolidated Appeal (CAP) – a key documentfor marshalling donor resources – was only fullyrevised at the end of July 2011. This was clearlya factor in the failure to scale up the responseearly on. In Somalia, for example, the original2011 CAP was set at $530m in late 2010. Thiswas revised to more than $1bn by August 2011.

Many agencies, including Oxfam and Save theChildren, had begun a small-scale response byDecember 2010, and tried to focus internationalattention on the impending crisis. But whilesome performed better than others, most agenciesdid not adapt their programming on a sufficientscale to meet the level of need over the followingsix months, and did not begin to respond at scaleuntil after the 2011 rains failed in May. Someagencies declared the situation a corporate prior-ity as early as February 2011, but this onlyhappened in Oxfam and Save the Children at theend of June and early July 2011 respectively.

Pressing domestic, regional and interna-tional developments, including the conflict inSomalia, the Arab Spring uprisings, globalrecession, other crises such as the Japan earth-quake/tsunami, or donor fatigue, may havedelayed the international community’sresponse to the drought.

While the early warnings were clear, thescale (numbers of people) and depth (severity)

Dangerous delay in responding to Hornof Africa early warnings of droughtSummary of briefing paper1

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

US$

, bill

ions

May

10

Jun

10

Jul 1

0

Aug

10

Sep

10

Oct

10

Nov

10

Dec

10

Jan

11

Feb

11

Mar

11

Apr

11

May

11

Jun

11

Jul 1

1

Aug

11

Sep

11

Oct

11

1 A Dangerous Delay. The cost of late response to early

warnings in the 2011 drought in the Horn of Africa. Oxfam,

Save the Children UK. Joint agency briefing paper. 18

January 2012. Available from:

http://policy-practice.oxfam.org.uk

February 2011: Further warnings:FEWSNET issues alert that poor rainsare forecast for March to May

Early warning signs:FEWSNET alert of poorrainfall and worsening foodsecurity

30 May: Kenyangovernment declaresthe drought a nationaldisaster

July 2011: UN declaresfamine in 2 regions ofSouth Central Somalia

Source: OCHA Financial Tracking Service

Research

10

of the crisis still caught many by surprise. Thisis partly because needs assessments carried outby UN agencies or governments – which are akey driver for donor interventions – arepublished several months after the assessmentis done and critically, do not incorporate fore-casts or predictions based on a changingsituation. Thus the UN appeal for Somalia,launched in November 2010, had relatively lowfigures for those in need of assistance in 2011and failed to sufficiently reflect the La Niñapredictions. Ultimately, the EWSs performed,but decision makers chose not to respond.Possible reasons for reluctance to respond earlymay include:• Fear of getting it wrong – with both

financial and reputational risk at stake.• Fear of being too interventionist – undermining

communities’ own capacities to cope.• Fatigue – “there are droughts every year” –

encouraging an attitude of resignation to the high levels of chronic malnutrition, and an inability to react to the crisis triggers.

All humanitarian actors – governments, UNagencies, donors, implementing NGOs – wantto be certain about the scope and depth of alooming food crisis before responding at scale.The international humanitarian system onlybecomes fully operational when IntegratedPhase Classification (IPC) phase 4 – ‘emer-gency’ – has been reached2. Responding on thebasis of forecasts instead of hard data requires ashift in dealing with uncertainty. Forecastsinvolve uncertainty. Yet this uncertainty is notunquantifiable – standard risk managementtechniques allow us to convert this uncertaintyinto risk, which can then be managed andminimised. Figure 2 shows a typical riskimpact/probability chart, which plots the proba-bility that a hazard will occur against its impact.

Using this logic, it would have been clearfrom around January 2011 that the high proba-bility of poor March–May rains in the Horn ofAfrica, magnified by the failure of the previousrains in late 2010, would constitute a critical riskthat needed to be addressed immediately.

Once the EWS has flagged a potential prob-lem, this should immediately activate a processof further investigation – detailed monitoringwhich can be used to design interventions – andthe operationalisation of emergency plans.These plans need to be clear on who should dowhat and when. There should be a commonapproach to using triggers, so that decisionmakers know exactly what they ought to bedoing as the situation deteriorates and theconsequences if they fail to act on those triggers.

All actors need to work together to developa system of triggers that:• recognises the national government (where

possible) as primary duty bearer for meetingcitizens’ food needs

• reflects the high levels of chronic malnutritionin some areas

• reflects the exponential rather than linear development of malnutrition

• does not lead to interventions that underminecommunities’ capacity to

• cope• is context-specific for different livelihoods

zones• is agreed between different actors

It is expected that there will be a range of trig-gers for different sorts of response. So, for

example, at an early stage the trigger might befor advocacy, but as the situation deteriorates, itmight be for a livelihood response, and subse-quently for a food/nutrition response.

Long-term programmes are in the best posi-tion to respond to forecasts of a crisis. There areestablished links with communities and thusthe vulnerabilities and the complexities areunderstood. There are staff and/or partners inplace, they are in a trusted position with donorswith funding available and their work has beennegotiated with government bodies. Long-termprogrammes should become more sensitive todrought risks and seek to reduce vulnerabilityby reducing the underlying risk factors. Whenrisk analysis is made integral to long-termprogramme design – by using Drought CycleManagement (DCM) or similar tools – droughtscan be seen as an integral part of the livelihoodsystem, rather than as an unexpected shock.

Early humanitarian response, which seeks toreduce disaster risk, is both effective and cost-effective in addressing the underlying factorsthat make people vulnerable.

Humanitarian interventions which shouldbe started on the basis of forecasts include:• Livelihood protection interventions, in

order to safeguard key assets.• Interventions involving a significant time

lag. If food distributions are likely to be necessary, the process of mobilising resourcesand arranging logistics should begin on the basis of early assessments, with quantities being revised at a later stage.

• ‘No regrets’ options: measures that build capacity and disaster preparedness but haveno negative effect even if the worst forecastsare not realised (either because the cost is very low or because they will build resilience).This would include activities such as- putting human resources systems in

place- talking to existing and potential donors

and drafting outline proposals for response

- building links with private sector partners,e.g. developing standing agreements with money transfer companies, starting tender processes

- building and strengthening the capacity of local partners

- starting to engage with all relevant stakeholders, such as partners, local authoritiesand the women, men and children in communities

- practical measures like assessing boreholeoperations, prepositioning stocks, marketassessments, mapping the capacity and coverage of traders, etc

• Flexible funding measures that involve a

level of expenditure commensurate with theprobability of the crisis occurring. Projects could be designed and initiated with limited but sufficient start-up funding, with a commitment to more substantial resourcesas the crisis develops.

Early response requires us to move away fromthe traditional distinction within the aid systembetween development and humanitarian work.This approach, with different staff, mandates,skillsets, timescales, budgets and beneficiaries,is not valid in regions like the Horn of Africa. Toincrease the effectiveness of the aid system, thisartificial gap must be bridged.

Skilled and experienced staff and partnersare needed who are able to build risk analysisinto their work and are thus able to adapt whatthey do, and how they do it, as the situation andneeds change.

Much greater investment is needed in long-term joint efforts to strengthen governmentcapacity, both in disaster risk management andcoordination, but also in improving the abilityof long-term development work in all sectors tobuild resilience.

Humanitarian and development strategiesare often developed separately, whereas a riskmanagement approach requires common think-ing and planning. Where structures areinstitutionally divided, an effective coordina-tion and integration approach with variousmechanisms for direct cooperation, jointprogramming and implementation (in combi-nation with shared learning cycles) can help tomerge development and response.

Neither humanitarian nor developmentfunding streams are ideally suited to the situa-tions of chronic vulnerability that occurregularly in the Horn of Africa, where the situ-ation is often in transition betweenhumanitarian emergency and development.Humanitarian programmes are short-term,which doesn’t allow for longer-term planning,but they are usually fairly flexible in terms ofprogramme approach and the ability to changeexpenditure. Development programmes arelong-term but less flexible. Implementing agen-cies are required to predict their expenditure atthe start, with often only a small contingency(for the EU, this is a maximum of five per cent).This is designed to boost accountability, but itdoes hinder flexibility and agile programming.Some emergency aid donors have made consid-erable efforts to be more flexible, withinnovative funding mechanisms to supportrecovery and resilience. Other donors shouldemulate these models.

Donors – including the UN for the CentralEmergency Response Fund – need to revisittheir mandates and protocols for fundingstreams and continue to push the boundaries,so that they can disburse sufficient fundsquickly to support early response.

The ‘bottom line’ of this analysis is plainlystated: predictions about the impact of the2010–11 drought in the Horn of Africa wereclear, and unfortunately, much of what hashappened was preventable. The scale of deathand suffering, and the financial cost, could havebeen reduced if early warning systems had trig-gered an earlier, bigger response.

2 http://www.ipcinfo.org/

Prob

abili

ty o

f occ

urre

nce

Source: Mind Tools

Figure 2: Typical risk impact/probability chart

Low-level risk

Impact of riskLowLow

High

High

Medium-level risk

Critical risk

Research

11

Community casemanagement of severeacute malnutrition insouthern BangladeshSummary of study1

Bangladesh has the fourth-highest numberof children (approximately 600,000 at anyone time) suffering from severe acute

malnutrition (SAM) in the world. Currently,ongoing national programmes (such as theNational Nutrition Programme) do not includean effective mechanism of identifying or treat-ing young children who suffer from SAM.

A recent prospective cohort study aimed toexamine the effectiveness and feasibility ofadding the diagnosis and treatment of SAM tothe community case management (CCM) pack-age delivered by community health workers(CHWs) outside health facilities in Barisal,Bangladesh. Research goals included: • To compare the effectiveness (i.e. the rate of

recovery) of treatment of SAM provided by CHWs with that provided by the standard of care for SAM in Bangladesh.

• To compare the cost effectiveness of CCM ofSAM provided by CHWs with that of the standard of care for SAM in Bangladesh.

• To estimate the coverage of CCM of SAM provided by CHWs.

• To examine the quality of care (error-free case management) delivered by CHWs for cases of SAM.

Barisal Division in southern Bangladeshcomprises six districts and with a population ofabout eight million people, is among the poor-est in the country, with alarmingly high rates ofacute malnutrition among children under five.Save the Children USA (SC US) have beenworking in the Division since June 2004 andbetween 2004 and 2010, have implemented asix-year Development Assistance Programmenamed ‘Jibon o Jibika’ (‘Life and Livelihoods’ inBangla) in three Districts. As part of thisprogramme, SC US employed a cadre of CHWs,all local women educated to grade eight, todeliver preventive and curative care to childrenin the target Districts. Interventions includedCCM of basic childhood illnesses such as diar-rhoea and acute respiratory infection (ARI),monthly growth monitoring and promotion(GMP) sessions, and household-level educationand counselling around infant and young childfeeding, health, and sanitation.

SC US and Feinstein International Centre(FIC) at Tufts University were given permissionby the Government of Bangladesh (GoB) andthe Institute of Public Health Nutrition (IPHN)to pilot the community case management ofSAM (CCM of SAM) in Burhanuddin Upazilain one of the JoJ target Districts (Bhola District).In a neighbouring Upazila (Lalmohan) in thesame District, the Upazila Health Complex

(UHC) was supported to provide inpatienttreatment for children with SAM according toNational Guidelines and to compile monitoringdata on referrals and outcomes of treatment.This Upazila received exactly the same supportby the ‘Jibon o Jibika’ programme apart fromsupport for the CCM of SAM.

In both Burhanuddin (the interventionUpazila) and Lalmohan (the comparisonUpazila) a mid-upper arm circumference(MUAC) measure and an oedema check for allchildren <3 years old was introduced into allroutine CHW activities. These included themonthly GMP sessions and household visits forcounselling and treatment of sick children.CHWs also discussed SAM and its conse-quences with different groups of communitymembers in ongoing counselling and mobilisa-tion activities.

MethodThe study ran between June 2009 and June2010. All children more than six months in agethat were identified as suffering from SAM byone of the 261 CHWs working under the SC USprogramme in Burhanuddin Upazila were eligi-ble for the intervention. Any child identifiedwith SAM with appetite and no medicalcomplication was treated directly by the CHWwith ready to use therapeutic food (RUTF). Anychild with SAM with medical complications,such as the absence of appetite, was referred tothe UHC to receive inpatient stabilisation care.In the comparison Upazila, all children identi-fied with SAM by CHWs were referred to theUHC.

Informed consent was obtained from allparticipating caregivers before recruitment.This involved the CHW discussing a verbalconsent form with groups of mothers beforeeach growth monitoring session and with indi-vidual caregivers at household visits. This formexplained the objective of the study and theprocedures for any child identified with SAM.

Children were discharged from treatment asrecovered once MUAC was assessed as morethan 110 mm and they had gained at least 15%of their admission weight for two consecutiveweeks. Children admitted with nutritionaloedema were discharged once oedema wasabsent for two consecutive weeks and theirMUAC was assessed as more than 110 mm.

All CHWs in the intervention and thecomparison Upazilas participated in a two-daytraining which covered the causes and conse-quences of SAM, the standardisedmeasurement of MUAC and how to check for

nutritional oedema. CHWs in the interventionUpazila were also trained on the classificationof SAM and the use of nutritional and medicalprotocols for its treatment. Subsequently,CHWs in the intervention Upazila met withtheir supervisors every month to discuss prob-lems, submit monthly reports, and receive anew stock of therapeutic food and medicines.

At the UHC in both the intervention and thecomparison Upazilas, core medical staff partici-pated in a two-day training that covered thecauses and consequences of SAM, the standard-ised measurement of MUAC and how to checkfor nutritional oedema, and the nutritional andmedical protocols for the inpatient treatment ofSAM. In both Upazilas, SC US supplied theequipment and all ingredients for therapeuticmilk. In the comparison Upazila, SC US alsoprovided one additional care assistant whosesole job was to care for children with SAM andcounsel caregivers on child feeding and carepractices.

In the intervention Upazila, CHWs weretrained to use a simple algorithm that classifiedchildren into two groups: SAM with complica-tions and SAM without complications. Anychild with SAM with complications wasreferred to the UHC to receive one to four daysof inpatient treatment with therapeutic milksand medication. Once complications wereunder control, children were referred back tothe CHW to complete treatment. Any child withSAM without complications was seen weekly intheir homes by a CHW and treated with RUTF.

All dietary treatment for any child admittedto the UHC was administered according to theBangladesh National Guidelines for inpatientmanagement of SAM. In the interventionUpazila, for children suffering from SAM withcomplications, this included an initial phase(phase 1) of treatment in the UHC. Locally-prepared Formula 75 (F75) containing 75kcal/100 ml/day was given over 12 feeds perday. The child was discharged back to theirCHW where treatment continued with RUTF athome when the following conditions were satis-fied: good appetite, oedema reducing andinfection under control.

For all children treated by the CHW, RUTFwas provided as a weekly ration in proportionto a child’s weight. The CHW used a simplechart to calculate the correct ration size which

1 Sadler.K et al (2011). Community Case Management of

Severe Acute Malnutrition in Southern Bangladesh. Save the

Children and Feinstein International Centre. June 2011

Research

Measuring MUAC

SCU

S/F

IC,

Bangla

desh

, 2009

12

Research

Mortality risk factorsin severely-malnourished children hospitalisedwith diarrhoeaSummary of published research1

provided 175-200 kcal/kg/day and 4-5gprotein/kg/day.

All medical treatment followed proto-cols as specified in the National Guidelinesfor the Management of SeverelyMalnourished Children in Bangladesh.This includes a single oral dose of folic acid(5 mg) and the broad-spectrum antibioticCotrimoxazole oral (Trimethoprim 5mg/kg and Sulphamethoxazole 25mg/kg) given twice a day for five days.Albendazole and vitamin A were onlygiven where there was no record of thechild receiving these treatments during thetwice yearly Vitamin A+ campaigns thatare common in the target area. All medica-tion was prescribed by the UHC staffduring inpatient management and by theCHW during outpatient management.

For cases of SAM without complicationsin the intervention Upazila, the antibioticwas administered by the caregiver athome. The CHW instructed each caregiveron when and how to give the drug. Forcases of SAM with mild pneumonia ineither the intervention or the comparisonUpazila, the trained CHW provided treat-ment with Cotrimoxazole following CCMof ARI and Diarrhoea guidelines.

ResultsResults show that when SAM is diagnosedand treated by CHWs, a very high propor-tion of malnourished children can accesscare and they are very likely to recover.The main outcome measures including thehigh recovery rate (92%) and low mortalityand default rates (0.1% and 7.5% respec-tively) are all considerably better than theSphere international standards for thera-peutic feeding programmes and comparefavourably with other community-basedmanagement of acute malnutrition(CMAM) programmes across the world.The also compare favourably to previouswork that has examined the outpatientrehabilitation of children suffering fromSAM in Bangladesh. The level of coverageseen in this program,e was 89% (CI78.0%–95.9%) by April 2010; this is one ofthe highest rates of coverage ever recordedfor similar programmes. In contrast, moni-toring data in a comparison Upazila (anadministrative subdivision of a district),where the standard of care (facility-basedtreatment) was the only mechanism fortreating SAM, showed that most childrenreferred never made it to the facility or, if

Acase-control study conducted in theDhaka Hospital of the InternationalCentre for Diarrhoeal Disease

Research, Bangladesh (ICDDR,B) to identifythe risk factors of mortality in severelymalnourished children hospitalised withdiarrhoea has recently been published. Onehundred and three severely malnourishedchildren (weight-for-age <60% of median ofthe National Centre for Health Statistics refer-ence) who died during hospitalisation werecompared with another 103 severelymalnourished children who survived. Thesechildren were aged less than three years andadmitted to the hospital during 1997. Onadmission, characteristics of the fatal casesand non-fatal controls were comparable,except for age. The median age of the casesand controls were six and eight monthsrespectively (p=0.05). Patients with low pulserate or imperceptible pulse had three timesthe odds of death compared to the controlgroup (p<0.01). The presence of clinical septi-caemia and clinical severe anaemia had 11.7and 4.2 times the odds of death respectively(p<0.01). Patients with leukocytosis(>15,000/cm3) had 2.5 times the odds of death(p<0.01).

Using logistic regression, clinical septi-caemia {adjusted odds ratio (AOR) = 8.8,confidence interval (CI) 3.7-21.1, p =<0.01},hypothermia (AOR = 3.5, CI 1.3-9.4, p < 0.01),and bronchopneumonia (AOR =3.0, CI1.2-7.3,p <0.01) were identified as the significant riskfactors of mortality. Severely malnourishedchildren (n=129) with leukocytosis, impercep-tible pulse, pneumonia, septicaemia, andhypothermia had a high risk of mortality.

The strength of this study is that it was awell-defined matched case-control studywhere the cases and controls were selectedfrom among severely malnourished childrenwith diarrhoea. However, the limitation of thestudy is that the data used were not represen-tative of mortality among all severelymalnourished children with diarrhoea inBangladesh.

The authors conclude that septicaemia,pneumonia, and hypothermia were high riskfactors of death among diarrhoeal childrenwith severe malnutrition and that this fatalitycould be prevented by early use of antibioticsand supportive care.

1 Roy. K et al (2011). Risk factors of mortality in severely-

malnourished children hospitalised with diarrhoea. Journal

of Health, Population and Nutrition 2011, June 29 (3);

229-235

they did, they went home before complet-ing treatment.

There are a number of reasons thatexplain these positive findings. First,results show that CHWs were able to iden-tify and treat SAM very early in the courseof the disease. This meant that childrenpresented with fewer complications, wereeasier to treat and there was rarely a needto refer a child for inpatient treatment. Theprogramme design supported this earlyidentification of cases through decen-tralised and multiple pathways totreatment including the use of MUACbands by CHWs at monthly growth moni-toring sessions and during home visits tosick children and the use of a ‘watch-list’ ofsick children by CHWs in their villages. Inaddition, study findings show that therewas a good interface between the commu-nity and the programme. Mothers andcommunity-level health practitioners, suchas village doctors and other community-based stakeholders, were aware of SAM,trusted CHWs to provide effective treat-ment, and referred their own children andothers in their villages when they were sickor losing weight. Study findings alsodemonstrated a very high quality of caredelivered by CHWs. When assessedagainst a treatment algorithm theyachieved, on average, a rate of 100% error-free case identification and management.

Cost effectiveness was also analysed aspart of this study. The CCM of SAM inBangladesh cost $165 per child treated and$26 per DALY (disability-adjusted lifeyear) averted. This is a similar cost-effec-tiveness ratio to other priority child healthinterventions such as immunisation andtreatment of infectious tuberculosis. It isalso at a level considered ‘highly cost-effec-tive’ according to WHO’s definition thatdefines an intervention as cost effective if itaverts one DALY for less than the percapita GDP of a country.

This study has demonstrated that sucha model of care in Bangladesh is feasibleand could be an effective and cost-effectivestrategy to ensure timely and high qualitytreatment for a condition that is typicallyassociated with high levels of mortality.This is an important finding in a countrythat has the fourth-highest number of chil-dren suffering from SAM in the world, yetto date has had no effective mechanism ofidentifying and treating them.

An informal group discussion

SCU

S/F

IC,

Bangla

desh

, 2009

13

Voices from the field: Optimisingperformance for humanitarian workers By Jared Katz, Deborah Nguyen, Carla Lacerda and Gerald Daly

The authors extend special thanks to Olesya Dudenkova for helping to get the project offthe ground, and to everyone who took the time to support this project along the way.

Arecent study1 set out to understand howhumanitarian workers remain effectivein challenging environments while also

maintaining personal life balance.

Discussions were held with twenty-sixprofessional humanitarians from eight interna-tional organisations currently working on thefrontline of global emergencies. The interviewswere framed on a McKinsey model of ‘CentredLeadership’, which identifies five capabilitiesthat, in combination, generate high levels ofprofessional performance and life satisfaction.They are finding meaning in work, convertingemotions such as fear or stress into opportunity,leveraging connections and community, actingin the face of risk, and sustaining energy. Thefollowing is a compilation of the best personalpractices of these 21st century humanitarians.

Having meaning in work and lifeThe humanitarians interviewed most oftenfound meaning in the understanding of andconnection with those receiving humanitarianaid. Humanitarians have a combination ofmotives that may not be altogether altruistic, asthe work can be financially motivating and thepositions themselves bestow power andrespect. However, those who find meaning intheir work convey energy and enthusiasmbecause the goal is important to them person-ally, because they actively enjoy its pursuit, andbecause their work plays to their strengths. Onehumanitarian commented: “The ultimate goalmust have meaning for humanitarians and thepeople they are serving; otherwise they’ll lackthe energy and motivation to get things done.”

‘Stay on target’ – Maintaining meaningOften, the noble motivations humanitariansmay have had when entering this field becomediluted, as they grow dependent on the moneyand benefits they receive from the work. Manyinterviewees stressed that staying out of thefield to climb the bureaucratic ladder causedtheir work to become abstract and meaningless.Maintaining a sense of meaning throughout ahumanitarian’s career can be aided by periodicvisits to the field, and the connection with realon-the-ground issues.

When unable to get into the field, respon-dents found other methods of maintaining theconnection. Some stayed in close contact withcolleagues in the field to remember the mean-ing of their work. “It is important to talk to staffthat work in the field, you will catch theirenthusiasm.”

‘Balancing act’ – Don’t neglect meaningoutside of workFinding meaning in humanitarian work needs tobe balanced with maintaining meaning outsideof work. Maintaining personal relationships isoften the best remedy against losing yourselfand becoming over defined by your job.

Balancing work and personal life is difficult.“It is a never ending challenge, and particularlyso during an emergency situation where you

are most likely to be separated from yourfamily.”

‘Rolling with the punches’ – Resilience andrecoveryHaving a strong sense of meaning and apurpose greater than oneself fosters humility,which strengthens the ability to withstand andrecover quickly from difficult situations.Emergencies cannot afford to have workers whoare knocked off balance by their own shortfalls,but who keep their “sights on target” and canadapt and react quickly to tough situations.

Humanitarianism is not about you, but whatyou do: “You are not the centre of anythingwhen you are a humanitarian. It’s about yourwork; it’s about achieving your objective.” Byrealising you are only a means to a greater end,you can limit frustrations and losses by notgetting knocked off balance with failures. Youcan recognise when you have made a mistake,“analyse your mistake and criticise yourself,but forgive yourself,” and move forward.

Mental Framing: Converting emotionsinto opportunityOptimists have an edge over pessimists.Leaders who do not naturally see opportunityin change and uncertainty create stress andlimit creative solutions. It is no surprise that75% of respondents believe they are optimistsand use positive framing on a regular basis;they have found ways to convert stress intoopportunity. “One of the reasons for my successis positive framing, I always see the bright side,I see the glass half full.”

‘Making a purse out of a sow’s ear’ – value inpositive framingEffective leaders grappling with failure seeopportunity and creative solutions when otherssee defeat. Early on in emergencies, confusion isusual and decision-making is based on weakinformation, so projecting a positive and calmmindset supports team spirit. Leaders oftenproject a positive frame because they knowtheir team needs it.

‘Teflon has its uses’ – Protecting yourself andyour teamSome people might choose deliberately toshelve emotions that are triggered by theintense suffering of people struck by a disaster.In doing so, they often are able to remain effec-tive in the midst of the storm. The ‘Teflonmethod’ is such a shelving technique, allowingnegative or intense situations to slide off andnot affect one’s work during a crisis. While the‘Teflon method’ can be effective in the short-term, one respondent cautioned on the need toshare with friends traumatic experiences asthey occur. Their reasoning was that anunprocessed negative experience mightambush a humanitarian worker years after atraumatic event has been experienced.

‘I’ve got high hopes’– A fine line Between opti-mism and naivetyA common piece of advice was the importance

of having realistic expectations of what can beaccomplished. Humanitarian workers do not‘save the world’, but rather the best they canhope for is to help a disaster-impacted commu-nity or help with the formulation of a policy.Although there is nothing wrong with idealisticaspirations, it is important they are notconfused with naivety, which can lead to disap-pointment or even depression.

Leveraging connections and community‘Stay connected’ – Managing complexnetworks in emergenciesMaintaining real-time updates in emergencysituations by developing community and inter-organisational networks can tremendouslyincrease efficiency and help decision-makingunder conditions of uncertainty. “In order tostay connected you have to spend a lot of timein the field and understand the local context;what is on paper is not always what is reallyhappening. Talk to donors, beneficiaries andother agencies to find out what needs to beworked on.” This approach provides updateson the politics, as well as other developments infast-paced crisis environments. The role ofinformation technologies to achieve effectivecommunication and decision-making goals inemergencies is a skill-set to be fully honed.

‘Soak it in’ – Learning from othersIn a complex and high-pressure environmentwith little structured training, mentors providean opportunity for young humanitarians bothto learn from their experiences and provideinspiration to emulate their actions and makethem more effective humanitarians. A shortfallin humanitarian mentoring system wassummed up as: “Mentors don’t come easily,you have to make them happen.”

Only one organisation, out of all those repre-sented in this study, had a structured mentoringsystem.

With little structured mentoring, it is impor-tant to learn as much as possible from everyoneyou come across. “At work you meet differentpeople, some inspire you, and some have moreexperience. I try to absorb as much as I can onhow they deal with different issues: the waythey manage a conflict between two colleagues,or how they listen to their team. I do not followone model.”

A unique angle of this issue is the newconcept of reverse mentoring. Humanitarianswho have been in the business a long time canhave a narrowly defined skill-base and be igno-rant of emerging humanitarian trends. Often itis younger colleagues who are most savvy inthe newly emerging skill areas (e.g. socialmedia in times of emergency). One respondentadvised: “If older humanitarians wish to stayrelevant and keep with the pulse of the 21stcentury needs, I suggest they ask to be reverse-mentored by a younger colleague.”

1 Development in Practice, Vol. 22 Issue 2.

Research

An Argentinean aid worker holdsa girl injured in the earthquakein Leogane, Haiti

UN

, H

aiti, 2

010

14

Acting in the face of riskRisk aversion and fear are widespread inhumanitarian crises, as mistakes are oftenpublicised and punished more than successesare rewarded.

‘No ‘I’ in Team’– Trusting in othersMany humanitarians do not delegate due to lackof trust and perceived consequences of pooractions, or the fear of losing their jobs by givingaway their ‘technical expertise’. Encouragingothers to act and take risks on your behalf isextremely difficult. However, it should beconsidered imperative for leaders in humanitar-ian organisations as this gives people confidenceand demonstrates trust, which creates an evenhigher demand for people to act responsibly.

‘Living on the Edge’– Decision making: risktaking and intuitionIn high-risk environments where there are morequestions than answers, intuition plays a majorrole in humanitarian decision-making. “Yourintuition tells you what is around the cornerand how to take action with a number ofunknowns.” However, decisions should alwaysutilise as much fact as possible and “commonsense based on an understanding of the politicsof where you are. Without that, you’re floun-dering around in the dark.”

‘Rules were meant to be broken’– Organisationalrules vs. humanitarian imperativeOrganisational rules set a general frameworkfor action and decrease the risk of negligenceand failure. But rules are sometimes toodetached from the reality on the ground, and inemergency situations, when lives are at stakeand the window for action is narrow, you some-times have to break the rules to do what makesthe most sense. “We are living by the rule ofhumanitarian imperative, which might be inconflict with agency regulations.”

Competent humanitarians need to have thepersonal strength to take critical initiatives bothwith creativity and speed. “In real emergencysituations you have to balance getting the jobdone, ethics, transparency, and are you going tolive with this the next day.”

‘The road to hell is paved with good intentions’– personal values in testing environments“When you are first starting off it’s all aboutideals, humanitarianism and doing the rightthing. With time, those ideals don’t hold up.You have to find your own way of reconcilingthese really big disparities. There’s a certainamount of hypocrisy that has to be reconciled.”

Historical and social factors make each coun-try different, and institutional structures, ethnicand religious rivalries can all test personalvalues. Corruption, for example, may not beseen as much of a problem in some societies,and bribery can be socially acceptable as ameans of getting things done quickly. It is veryeasy to empathise with the communities oneworks with or with a government official,particularly when it gives access to key decisionmakers and project benefits are immediate.

In these situations, humanitarians must stickto the ‘limits’ of their personal values. “A toughsituation was witnessing a culture of corruptionfor one of the organisations I worked in. I wasinstructed to utilise stolen equipment, evenafter I pointed out to them that it was stolen.”One interviewee felt that, “a morally ‘purist’approach may not always be possible” whileanother felt you should “never break personal

principles; you have to weigh whether theconflict breaks personal principles or if it is justan approach that can be resolved by swallow-ing some pride and compromising.”

While values are context-bound and varyacross cultures and individuals, there are limitsin all cultures. Humanitarians should notsupport cultural norms that undercut the long-term stability and development a society hopesto achieve. “Humanitarians need to under-stand that an emergency is a rupture in thenormal development of a country – and the jobis to help it resume that development as soon aspossible – so we have to understand long-termdevelopment and cultural contexts as well asshort-term emergency efficiency contexts.”

Sustaining energy ‘What’s in there, you ask? Only what you takewith you’ – Emergency response is not foreveryone“While the mind can be very powerful, it canvery easily break (or burn-out) in high-pressuresituations. We owe it to ourselves to not walkblind into a wall of fire.” Being aware of yourstrengths and weaknesses helps you cope withthe challenges of the job and avoid situationsyou cannot handle. “You need to understandwho you are before you can throw yourself intothis type of work. Nobody is effective on theground if they collapse when personal prob-lems surface due to extreme psychologicalstress.” For some humanitarians, the pressuresinduced by working in conflict and emergencyenvironments are used as a driving force. “Thepressure itself helps me. You see results faster, itmakes you motivated, it helps you pull throughand you have a goal to achieve. There is anadaptation to work under adrenaline. But youneed to like pressure … if you don’t, you’re notcut out for humanitarian work.”

‘You are what you eat’ – Preparation beforeentering the fieldMany respondents stressed the interdependenceand importance of both physical and mentalpreparation. “Psychologically preparing your-self for the length of your assignment is critical.“I was in Darfur as an emergency coordinatorfor a few months, which was fine, because Iknew it was a few months. It’s psychologicallyimportant to know duration, because if you’rein an intense location and don’t know how longyou’ll be there, it’s draining.”

‘Lean on me’ – Relying on colleagues to monitor stressStaff burnouts harm humanitarian relief opera-tions because an individual’s quality of workcan deteriorate to the point that there is a gap inoperational capacity and a replacement isneeded. Due to the lack of training and prepa-ration, most humanitarians develop their ownad-hoc strategies to maintain themselves incrisis situations.

Relying on and communicating withcolleagues working under the same externalpressures is fundamental in identifying yourown levels of stress and coping with the pres-sures. “Make sure everyone on your team isdoing OK … make a pact with a friend whowould tell you if you are near to burnout.”Beyond symptoms of extreme stress, respon-dents identified indicators to know whencolleagues are nearing a burnout point.“Reaching the edge, humanitarians often resortto over use of alcohol as a negative means todeal with stress, and people tend to becomeextremely ironical and sarcastic.”

‘Leave the ball on the field’ – The importance ofdisconnecting from work“People who have burnouts don’t detach them-selves from a cause and don’t move on.” It isimportant for humanitarians to be efficientwhile at work, but to limit themselves, and oncethey leave, try to switch off from their job.”While working in an emergency context, “Youneed to relax, know yourself, have something todo after work, have a comfortable place tostay.” Being able to find some peace after a hardday is easier said that done, but if you canempty your mind by doing something that hasnothing to do with work, it helps to get thestress out of your system. This is particularlydifficult for humanitarians, because shuttingone’s mind to suffering is not easy. Mechanismsare diverse, and range from meditation, exer-cise, to reading a novel.

‘The horror’– Dealing with trauma, one wayor anotherThe period following a difficult mission is just assignificant as preparation in avoiding burnouts.One organisation had psychologists who dosystematic debriefs of staff who are departingpsychologically damaging work situations, butthis support is not always effective for everyone,and informal channels need to be explored.Talking with colleagues, friends and family canhelp to express the traumas experienced.

ConclusionsSomewhere within this physically and psycho-logically demanding and politically complexenvironment is where humanitarians findthemselves working and living day-by-day. Asthere are very limited training mechanisms,each humanitarian is often making their ownway through the field, learning from their ownmistakes and successes. Humanitarians havedevised innovative (and sometimes unusual)practices to overcome the old and modern chal-lenges of working in the humanitarian field. 1. Maintaining meaning in work and life inorder to sustain personal effectiveness andsatisfaction.2. To convert stress into opportunity, positiveframing was seen to project confidence andtranquillity on the team and encourage outside-the-box solutions. 3. Leveraging connections included develop-ing informal community and inter-organisational connections to maintain real-time updates in emergencies, and an ad-hocsystem of mentoring as a source of informationand inspiration. 4. Learning to act in the face of risk is achievedthrough having a thorough understanding ofthe local context to limit failure, knowing howto work around organisational rules, andunderstanding the limits of personal values.5. Sustaining energy is the most difficult aspectto control in emergency situations. It is a factorof personal character and self-awareness, phys-ical and psychological preparation, relying oncolleagues to monitor stress, disconnectingfrom work, and effectively dealing with traumawhen it occurs.

Through this study, we have seen that ahumanitarian’s unique, individual practicescan be combined to fit together within a largerframework maximising work effectiveness andpersonal satisfaction. Each humanitarian findstheir own way to cope with the challenges theyencounter, and keeping the balance is key.

For more information, contact: DeborahNguyen, email: [email protected]

Research

15

Summary of published research1

Exclusive breastfeeding (EBF) is reportedto be a life-saving intervention in low-income settings. A recent

cluster-randomised trial set out to assess theeffect of breastfeeding counselling by peercounsellors in Africa.

Twenty-four communities in Burkina Faso,24 communities in Uganda, and 34 communi-ties in South Africa were assigned in a 1:1 ratio,by use of a computer-generated randomisationsequence, to the control or intervention clusters.In the intervention group, the research teamscheduled one antenatal breastfeeding peercounselling visit and four post-delivery visitsby trained peers. The data gathering teamswere masked to the intervention allocation. Theprimary outcomes were prevalence of EBF anddiarrhoea reported by mothers for infants aged12 weeks and 24 weeks. Country-specific preva-lence ratios were adjusted for cluster effects andsites. Analysis was by intention to treat.

A total of 2,579 mother-infants pairs wereassigned to the intervention or control clustersin Burkina Faso (n=392 and n = 402, respec-tively), Uganda (n=396 and n=369,respectively), and South Africa (n=535 and 485,respectively).

The prevalences of EBF at 12 and 24 weeks ofage are given in Table 1, and the prevalences ofdiarrohea in Table 2. These show a much higherprevalence of EBF in the intervention groups inBurkino Faso and Uganda at both 12 and 24weeks. The prevalence and differential betweenthe intervention and control groups is muchlower in South Africa. An impact on diarrhoeaprevalence was not found.

The results for South Africa must be viewed inthe context of EBF being rare at baseline. Mostpeople in rural and urban South Africa purchasemost of their food and the Department ofHealth’s routine child health services have ahistory of promoting commercial infant formulaas part of the protein energy malnutritionscheme, while the International Code ofMarketing of breastmilk substitutes is not yetlegislated in South Africa.

The authors conclude that low-intensity indi-vidual breastfeeding peer counselling isachievable and, although it does not affect thediarrhoea prevalence, can be used to effectivelyincrease EBF prevalence in many sub-SaharanAfrican settings.

The authors of the study caution that thecommunity based approach could possibly haveresulted in socially desirable answers. Theresults were based on self-reports so that a biastowards desirable answers cannot be ruled out.There was also some questionnaire fatigue in theUgandan site, i.e. reluctance to fully engage inanswering similar questions after a few inter-views.

Overall, however, the authors conclude thatlow-intensity individual breastfeeding peercounselling is achievable and, although it doesnot affect the diarrhoea prevalence, can be usedto effectively increase EBF prevalence in manysub-Saharan African settings.

1 Tylleskar. T et al (2011). Exclusive breastfeeding promotion

by peer counsellors in sub-Saharan Africa (PROMISE-EBF): a

cluster-randomised trial. www.thelancet.com Vol 378, July

30, 2011, pp 420-427

Diarrhoea prevalence at 12 weeks EBF (24 hour recall) at 24 weeks

Intervention Control Prevalenceratio

95% CI Intervention Control Prevalenceratio

95% CI

Burkino Faso 5% (20/392) 9% (36/402) 0.57 0.27-1.22 7% (26/392) 8% (32/402) 0.83 0.45-1.54

Uganda 10% (39/396) 9% (32/369) 1.13 0.81-1.59 13% (52/396) 16% (59/369) 0.82 0.58-1.15

South Africa 8% (45/535) 7% (33/485) 1.16 0.78-1.75) 10% (54/535) 7% (33/485) 1.31 0.89-1.93

Table 2: Diarrhoea prevalence at 12 weeks and 24 weeks of age

EBF (24 hour recall) at 12 weeks EBF (24 hour recall) at 24 weeks

Intervention Control Prevalenceratio

95% CI Intervention Control Prevalenceratio

95% CI

Burkino Faso 79% (310/392) 35% (139/402) 2.29 1.33-3.92 73% (286/392) 22% (88/402) 3.33 1.74-6.38

Uganda 82% (323/396) 44% (161/369) 1.89 1.70-2.11 59% (232/396) 15% (57/369) 3.83 2.97-4.95

South Africa 10% (56/535) 6% (30/485) 1.98 1.30-3.02) 2% (12/535) <1 % (2/485) 9.83 1.40-69.14)

Table 1: EBF prevalence at 12 weeks and 24 weeks of age

The Bolsa Familia programme (BFP) inBrazil is the world’s largest conditionalcash transfer programme. It reaches

5,564 municipalities in the 27 states of Braziland about 11 million families (25% of theBrazilian population). The programme guar-antees direct cash transfers to: families inpoverty or extreme poverty (householdincome per capita below US$44 and belowUS$22 respectively in 2005-6), families withchildren 0-15 years of age and families witha pregnant or lactating woman.

In 2008, the age group for the childrenwas extended to 17 years. In most cases, thecash transfer is paid to the reference femaleof the family group. The per capita incomecut-offs and the values of cash transferredare readjusted every two years or so, bydecree. The value per family depends on thepoverty threshold and family composition.No nutrition supplement is distributed.

Once a family enrols, it must comply withcertain health and education conditions toremain in the programme: i) a minimumschool attendance of 85% of the monthlyschool hours for children 7-17 years old, ii) ahealth and nutrition agenda for beneficiaryfamilies with pregnant women, breastfeed-ing mothers or children under 7 years of age(pre-natal care, vaccination, health andnutrition surveillance).

A study that has just been published setout to examine the association between theBFP and the anthropometric indicators ofnutritional status in children. Using theopportunity provided by vaccinationcampaigns, the Brazilian governmentpromotes Health and Nutrition Days to esti-mate the prevalence of anthropometricdeficits in children. Data collected in 2005-6

Effects of a conditional cashtransferprogramme onchild nutrition inBrazilSummary of published research1Exclusive breastfeeding promotion by

peer counsellors in sub-Saharan Africa

Mother and childhealth nutritionprogramme in

Uganda

WFP/M

arc

Hofe

r, U

ganda,

2010

A family enrolled in the cashtransfer programme

©fo

tos

de B

runo S

pada/M

DS

16

Research

MUAC and weight-for-height in identifyinghigh risk children Summary of research1

for 22,375 impoverished children under 5years of age were employed to estimatenutritional outcomes among recipientsof the BFP. All variables under study,namely child birth weight, lack of birthcertificate, educational level and genderof family head, access to piped water andelectricity, height for age, weight for ageand weight for height, were convertedinto binary variables for regressionanalysis.

The subsequent analysis found thatchildren from families exposed to theBFP were 26% more likely to have normalheight for age than those from non-exposed families. This difference alsoapplied to weight for age. No statisticallysignificant deficit in weight for heightwas found. Stratification by age grouprevealed 19% and 41% higher odds ofhaving normal height for age at 12-35months and 36-59 months of age,respectively in children receiving theprogramme and no difference at 0-11months of age.

The authors of the study note thatothers studies have found that the Giniindex, an indicator of income distribu-tion, remained stable in Brazil for manyyears but has dropped consistently since2001. Almost one quarter of the drop isattributable to the BFP. Furthermore,propensity score analysis used in thebaseline study for the BFP showed largerfamily expenditures among enrolledfamilies than in the comparison groups,especially on food (US$172 more a yearon food items).

The authors conclude that the BFPcan lead to better nutritional outcomesin children between 12-59 months ofage. However, longitudinal studiesdesigned to evaluate the impact of theBFP are necessary to determine if thenutritional effects observed in the studycan be attributed to the conditional cashtransfer programme. Furthermore, thereis a need to guarantee families in the BFPincreased access to goods and servicesconducive to improved nutrition, whichshould in turn result in improved health.Similarly, to guarantee programme effec-tiveness, the Brazilian government needsto provide more and better services inthe spheres of basic education, health,social protection and inclusion in thelabour market.

1 Paes-Sousa. R, Santos.L and Miazaki. E 2011).

Effects of a conditional cash transfer

programme on child nutrition in Brazil. Bulletin

of the World Health Organisation 2011; 89:

496-503. Published online: 29th April 2011

A family enrolled in the cashtransfer programme

©fo

tos

de B

runo S

pada/M

DS

The World Health Organisation (WHO) andUNICEF propose to use two independent criteriafor diagnosing non-oedematous severe acute

malnutrition (SAM) in children aged 6–60 months:mid-upper arm circumference (MUAC) less than 115mm and weight-for-height z-score (WHZ) less than -3(based on WHO growth standard). WHZ has been usedfor years in clinical settings for diagnosing SAM, butthe use of MUAC was introduced more recently withthe development of community-based managementof SAM. In practice, large-scale programmes increas-ingly use MUAC as single diagnostic criteria as it isclosely related to the risk of dying and is easy to imple-ment at the community level after minimum trainingby health workers or even by volunteers. MUAC andWHZ, however, do not identify the same set of chil-dren as having malnutrition, and using only one of thediagnostic criteria proposed by WHO may potentiallyleave some high-risk children untreated.

A newly published study used an old data set fromSenegal to examine the risk of dying of childrenhaving either a low MUAC or a low WHZ or a combina-tion of both in the absence of treatment. The specificobjective was to test whether combining both MUACless than 115 mm and WHZ less than -3 wouldimprove the identification of high-risk children.

The original study, which was conducted in 1983and 1984, followed an open cohort of about 5000 chil-dren, comprising all children under 5 living in 30villages. The team visited the children four times at six-month intervals in May and November of 1983 and1984. At each visit, comprehensive anthropometricmeasurements were taken (weight, height, headcircumference, arm circumference, triceps skin foldand subscapular skin fold).

For this analysis, the data from all children aged6–59 months at the time of anthropometric assess-ment were used. WHZ was calculated using the WHOgrowth standards. Receiver operating characteristic(ROC) curves (sensitivity vs. 1-specificity) were calcu-lated for WHZ and for MUAC using death within sixmonths as outcome. Only the part of the curve with apercentage of false positive less than 5% wasexplored, as these indicators are always used in acontext of limited treatment capacity and a highspecificity is required. The anthropometric index withthe highest ROC was considered as more adapted toidentify high-risk children.

Specificity and sensitivity of ‘MUAC less than 115mm AND WHZ less than -3’ and also for ‘MUAC lessthan 115 mm OR WHZ less than -3’ were also esti-mated and positioned in relation to the ROC curve ofMUAC and WHZ.

In total, 12,638 measures were made on 5,751 chil-dren, and 303 deaths occurred within six months ofthe nutritional assessment. The criteria ‘WHZ less than-3 AND MUAC less than 115’ had a specificity of 99.0%and a sensitivity of 5.9%. For ‘MUAC less than 115 mmOR WHZ less than -3’, the specificity was 96.9% and thesensitivity was 13.2%. Both points combining WHZless than -3 and MUAC less than 115 mm were posi-tioned above the WHZ ROC curve but below theMUAC curve.

For a MUAC, the threshold of 112 mm had a speci-ficity of 99.1, comparable with the specificityobserved for ‘MUAC less than 115 AND WHZ less than-3’. However, the sensitivity of MUAC 112 mm was

6.0%, slightly higher than for these two indicescombined. For a MUAC of 119 mm, the specificity was96.9%, comparable with the specificity of ‘WHZ less than-3 OR MUAC less than 115 mm’. However, the sensitivitywas higher at 14.9%.

This study confirms that MUAC has a better abilitythan WHZ to assess the risk of dying. This is consistentwith a previous analysis of the same data set, whichshowed that MUAC was superior to weight for- height(percentage of the National Centre for Health Statisticsmedian) to identify high-risk children. This is also consis-tent with other reports from the literature.

In addition to confirming previous findings, thisstudy also shows that using WHZ equal to or <-3 andMUAC equal to or <115 mm together did not improvethe identification of high-risk children. The position oftwo possible combinations of WHZ equal to or <-3and/or MUAC equal to or <115 mm compared with theMUAC ROC curve suggests that using the two indicestogether may lead to poorer results than using MUACalone. Arguably, using WHZ equal to or <-3 AND MUACequal to or <115 mm increases the specificity, but thiscan be obtained by using a lower MUAC cut-off (112mm) with a greater sensitivity. Conversely, using WHZequal to or <-3 OR MUAC equal to or <115 mm increasessensitivity, but it also decreases the specificity. Using aMUAC cut-off of 119 mm, with the same specificity asWHZ equal to or <-3 OR MUAC equal to or <115 mm, willresult in a higher sensitivity.

Several explanations have been given to explain theapparent superiority of MUAC to assess the risk of dying.A first hypothesis is that MUAC, which grows continu-ously with age, selects younger children with a higherrisk of dying when used with a cut-off not adjusted forage. Another explanation might be that MUAC is closelyrelated to muscle mass. A close association betweenMUAC and muscle mass has been suggested by corre-sponding measures of body composition by Dualenergy X-ray absorptiometry. The underappreciatedmetabolic role of muscle both in health and disease mayexplain its association with survival.

Finally, it has been shown that WHZ differences arelargely influenced by leg length, unrelated to the nutri-tional status of the child, which may also decrease itsability to identify high risk children.

This study was based on MUAC data carefullycollected by skilled investigators. Its findings may not beextrapolated to other settings where these conditionsare not met. Appropriate training of field workers andstandardisation of measures are probably needed to useMUAC successfully in field conditions to identify high-risk children. The use of colour bandedMUAC straps could also help to minimise measurementerrors.

In conclusion, this study shows that there is no bene-fit for programmes in using both MUAC equal to or <115mm and/or WHZ equal to or <-3 to identify high-riskchildren. If a higher sensitivity is required for program-matic reasons, for instance to take into account a poorfood security, it seems preferable to increase the MUACcut-off rather than combine it with WHZ. In the sameway, if a higher specificity is required, in case of limitedtreatment capacity, lowering the MUAC cut-off shouldbe preferable.

1 Briend. A et al (2011). Mid-upper arm circumference and

weight-for-height to identify high-risk malnourished under-five

children. Maternal and Child Nutrition 2011

17

Research

Summary of position paper1

Arecent position paper has beenproduced to guide country-level healthclusters on how to apply Inter-Agency

Standing Committee (IASC) civil military coor-dination principles to humanitarian healthoperations. The paper is intended to serve asthe basis for discussions with a wide range ofstakeholders including health cluster partners,military representatives, civil defence and civilprotection actors and other humanitarian clus-ters. The relation between health humanitarianactors and non-state military groups is outsidethe scope of the paper.

The paper reviews the existing guidance oncivil-military coordination and attempts to clar-ify how it applies to the health sector. It alsoidentifies some gaps in the guidance andemerging challenges.

The document’s target audience is healthcluster participants involved in civil-militarycoordination. It is also intended to stimulatediscussion within the overall humanitariancommunity and military counterparts.

The document is informed by and builds onthe more general efforts of the United Nations(UN) and other humanitarian organisations toidentify appropriate civil-military coordinationmodalities during humanitarian crises. Theposition paper is a work in progress that maybe revised to take account of inputs from globalhealth cluster (GHC) partners and otherhumanitarian agencies, as well as develop-ments in the area of civil-military coordination.

The scenarios in which humanitarian healthagencies operate are complex in terms of inter-nal dynamics and interactions with externalparties involved in the response. Over the lastdecade, military actors have been increasinglyinvolved in relief activities in various settings,including sometimes providing direct assistanceto crisis-affected populations. From a humani-tarian perspective, this poses specific questionsregarding the extent to which their involvementhas a positive impact and, conversely, whetherand how this involvement might affect humani-tarian organisations’ ability to respondimpartially to the needs of the population. Civil-military coordination problems are particularlyrelevant for the health sector. Health activitieshave historically been part of counterinsurgencymilitary strategies. More importantly, rehabili-tating the health sector is increasingly seen askey to ensuring the country’s stability.

Identifying a way to engage with the military –one that does not dangerously confuse the twomandates – is at the core of the civil-militarycoordination challenge.

Limitations in current guidelinesThe IASC’s current guidelines clearly outlinethe principles that should inform the relationsbetween military and civilian actors. Some limi-tations in the guidelines emerged during thepreparation of the position paper:• The guidelines primarily address the UN

peace-keeping environment. Multi-stake-holder peace operations pose new challengesthat the guidelines address only partially.

• The multiplication of actors involved in relief activities has resulted in an ever-increasing variety of operational scenarios for civil-militarycoordination. For example:- National armies and civil defence and

civil protection units intervening in their own country, assisted by an internationalresponse effort (e.g. Pakistan earthquake in 2007 and floods in 2010)

- Civil-military units with a reconstructionmandate endorsed by the national government (e.g. provincial reconstruc-tion teams in Afghanistan)

- Private security firms protecting the offices, homes and staff of humanitarian organisations.

Proposed revision of the civil-militarycoordination guidelines The GHC encourages the revision of the civil-military coordination guidelines to respond tothe new challenges posed by emerging complexscenarios including the following:

The proliferation of non-traditional actors inthe humanitarian arena has blurred the lines ofdistinction between humanitarian action basedon the principles of humanity and impartialityand other activities inspired by different agen-das. This calls for an analysis of how theinteractions between different actors and agen-cies can affect humanitarian principles.

National civil defence and civil protectionagencies, which are part of the internationaldisaster response system, raise a number ofquestions for coordination which must beaddressed in relation to the specific nature ofthe entities involved. There is no internationallyagreed definition of civil defence or civil protec-tion actors in terms of how they operate, whatis their mandate or nature of the relationshipwith military or security forces of their coun-tries. While in some countries and regions,these terms may have developed distinct mean-ings, these terms are sometimes usedinterchangeably. Given their increasing impor-tance in humanitarian response, improvedcoordination is needed between humanitarianhealth actors and civil protection actors in thefield and globally. How this should happen in aspecific setting depends on the specific natureof the civil defence and civil protection actors inthat setting. It may be appropriate to includesome of these actors in the health cluster coor-dination mechanism itself, where these arecivilian actors explicitly operating on the basisof humanitarian principles. It is important tonote that even such entities may regularly relyon their own national military forces for trans-portation and other logistical support whenresponding internationally and that this shouldbe considered in determination of coordination

Different mandatesHumanitarian organisations and military forceshave different mandates:• Humanitarian organisations endeavour to

provide life-saving assistance to affected populations based on assessed and docu-mented needs and on the humanitarian principles of humanity, independence and impartiality.

• Civil defence and civil protection units are usually deployed in a humanitarian crisis on the basis of an agenda of the governmentto which they belong. As there is no agreed international definition for these categories, the different mandates, modes of operation and natures (civilian or military) of these actors must be considered when identifyingwhether and how the humanitarian mecha-nisms on the ground will engage and coor-dinate with these actors.

• Militaries may be present in the context of ahumanitarian crisis as combatants, they may have a specific mandate granted by theSecurity Council (peacekeeping, peace-enforcement or combat), or they may deployinternationally at the invitation or with permission of the affected government. Military forces may be deployed abroad or inside their own borders. While the specific mandate will differ in different settings, it isimportant to recognise that militaries are deployed with a specific security and political agenda or in support of a security and political agenda.

These fundamental differences at the core of themandates – the needs of the population on theone hand and political/security goals on theother – guide the respective decision-makingprocesses of humanitarians and the military.This can result in minor differences that stillallow for cooperation (e.g. when responding toa natural disaster in a non-conflict setting) ormajor differences (e.g. those that may occur incombat settings). Any confusion between thedifferent mandates carries the risk that human-itarian aid agencies may be drawn, or perceivedto be drawn, into conflict dynamics.Humanitarian agencies that are perceived asacting according to agendas other than theirhumanitarian mandate may lose their credibil-ity in the eyes of other local actors, as well as thetrust of the population they are there to serve.This can severely affect their ability to operateand ultimately, create security risks for theirstaff and for the aforementioned populations.

Civil-military coordination duringhumanitarian health action

Army helping civilianevacuation in Pakistan2005

Cre

dit:

AP/W

WP,

Pakis

tan,

2005

18

Research

approaches. In other cases where there isa stronger link to a political or militaryagenda (including where the entitiesthemselves are comprised of militarypersonnel), the approach should moreclosely resemble the approach to coordi-nation with military actors in the setting.

Private security providers havebecome part of the crisis response land-scape. Humanitarian organisationsfrequently use the services offered bythese companies, ranging from securitytraining to facility protection servicesand more rarely, the armed escort ofhumanitarian convoys. Internationalguidelines contain little guidance on theuse of private security providers. Whendebating whether to use such services,humanitarian agencies should apply thegeneral principle that interaction withthe military must not affect the actual orperceived independence of humanitarianhealth action.

The scenario where national armiesand civil defence and civil protectionunits are intervening in their own coun-try, assisted by an international responseeffort, raises specific issues that gobeyond the scope of the traditional civil-military coordination modalities.National armies are often leading or arethe main actors of a national civil defenceand civil protection system. Internationalassistance may be deployed, uponrequest of the national government, tosupport the national response effort butthe final decision on what to do and howto carry out the relief effort rests with thenational authorities. In this framework,certain civil-military coordination princi-ples (e.g. last resort, no direct assistance)are difficult to apply and to some extentnot useful to guide relations withnational military forces. Certain dilem-mas remain, particularly when nationalmilitary forces are also involved inresponding to internal political crisis andunrest (e.g. northern Pakistan).

In recent years, there has been anincreasing tendency to include humani-tarian assistance as part of or in theservice of broader agendas of a militaryor political nature. This trend has beenformalised with the ‘comprehensiveapproach’ concept embraced by NATO,which aims at combining military, politi-cal and humanitarian activities in theoverall goal of the stabilisation of a coun-try. This concept – first operationalisedwith the provincial reconstruction teams(PRTs) in Iraq and Afghanistan – maybecome the model for future civil-mili-tary coordination. However, thisblending of strategies and tactics servesto undermine the international humani-tarian community’s core humanitarianprinciples. The integrated missionconcept developed by the UN follows asimilar trend. Although there are signifi-cant attempts to protect thehumanitarian space within integratedmissions, the concept foresees the inte-gration of different agencies andcomponents into an overallpolitical/strategic crisis management

framework. This can blur the linesbetween the UN’s different political andhumanitarian branches, with predictablynegative results.

The military’s involvement in theprovision of indirect and direct healthactivities is multi-faceted:• Armed forces deployed abroad

traditionally offer some form of health services to the local populationthrough their military medical units.

• Health activities are an important component of counterinsurgency strategies.

• NATO’s ‘comprehensive approach’ includes health recovery activities as an integral part of its military inter-vention strategy (for example, in Afghanistan).

Evidence from the field suggests thatmost of these health actions go unre-ported and uncoordinated with theoverall health national framework. TheGHC is concerned that these health serv-ices may not be appropriate to thecontext and that ad hoc health actionsmight raise the expectations of the localpopulation and create inequalities andinequities in the provision of health serv-ices.

The GHC reiterates the guiding prin-ciple that health activities should bebased on assessed health needs andguided by humanitarian principles, notby objectives that are either political ormilitary in nature. It recommends thathealth activities should not be used as acomponent of a “winning hearts andminds” strategy.

The GHC recommends that whenevermilitary actors are involved in the provi-sion of health services, any such actionshould follow the health priorities andplans approved by the national govern-ment/local health authorities, andadhere to the international humanitarianresponse plans.

Local actors and populations viewinternational aid organisations more andmore as part of a ‘western agenda’ andless and less as neutral and impartialagencies responding to humanitarianneeds. As a possible consequence, thenumber of security incidents targetingaid workers has been on the rise since1997, and attacks on medical workersand facilities are a common feature ofarmed conflicts. This is an element of thelarger phenomenon of the shrinking ofthe humanitarian space, which meanshumanitarian agencies are less able toaccess affected populations and providemuch-needed assistance.

The GHC is concerned that continuingcoordination with military forces mightfurther skew local actors’ and popula-tions’ perception of the impartiality ofhumanitarian health actions.

1 IASC Global Health Cluster (2011). Civil-military

coordination during humanitarian health action.

Provisional version – February 2011

1 Basinga. P et al (2011). Effect on maternal and child health services

in Rwanda of payment to primary health-care providers for

performance: an impact evaluation. Lancet 2011, 377: 1421-28

Operational researchin low-income countries: what, why, and how? Summary of research1

ALancet published article puts forward a defini-tion of operational research, articulates itsrelevance to infectious-disease-control

programmes, and describe some of the enablingfactors and challenges for its integration intoprogramme settings and into changing policy andpractice.

From a health programme perspective, a pragmaticdefinition of operational research is the search forknowledge on interventions, strategies, or tools thatcan enhance the quality, effectiveness, or coverage ofprogrammes in which the research is being done.

What is operational research?Operational research involves three main types ofmethod: descriptive (cross-sectional, if a stronganalytic component is also present), case–control, andretrospective or prospective cohort analysis. Basicscience research and randomised controlled trialsshould not be included as operational research. Therandomised controlled trial determines efficacy of anintervention in a strictly controlled environment withinclusion and exclusion criteria, whereas operationalresearch should assess effectiveness within routinesettings. Both types of research play an important partin the generation of new knowledge: the randomisedtrial provides clear-cut data on the efficacy of an inter-vention in defined groups of patients, whereasoperational research determines how such interven-tions are translated into benefit in the heterogeneoussetting of routine care.

The key elements of operational research are that theresearch questions are generated by identifying theconstraints and challenges encountered during theimplementation of programme activities (prevention,care, or treatment), and the answers provided to thesequestions should have direct, practical relevance tosolving problems and improving health-care delivery.A strong connection exists between good monitoringand evaluation of infectious-disease programmes andoperational research.

Good quality data on cases and treatment outcomescan be used to do operational research, which in turncan help to improve the routine data collected in thefield. Nothing is more encouraging to healthcare work-ers than to see their work in recording and monitoringdata on treatment cards and registers being used toanswer important questions, provided that thisperformance is recognised and applauded.

There are at least three reasons why operationalresearch is relevant to health. To improve programmeoutcomes in relation to medical care or prevention, toassess the feasibility of new strategies or interventionsin specific settings or populations, and to advocate forpolicy change.

19

The study question must be of importance toprogramme implementation for ‘buy in’. Muchof the internationally published research donein Africa has been generated by academic insti-tutions and researchers, predominatelyreflecting their interests or based around basicscience or questions of intervention efficacy.Although useful, this type of research needs tobe balanced by increasing the work done byoperational organisations (e.g. non-governmen-tal organisations (NGOs)) that will havedifferent perspectives. Different actors willnaturally have comparative advantages forparticular kinds of research that can benefitprogrammes. For example, an academic institu-tion might be best placed to design andimplement a randomised clinical trial or avaccine study, whereas an implementing organ-isation might be best suited to take the lead infeasibility and acceptability studies. If researchis disconnected from health-service deliverywith little or no input from programme staff inthe research design and process it may be beingresented as an additional and often unwantedextra on existing busy and often overburdenedservices.

Enabling factors and challengesFactors that enable operational research and itstranslation into policy and practice include:• Research questions are generated from

within programmes • Research planning, agenda setting, objectives,

targets, and budgeting are included within programme plans and as agenda items in programme management meetings

• Research projects use simple designs and are focused to answer implementers’ questions

• Close collaboration and partnership has been established between researchers and programme managers

• Research is done within existing systems and not done in parallel

• A competent research officer works along-side the programme manager

• Training, mentorship, and on-the-job super-vision is sustained over time

• Sufficient programme capacity exists to hostworkshops, present, and discuss research findings, and ensure their translation into policy and practice

• Programme staff have access to scientific literature through subscribed journals or theinternet

• Sufficient numbers of programme staff are available with the capacity to do operational

research, write up manuscripts, and publishrelevant research

• Funding for applied research is available and individuals develop a desire to participatein research and are mentored

• NGOs and other stakeholders are recognisedand have a contributory role in operational research

• Good quality, appropriate, and relevant research gets translated into policy and practice and thereby has a spin-off effect to stimulate more research

One challenge is that foreign academic institu-tions often have the funding, time, andmandate for research and thus the associatedpower in decisions about what gets done. Localinstitutions should also be supported withmoney and staff for operational research, thusallowing them the necessary independence tomake decisions, take responsibility, and estab-lish partnerships that are more equal inresources and decision-making power.

The authors suggest building a researchagenda into district and national programmes,based on local needs, but primarily reflectingthe research priorities of the country. Within acountry, it is important to have a coordinationmechanism to provide a clear strategy of whosets research priorities and how choices aremade at national level.

A bibliographic analysis of tuberculosisresearch done globally between 1997 and 2006showed that Africa, which has the highesttuberculosis case rate burden in the world,contributed only 7% of global research output.

Local programmes have a tendency tooutsource research to academic institutions,which then set up parallel research systems oraffiliated sites. Although this might be a highlyefficient means to produce quality research andscientific publications, if there is no satisfactorymechanism for integration, collaboration, andcommunication with the programme, thisapproach might hinder the development ofoperational-research capacity by drawingnational researchers away from nationalprogrammes. Furthermore, because researchinstitutions and technical agencies (either inter-national or national) normally have no mandateor responsibility for implementing researchfindings after their studies are completed, theresults often end up being sent or presented tobusy programme managers, who have noownership of the research and who are there-

fore unlikely to direct the effort needed to trans-late the research into policy and practice.

A change to a partnership model wouldenable greater involvement, co-ownership, andresponsibility of programme staff along withresearchers and policy makers. For example,the research question should be developed bythe entire team, including those involved withquestionnaire development, collection andanalysis of data, and dissemination of theresults. Planning at this stage also requires aclear engagement with the people who makedecisions on policy so that they are aware ofwhat is being asked, supportive of the research,and interested to learn about the findings.

Capacity and time for research activities, suchas writing study protocols or dealing with peerreview are often lacking within most programmesettings but are essential to see research tocompletion. However, if they are planned as anessential part of the programme, they can beaccomplished (see Box 1 for an example).

The failure to publish research is not justconfined to the low-income countries: a recentreport found that only 53% of 79 research stud-ies reported in conference abstracts werepublished in peer-reviewed journals after nineyears. Common reasons for so-called research‘waste’ include the wrong choice of researchquestion, poorly designed studies, and failureto publish relevant research promptly or at all.

There is a perceived need to create a so-called critical mass of trained researcherswithin health programmes to ensure that suffi-cient numbers of researchers continue in post tosustain future research. The Japanese ResearchInstitute of Tuberculosis, the JapaneseFoundation for AIDS Prevention, theInternational Union Against Tuberculosis andLung Disease, and the US Centres for DiseaseControl and Prevention are among some of theinstitutions that support international trainingon operational research.

There is also a need to assess whethercurrent capacity-building initiatives are havingan effect, such as tracking personnel after train-ing (e.g. through databases) to document theoutcomes, explore any barriers, and capturesuggestions for improving the situation. Manyresearchers from low-income countries, evenafter obtaining PhDs, do not take up researchwhen working back in disease-controlprogrammes. This might be because they endup in senior-level management posts, the infra-structure to plan or do research is lacking, orthere are simply no opportunities

Creating opportunitiesThere are various ways in which these opportu-nities could be created. First, small grants couldbe offered to pursue locally applied research.Second, junior and senior operational researchfellowships could be created for colleagues inlow-income countries with active mentoring byinternational researchers, institutions, orNGOs.

Bureaucracy should be kept to a minimumwith the main focus on deliverable outcomesthat would include publications with specific

Box 1: Malawi National Tuberculosis Control programme

The experience of integrating operational researchwithin the Malawi National Tuberculosis Control(NTP) programme is an example of how researchelements can be successfully built into a nationalprogramme, and be of great value in shaping policyand practice.

Between 1996 and 2004, many studies weredesigned, planned, and budgeted within the NTP.The programme invested in a full-time operational-research officer and a data-management officer whoworked alongside the programme manager andprovided on-the-job training and supervision ofresearch. Planned and continuing programmaticresearch was always an agenda item at the regular 6-weekly programme management group meetings.The NTP held an annual scientific review and dissem-

1 Zachariah et al (2009). Operational research in low-income

countries: what, why, and how? The Lancet Infectious

Diseases, Volume 9, Issue 11, Pages 711 - 717, November

2009

ination meeting, and operational-research findingswere presented by local and international investiga-tors to all stakeholders. A medical editor wasrecruited to give an annual workshop on paper writ-ing skills and on how to get the research published.

All publications in national and international journalswere collated each year, and the resulting bookletwas distributed to health-care workers around thecountry at national meetings and during supervisionso that everyone had potential access to the resultsof locally generated research. All these componentswere built into the NTP plans with explicit budgetstreams, and these helped to develop the capacity toask pertinent questions and to carry them throughto publication for dissemination.

Research

20

benefits to programmes and communities.Targets for research output should be set (e.g.one or two research papers each year submit-ted to a peer-reviewed journal), with financialand technical support continued when targetshave been met and termination of support iftargets have been missed. Young nationalresearchers should have the opportunity topresent scientific abstracts and participate atregional and international conferences, andmore attention should be paid to teaching theprinciples of how to write scientific papersand to mentoring.

Attention must also be paid to the problemof poor access to up-to-date scientific literature,and despite laudable initiatives (such as theHealth InterNetwork Access to ResearchInitiative), this remains a barrier in low-incomecountries. Free and open access for all articles ofinterest to low-income countries is urgentlyneeded. For example, Médecins Sans Frontières(MSF) has negotiated with publishers to allowfree access to all articles written by its staff.

The distinct role for NGOs in operationalresearch should be recognised for two mainreasons. First, NGOs such as MSF often workin conflict settings, with marginalised andvulnerable populations, or with neglecteddiseases. Academics rarely have access to suchsettings, and national programmes mightdecide they do not have sufficient resources tostudy them. Research in these areas is, never-theless, needed to better understand how tomanage questions such as mental health issuesin war zones, treatment and diagnosis ofneglected diseases, or offering of HIV/AIDScare in slum settings. Secondly, NGOs are, bymandate, implementers and can thus beinvolved in the translation of research findingsinto policy and practice. If they have skills inresearch and advocacy as well as sufficientfinancial and human resources, then theyprobably have the potential to actively engagein operational research and help change prac-tice.

However, NGOs are sometimes not theappropriate entities for designing or imple-menting research. They might lack theinstitutional support, culture, and skills forinteracting with national programmes anddecision makers. NGO focus might be on solv-ing localised, short-term problems, they mighthave had little exposure to systems thinking.and they might lack the training and capacityto do rigorous research. They might also havea rapid turnover of staff, which hinders thesustainability of research and the ability tobuild up trust and understanding with coun-try partners. These points might explain whyNGOs rarely undertake research, are rarelyasked by country programmes to do so, andwhy the research they do undertake is some-times badly done, with little or no programmeimpact.

The authors conclude that they have madethe case for the importance of operationalresearch as a necessary component of healthprogramming in low-income countries. Whatis needed now is further development of oper-ational-research capacity, allocation of specificresources, and the need for different partici-pants such as international and nationalacademic institutions, national programmemanagers, and NGOs to work together inpromoting operational research.

1 Bezner Kerr R, Berti P and Shumba L (2011). Effects

of a participatory agriculture and nutrition education

project on child growth in northern Malawi. Public

Health Nutrition 14 (8), 1466-1472

over half of the smallholder families expe-rience food insecurity every year. Malariais endemic and child malnutrition rates aresimilar to the national average at 48%.Agricultural interventions involved inter-cropping legumes and visits from farmerresearchers, while nutrition educationinvolved home visits and group meetings.

The study employed a prospectivequasi-experimental approach comparingbaseline and follow up data in interventionvillages with matched subjects in compari-son villages. Mixed model analyses wereconducted on standardised child growthscores (weight- and height-for-age Z-scores), controlling for child age andtesting for effects of length of time andintensity of village involvement in theintervention.

Participants in intervention villageswere self-selected and control participantswere matched by age and household foodsecurity status of the child. Over a six yearperiod, nine surveys were conductedtaking 3,838 height and weight measures ofchildren under the age of 3 years.

The study found that there was animprovement over initial conditions of upto 0.6 in weight-for-age (WAZ) Z-score(WAZ: from -0.4 (sd 0.5) to 0.3 (sd 0.4)) forchildren in the longest involved villages,and an improvement over initial conditionsof 0.8 in WAZ for children in the mostintensely involved villages (from -0.6 (sd0.4) to 0.2 (sd 0.4)).

The authors conclude that longtermefforts to improve child nutrition throughparticipatory agricultural interventionshad a significant effect on child growth.

Effects of agricultural and nutritioneducation projects on child health inMalawi

Margaret Shonga, participating farmer, herhusband Donald Gondwe and their baby,standing in a field of sorghum

Rach

el Bezn

er

Kerr

, M

ala

wi, 2

011

Esnai Ngwira, participating farmer, andher children, standing beside her maizefield grown after legume residue buried

into the ground

Rach

el Bezn

er

Kerr

, M

ala

wi, 2

011

Research

Summary of published research1

A recent study set out to investigatewhether children in households involvedin a participatory agriculture and nutritionintervention had improved growthcompared to children in matched compara-ble households. The study also exploredwhether the level of involvement andlength of time in the project had an effecton child growth.

The Soils, Food and HealthyCommunities project (SFHC) was initiatedby Ekwendeni Hospital and aimed toimprove child nutritional status amongstsmallholder farmers in a rural area innorthern Malawi. In villages surroundingEkwendeni, a town in Mzimba district,

21

2 Methodology and age of children investigated not reported

for this indicator.1 WFP (2011). Study report. Causality study on causes of

persistent acute malnutrition in north Darfur (Kabkabyia)

2010-2011. Report prepared by Insaf Ibrahim for WFP

Sudan.

Research

Cluster A: children suffered no acute malnutritionat the time of conducting this study, i.e. childrenwere well-nourished with +1 WHZ or above.Cluster B: children suffering from moderateacute malnutrition at the time of conductingthis study with a WHZ between -2 to -3. Cluster C: children suffering from severe acutemalnutrition at the time of conducting thisstudy based on their WHZ score assessment,which was <-3.

Study findingsThe households in all three clusters were foundto eat at least three meals per day. Food aid wasreported to be consumed by a larger number ofpeople than indicated by the ration cardsshown to the study team. As a result, food aidreportedly only lasted between 20-30 daysinstead of the planned 60 days.

All of the households in clusters B and C,and a few households from cluster A, experi-enced food shortage several times during theyear. Food shortages were also frequently expe-rienced in all households that did not cultivateland, which was more prevalent in clusters Band C than in cluster A. Most affected house-holds responded to food shortages by eatingless favoured foods that were cheaper and oflower nutritional value, reducing the number ofmeals and portion size, as well as borrowingfood and money. In addition, household headsalso worked to generate income.

The study found that children started breast-feeding a few hours after birth and that most ofthese children were still breastfeeding duringthe time of the interviews. Complementaryfeeding for all the children in this study startedat the age of 6 months2.

All households in the three clusters spentmost of their income (more than 50%) onpurchasing food. Earned income was also spenton purchasing firewood and paying for medicalservices in addition to food to supplement theGFD. In clusters B and C in particular, datasuggested that households’ adults who couldwork, would only seek work when there wasno food and/or money in the households.

Water sources for all households in the threeclusters came mainly from the hand pumps andwater tanks, depending on which water sourcewas nearer to the households. Enough water forall the households was collected in plasticjerkins everyday and each jerkin containedabout 20 litres of water. Water samples weretaken for testing from all households includedin this study. These samples were positive forcontamination for different types of bacteriasuch as Cirtobacter, Klebsiella, E.coli,Salmonella and Vibrio.

For most of the households, latrines werepresent outside of the household and weredonated by aid organisations. Each latrine wasused by 2-5 households. Mothers in all house-holds mentioned that they would washchildren’s hands several times during the dayand would bath these children regularly.Mosquito nets and blankets were not observed

vegetables, fruits, and oleaginous plants culti-vated in Kebkabiya and Jebel Si. The town wasalso known as a trading point for livestock suchas sheep, goats, cattle from nomads, and fornon-food items from El Fasher city.

The unstable security situation in Kabkabyiain recent years has limited people’s access toagricultural lands. Since the onset of the conflictin the region, WFP has been providingKabkabyia residents, including the IDPs, withfood aid in the form of a general food ration(GFD). This aid, which is distributed once every60 days, initially met 100% of the kcal require-ment of the IDPs. With an improved foodsecurity situation over several years, the GFDration was reduced in 2010 to provide 50% ofthe requirements (1200 kcal). Other food aidinterventions such as food for education, foodfor work, blanket supplementary feedingprogrammes (BSFP), and food rations forhouseholds with malnourished children, e.g.supplementary feeding programmes (SFPs),have also been implemented. The FSMS dataindicated that the decreased GFD ration did notadversely affected household food security.

Study site and groupEl-Salaam area was selected as the study site inKabkabyia town as it had the second highestpopulation density and the highest GAM ratein the town (>21% according to a May 2010survey). El-Salaam area is composed mainly ofIDPs who live outside of camps but are notliving with relatives in the town.

For the purposes of the study, a householdwas defined as a group of people who routinelyate out of the same pot and lived in the samecompound or physical location. In order tosatisfy the objectives of this study, selectedhouseholds had to be located in El-Salaam areain Kabkabyia town and have a child between 6-23 months of age.

Selected households were then organisedinto three clusters based on children’s nutri-tional status:

Study of causes of persistent acutemalnutrition in north Darfur Summary of study1

In spite of national and international effortsto manage the devastating impact of theconflict in Darfur which began in 2003, chil-

dren in different parts of the region haveconsistently demonstrated high levels of globalacute malnutrition (GAM) often exceeding theWHO emergency threshold of 15%. This hasbeen found in areas where the food securitysituation appears to be ‘good’ according to theregional food security monitoring system(FSMS) established by the World FoodProgramme (WFP).

WFP recently conducted a study to shedlight on this situation. The main objectives ofthis study were to investigate the underlyingcauses of acute malnutrition in North DarfurState, specifically in Kabkabyia town where thestudy was conducted, and to suggest feasibleand realistic recommendations to address thesethat would guide future interventions andprogrammes.

ContextKabkabyia was selected as it had experienced ahigh GAM rate (>15%) among children under 5years for the past five years, while the FSMSindicated that the town was relatively better offin terms of food security compared to otherareas in North Darfur.

Kebkabiya town is located in the southwestof North Darfur State approximately 165 kmfrom El Fasher, the capital of the state. The townis divided into 16 quarters, which are them-selves divided into several sub-quarters.Following the first major attacks on ruralKebkabiya and Jebel Si in July 2003, there was ahuge influx of internally displaced persons(IDPs) into Kabkabyia town.

Currently, it is estimated that 119,793 indi-viduals are living in the town, with IDPsaccounting for almost 70-75% of the population.Up until the start of the conflict in 2003,Kebkabiya was known as a central tradingpoint for agricultural products such as cereals,

Children and their caregiversenrolled in the BSFP

WFP,

N D

afu

r, S

udan

22

in most of households, particularly forhouseholds in cluster B and C.

Most households in cluster A werefound to have been engaged in agricul-tural activities on their own land duringthe last rainy season where they cultivatedmainly millet. Households in clusters Band C did not cultivate land which madethese households dependent on food aidand on purchase of food from the localmarket.

Cluster A households ate a wider vari-ety of food items in the two weeks prior toconducting the interviews whencompared to households in cluster B andC during the same period. In cluster A,food types consumed included cereals,sugar, cooking oil, dry and fresh meat,milk, biscuits, dry okra, fresh vegetables,and sometimes fruits. In comparison,households in clusters B and C werefound to have rarely consumed freshmeat, vegetables and fruits.

Children in cluster A were between theages of 7-22 months and were found tohave been fed more frequently i.e.between 3-4 times, when compared tochildren in clusters B and C who were fedbetween 2-3 times a day. Children in clus-ter B and C were mainly fed asida andpoor quality molah made of dry meat, dryokra and kawal.

Heads of households in cluster A werefound to have more access to regularsources of income and were either receiv-ing monthly salaries from regularemployment or owned small businesswhich provided regular sources of incomeall year as well as access to cultivation.The household heads of cluster B and Cdepended on seasonal employmentopportunities. These household reportedexperiencing money shortage and subse-quently food shortage frequently duringthe year.

Water consumption/uses in all house-holds seemed to be adequate. Differencesbetween the three clusters were mainly inwater uses/quality/hygiene. Observationof water containers, especially waterjerkins, from all households in cluster Alooked clean unlike most jerkins fromcluster B and C.

Interview results suggest that left-overfood was not consumed by the targetedchildren in most of the cluster A house-holds. The few households in cluster Awhich fed targeted children left-over,reported feeding children the leftoverfood only after reheating. They alsoreported food was consumed shortly afterit was prepared/reheated. These ‘good’food handling practices were not preva-lent in the other two clusters, where leftover food was often fed to the targetedchildren.

Mothers from all households in clusterA reported washing their hands with soapand water more frequently during the day,7-10 times, compared to mothers in clus-ters B and C who used to wash their handsonly between 5-6 times. Soap consump-

Astudy just published in the Lancet set out to assessthe effect of performance-based payment ofhealthcare providers on the use and quality of

child and maternal care services in healthcare facilities inRwanda. Payment for performance (P4P) schemesprovide financial incentives to healthcare providers forimprovements in utilisation and quality of specific careindicators. They can affect the provision of heath care intwo ways: by giving incentives for providers to put moreeffort into specific activities and by increasing theamount of resources available to finance the delivery ofservices. However, P4P schemes can have a detrimentaleffect. For example, when P4P payments depend oncompletion of reports, providers might spend more timeon administrative duties and less time ensuring thatpatients receive the best quality care. In this study, theresearchers assessed the potential of a P4P scheme toincrease use and quality of key maternal and child healthservices. The impact evaluation was done prospectivelyin parallel with the rollout of a national P4P programmein Rwanda.

One hundred and sixty-six facilities were randomlyassigned at the district level either to begin P4P fundingbetween June 2006 and October 2006 (interventiongroup, n=80) or to continue with the traditional input-based funding until 23 months after study baseline(control group, n=86). Randomisation was done by tossof a coin. The researchers surveyed facilities and 2,158households at baseline and after 23 months. The mainoutcome measures were prenatal care visits, institutionaldeliveries (births), quality of prenatal care, child preven-tive care visits and immunisation. The study teamisolated the incentive effect from the resource effect byincreasing comparison facilities’ input-based budgets bythe average P4P payments made to the treatment facili-ties. The team estimated a multivariate regressionspecification of the difference-in-difference model, inwhich an individual’s outcome is regressed against adummy variable, indicating whether the facility receivedP4P that year, a facility-fixed effect, a year indicator, and aseries of individual and household characteristics.

The model estimated that facilities in the interventiongroup had a 23% increase in the number of institutionaldeliveries and increases in the number of preventive carevisits by children aged 23 months or younger (56%) andchildren aged between 24 months and 59 months(132%). No improvements were seen in the number ofwomen completing four prenatal care visits or of chil-dren receiving full immunisation schedules. The teamalso estimated an increase of 0.157 standard deviations(95% CI 0.026-0.289) in prenatal quality as measured bycompliance with Rwandan prenatal care clinical practiceguidelines. The P4P scheme in Rwanda had the greatesteffect on those services that had the highest paymentrates and needed the least effort from the serviceprovider.

Researchers concluded that P4P financial performanceincentives can improve both the use and quality ofmaternal and child health services and could be a usefulintervention to accelerate progress towards MillenniumDevelopment Goals for maternal and child health.

Effects of performancepayments to healthworkers in RwandaSummary of published research1

1 Basinga. P et al (2011). Effect on maternal and child health services

in Rwanda of payment to primary health-care providers for

performance: an impact evaluation. Lancet 2011, 377: 1421-28

Research

tion was reported to be more prevalent inhouseholds in cluster A than in house-holds in clusters B and C.

All children in cluster A where foundnot to have not experienced any illnesssuch as diarrhea, vomiting, fever orcommon cold within the last 30 days priorto conducting of the interviews for thisstudy. On the other hand, all childrenincluded in Clusters B and C were sickwith diarrhoea, vomiting and fever withinthe last 14 days prior to conducting of theinterviews. Food consumption of childrenin these clusters during the illness periodwas described as very poor. Motherreported that these were children mainlydependent on breastfeeding during thebouts of illness. These findings were moreevident in cluster C (severely malnour-ished children).

Discussion and RecommendationsAlthough the sample size of householdsincluded in this study was small andtherefore, findings cannot be generalizedto the larger population in Al-Salaam areaor Kabkabyia town these findings are stilluseful for planning purposes.

Findings suggest that agencies shouldconsider job creation interventions, e.g.income generating activities, that would liftvulnerable populations out of poverty.Training on proper finance management atthe household level should also be consid-ered in an effort to change the noted cultureof “I only need to look for work when thereis no money or food in the house”.

It is also important to look at theadequacy of food aid rations received bydisplaced people. The study finds a signif-icant discrepancy between number ofpeople living in a household with thenumber registered on the ration card, sothat the ration does not last as long asplanned. The ongoing re-verification exer-cise of the IDPs in Darfur should help inaddressing such discrepancies and shouldalso assist WFP in determining whichhousehold are more vulnerable thanothers and therefore allow for provision offood aid required accordingly.

There is also a need for more educationand awareness raising programmesaround issues of hygiene and sanitation,as well as more provision of soaps/deter-gents or water purifiers as necessary to thehouseholds.

The issue of soap shortage in most ofthe households with malnourished chil-dren should also be addressed. This can bedone through increasing the soap rationreceived, which should be linked to theresults of the proposed verification exer-cise in order to properly match thenumber of people actually living in thehousehold with the number of soap barsto be received.

There also needs to be awareness rais-ing activities for mothers and childcaregivers regarding symptoms andmanagement of child malnutrition with anemphasis on child feeding practices.

23

Summary of working paper1

In 2010, Concern Worldwide developed a humanitarianprogramme in response to the 2009/10 Niger droughtand food crisis. In an attempt to prevent asset depletion

and reduce malnutrition among drought-affected house-holds, the programme provided unconditional cashtransfers to approximately 10,000 households during thehungry season, the five-month period before the harvestand typically the time of increased malnutrition.Programme recipients were to receive an average of 22,000CFA (US$ 45) per month for five months, to a total of US$215. In an effort to facilitate the disbursement of cash inremote areas, Concern decided to implement a pilot studyacross 116 villages in six communes of the Tahoua region.

Three interventions were chosen for the pilotprogramme. The reference was the standard manual cashintervention, whereby beneficiary households received

unconditional cash transfers of 22,000CFA ($US45) per month. The total value ofthe transfer over the five-month periodwas approximately two-thirds of the totalannual gross domestic product (GDP) percapita. Payments were made on amonthly basis, whereby cash was countedinto envelopes and transported viaarmoured vehicles to individual recipi-ents. Rather than distributing the cash ineach village, a central village location waschosen for groups of 4-5 villages.Programme recipients had to travel totheir designated location on a given dayto receive the cash transfer.

The two additional interventions werevariants of the basic intervention. One ofthese aimed to reduce the costs of distrib-uting cash to remote, sparsely-populatedand in some cases, insecure rural areas.Programme recipients in the secondgroup (zap) received their cash transfervia mobile phone (m-transfer). Afterreceiving the electronic transfer, recipientshad to take the mobile phone to a m-trans-fer agent located in their village, a nearbyvillage or a nearby market to obtain theirphysical cash. Since less than 30 percent ofhouseholds in the region owned mobilephones prior to the programme, Concernalso provided programme recipients withmobile phones, as well as the zap account,and paid for the transfer charges. Thesecond intervention thereby differed fromthe manual cash intervention with respectto the transfer delivery mechanism, aswell as the provision of the handset andthe m-transfer technology.

In an effort to disentangle the impact ofthe change in delivery mechanism fromthat of receiving a mobile phone, the thirdgroup (placebo) mirrored the manual cashintervention, but also provided a mobilephone. Like the manual cash group,programme recipients received $US45 inphysical cash on a monthly basis and hadto travel to a meeting point to receivetheir cash. However, like the zap group,programme recipients also received amobile phone (which was ‘zap’ enabled),yet did not receive their transfer via themobile phone.

Prior to the introduction of theprogramme, Concern Worldwide identi-

fied 116 ‘food deficit’ villages in theTahoua region, i.e. those classified by theGovernment of Niger as having producedless than 50 percent of their consumptionneeds during the 2009 harvest. Of these,some villages were prioritised for the zapintervention based upon their populationsize and proximity to skirmishes near theNiger-Mali border, thereby reducing thesample size to 96 villages. The remainingeligible villages were randomly assignedbetween the basic (manual cash), placeboand zap interventions. In all, 32 villageswere assigned to the cash group, 32 to theplacebo group and 32 to the zap group.

Within each food deficit village, house-hold-level eligibility was determined bytwo primary criteria: the level of house-hold poverty (determined during avillage-level vulnerability exercise) andwhether the household had at least onechild under five years. The number ofrecipient households per village rangedfrom 20 to 75 percent of the village popu-lation. In all villages, the cash transfer wasprovided to the woman.

A comprehensive household survey ofmore than 1,200 programme recipientswas conducted in all 96 villages. The base-line survey was conducted in April 2010,with a follow-up survey in December2010. The research team located over 98percent of households for the follow-upsurvey. The household survey collecteddetailed information on household foodsecurity, demographics, asset ownership,agricultural production and sales, mobilephone ownership and usage, uses of thecash transfer and village and household-level shocks. A second dataset wascollated on weekly agricultural priceinformation from over forty-five marketsfor a variety of goods between May 2010and January 2011, as well as the date ofeach cash transfer in each village. Thesedata were used to test for different effectsof the cash transfer delivery mechanism(zap or manual cash) on local marketprices.

1 Aker. J et al (2011). Zap it to me: The short term

impacts of a mobile cash transfer programme.

Centre for Global Development. Working Paper 268.

September 2011.

Research

Hadijatou with her cash, Toro village

Conce

rn,

Nig

er, 2

010/1

1

Figure 1: Examples of communication materials used in the programme

‘Zap’ it to me: short-term impacts ofa mobile cash transferprogramme

The average per recipient costover the life of the project wasUS$12.76 in cash/placebovillages and US$13.65 in zapvillages, or US$0.90 more perrecipient. While there was arange of benefits from the zapintervention, the researchfocused on two in particular forthe cost-benefit analysis: themonetary value of the reducedopportunity costs of programmerecipients’ time (a value of US$0.91) and the increased cultivationof cash crops. Using averagehousehold okra production andthe market price for okra duringthe programme period, the aver-age value of this okra productionin zap households would havebeen US$5. This suggests that thecost-benefit ratio is greater thanone, meaning that the additionalcosts of the zap interventionyielded an equivalent or highermonetary benefit for zapprogramme recipients. If theprogramme yields benefits in thelonger-term, perhaps by allowinghouseholds to send and receivemore informal transfers or accessformal financial services, thiscould potentially yield a higherrate of return.

An intervention that provided acash transfer via the mobilephone strongly reduced the costsof programme recipients inobtaining the cash transfer, andreduced the implementingagency’s variable costs associatedwith distributing cash. Thissuggests that mobile telephonycould be a simple and low-costway to deliver cash transfers. Inaddition, those in the m-transfergroup bought more types of foodand non-food items, increasedtheir diet diversity, depleted theirnon-durable assets at a slowerrate and produced a more diversebasket of agricultural goods.These differences are primarilydue to the m-transfer interven-

24

Numerous countries still implement growth monitoring (GM) astheir main community-based nutrition activity. A health surveyin 2003 showed that 154 countries worldwide used growth

charts, with two-thirds of the charts covering preschool-aged children.In the mid-1980s, several consultations suggested that GM should bedesigned with additional promotional activities to become growthmonitoring and promotion (GMP.) GMP was envisioned as a corner-stone activity that would help target at-risk children for secondaryinterventions, as a way of empowering caregivers and households totake an active role in preventing malnutrition of their children, and as away to encourage the use of other services available through primaryhealth clinics.

Differing opinions about the impact and outcomes of GM and GMPhave led to different conclusions from evaluations and assessments ofcommunity-based programmes including GM. This has led to a rela-tive lack of clarity and common ground in discussions about the valueand place of GM and GMP in addressing the problem of undernutri-tion in children.

A recent review attempts to provide answers to questions aboutGM, such as its added value and possible place within community-based programmes.

The reviewed literature showed that the objectives and expectationsof GM and GMP vary, and programme evaluations are performedbased on different frameworks. Furthermore, multiple reasons for thelack of impact of GMP have been cited in evaluations. These include afocus on nutrition status rather than faltering growth, a misplacedemphasis on curative rather than preventive actions, enrolment of chil-dren in GMP programmes after (instead of during) infancy, the use ofGM as an isolated activity instead of a cornerstone activity, the lack ofindividualised advice, the lack of positive feedback for mothers whosechildren are growing adequately, the lack of community participation,an oversimplification of the GMP process, and poor quality of imple-mentation.

As a result of these evaluations, agencies behind large-scale imple-mentation of GMP were criticszed. At the same time, largeprogrammes in Tanzania (Iringa), India (Tamil Nadu Integrated

Revisiting the concept ofgrowth monitoring and itspossible role in community-based programmes

1 Mangasaryan. N, Arabi. M, and Schultink. W (2011). Revisiting the concept of growth

monitoring and its possible role in community-based programmes. Food and Nutrition

Bulletin, vol. 32, no. 1 © 2011, The United Nations University.

http://www.foodandnutritionbulletin.org/

Research

Summary of review1

tion, and not to the presence ofthe mobile phone, suggestingthat a programme that simplydistributes mobile phones mightnot yield the same impacts. Theseeffects appear to be due to thereduced costs of the programmeand the greater privacy of the m-transfer mechanism, which arepotentially linked with changesin intra-household decision-making.

The m-transfer approach maybe limited in its application to allcontexts. First, it will only beeffective in cases where telecom-munications infrastructurecurrently exists, which couldlimit its utility in remote areas.Second, in areas with high ratesof illiteracy – as is the case inNiger – programme recipientsmight not able use the m-transfertechnology on their own, imply-ing that they might need helpfrom other family members,friends or m-transfer agents. Thiscould potentially limit the use ofthe technology by programmerecipients for informal privatetransfers or in accessing othermobile financial services, butcould be beneficial for the house-hold as a whole. And finally, theshort-term impacts of theprogramme might not persist inthe longer-term. Despite thesecaveats, the widespread growthof mobile phone coverage,cheaper mobile phone handsetsand m-money services in devel-oping countries suggests thatthese constraints could be easilyovercome. In addition, the bene-fits of the programme in a contextsuch as Niger - a country withlimited investment in power,roads and landlines, low literacyrates and one of the highest ratesof financial exclusion in sub-Saharan Africa - suggests that theapproach could thrive in lessmarginalised contexts.

Women in Abala Sani village at amobile phone training session

Conce

rn,

Nig

er, 2

010/1

1

Growth monitoring

© U

NIC

EF/N

YH

Q2008-1

444/B

onn

ity of reaching a high quality of GMP activities.In many settings where a concrete nutritionalproblem is affecting most of the population,such as micronutrient deficiencies or lowbreastfeeding rates, a targeted intervention maybe a first priority for reaching quick improve-ments before deciding on more comprehensivecommunity-based programmes, which couldinclude GMP.

Although it is not strictly necessary for inclu-sion in any community-based programme,under certain conditions having quality GMPcan add desirable aspects to these programmes.The approach of regular monitoring of childgrowth provides the opportunity for bettercommunity actions to prevent undernutrition.

High quality GM can:• Provide an opportunity to prevent

undernutrition before it occurs. GMP helps community workers identify infants and children who have growth faltering (or are at risk for faltering) and promotes timely actionsto improve the situation within a short time frame

• Assist in focusing attention and resources on children at risk.

• Motivate families and caregivers to change and improve practices.

• Help target and tailor counselling messages.

• Produce ancillary benefits. GM sessions provide opportunities for immunisation, screening and treatmentfor diarrhoea, malaria, and pneumonia,counselling on various health and nutrition topics and the provision of other community-level health or preventive services as needed.

These additional benefits that are perti-nent to GMP do not receive enoughattention during most of the evaluationsof community-based programmes.

In general, the level of commitmentfrom the health system required forsuccessful implementation of GM andGMP has proven difficult to maintain ata large scale, with the exception of fewwell-supported and well-supervisednational programmes. Supportivesupervision of community health work-ers requires ample allotment of time andfunding, which may not be realistic

within a strained healthcare system.

Appropriate implementation of GMP isdependent on the motivation of health workers.Experience shows that community workers canbe effectively motivated to accurately measure,plot, and diagnose growth faltering but areoften undervalued, under supervised, andpoorly paid. The ratio of trained staff to thetarget population may also be inadequate.

The quality of training of community work-ers requires significant resources and efforts. Inan evaluation of nine projects (governmentallyand non-governmentally implemented) inAfrica and Asia that included GM, most of thesettings had adequate infrastructure to supportGM but training was incomplete, leaving only asmall proportion of the staff able to adequatelytake weight measurements.

In addition, the low educational level ofcommunity workers in some settings impedes

Nutrition Project), Madagascar, and Senegalshowed that children whose growth is moni-tored and whose mothers receive nutrition andhealth education and have access to basic childhealth services have a better nutritional statusand/or survival than children who do not.

The debate about GMP has remained ongo-ing. In 2003, a report by Save the Children UKquestioned the evidence behind communitynutrition projects in Bangladesh, Ethiopia, andUganda. It also stated that “growth monitoringand promotion interventions are bound to failunless they are explicitly linked to efforts toaddress the underlying causes of malnutrition.”

A systematic review of the evidence for theimpact of GMP in 2007 provided a comprehen-sive view of various programmes worldwideand provided evidence that significantreductions in malnutrition can beachieved through intensive health andnutrition education and basic health-care without GM.

After the launch of the new growthstandards by WHO in 2006, a momen-tum was created to revisit GM activitiesand rethink the best use of the years ofexperience. As countries have begun toadopt the new standards, many ques-tions have been raised concerning theprogrammatic uncertainties of GM atthe community level.

Many countries face a challenge indealing with the question of whether ornot to implement GM and GMP.Despite all the developments in nutri-tion programming in the past 10 years,GM still seems to be a convenient deliv-ery mechanism for communityinterventions. However, national plan-ners need better guidance ontransitioning to alternative options thatare not based on monitoring growth inthe communities, if GM has not provento be effective in contributing toprogrammes for prevention of under-nutrition.

Part of the confusion about the placeof GM in nutrition programmesappears to be due to lack of agreementon the definition and expectedoutcomes of GM and GMP. The authorsof this review suggest the followingclarifications:

Growth monitoring is a process of followingthe growth of a child compared with a standardby periodic, frequent anthropometric measure-ments and assessments. The main purpose ofGM is to assess growth adequacy and identifyfaltering at early stages before the child reachesthe status of undernutrition.

Community based growth monitoring is notitself an intervention that can treat growthfaltering when it is identified. It is rather anactivity which, in addition to making a child’sgrowth visible, may become an important pointof contact with the caregiver and stimulatediscussions at the community level. If imple-mented as a stand-alone activity, GM does notprovide any benefits apart from knowledgeabout a child’s growth status.

GMP is defined as tailored counselling basedon the GM results and follow-up problem solv-

25

Combining GMP and additional interven-tions needs to be planned carefully to ensurethat the quality of both is maintained. In somesettings, workers may become overburdened byadditional tasks and focus most of their attentionon delivery of services rather than effectivecounselling and problem-solving with mothers.

Evidence is accumulating on the types ofcommunity interventions that are effective,practical, and sustainable. These interventionsare not necessarily linked to GM, which raisesthe question of whether there is a need for thisactivity if the community-based programmescan be designed and implemented successfullywithout monitoring the growth of each child.

The decision to include GM and promotionsessions in community-based programmesneeds to be made at the national and sub-national levels after careful consideration ofpriorities, available resources, and the feasibil-

ing with caregivers. This allows looking intogrowth monitoring-specific outcomes andbenefits, as compared with general counsellingand other interventions that could be deliveredoutside the GM session as well.

A community-based programme shouldinclude a number of interventions such asgeneral counselling to caregivers (either indi-vidually or in groups) and delivery of differentservices within the context of the community-based programme. These interventions andservices could be delivered during the sameGMP session, using the opportunity of thecontact with caregivers. These services,however, are not dependent on measuring thegrowth of children and can also be deliveredoutside the GM context.

Research

Growth monitoring

UN

ICEF/N

YH

Q2011-0

273/D

orm

ino

26

their capacity to interpret and analysegrowth measurement results, identifyat-risk children, and analyse possiblecauses of growth faltering.

Although good coverage has beenshown in small-scale programmes,reaching all targeted children isgenerally difficult to achieve, andattendance is often less than desired.The frequency of GMP attendanceoften declined in children of older agegroups, and children who were mostat risk attended less often than better-off children. Health managersworldwide attribute low attendanceto a lack of interest by mothers aftercompletion of vaccination, weakawareness campaigns to motivatemothers, and the inability of parentsto respond to information providedduring the sessions (due to illiteracy,inability to understand the growthchart, or lack of access to foods).

A review of GMP in seven coun-tries concluded that GMP is notimplemented appropriately andattributed its failure to a lack ofadequate investment and to the factthat GMP is often implemented inisolation from other necessary nutri-tional actions. The keyimplementation problems were lowcoverage (often the poorest childrenhad the worst coverage), no action orlow-quality action taken based on theanalysis of GM data, and no agree-ment on the human, organisational,and financial resources needed forsuccessful GMP.

In general, GMP has been shownto be successful in cases where it wasadded to an existing well-managedand well-supervised health system,where health workers and commu-nity workers were adequately trainedand recognised for their work, whereaccurate equipment and materialswere available, where communitieswere involved in the GMP process,and where culturally appropriatecommunication was developed

Important questions to answer inevaluations of GM or GMPprogrammes could be:

“Does the measurement facilitatedialogue and counselling?”

“To what degree does information aboutchild growth affect the quality of

counselling?”

“To what extent can community workersprovide quality tailored counselling

based on growth status?”

Such questions need to be answeredby looking at different outcomes,including the caregiver’s awareness ofthe child’s growth status, knowledgeabout necessary care practices, confi-dence and satisfaction with theacquired information during coun-selling sessions and child carebehaviours.

because of its centrality in the country’s electoralpolitics. For the government, food security can be guaranteed by the use of high-yielding hybrid maize varieties.

• Seed companies are keen to promote hybrid maize seeds since they are their main productand through the subsidy programme, seed companies are guaranteed a ready market.

• Donors are interested in promoting a private sector-driven input supply system through the promotion of agro-dealers to fill the vacuum following the dismantling of the state-driven input supply system through liberalisation. This has made hybrid maize thedominant seed that is made available to farmersthrough the input subsidy programme.

• Most communities argue that crop diversifi-cation cannot be a success due to the culturalorientation that equates food to maize. Food- stuffs made from alternative cereals, such as millet and sorghum, are widely perceived as ‘inferior’ to the extent that households resortingto such foodstuffs are taken to be desperate.

• Most international humanitarian non-governmental organisations (NGOs) perceive crop diversification as desirable but not attainable as long as weather index insurance schemes are exploited as commercial venturesby the private sector. They condemn the insurance schemes as a barrier to crop diversi-fication mainly due to their institutional arrangements. Farmers take loans from a designated bank, procure seeds from a speci-fied company, and dispose of their produce to a designated buyer. Not only are the farmers bound up in a particular institutional arrange-ment that effectively curtails their freedom, butthe seed companies involved are known to promote almost exclusively hybrid maize.

The major lesson from the study is that policyinfluence is not merely a question of generatingrobust scientific evidence and making it availableto policymakers. It is as much about generatingnew evidence as it is creating strategic partner-ships, coalitions and alliances. Policy engagementand influence entails engaging with policymakersfor them to understand the implications of new oreven old evidence as vital input into their deci-sion-making processes. Therefore, the manner inwhich evidence is packaged, how it is communi-cated to policymakers, by whom and when,matters a great deal for researchers to effectivelyengage with the policy processes and influencethe final outcomes. The main implication of theKNOTS framework is that processes ofdiscussing, negotiating, approving and imple-menting policies are as important as the scientificcontent of the policies themselves.

The experience with crop diversification showthat dominant stakeholders almost always havetheir way and that implementation of crop diver-sification has been constrained by a dominantnarrative that equates food security with maize.The input subsidy programme, which is the singlemost important resource in the agricultural sector,has been captured by politicians primarily toadvance their own political goals. Alternativenarratives had had no significant effect on themainstream narrative of crop diversification thatequates food security with availability of maize.The case study sheds light on the dynamics ofpolitical, economic and social processes thateither promote or block pro-poor change.

1 Chinsinga.B, Mangani. R, and Mvula.P (2011).The political

economy of adaptation through crop diversification in Malawi.

IDS Bulletin Volume 42, Number 3, pp 110-117, May 2011.

The seriousness of the problem of climatechange and its negative effects on liveli-hoods is widely recognised in Malawi, even

though no single coherent policy frameworkexists. However, there are several sector policiessuch as crop production (1990), national environ-mental management plan (1994), national forest(1996), national irrigation (2000), amongst others.In addition, Malawi developed and adopted theNational Adaptation Programme of Action(NAPA) in 2006, which serves as a framework forclimate change adaptation efforts in the country.The objectives of NAPA are to improve commu-nity resilience, restore forests, improveagricultural production and improve prepared-ness for floods and droughts.

A recent article examines the opportunities andchallenge of climate change adaptation in Malawiusing the case of crop diversification. It drawsfrom an innovative experiment about policyengagement and influence between two sets ofresearchers: those working with the Research toPolicy for Adaptation project and ParticipatoryAction Researchers. The former are mainlyexperts in policy processes, whereas the latter areexperts in climate change adaptation. The engage-ment between these two groups was guided bythe conceptual framework for policy processesdeveloped in the Knowledge Technology andSociety (KNOTS) team at the Institute ofDevelopment Studies, UK.

The framework analyses policy processes fromthree perspectives: narratives and evidence, actorsand networks, and politics and interests. The basicthrust of the framework is the way in which poli-cies are talked about, and the associated values,power relations and politics that frame policies ina particular way. The framework draws attentionto the fact that policymaking and processes cannotbe reduced to universal recipes that are supposedto work irrespective of the time and place inwhich they are adopted. In other words, policiesare more effective when they are informed by anunderstanding of power relations, incentives andchange processes.

The key findings of this research included thefollowing:• While constantly making references to the ideals

of crop diversification, the main preoccupationof the government is to achieve food security

Political economy ofadaptation throughcrop diversificationin Malawi Summary of article1

ResearchA government agricultural extension

officer showing a farmer how to care fora cassava plant (FAO supported project)

©FA

O/J

on S

paull

27

Amass or ‘blanket’ supplementary feed-ing programme (BSFP) wasimplemented by the World Food

Programme (WFP) and partners in five north-ern districts of Kenya between January andApril 2010. It was undertaken due to fears of anincrease in the incidence of malnutrition as aresult of seasonal food insecurity exacerbatedby persistent drought. The five programmedistricts of Mandera, Marsabit, Samburu,Turkana and Wajir cover 45% of Kenya's totalland area (Figure 1) but at the time, containedonly 4.5% of the population of 28.87 millionrecorded in the 1999 census2.

An attempt to evaluate the impact of thefood on children's anthropometric status wasput in place in three districts. A recentlypublished study set out to assess the quality ofthe data on the cohort of children studied in theevaluation and to propose methods by which itcould be improved to evaluate future blanketfeeding programmes. Reasons for the poorquality of the evaluation are identified.

BSFP interventionThe primary stated aim of the programme wasto protect the nutritional status of an estimated300,000 children aged 6–59 months, or 20% ofthe 1999 census population3. All children <110cm in height were eligible for a ration of foodplus any taller children whose mother insistedthat they were <60 months of age.4,5

Rations of food provided by WFP weredistributed on four occasions, each about 30days apart, beginning in January 2010. Theyconsisted of 7.5 kg of corn-soy blended flour(CSB) and 0.75 kg of vegetable oil to provide anaverage of 1,000kcal/day/child. The food wasdistributed by non-governmental agencies(NGOs) at the sites of pre-existing feedingprogrammes and at some extra sites, toimprove local access.

An evaluation was undertaken to try todetect evidence for an effect of the rations onthe anthropometric status of children in threemain ways:• By comparing the anthropometric indices

of newly recruited children at the second and third food distributions with children enrolled at the first distribution, to assess if their anthropometric status was getting worse during a period when food security was supposedly poor or deteriorating.

• By comparing the weight change of children who received two, three or four rations of food during the programme, in order to detect a dose-response relationship.

• By comparing the weight change of singleton children with children matched for age and sex in households of two or more children, based on the premise that if a ration was shared it would be less effectivethan if it was given to an only child, and assuming that the ration was not shared outside the household.

Evaluation sites selection and processThe rations of food were distributed initially at540 sites in the districts of Mandera (99 sites),Marsabit (55), Samburu (101), Turkana (162)and Wajir (123) (Figure 1) by a group of eightNGOs led by Save the Children, UK (SCUK).

An arbitrary number of 25 food distributionsites were randomly selected for study in eachof two adjacent districts, Mandera and Wajir, 26

sites operated by SCUK and 24 by IslamicRelief. At the request of the National NutritionTechnical Forum, 25 sites were also randomlyselected in Turkana district, which is in a differ-ent livelihood zone and contains a differentethnic group, the Turkana (most people inMandera and Wajir are Somali). Four agencieswere responsible for collecting data in Turkana:Merlin (10 sites), Samaritan's Purse (4), IRC (1)and World Vision (10). Because of a delay infunding, World Vision did not collect data attheir 10 sites.

The staff of each NGO was responsible bothfor distributing the food and for collecting datafor the evaluation. All members of staff weregiven one day's theoretical training by SCUKon the BSFP, community mobilisation, organis-ing distribution sites and on the evaluationmethods, including sampling children andadministering questionnaires. The five NGOsthen organised two days practical training fortheir field staff according to an agenda speci-fied in the programme guidelines6. All theNGO staff were nurses or nutritionists whowere supposedly practiced at making anthro-pometric measurements, so no specific trainingon anthropometry was arranged.

The aim was to recruit up to a 10% sample ofchildren at the first food distribution at eachstudy site and then at the same sites, recruit allnew children who were brought to claim a

Practicalchallenges ofevaluatingBSFP innorthernKenyaSummary of published study1

1 Hall A, Oirere M, Thurstans S, Ndumi A, Sibson V, 2011.

The Practical Challenges of Evaluating a Blanket

Emergency Feeding Programme in Northern Kenya. PLoS

ONE 6(10): e26854. doi:10.1371/journal.pone.0026854 2 Kenya National Bureau of Statistics (2007). Statistical

Abstract Nairobi: Kenya National Bureau of Statistics.3 Save the Children (2009) Blanket supplementary feeding

programme monitoring and evaluation guidelines. Nairobi:

Save the Children. p58.4 See footnote 25 World Health Organization (2000) The management of

nutrition in major emergencies. Geneva: World Health

Organization. p236.6 See Footnote 37 StataCorp (2010) Stata Statistical Software: Release 11.0.

College Station, Texas, USA: StataCorp.8 ACF USA (2009) Anthropometric and retrospective

mortality surveys in the Districts of Mandera, Kenya.

NairobiKenya: Action contre la Faim.9 See Footnote 310 See Footnote 611 WTO (2011) Macros to analyse growth data for the age

group 5-19 years.Geneva: World Health Organization.

Available: http://www.who.int/growthref/tools/en/

Accessed 2011 Oct 9.

Figure 1: Five districts of Kenya in which the blanketsupplementary feeding programme was implemented in January 2010

0 50 100 200 300 400

Marsabit

Turkana

Wajir

Mandera

Sambutu

Research

Beneficiaries of BSFP in Kenya

SCU

K,

Kenya,

2011

28

taken to be acceptable.• The difference in length or height of each

child between the first and last measurements. An acceptable range was taken to be −1 cm to +4 cm. This is a combination of measurement error and rounding (which was evident in the data) of ±1.0 cm; changes in measuring children from supine to standing of 0.7 cm, plus a possible gain in height of up to 2.7 cm rounded up to 3.0 cm. This height gain is the maximum possible gain for a nearly 5 year old boy who is 3 S.D. above the median in height according to WHO growth references. A change greater than 4 cm or less than −1 cm should not have beenpossible.

Findings of studyOf the 3,544 children enrolled, 483 (13.6%) didnot return to collect a fifth ration. Of the 3,061children who did return, 196 (6.4%) had adifferent name and 200 (6.5%) had a possiblydifferent name, indicating that perhaps up to13% of mothers had brought a different child tocollect the last ration. There were three namesmissing.

Figure 2 shows the age distribution of 3,397children aged 6–59 months whose age wasrecorded at enrolment compared with theexpected age distribution based on the 2009census in the same three districts12. Theexpected number of children aged 6–11 monthswas estimated by dividing by two the numbersrecorded for children aged 0–11 months. Figure2 shows that there were 89% more childrenthan expected aged 12–23 months and 56%fewer children aged 48–59 months, suggesting

ration at the second and third food distribu-tions, as they should not yet have received anysupplementary food.

A power calculation using Stata 117 indi-cated that a sample size of 3,022 children coulddetect a 4% difference in the prevalence ofwasting from 26% (the average prevalencereported in three previous surveys)8 over theperiod of intervention. This allowed for adesign effect of 2 due to the clustering of chil-dren around distribution sites, 25% drop-out,and assuming a power of 80% and a two-sidedstatistical significance of P<0.05.

Each child was weighed to a precision of 0.1kg on electronic scales (Uniscale, UNICEF) andmeasured to a precision of 0.1 cm, supine if <87cm and standing if ≥87 cm on locally madestadiometers, according to Kenya Governmentguidelines9.

Each caregiver was given a ration card for thechild with a unique identification numbercreated from the site code and the child's serialnumber. These numbers were also recorded in aregister book for each site and on the data formsfor each child at each visit to collect a ration andwere used to link data. Ration cards were givenonly at sites taking part in the evaluation.

The date of birth and date of visit were used toestimate each child's age in months at enrol-ment and z-scores of height-for-age,length-for-age and weight-for-height werecalculated using a macro for Stata 1110

published by the WHO11. This flags values ofweight-for-age that are >5 and <−6 S.D., valuesof height-for-age that are >6 and <−6 S.D. andvalues of weight-for-height that are >5 and <−5S.D. because the underlying data are likely tobe wrong.

Assessment of evaluation data quality In order to assess the quality of the datacollected in the evaluation, five indicators wereused:• The name of the child recorded on both

occasions on different data forms. The names were judged to be the same, differentor possibly different.

• The age distribution of children aged 6 – 59months, which should be similar to the distribution reported in the last district census.

• The number of z-score values that were flagged by the WHO anthropometry macroin Stata, and the difference in months between the age estimated at the first and last visits. For the purpose of analysis, a difference of ±3 months was arbitrarily

a bias towards younger children. Only 93 chil-dren (2.63%) were older than 60 months (notshown in Figure 2), which seems unlikely if theentry criterion to the programme was based ona height of <110 cm rather than age and shouldhave included older but stunted children. Therewere no statistically significant differences inthe mean reported age of children enrolled atthe first, second or third food distributions.

The WHO macro to calculate anthropometricindices flagged baseline values of weight-for-height, height-for-age and weight-for-age for237 children (6.67%). of these, 67 (2.56% of thetotal) were weight-for-height, suggesting that ameasurement of weight or height was incor-rect. The same values were flagged for fewerchildren at the fifth food distribution: 113(3.18%) had any index flagged while 35 (1.17%)had the value of weight-for-height flagged.

Figure 3 shows the distribution of the differ-ence in age in months recorded for 3,061children at enrolment and at the fifth fooddistribution, an average of 97 days later (range16–135 days), depending on when childrenwere enrolled. Only 21.23% of children wererecorded as having the same month of age,23.72% were 1 to 3 months younger or older,25.22% were 4 months or more younger, and29.79% were four months or more older. Insummary, 44.95% of children were within ±3months of the same age and 55.05% were ≥4 or≤4 months different in age.

Figure 4 shows the distribution of the differ-ence in height of 3,032 children measured atenrolment and the fifth distribution of whom66.09% were within a range of >−1 to <4 cm,15.77% were >1 cm shorter, and 18.14% were ≥4cm taller.

Of the 2,640 children who were consideredby their name to be the same on both occasions,data on only 902 children (34.17%) wereconsidered to be acceptable based on both theirstated age (±3 months different) and length orheight (>-1 or ≤4 cm different) at the twoinstances they were seen. This meant that dataon nearly two thirds of children were of ques-tionable quality. Because of thesediscrepancies, no further analysis was done toassess the impact of the feeding programme.

12 Kenya National Bureau of Statistics (2011) Kenya National

Bureau of Statistics The 2009 Kenya Population and

Housing Census.Volume 1C.Population distribution by age,

sex and administrative units. Nairobi. 546 p.

Research

Figure 2: The observed and expected numbers of children (n = 3,555) by age group and sex

Expected

Observed

Expected

Observed

Expected

Observed

Expected

Observed

Expected

Observed

Number of children

Girls Boys

600 400 200 200 400 600

6-12

12-23

24-35

36-47

48-59

Age

gro

up (m

onth

s)

Figure 3: The distribution of the differences in ages of children in months estimated at the first and fifth fooddistributions

25

20

15

10

5

0

Perc

enta

ge

≤-13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 ≥13

Difference between age in months recorded at first and fifth food distributions

old for health programmes. Ideally this crite-rion would be used for all programmesbecause it is simple, objective and transparent,and would include stunted children older than5y, who could also benefit from most interven-tions. The disadvantage of using height is thatthere is no easily calculated denominator toestimate both the numbers of eligible childrenand coverage, whereas a denominator basedon age can be estimated from census data. Asan estimate of coverage is an important indica-tor of the effectiveness of an intervention, aseparate survey would be necessary at addi-tional expense.14,15

Third, every study child ideally should beidentified either using a digital photograph orperhaps using a fingerprint reader, either in apersonal digital assistant (PDA) or connectedto a laptop computer, to confirm their identityat subsequent contacts. Battery powered PDAscould also be used to collect, store and comparedata in the field, so that widely differinganthropometric measurements could beflagged and checked immediately. Suchdevices require a capital outlay, a softwareprogrammer and field testing before deploy-ment, which is expensive. But this couldimprove data quality and speed up the processof analysis and reporting, as well as increasingthe validity of the evaluation. If suitable equip-ment is not available, then key data should becopied onto ration cards to act as a check,including the estimated date of birth and thefirst height and weight. Neither process wouldguarantee that the same child is seen on alloccasions, but any substitute could be identi-fied on the spot.

Fourth, the staff doing the evaluation shouldbe different from the staff delivering the rationcards, food or other interventions, so that bothjobs are done as well as possible in an oftenchaotic and busy environment in whichagitated parents demand attention. A dedi-cated evaluation team would requireadditional funding, an issue not addressedhere, but the compromised evaluation alsorepresents a waste of resources, as well as thetime of staff and mothers. The evaluation staffalso require specialised training in anthropo-metric measurements, even if they have donethem many times before, because both accu-racy and precision are required and should notbe assumed.

Finally, data analysis should be done asquickly as possible in the field, so that system-atic errors such as rounding can be identifiedand rectified by re-training or reorganisation ofprocedures. If all data are entered in the fieldonto PDAs, the confirmation of each entrywould duplicate the process of double dataentry. Data could also be merged from differentfield teams and analysed quickly in the fieldusing batch files written for statistical software.By reporting the problems and lessons learnedfrom this evaluation of a BSFP, it is hoped thatfuture evaluations will be better planned andimplemented and may provide plausibleevidence of a benefit to children's nutritionalstatus.

Figure 4: The distribution of the difference in height of children at the first and fifth food distributions

20

15

10

5

0

≤ -1

0

< -9

to >

-10

< -8

to >

-9

< -7

to >

-8

< -6

to >

-7

< -5

to >

-6

< -4

to >

-5

< -3

to >

-4

< -2

to >

-3

< -1

to >

-2

< 0 t

o > -1

No di

ffere

nce

≥ 0 t

o < 1

≥ 1 t

o < 2

≥ 2 t

o < 3

≥ 3 t

o < 4

≥ 4 t

o < 5

≥ 5 t

o < 6

≥ 6 t

o < 7

≥ 7 t

o < 8

≥ 8 t

o < 9

≥ 9 t

o < 10 ≤ 10

Perc

enta

ge

Difference in length or height (cm) between first and fifth food distributions

29

Any attempt to estimate the impact ofsupplementary food on weight gain requiresthat each child is measured twice, at the startand end of the programme to obtain pairedmeasurements, and that accurate data on age isobtained if anthropometric indices other thanweight-for-height are to be calculated. The datacollected during the present evaluation innorthern Kenya indicated that a large propor-tion of children were not the same at the firstand fifth food distribution and that the age ofmany children was not given nor estimatedconsistently and so was probably inaccurate.There are a number of likely reasons for this,including the possibility that mothers did notbring the enrolled child to collect the fifth andfinal ration, systematic bias or errors in esti-mating age, inconsistent estimates of age onseparate occasions, and errors in makinganthropometric measurements, in recordingdata, and in data entry.

The age distribution shown in Figure 2 isunlike a typical age pyramid and suggests thatmany mothers, who made up 92% of the care-givers at enrolment, were either not giving ornot estimating correctly the age of their child,perhaps to ensure that they obtained a ration offood. The fact that about 40% of all childrenwere either shorter by 1 cm or more or taller by3 cm or more suggests that a substantialproportion of caregivers did not bring the samechild to collect the fifth ration, although somedifferences could be measurement errors madeby busy staff.

It is to be expected that parents will makeevery effort to obtain a ration of food for theirchildren during a food shortage and the manyyears of food insecurity in this part of Kenyahave led to a degree of dependency on human-

itarian assistance. An eligibility threshold of<110 cm in height13 was applied to try to elimi-nate a reason for parents to be untruthful aboutgiving the correct age of their child. This didnot seem to work, perhaps because communitymobilisers did not understand or make it clearto parents, because mothers did not under-stand, or because mothers were mistrustful of adifferent criterion of eligibility for health serv-ices from the usual, which is age.

Factors that compromised data qualitySeveral things compromised data quality:• The design and implementation was

complicated by the request to evaluate the impact in a different a livelihood zone, among a different ethnic group, and by additional NGOs. This increased the samplesize, increased the cost, and increased the number of agencies involved from two to six, with consequences for staff training anddata quality. As this was the first attempt inKenya to evaluate the impact of a blanketfeeding programme using such methods, it might have been best to focus efforts in twocontiguous districts among a single ethnic group and in one livelihood zone, to keep itas simple as possible.

• The pressure to begin distributing rations reduced the time available for training fieldstaff of six NGOs to two days by eight different trainers, adding other factors that may have compromised data quality.

• The personnel doing the evaluation were also responsible for registering and distributing rations to about 500–2,500 beneficiaries at each site during the first fooddistribution, so the staff were overburdened.

• Supervising the collection of data at food distribution sites spread over an area of 150,000 km2 posed an insuperable problem to the lead NGO. Wajir town is 1,100 km byroad to the main town in Turkana. So after being quickly trained, the NGO staff were unsupervised by the lead evaluation agency.

Lessons learnedThe experiences described here offer usefullessons that could be applied to improve thequality of data in future evaluations of blanketfeeding programmes in Kenya and elsewhere.

First, the evaluation should be put in placeas the intervention is being planned so that theevaluation is a part of the programme, not anexternal component. Both require preparation,swiftness and adequate funding.

Second, community mobilisers need toexplain clearly and effectively to potentialbeneficiaries the criterion for eligibility: height<110 cm, not age <59 months - the usual thresh-

13 See Footnote 414 Myatt M, Feleke T, Sadler K, Collins S (2005) A field trial of

a survey method for estimating the coverage of selective

feeding programmes. Bull World Health Organ 83: 20–26.15 Sadler K, Myatt M, Feleke T, Collins S (2007) A comparison

of the programme coverage of two therapeutic feeding

interventions implemented

Beneficiaries ofBSFP in Kenya

SCU

K,

Kenya,

2011

Research

30

Kate Greenaway is SeniorTechnical Advisor, HIV Unit,Catholic Relief Services,Baltimore, MD

Elizabeth Jere is Senior TechnicalAdvisor, STEPS OVC, CatholicRelief Services, Lusaka, Zambia

Milika Zimba is ProgrammeManager, Children's AIDS Fund,Lusaka, Zambia

Cassim Masi is ExecutiveDirector, National Food andNutrition Commission, Lusaka,Zambia

Beatrice Mazinza Kawana isDeputy Executive Director,National Food and NutritionCommission, Lusaka, Zambia

This research was funded by a USAID Cooperativeagreement (award number 690-A-00-06-00093-00).The authors would like to acknowledge the supportof the Ministry of Health (Zambia), the NationalFood and Nutrition Commission (Zambia) and theparticipating HIV care sites. We are also grateful tothe interviewees, focus group participants andresearch assistants for their participation in thisstudy. Finally, we acknowledge the Food andNutrition Technical Assistance (FANTA-2) team fortheir guidance and advice throughout the designand implementation process.

Examining theintegration ofFood byPrescriptioninto HIV careand treatmentin Zambia

Field Article

By Kate A. Greenaway, Elizabeth C. Jere,Milika E. Zimba, Cassim Masi and Beatrice Mazinza Kawana

There is increasing evidence that anti-retroviral therapy (ART) outcomesand nutrition interventions areclosely linked. Studies from sub-

Saharan Africa have established that lowBody Mass Index (BMI) at ART initiation is asignificant predictor of early mortality andthat malnutrition plays a substantial role indisease progression1,2,3. In late-stage HIVinfection, unintended weight loss iscommon: up to 25 percent of clients experi-ence dramatic, life-threatening weight loss.

FBP is a treatment approach that targetsmoderately and severely malnourished indi-viduals with ‘medicalised’ doses of specificnutrition supplements. While empiricalevidence about causal relationships betweennutrition support, weight gain andimproved treatment outcomes among ARTclients is lacking, there is evidence thatweight gain at three months on ART isstrongly associated with survival4 and thatnutrition supplements have a positive effecton ART adherence5.

Zambia has a generalised HIV epidemic,where more than 900,000 Zambians areliving with HIV (PLHIV), with 280,000 onART6. Research conducted in 2007 revealedstartlingly high rates of malnutrition amongadult PLHIV starting ART: 33.5% had a BMI< 18.5 kg/m2, and 9% had a BMI less <16.0kg/m2 (7).

Catholic Relief Services FBP Pilot inZambiaTo address malnutrition in people livingwith HIV (PLHIV), Catholic Relief Services(CRS) Zambia worked in partnership withZambia’s National Food and NutritionCommission (NFNC) to pilot a Food byPrescription (FBP) programme as an adjunctto a USAID-funded palliative care grant. Theevaluation was undertaken to understandthe practical implications of FBP implemen-tation and to gather information on clientoutcomes.

The Zambia FBP model prescribes anddispenses specialised nutrition commoditiesin response to clinical malnutrition (Figure1). Small daily ‘doses’, packaged in individ-ual sachets, are intended to reduceintra-household sharing, institutionalise theconcept that these foods are ‘medicines’, easecalculation of recommended daily allowance(RDA) and aid monitoring at the householdlevel.

The model requires that nutrition assess-ment, education, counselling and support(including food dispensing) be synchronisedwith HIV care and treatment services. The

pilot tested the model in three types ofsettings: clinical facilities (eight), hospices(ten) and home based care (two). Proceduresvaried by setting to accommodate pre-exist-ing systems and emphasising integrationrather than establishment of parallelsystems. Staff training, using the (draft)national FBP guidelines, was provided in the22 sites.

Clients were admitted to the FBP compo-nent of HIV care according toanthropometric criteria. BMI was most oftenused. Mid-upper arm circumference(MUAC) was used to assess pregnant andlactating women, as well as clients whoseheights could not be taken. Children wereassessed using weight-for-height z scores(WHZ). As dictated by the national protocol,adults with severe acute malnutrition (SAM)were prescribed both Ready to UseTherapeutic Food (RUTF) and High EnergyProtein Supplement (HEPS) in sufficientquantity to meet 100% RDA. Those withmoderate acute malnutrition (MAM)received HEPS to meet 50% RDA. Clientswere re-evaluated regularly and dischargedwhen anthropometric assessment indicatednutrition rehabilitation. At the time of theevaluation, the pilot had reached 5,360clients.

Methodology of pilot evaluationTo evaluate the pilot programme, purposivesampling was used to select 11 evaluationsites: six ART clinics, four hospices and oneHBC site representing locality (urban orrural), supporting organisation and size ofprogramme.

From the 1,671 clients enrolled at the 11selected sites, the evaluation team plannedfor a purposive sample of ten dischargedclients per site (total of 110). Difficulties incommunication and logistics resulted inidentification of 91 clients (84 adults and 7

1 Johannessen A, Naman E, Ngowi BJ, Sandvik L, Matee

MI, Aglen HE, Gundersen SG, Bruun JN: Predictors of

mortality in HIV infected patients starting antiretroviral

therapy in a rural hospital in Tanzania. BMC Infect Dis

2008, 8(52).2 Stringer JS, Zulu I, Levy J, Stringer EM, Mwango A, Chi

BH, Mtonga V, Reid S, Cantrell RA, Bulterys M et al:

Rapid scale-up of antiretroviral therapy at primary care

sites in Zambia: feasibility and early outcomes. Jama

2006, 296(7):782-7933 Zachariah R, Fitzgerald M, Massaquoi M, Pasulani O,

Arnould L, Makambe S, Harries AD: Risk factors for

high early mortality in patients on antiretroviral treat-

ment in a rural district of Malawi. AIDS 2006, 20(18):

2355-2360.4 Madec Y, Szumilin E, Genevier C, Ferradini L, Balkan S,

Pujades M and Fontanet A: Weight gain at 3 months of

antiretroviral therapy is strongly associated with

survival: evidence from two developing countries. AIDS

2009, 23(7): 853–861.5 Cantrell RA, Sinkala M, Megazinni K, Lawson-Marriott

S, Washington S, Chi BH, Tambatamba-Chapula B,

Levy J, Stringer EM, Mulenga L, Stringer JS: A pilot

study of food supplementation to improve adherence to

antiretroviral therapy among food-insecure adults in

Lusaka, Zambia. J Acquir Immune Defic Syndr 2008,

49(2):190-5.6 Ministry of Health and National AIDS Commission:

Zambia Country Report: Monitoring the Declaration of

Commitment on HIV and AIDS and the [sic] Universal

Access Biennial Report, Lusaka; 2010.7 Koethe JR, Lukusa A, Giganti MJ, Chi BH, Nyirenda CK,

Limbada MI, Banda Y, Stringer JS: Association between

weight gain and clinical outcomes among malnourished

adults initiating antiretroviral therapy in Lusaka, Zambia.

J Acquir Immune Defic Syndr 2010, 53(4):507–513.

Figure 1: Food by Prescription Client Flow Model

Anthropometric assessment:

Height, Weight, MUAC

Prescription Dispensing:ARVs, Ol drugs, RUTF,

HEPS, Clorin

Clinical care:Clinical and metabolicassessment, Food

prescripion

Community care:HBC, Support Groups,Livelihood & Safety Net

support

Counseling: Adherence,Nutrition, BMI or WHZ

calculation, FBP enrollment

Adapted from Zambia National FBP Guidelines (Draft)

Clients collect their prescribedHEPS at Wusikile Mine Hospital

K G

reeenw

ay,

Zam

bia

, 2009

Field Article

children). The guardian most conversant withthe child’s health/illness situation was inter-viewed.

Quantitative data was collected andanalysed from client records, monthly reportsand field visit reports. Variables of interest wereage, sex, weight, anthropometry on admissionand discharge, length of stay on treatment andreason for discharge. Data were edited andentered into Excel. Data were exported toStatistical Application Systems (SAS) forfurther cleaning and analysis. Analysesinvolved descriptive and inferential statisticalanalyses including frequencies and distribu-tions of all variables.

Structured individual interviews were thenheld with 91 clients, ten administrators and 38health care workers. Focus group discussions(FGDs) of five to ten participants, drew furtherinformation from clients (four groups) andservice providers (five groups) using open-ended questions and scenario methods, led by amoderator and recorded by a note-taker.Qualitative data were analysed manually by theevaluation team, who read through the inter-views to identify emerging themes.

ResultsIntegration – Clinical facilities: The majority ofART clinics achieved integration of FBP andART services as intended in Figure 1. Facilitieswith the weakest community outreachprogrammes had the highest number ofdefaulters.

Some facilities, having achieved competencewith FBP implementation, extended services tosatellite sites. These were significantly moredifficult to manage but reduced congestion at thehospitals and greatly increased FBP enrolment.

Integration – HBC: HBC programmes provideda decentralised, ‘one-stop’ service model.Clients were assessed, counselled, prescribedand dispensed rations by a trained caregiver atthe parish office. Home-based follow-up wasprovided by assigned HBC providers whoserole was to support both FBP and ART adher-ence. HBC service providers and clientsreported that integration was seamless.

Integration – Hospice: Service models variedconsiderably among hospices, with hybrids ofcentralised and decentralised models tried.Success with integration varied, possibly due toreliance on lay counsellors and volunteer care-givers (in contrast to technical staff employedby ART clinics) as well as less rigorous record-keeping (in many hospices), and lessmanagement oversight in some cases.

Service provision: Anthropometric assessmentwas often done selectively (on clients whoappeared malnourished) rather than as aroutine aspect of the standard of care. Whileweighing and recording weight is a standard-ised practice, BMI is rarely calculated andMUAC is seldom used. FGD respondents at allsites noted the need for additional training andsupervision to ensure adherence to admis-sion/discharge criteria and to improve skillsand consistency in nutrition assessment.

Respondents consistently reported thatactive supervision positively influenced staffcommitment to providing nutrition assessmentand education, and improved accuracy ofrecord-keeping. Many staff requested incen-

tives for providing FBP services. The pilot,however, was neither designed nor budgeted toaccommodate incentive requests.

Only 11% of clients reported that they werelinked to livelihood activities, illuminating theenormity of the gap in this aspect of FBPprogramming.

Food storage and dispensing: Lack of FBPcommodity storage space was cited as a signifi-cant challenge. CRS generally disbursedcommodities every two months to accommo-date storage limitations. Some alternate storagelocations, such as kitchens, did not meet storagestandards for temperature and humidity. Therewas no consensus among service providersregarding the ideal location for food dispensingbut agreement that each setting should evaluateits options with emphasis on creating the mostseamless, efficient pathway for clients.

Overall, supply chain management was asignificant challenge, with three primary diffi-culties noted:1. Month-to-month new enrolment numbers

varied considerably2. Length of client enrolment varied3. Short shelf-life of selected commodities

reduced prepositioning options.

Monitoring and Evaluation: The project soughtto align with ‘the 3 Ones’8 by contributing to asingle national reporting system, but wasobliged to create a parallel approach becausethe existing system, SmartCare, did not allowfor the capture of comprehensive nutrition data(e.g. BMI and WHZ). Lack of nutrition training,combined with the lack of tools and systems fordata collection, have resulted in a nationwidegap in the detection, tracking and treatment ofmalnutrition among PLHIV, especially adults.

Client weight gain and BMI: All sites showedan increase in client BMI between admissionand discharge. Among adult clients, the aver-age BMI on admission was 17.6 kg/m2 and theaverage BMI on discharge was 20.5 kg/m2. Theoverall average increase in BMI pre-FBP topost-FBP was 2.9 kg/m2. Most clients requiredthree to six months of nutrition rehabilitation toqualify for discharge.

Of the 22% of clients already dischargedfrom the programme at the time of the evalua-tion, 997 (84%) met discharge criteria, 127 (11%)died from various causes, 45 (4%) wereunknown or lost to follow-up and 18 (1%) wereremoved from treatment because of medicalcomplications.

Client health status: Clients were asked to ratetheir pre- and post-intervention health statususing the Eastern Cooperative Oncology Group(ECOG) performance scale9. The percentage ofclients who were ‘fully active’ went from 5%pre-FBP to 51%, post-FBP. Only 1% of clientsremained ‘completely disabled’ post-FBP,compared to 17% pre-FBP.

LimitationsSite records and quantitative datasets hadnumerous missing anthropometric data whichlimited the scope of analysis. Geographicdistance, communications challenges, delayedproject start-up and time constraints resulted ina disproportionate number of enrolled clients(thus fewer-than-planned rehabilitated anddischarged) represented in the evaluationsample.

8 The ‘3 Ones’ is a set of three key elements that underpin a

coordinated national response: One agreed HIV/AIDS

Action Framework that provides the basis for coordinating

the work of all partners. One National AIDS Coordinating

Authority, with a broad-based multi-sectoral mandate; and

One agreed country-level Monitoring and Evaluation

System. (UNAIDS, 2004)9 Oken, MM, Creech, RH, Tormey, DC, Horton, J, Davis, TE,

McFadden, ET, Carbone, PP: Toxicity And Response Criteria

Of The Eastern Cooperative Oncology Group. Am J Clin

Oncol 1982, 5(6):649-655.

The short six-month project implementationperiod was sufficient to measure integrationactivity but necessitated pooling of clientsacross several sites in order to obtain a suffi-cient sample, which may have maskedsite-specific patterns.

It should be noted that weight gain, BMI andECOG performance cannot be attributed exclu-sively to a FBP intervention. It is understoodthat nutritional status and activity level arelikely to improve with ART only, or with someother combination of ART and nutrition.

Conclusions and recommendationsIntegration of FBP into existing HIV care andtreatment was successfully adapted to facility,home-based care (HBC) and hospice servicedelivery settings. Integration did not interruptexisting service delivery and can be accom-plished using available human and materialresources. The ‘medicalisation’ concept wasappreciated and understood by both clients andservice providers and the selected rations weresuccessful in treating malnutrition. Weight gainand body mass index (BMI) improved whilepercentage of discharges cured (i.e. nutrition-ally rehabilitated) exceeded standards. Inaddition, activity levels and perception of well-ness improved dramatically.

The keys to success were on-going supportfor application of nutrition concepts and carefulrecord-keeping, and the identification of sitecoordinators who brought both technical nutri-tion knowledge and a high level of commitmentto the pilot project. However, on-going trainingis required in nutrition, record-keeping andreporting. Future implementers would benefitfrom formal incorporation of new (FBP) tasksthrough either scopes of work for key staff, orthe full integration of FBP responsibilities intostandard job descriptions. These additionaltasks may have implications for remuneration.Furthermore, integration of FBP commoditiesinto the medical stores procurement and distri-bution system would reduce duplication ofeffort and promote national ownership.

The national ART M&E systems must beexpanded to capture nutrition data. To fostertimely discharge, linkages to wrap-around foodsecurity and livelihood programmes should bedesigned from the early stages of projectconceptualisation.

Children and pregnant/lactating womenwere under-represented, suggesting thatMaternal & Child Health (MCH) andPrevention of Mother to Child Transmission(PMTCT) programmes should be more inten-tionally included in scale-up plans. With regardto the use of MUAC, it was suggested that it beused for screening only, applying an increasedcut-off to trigger referral of potential clients forassessment by a clinician.

For more information, contact Kate Greenaway,email: [email protected]

31

32

Name: Centres for Disease Control and Prevention (CDC), International

Emergency and Refugee Health Branch (IEHRB)

Address: Centres for Disease Control and Prevention, 1600 Clifton Rd. Atlanta, GA

30333, USA

Phone: +1 770-499- 3910

Email: [email protected], [email protected]

Website: http://www.cdc.gov/

Director: Dr Thomas Frieden

No. of HQ staff: 35 Atlanta based staff in the International Emergency and Refugee

Health Branch

No of staff worldwide: 15,000 total employees, across more than 50 countries

Agency Profile

The ENN recently conducted an agencyprofile interview with Leisel Talley andCarlos Navarro-Colorado from CDC’sInternational Emergency and Refugee

Health Branch (IEHRB). CDC is a US federalagency dedicated to the prevention and controlof disease, injury and disability. Leisel hasworked for CDC for 11 years while Carlos is newto the branch, with a little over a year’s time atCDC. The IEHRB branch has approximately 35staff who cover a range of disciplines includinginfectious disease, malaria, child protection, warrelated injuries, WASH, immunisation , mentalhealth, reproductive health, survey and surveil-lance methodologies and statistics. Oleg Bilukha(who couldn’t make the interview and hasworked for CDC for 11 years), Leisel and Carlosare the three individuals who spend most oftheir time working on nutrition in this multi-disciplinary team.

Leisel explained that IEHRB activities varyyear to year depending on the number and typeof emergencies. Their work may encompassdirect support in emergencies, operationalresearch and development of programmatictools, providing technical advice, participatingin technical forums and teaching (in universities,US government, US agencies or UN agencies). Inorder to provide direct support, CDC has to beinvited into a country to work, i.e. by the USgovernment, national Ministry of Health (MoH),international non-governmental organizations(INGOs) or UN agencies. Staff from the branchmay be seconded during emergencies to agen-cies with whom CDC has agreements, especiallyat the beginning of an emergency before agen-cies can identify and recruit longer-term staff.Often, a senior epidemiologist is deployedtogether with an Epidemic Intelligence ServiceOfficer (EISO) for mentoring purposes whilestrengthening CDC’s response capacity. SinceCDC is a US government agency, staff areprovided as in-kind technical assistance to agen-cies and programmes.

Operational research is always conducted inpartnership with other agencies. A current exam-ple is the ongoing evaluation of a blanketsupplementary feeding programme in Turkanaand Wajir, northern Kenya in partnership withWFP and several field partners. This involvesfollowing a cohort of children to determine theimpact of the intervention and, through a case-control study, the determinants of malnutritionwhile enrolled in the programme. The IEHRBmay be approached by agencies to conductresearch or individuals in the branch may proac-tively approach agencies for a specific researchproject. For example, UNICEF recentlyapproached IERHB to work with them on assess-ing the impact of Plumpy’doz in a Darfurfeeding programme. IERHB is also workingclosely with the ENN on a study of defaulting

from emergency supplementary feedingprogrammes. The branch is currently workingon a funding announcement that will allow asubstantial volume of operational research onacute malnutrition. IERHB is hoping to use thisopportunity to strengthen existing partnershipsand forge new ones.

Leisel recalled that when she started workingin the branch there were only seven staff, whichmade it difficult to respond to emergencies. Nowthat the branch has grown to more than 30 staffmembers, it has been able to expand into otherareas like operational research, teaching anddevelopment of programmatic tools. It alsomeans that they can be much more proactiveabout work and participate in relevant expertgroups and discussions in most key areas ofemergency public health.

Carlos explained how branch members havea solid understanding of public health principlesand emergency relief, no matter what theirprofessional background. Any of the three indi-viduals mentioned may therefore findthemselves working in areas outside of nutrition.For example, Leisel recently worked onsampling aspects of a survey of violence againstchildren in Tanzania. Oleg routinely works onwar-related injuries and Carlos has beeninvolved in disease outbreak investigations inSouth Sudan and Kenya. The multi-disciplinarynature of the team means that everyone learnsfrom everyone else. Carlos also felt that workingin this type of multi-discipline team helps withprofessional development and leads to strongtechnical support, as well as locating nutritionunder a broader public health umbrella.

At this point in the interview I rememberedsomething that had always perplexed me aboutCDC. Why was it located in Atlanta in the stateof Georgia? Leisel explained that when CDC wasestablished early in the last century, malaria wasstill a substantial problem in the southern UnitedStates. SInce Atlanta was the largest city with thebest transportation in the region, it was viewedas the most appropriate location.

We then moved onto a discussion about thebranch’s role in the current Horn of Africa crisis,both in the field and from Atlanta. Leiselrecounted how the Food Security and NutritionAnalysis Unit - Somalia (FSNAU) had contactedCDC in July 2011 asking for support to improvethe quality of their nutrition data from Somaliaand to validate the findings. Oleg has beeninvolved in SMART training in Ethiopia for anumber of humanitarian agency staff. Carloswas sent to the region fairly early on in the crisisand worked in the Dadaab camps in Kenya,mostly supporting the UNHCR public healthoffice with nutrition and public health advice.He also helped remotely from Atlanta with thedesign of a nutrition survey in Dolo camp in

Ethiopia, as well as supporting analysis ofcommunity based management of acute malnu-trition data in order to improve reporting andresponse. The whole branch has generally beenvery involved in the region supporting a numberof activities, like measles and cholera surveil-lance, in coordination with in country CDCoffices and programmes. However, the difficultand fluctuating security situation in the regionhas meant that a number of surveys that CDCwas planning have had to be cancelled.

We also spoke about challenges that thebranch faces. As with most agencies in thehumanitarian sector, the current global financialcrisis is cause for concern. CDC’s budget isdetermined on an annual basis and the currenteconomic situation may result in reduced budg-ets across the US government. This situation hasnot seriously affected the branch, though. Thereare also technical challenges such as trying tokeep the right balance between methodologicalrigour of surveys, surveillance and impactassessment and the reality of what is needed andpossible on the ground during emergencies.Another challenge, not unique to CDC, relates tohow to evaluate programme outcomes when inmost situations, randomised controlled trials aresimply not viable. There is a spectrum of opinionwithin the branch about other means of provingimpact and outcomes and a strong engagementin helping evolve epidemiological methods inemergencies. The high level of verification andquality control within CDC required to formallyrelease results can be lengthy at times butensures a very high quality of work.

I asked Carlos and Leisel where they hoped orexpected the branch and in particular nutritionactivities in the branch, to be in five years time.While there isn’t a specific five-year plan, there isthe hope that the new funding announcementwill allow the branch to partner with a numberof agencies to conduct operational research toimprove the evidence base for nutrition-relatedinterventions, thereby improving interventioneffectiveness. This should contribute to thecontinuous development of the evidence base,building on CDC’s unique combination ofmethodological expertise and public healthapplied work. In addition to this, they expect tobuild a critical mass of emergency trainedepidemiologist within CDC that will facilitateresponding to field requests.

What seemed obvious from talking withLeisel and Carlos was that the nutrition teamhave consistently high demands and expecta-tions placed upon them and have to react rapidlyto requests from partner agencies with no way ofknowing when these requests will be made. Ittherefore seemed like a bit of intended under-statement when Leisel concluded our interviewby saying that “we were certainly very pleasedwith the recent expansion of the branch”.

33

During the past 10 years, the management of acute malnutrition has undergone a major paradigm shiftthat has changed the previous inpatient ‘clinical’ model of care into a community-based “publichealth” model of care. Since 2007, this new model, called Community-based Management of AcuteMalnutrition (CMAM) has expanded rapidly and is now implemented in over 55 countries worldwide

In the old clinical model, the main determinant of impact was the quality of the inpatient medical and nutri-tional care provided in the centres and hospitals. By contrast, in the CMAM model, the key determinants of impactare the degree to which interventions access people early in the course of their disease and the ability to reach asmany of those affected as possible. This is a profound shift that requires an equivalent change in the protocols andindicators used to implement and monitor programmes. Previously in the clinical model, impact was achievedusing in-depth medical and nutritional protocols and results were monitored using clinical outcomes indicators.Now, the simplicity and robustness of the CMAM treatment protocols are such that so long as the simple basicssuch as ready-to-use therapeutic food (RUTF) are available and those afflicted by acute malnutrition present earlyand in sufficient numbers, then impact is assured. In the new CMAM public health model, the focus on clinicalguidelines has been replaced by protocols to ensure that those who are affected are admitted into programmesearly and the clinical outcome indicators have been supplemented by the direct assessment and monitoring ofcoverage.

The Semi Quantitative Evaluation of Access and Coverage (SQUEAC) coverage assessment method is a new setof tools that draws together access and coverage, the two essential determinants of quality CMAM programming.SQUEAC combines an array of qualitative information about access and the perceptions of CMAM programmeswith small sample quantitative surveys. These surveys test hypotheses generated during the qualitative work andestablish levels of programme coverage in key geographical areas. This combination both identifies key issuesaffecting presentation and programme uptake, whilst also establishing the actual levels of coverage attained.Vitally, all this can be done in real time, allowing the tool to be of immediate practical use to tweak programmedesign and implementation in response to the information obtained.

The key to the success of SQUEAC is diversity, triangulation and iteration, that gradually builds up a picture ofthe “truth” about programme coverage, whilst at the same time, indicates what practical measures can be under-taken to improve access and coverage. The beauty of the technique is that it combines information that is oftenroutinely collected but rarely used, with other data specifically collected by fast, low resource methods. Directlyharnessing existing routine monitoring data to improve impact and program effectiveness greatly increases thecost efficiency of the additional time spent collecting new data, thereby decreasing the time and resource overheadsrequired to implement SQUEAC.

SLEAC (Simplified LQAS Evaluation of Access and Coverage) was designed to fill a gap in the early SQUEACmethodology regarding the method's ability to assess overall programme coverage. This gap has now beenaddressed in SQUEAC and SLEAC has been adapted as a tool for assessing coverage over wide areas. As CMAM shifts from a donor-funded emergency intervention into a routine part of primary health careprogramming, the resources available to implement these programmes will inevitably decrease. In this environ-ment, low resource methods to increase timely access, monitor coverage, and allow programme design to beproactively refined in view of these, are essential if CMAM is to maintain its effectiveness. In my opinion,SQUEAC is a major step forward towards achieving these goals.

Adapted from foreword by Steve Collins. Technical Reference for SQUEAC and SLEAC Methods, 2012. FANTA.

Focus on coverage assessment

Wide area survey during SQUEAC's stage 3 in Gassi (Mali)

Jose

Luis

Alv

are

z M

ora

n,

Mali

Remote monitoring of CMAMprogrammescoverage: SQUEAC lessonsin Mali andMauritania

34

Focus on coverage assessment

Jose Luis Alvarez Moran is aMedical Doctor with a PhD inInternational and Public Health. Heworks as an assistant in Rey JuanCarlos University and is currentlyconducting nutrition surveys for

Action Against Hunger.

Brian MacDomhnaill is an inde-pendent expert in monitoring andevaluation of health programmes.He has worked in Ghana, Brazil,Angola, Mauritania and Djibouti.

Saul Guerrero is the Evaluations,Learning and Accountability (ELA)Advisor at Action Against Hunger(ACF-UK). Prior to joining ACF, heworked for Valid International Ltd.in the research, development and

roll-out of CTC/CMAM. He has worked in Afghanistan,Algeria, Chad, DRC, Ethiopia, Indonesia, Kenya, Liberia,Malawi, Mali, Mozambique, Nepal, Niger, Nigeria,Sierra Leone, North & South Sudan and Zambia.

The authors would like to thank Chantal AutotteBouchard, David Kerespars, Dr. Theophane Traore,INSTAT, and the ACF teams in Mali, Mauritania andSpain (Elisa Dominguez in particular) for their support.To Ernest Guevara (Valid International) and Mark Myatt(Brixton Health) for their valuable comments and tothe European Commission Office for Humanitarian Aid& Civil Protection (ECHO) for their financial support.

By Jose Luis Alvarez Moran, BrianMac Domhnaill and Saul Guerrero

Action Against Hunger (ACF)currently supports communitybased management of acutemalnutrition programmes

(CMAM) programmes in over 20 countriesaround the world, with a long-standing pres-ence in the Sahel region of West Africa,including Mauritania, Niger, Mali and Chad.Most of these interventions are integratedCMAM programmes, operated by Ministriesof Health and local partners with technicaland logistical support from ACF teams on theground. Monitoring the impact of these inter-ventions, and their coverage in particular, isof paramount importance to the organisation.Increasing coverage was instrumental in theshift from inpatient care in the form of thera-peutic feeding centres (TFCs) to outpatientmodels (CMAM) and remains one of themost widely accepted indicators ofprogramme performance and impact. Whilstother indicators (e.g. cure rates, length ofstay, average weight gain) provide an insightinto the efficacy of treatment, only whencombined with coverage do they provides anaccurate and reliable indication of the needsmet by a programme. Since December 2010,ACF has been increasingly relying on theSemi-Quantitative Evaluation of Access &Coverage (SQUEAC) to measure programmecoverage and identify the factors affecting theperformance of CMAM programmes1.

According to a recent UNICEF estimation,there are 55 countries currently implement-ing CMAM in one form or another2. The scaleof CMAM programming, limited non-governmental organisation (NGO) resources,and deteriorating security conditions inmany regions (including in the north-westand Horn of Africa) is increasingly forcingsupport organisations such as ACF to operateremotely with limited access to programmeareas. The extent of the constraints varies,from limited access to areas within a district(e.g. ACF supported programme inGuidimaka, Mauritania), to limited access toparts of a country (e.g. ACF supportedprogramme in Gao, Mali) to limited access toan entire country (e.g. ACF supportedprogrammes in Somalia). All of these envi-ronments present challenges, in particular forthe implementation of monitoring and evalu-ation activities with a strong field componentsuch as SQUEAC.

Monitoring coverage remotelyExperiences in using SQUEAC remotely havebeen limited, with the most notable experi-ence provided by Valid International andOxfam-Novib in Somalia3 (see Box 1).Recently, ACF carried out SQUEAC investi-

gations in Mauritania (February 2011) andMali (July-August 2011). In both cases, lack ofsecurity prevented the SQUEAC lead investi-gators from travelling to the programmeareas. In the case of Gao (Mali), the leadinvestigator was unable to visit the district inwhich the programme operated but was ableto visit a neighbouring district. In the case ofGuidimaka (Mauritania), the investigatorwas able to visit the district but could nottravel to most areas outside of the districtcapital.

The analysis presented here will drawlargely from these two experiences. A briefsynopsis of the SQUEAC methodology andits key features in more conventional settingsis included in Box 2. The article focuses itsattention on two general stages of usingSQUEAC to monitor programme coverageremotely: planning and implementation. Itwill conclude with some lessons learned andprovide practical suggestions for other prac-titioners wishing to undertake similarexercises in the future.

ACF’s remote experiences in Mali andMauritaniaCMAM programmes supported by ACF inGao (Mali) and Guidimaka (Mauritania) arelargely inaccessible to expatriate staff due tosecurity threats posed by AQMI (Al-Qaida auMaghreb Islamique) in the region. Securitythreats do not prevent local teams fromimplementing programme activities, butmonitoring supervision is more difficult sincethe local teams often need to travel to moreaccessible areas to meet with technicalsupport and management staff. The decisionto evaluate the coverage of both theseprogrammes forced the organisation toexplore different means of employingSQUEAC.

Both investigations faced similar accessi-bility problems and relied on the work ofexternal SQUEAC lead investigators broughtinto the programme especially to carry outthe investigations. The lead investigators hadconstant remote support from ACF’sEvaluations, Learning & Accountability

1 For more on the SQUEAC method and its use see Myatt,

M. SQUEAC: Low resource method to evaluate access

and coverage of programmes. (Field Exchange,

Emergency Nutrition Network, Issue 33, June 2008,

p.3) & Schofield, L. et.al. (2010) SQUEAC in routine

monitoring of CMAM programme coverage in Ethiopia

(Field Exchange, Issue 38, April 2010, Emergency

Nutrition Network, p.35).2 UNICEF & Valid International (2011) Global Mapping

Review of Community-based Management of Acute

Malnutrition with a focus on Severe Acute Malnutrition

(Nutrition Section, Nutrition in Emergency Unit, UNICEF

HQ-NY and Valid International, March 2011)3 Valid International. Personal Communication.

Box 1: Monitoring coverage remotely in Somalia: The Valid International experience

Valid International has supported the set-up, moni-toring and evaluation of a community therapeuticcare (CTC) programme in Mogadishu, Somalia forthe past 2 years. The monitoring and evaluationsupport was built around the assessment of cover-age using SQUEAC as a framework. Hence,components of the SQUEAC toolbox were put inplace right from programme set-up. This allowedfor a more organic SQUEAC process that followedthe programme cycle of implementation.

This was deemed suitable in the context ofprogramming in Mogadishu where access to theprogramme sites by external persons is an issue.

Institutionalising a routine system of coverage eval-uation was the most suitable way and SQUEACproved to be an effective framework. This allowedfor a mechanism by which the use of differentcomponents of the SQUEAC toolbox at various peri-ods or steps, rather than in ‘just one go’ typical ofother investigations. This also allowed for remoteexternal support to be provided appropriately andas needed.

This is the approach that Valid International istaking in contexts such as Mogadishu but is anapproach that is ideal even in developmental andmore stable conditions.

Data collection team measuringMUAC in the village of

Hamakouladji, Sony Alibermunicipality (Gao, Mali)

K G

reeenw

ay,

Zam

bia

, 2009

35

which included joint analysis of existingprogramme data and the development ofpreliminary hypotheses. The availability ofprevious SQUEAC experience was helpful inplanning SQUEAC remotely, particularly fordeveloping a hypothesis about coverage withlimited access to the programme.

Key lessons learnedACF’s experiences in implementing SQUEACremotely in Mali and Mauritania provided fivekey lessons:

Advanced planningWhen undertaking SQUEAC remotely, forwardplanning is essential. This is partly due to timeconstraints. When working remotely, activitiestake longer, but since the exercise must becompleted in a similar timeframe (to remainpractical and cost-effective), time must bemanaged more strategically than in ‘conven-tional’ environments. ACF’s experience showedthat both the coordination and data collectionteams must be well coordinated to ensure anoptimal use of each team’s time. For example,with advanced planning, the coordination teamis able to carry out some parts of the analysiswhilst the field team simultaneously collectsfield data. In that respect, the SQUEAC method-ology is appropriate for such environments, as itis not always a linear process (between inputsand outputs) and is flexible enough to allow formultiple activities to be implemented, some-times in parallel. Advanced planning is alsoessential to ensure an adequate recruitmentprocess for reliable enumerators that can takesignificantly longer when undertaken remotely.

Data collationThe first stage of a SQUEAC investigationinvolves collating/collecting programme datato build a picture of what programme coverageis and where the areas of high and low coverageare likely to be. This process of data collationnormally takes place during the SQUEACinvestigation period, partly on the assumptionthat these data are readily available fromprogramme reporting, databases and otherinformation management systems. Collating allthis information in remote programmes can bea long process, especially for integrated (MoH-led) programmes where information is oftenheld at the Service Delivery Units (e.g. healthcentres). The experience from Mali shows thatcollating such information prior to the start ofSQUEAC can ensure that Stage One focusesmostly on analysing the data (and requestingadditional data) rather than on collating it. Inthis respect, having a multi-layered team (withcoordination and a data collection team in thefield) enables some elements of theanalysis/collection of (last minute) data to beundertaken in tandem.

The Mali and Mauritania experiences showthat some data can and should be collected inadvance (see Box 3). Some of this information isconsistently collected through routine monitor-ing data (including admissions, defaulters and

deaths) but other atypical and non-routine datarequire specific mechanisms to collect them.Integrating these last ones in the basicprogramme monitoring data would facilitatethe implementation (and mainstreaming) ofcoverage investigations.

Multi-layered teamSQUEAC investigators will still need to deter-mine what and how additional qualitative andquantitative data are to be collected, as well asmeans of analysis. In conventional SQUEACinvestigations, these processes generally occurin the same place and are carried out by thesame teams (enabling a more real-time,active/reactive process of data collection). Inremote investigations, a separation of the twoprocesses may be necessary, employing a multi-layered team approach. The model used in Malireplicated the two-tier CMAM implementationapproach used by ACF to support the CMAMprogramme. In other words, most technical/strategic/analytical processes were carried outremotely by one team (the coordination team)while a second team (the data collection team)had access to the programme area and was in

(ELA) Advisor based in London. In both coun-tries, two teams were formed: a coordinationteam (including the lead investigator, the ACFMedico-Nutritional Coordinator, and the logis-tics department at capital level) and a datacollection team (composed of the investigator’sassistant and local enumerators recruited forthe purposes of SQUEAC).

The type of training received by the leadinvestigators prior to their respectiveSQUEACs was different. The lead investigatorfor Mauritania received a three-day SQUEACintroductory training prior to departure, andremote technical support throughout the inves-tigation period. The lead investigator for Malireceived a 5-day, on-the-job training in-country,

Box 2: SQUEAC: a summary

In 2007, Valid International in collaboration withFANTA/AED, UNICEF, Concern Worldwide, WorldVision International, ACF-UK, Tufts University andBrixton Health, developed the Semi-QuantitativeEvaluation of Access & Coverage (SQUEAC). TheSQUEAC methodology was designed as a low-resource method capable of evaluating programmecoverage and identifying barriers to access. SQUEACis not a survey method but a toolkit designed toprovide programme practitioners with differentmeans to evaluate the proportion of the targetpopulation covered by a nutrition programme.

Whilst the need to increase nutrition programmecoverage was one of the central pillars behind theshift from centre-based treatment to community-based models, measuring programme coveragedirectly has often proven difficult. Existing tools,such as the Centric Systematic Area Sampling(CSAS) technique, were robust and reliable, yet bytheir very nature, resource-intensive and oftencostly. This effectively led to their use as evaluativetools rather than monitoring mechanisms.

SQUEAC investigations are generally carried out inthree distinct stages :

Stage One identifies areas of high and low coverageand reasons for coverage failure using existingprogramme data (e.g. admissions, exits) and easy-to-collect data. Whilst much of this data analysiscan be collected remotely, access to programmeareas is normally required to allow for the collectionof additional data (qualitative data in particular) usedto triangulate existing information.

Stage Two is designed to test the hypotheses(about areas of low and high coverage and reasonsfor coverage failure) developed in Stage One.Testing can be carried out using small studies,small surveys and/or small-area surveys. All ofthese alternatives normally require access to theprogramme area.

Stage Three uses Bayesian techniques to estimateprogramme coverage. The technique relies onpreviously collected data to develop a ‘prior’4

about programme coverage. A wide-area survey isthen carried out to collect data to develop a “likeli-hood”5 (which, together with the “prior”, helpsprovide a “posterior”6 or final estimate ofprogramme coverage). Wide area surveys requireaccess to all survey areas of the programme. WhilstStage 3 can potentially be left out of the SQUEACprocess, it is an essential component if overallcoverage estimate is required.

There are no pre-set timeframes for a SQUEACinvestigation, but under stable conditions in whichinformation can be accessed and tested relativelyeasily, a full SQUEAC can last between 14-28 days.Whilst SQUEAC was designed to be implemented byprogramme staff directly, SQUEAC investigations arestill commonly implemented under the supervisionof SQUEAC lead investigators.

Box 3: Key CMAM programme data to becollected/collated prior to the start of a SQUEACinvestigation

• Programme admissions (by month, by site, byhome location)

• MUAC on admission• Critical events calendar (i.e. annual calendar

showing key events that influenced programme coverage positively or negatively)

• Seasonal diseases calendar• Stock Break Calendar (i.e. annual calendar

showing periods of disruption in RUTF supply)• Referral effectiveness• Volunteers activity data• Programme exits

Box 4: Key themes for qualitative data collecting inSQUEAC

• Local aetiologies• Community awareness • Participation in the programme• Barriers to access• Perceived coverage• Accessibility and insecurity

Focus on coverage assessment

4 In Bayesian inference, the prior is a probabilistic represen-

tation of available knowledge about a quantity. In SQUEAC,

the prior is a probabilistic representation of knowledge

relating to program coverage. SQUEAC uses a Beta

distributed prior.5 In Bayesian inference, the information provided by new

evidence. The likelihood is use to modify the prior to arrive

at the posterior. In SQUEAC, this is the information

provided by a survey (the likelihood survey).6 In Bayesian inference, the posterior is the result of

modifying prior belief using new evidence

Lead investigator's assistant collectingqualitative data with mothers of thevillage Monzonga, Ansongo (Mali)

36

charge of carrying out qualitative and quantita-tive data collection processes (see Figure 1). Forthis two-tier arrangement to succeed, regularcommunication (prior to and during) the inves-tigation was crucial.

Regular communicationEven when a programme area is not equallyaccessible to all, it is important to bring all theteams working in a SQUEAC to an accessiblelocation for discussion about the activities andprocesses involved (e.g. calculating weight forheight, measuring mid upper arm circumfer-ence (MUAC), presence of oedema, etc.).Face-to face communication should occur atleast once with the lead investigator. Duringthese meetings it is important to involve every-one in the development of a map of theprogramme area. The development of a mapjointly with the team not only ensures that thespatial dimension of the exercise is understood,but it is a critical step in ensuring that the leadinvestigator gets an opportunity to discuss andexplore questions about the programme area.Working with a map will help in the implemen-tation of SQUEAC and also assist thesupervision of the teams.

Once SQUEAC begins, regular communica-tion becomes essential. New technologies, suchas internet, emails and mobile telephones, areable to provide a real-time link between thosewith direct access to the field and the coordina-tion team working remotely. In Mali, otherplatforms such as radio proved helpful inenabling field teams to notify remote communi-ties of their planned field visits. Newtechnologies allowed for a timely transfer ofinformation between field teams and coordina-tion teams. More importantly perhaps, newtechnologies enabled both teams to remain intouch and in the process steer the process ofdata collection and data analysis.

Linking data analysis and data collectionand steering the process of data collection isparticularly important when it comes to collect-ing qualitative data. Qualitative data collectionin SQUEAC can set out to assess factors that areknown to influence coverage (see Box 4) but itmust ultimately be an iterative process,adapted to newly emerging information andtrends. Communication between those collect-ing qualitative data, and those responsible for

analysing it and identifying new lines ofenquiry, is therefore essential, as is the triangu-lation of qualitative data.

Supervision & motivationThe process of qualitative and quantitative datacollection in SQUEAC often merits close super-vision to ensure that data are adequatelytriangulated (by source and method) and toensure that sampling is comprehensive andexhaustive. In Mali and Mauritania, supervi-sion could not be undertaken directly by thecoordination team or the lead investigator dueto a lack of access to (most) programme areas.Some of the issues already discussed (e.g. regu-lar communication, advanced planning andrecruitment of adequate field teams, etc.)combined with well managed workloads andclear roles and responsibilities can helpminimise risks of remote coverage The selec-tion of a strong and reliable assistant(s) is anessential part of remote SQUEAC implementa-tion. Spending sufficient time to transmit themethodology and the processes involved canensure that the assistant(s) will be able to steerthe teams in the right direction. Constantcommunication is also essential.

The experiences from Mali and Mauritaniaprovide some examples of how how proac-tively to strengthen supervision andmotivation. In Mali, teams carried out dailyphone conversations at the start of the day todiscuss the daily plan of action and at the endof every day to follow up, strengthen the teammotivation and address everyday field prob-lems. In Mauritania, data collection teamsreturned to base whenever possible to debrief,relay data, and discuss challenges.

Proactively investing in recruitment andtraining cannot always ensure a successfuloutcome. In Mauritania, the data collected aspart of the wide-area survey (Stage 3) wasfound, upon checking, to be unreliable. A deci-sion was made to send a second team ofenumerators to re-verify the data. This was onlypossible because of the contingency planningdeveloped to accommodate the remote natureof the exercise.

ConclusionsACF’s experiences in Mali and Mauritania haveshown that physical lack of access toprogramme areas is not an insurmountablebarrier to monitoring the performance of the

Figure 1: Development & testing of hypothesis by multi-layered team in Mali

Continue withSQUEAC stage 3

Distance is the key barrierfor nomadic groups

Accessibility (including proximity tohealth centres, infrastructure, river

crossing) is the main reason for coverage failure.

Hypotheses 1:Nomadic vs.

sedentary

Hypotheses 3:Security problems

in some areas

Hypotheses 4:Stock breaks inselected health

Hypotheses 2:Distance to health

centre

Return to dataanalysis

No Yes

In general terms

Do thefindings confirm the

hypothesis?

Re-structure hypotheses

Field teams carry out small-area survey to confirm or deny hypotheses (Stage 2)

Yes, yet access is broader than proximity. Topography(river crossing in particular) is a key factor.

Request qualitative data from field teams (e.g. interviews with local authorities, healthstaff, and beneficiaries)

Nomadic areasseems to have

lower coverage

Limited impact on health centre

attendance

Stock breaksgeneralised (not

site-specific)

Far away villagesseem to have

lower coverage

Request additional data from field teams (Gao)

Analysis of programme data

available

Coordination team propose several hypotheses.

What are the areas of high and low coverage and the reasons for coverage failure?

Focus on coverage assessment

K G

reeenw

ay,

Zam

bia

, 2009

37

intervention. Implementing remote SQUEACinvestigations is feasible and can provide suffi-ciently reliable data about programme coverageand the factors affecting it. Remote coverage inves-tigations do not require additional time orresources if there is enough advance planning,support from the local base and a contingency planhas been provided. They do require that standardSQUEAC processes be accentuated or strength-ened. These include: advanced planning,preparation of data for its analysis, separating datacollection and data analysis processes, using newtechnologies to ensure regular communicationbetween both sets of activities, and addressing theissue of supervision and motivation proactivelyand reactively as the investigation develops.

Like other aspects of remote technical support,implementing SQUEAC investigations remotelydoes require a greater degree of reliance on fieldteams. Trust is of the essence, but CMAMprogrammes can minimise potential risks byinvesting time in the selection of these teams andby allocating manageable daily workloads.Although SQUEAC was designed to be imple-mented by programme staff, the involvement ofexperienced lead investigators/technical advisorsoften proves valuable in the process of data analy-sis, by bringing a measure of objectivity to keyprocesses (e.g. interpretation/weighing of findingswhen building a prior). In remote SQUEAC, thepresence and input of external technical advisorscan help bridge the gaps left by lack of access andlimited data accessibility. As the experience fromMali showed, such input during the early part ofthe process (Stage One) was particularly helpful inensuring that subsequent processes wereadequately implemented. Finally robust datacollection, always important for SQUEAC, isessential for remotely managed programmes. Byintroducing local teams to SQUEAC, it becomeseasier for programmes to adopt SQUEAC-basedmonitoring frameworks that can facilitate futureSQUEAC investigations and programme monitor-ing as a whole.

SQUEAC was developed as a way for nutritionprogrammes to monitor their own performance.For programmes operating in areas with limitedaccess/mobility, the need for reliable self-evalua-tion tools is particularly pressing. Carrying outSQUEACs in such contexts is possible with onlyminimal changes to the methodology. The realchallenge lies in creating the capacity within theseprogrammes to collect, document, analyse andreport routine data in a manner that enables themto carry out future exercises with minimum exter-nal support.

For more information, contact Saul Guerrero,email: [email protected]

Focus on coverage assessment

Causal analysisand the SQUEACtoolbox

Mara Nyawo is a nutritionistspecialising in nutrition surveysand surveillance. She has nineyears experience working in emer-gency and chronic emergencysettings in Africa and is currentlyworking for UNICEF in Sudan.

Mark Myatt is a consultantepidemiologist. His areas ofexpertise include surveillance ofcommunicable diseases, epidemiology of communicablediseases, nutritional epidemiology,spatial epidemiology, and survey

design. He is currently based in the UK.

The authors wish to thank the Sudan Federal Ministryof Health, Kassala State Ministry of Health, GOAL, andUNICEF's Kassala Office for help with organisation,facilities, accommodation, and logistics.

By Mara Nwayo and Mark Myatt

In this article we report our experiencesusing the SQUEAC1 toolbox to under-take a causal analysis of severe wasting(SAM) in a rural area of Eastern Sudan.

The work reported here took place during atrainers-of-trainers course in SQUEAC andSLEAC2 coverage assessment methods. Thecourse was organised by UNICEF and heldin the city of Kassala in Eastern Sudan inSeptember 2011. Course participants weredrawn from United Nations (UN) organisa-tions, non-governmental organisations(NGOs), and state and federal ministries ofhealth. None of the course participants hadprior experience with SQUEAC, SLEAC, orthe CSAS3 coverage assessment method.

A semi-quantitative model of causalanalysis was proposed and tested. Theelements of this model are outlined in Figure1. It is important to note that many of theactivities required to undertake the causalanalysis are existing SQUEAC activities. Theapproach uses SQUEAC tools to identify riskfactors and risk markers for subsequentinvestigation by case-control study. Amatched case-control design was proposedand tested as this requires a smaller samplesize than an unmatched design for the samestatistical power. Matching was done onlocation and age. Cases were children agedbetween six and fifty-nine months with amid-upper-arm-circumference (MUAC)below 115 mm and/or bilateral pittingoedema. Controls were nearby neighbours ofcases and of similar age (i.e. within ± threemonths) with a MUAC greater than 124 mmwithout bilateral pitting oedema. Data werecollected on 35 sets of matched cases (n = 35)and controls (n = 78). The overall sample sizefor the study was, therefore, n = 113.

Collection of causal data using the

SQUEAC toolboxTrainees had no difficulty collectingcase-histories from the carers of SAMcases in the programme and from carersof non-covered SAM cases found in thecommunity during SQUEAC small-areasurveys. Trainees also had no difficultycollecting causal information from avariety of informants (e.g. medical assis-tants, community based volunteers(CBV), traditional birth attendants,traditional health practitioners, villageleaders, etc.) using informal groupdiscussions, in-depth interviews, andsemi-structured interviews. They alsohad no difficulty in collating andanalysing the collected data usingconcept-maps and mind-maps (seeFigure 2). Trainees had little difficultyexpressing findings as testable hypothe-ses. These are all core SQUEACactivities. Trainees selected potential riskfactors and risk markers for furtherinvestigation with minimal interventionfrom the trainer.

Translation of findings to datacollection instrumentsSome trainees had difficulty in design-ing instruments (i.e. question sets) to teststated hypotheses. The problemappeared to be in formulating unam-biguous questions and in breaking downcomplex questions into small sets ofsimple linked questions. Future devel-opment work should explore whetherrole-playing might help with this activ-ity. Trainees found little problemidentifying, adapting, and using prede-fined question sets (e.g. for a householddietary diversity score and for infant andyoung child feeding (IYCF) practices)when these were available. Future devel-opment work should focus on building alibrary of pre-tested and ready-to-usequestionnaire components likely to be ofuse. Trainees had little difficulty field-testing their data collection instrumentsand adaptations were made and testedin the field and again at the surveyoffice.

Case-finding and questionnairemanagementTrainees quickly developed the skillsrequired for active and adaptive case-finding (this was expected fromprevious SQUEAC trainings).Identification of matched controls wasperformed well under minimal supervi-sion. The management of questionnairesfor a matched case-control study wasalso performed well under minimalsupervision.

Applying the case-control question-naire to cases, identifying appropriatelymatched controls for each case, applyingthe case-control questionnaire tocontrols, and the management of studypaperwork added a considerable data-collection overhead above that already

1 Semi-quantitative Evaluation of Access and Coverage2 Simplified LQAS Evaluation of Access and

Coverage (LQAS: Lot Quality Assurance Sampling)3 Centric systematic area sample

Surveyor filling up the questionnairein the village Zinda, Gabero municipality (Mali)

K G

reeenw

ay,

Zam

bia

, 2009

Routine Data

38

Focus on coverage assessment

required by the SQUEAC likelihood survey4. It is estimated thatsurveyor workload for the likelihood survey may increase by50% or more.

Data-entry and data-checkingGreat difficulty was experienced and much time wasted work-ing with EpiInfo for Windows. This software proved bothdifficult to use and unreliable. Data were lost on two occasions.Switching to EpiData proved necessary. This software provedmuch easier to learn and use. Future development work shoulduse a simple and reliable data-entry system such as EpiData.This software can be run from a USB flash drive and does notrequire software to be installed.

Data-analysisNo attempts were made to teach the details of the techniquesrequired for data management and data analysis. This compo-nent was not tested because the computers available wereconfigured so as to prevent the installation of software (theintention had been to test this activity using a free studentversion of a major commercial statistics package). Data wereanalysed using the MSDOS version of EpiInfo (v6.04d) and thecLogistic add-in software. This command-line driven softwaremay not be suitable for use by workers used to using moregraphical software.

The process of data analysis (i.e. conditional logistic regres-sion with backwards elimination of non-significant variables)was demonstrated to a local supervisor with some experiencewith the analysis of cross-sectional survey data (e.g. SMART5,IYCF, MICS (Multiple Indicator Cluster Survey)). He managedto replicate the demonstrated analysis using EpiInfo andcLogistic. He later demonstrated the analysis to the traineegroup and independently reproduced the analysis usingSTATA. The results of the analysis (from cLogistic) are shown inFigure 3.

Further work is required to identify useful software and todevelop a practical manual including worked examples. Themanual could be a self-paced programmed learning course.This would allow both self-teaching and classroom-basedteaching. The manual should cover data-entry and checking,data-management, data-analysis, and reporting.

SummaryThe data collected in this exercise were sufficient to identify riskfactors and risk markers (i.e. diarrhoea, fever, early introductionof fluids other than breastmilk – a marker for poor IYCF prac-tices) that were significantly associated with SAM. Thissuggests that it is possible to use the SQUEAC toolbox to collectcausal data using the level of staff selected for training asSQUEAC supervisors and trainers. Data analysis may, however,require staff with a stronger background in data-analysis.

Consideration should be given as to whether a case-series or setof case-reports collected from carers of cases in a communitybased management of acute malnutrition (CMAM) programmeand non-covered cases found in the community duringSQUEAC small-area surveys could provide a useful causalanalysis. Collected data could be organised and presented usinga mind-map (as in Figure 2). This would be simpler and cheaperthan a case-control study and would probably be more robustthan currently utilised methods which tend to use a singleround of focus groups (typically excluding carers of SAM cases)and a ‘problem-tree’ analysis.

The work reported here supports the further development andtesting of the proposed model for a causal analysis add-in toSQUEAC. This article is intended to inform the emergency anddevelopment nutrition community of our experiences with thismodel so as to allow us to judge the level of interest in furtherdevelopment of the method.

For further information, contact: mark[at]brixtonhealth.com

4 The survey conducted in the (optional) third stage of a SQUEAC investigation

which, when combined with other data, provides an estimate of overall

programme coverage5 Standardised Monitoring and Assessment of Relief and Transitions.

http://www.smartmethodology.org/

Figure 2: Mind map of potential risk factors and risk markers for severe wasting created usingstandard SQUEAC tools

Figure 3 : Results of the analysis in cLogistic

Figure 1 : The proposed model of causal analysis

Hygiene

FeedingCare practices

Case-histories,in-depth interviews,and group discussion

with carers of SAMcases

Semi-structuredinterviews with

MoH, clinic Staff,and CHWs

Group discussionswith CHWs, CBVs,and community

leaders

In-depth interviewswith THPs / TBAs

Qualitativeanalysis

Existing SQUEAC activity

Additional activity for causal analysis

Time / Place / Person

Host / Agent / Environment

Folk terms

Folk Aetiologies

Local Aetiologies

Treatment seeking behaviour

Matched Case-control study

Reportof qualitative

analysis

CAUSALANALYSIS

Reportof quantitative

analysis

Early marriage

Poor IYCF

Early weaningBirthspacing

Abrupt weaning

Early introductionof fluids

Low diversityof diet

No medicinesat clinics

DISEASE

SEVEREWASTING

DISEASE

Diarrhoea

Fever

Food availability

Care practicesCBV, TBA, THP, &c.

CBV, TBA, THP, &c.

CARERS

No breastmilk

Hygiene

Feeding

Poormaternal diet

Fluids / solidsrestricted

Diarrhoeaand fever

Diarrhoea

Fever

Poor appetite

Low diversityof diet

Diarrhoeaand fever

Father awayfor labour

Conditional logistic regression DIA + FEVER + FLUID

Score statistic = 24.0557 3 df (P = .0000)

Likelihood ratio statistic = 28.9999 3 df (P = .0000)

Dependent Variable = CASE

Standard

95% Confidence limits

Likelihood ratios

Coefficient Error Coef/SE "P value"

DIA 1.5247 .5675 2.6866 .0072FEVER 2.1704 .6589 3.2941 .0010FLUID -.3233 .1717 -1.8829 .0597

H0: coeff = 0 lr statistic (1 df) P-value

DIA 8.6471 .0033FEVER 15.6792 .0001FLUID 4.3655 .0367

Coefficient Odds ratio

lower limit

upper limit

lower limit

upper limit

DIA .4124 1.5247 2.6370 1.5104 4.5938 13.9713FEVER .8790 2.1704 3.4617 2.4085 8.7615 31.8721FLUID -.6599 -.3233 .0132 .5169 .7237 1.0133

This analysis shows that diarrhoea (DIA) and fever(FEVER) are strongly and positively associated withSAM. The variable FLUID is the age (in months) at whichthe mother reports that fluids other than breastmilk wereintroduced into the child's diet. Increasing age is nega-tively associated with SAM (i.e. early introduction of fluidsother than breastmilk increases the risk of SAM).

Analysis of other data that were collected during the case-control study revealed that around 63% of carers whose

child had a recent episode of diarrhoea had (inappropriately)restricted the intake of both fluids and solids.

This analysis suggests the following interventions:• Promotion of ORS• Promotion of hand-washing and other hygienic practices• Improved provision of antimicrobials at PHC facilities• Increasing water availability (supporting hygiene promotion)• Promotion of appropriate IYCF practices

Figure 1: The SLEAC surveys identify low coverage areas for investigation usingSQUEAC

39

Ernest Guevarra leads Valid International'scoverage assessment team. He has formaltraining as a physician and a public healthpractitioner and invaluable informal trainingas a community worker from the communitieswith whom he has worked. He has worked in

the Philippines, Uganda, Somalia and Sri Lanka.

Saul Guerrero is the Evaluations,Learning and Accountability (ELA)Advisor at Action Against Hunger (ACF-UK). Prior to joining ACF, he worked forValid International Ltd. in the research,development and roll-out of CTC/CMAM.

He has worked in many countries in Asia and Africa

Mark Myatt is a consultantepidemiologist and senior researchfellow at the Division ofOpthalmology, Institute ofOpthalmology, University CollegeLondon. His areas of expertise

include infectious disease, nutrition and survey design.

The authors would like to thank the Sierra Leone Ministry of Health and Sanitation for managing the survey and for allocating key ministry personnel to conduct the survey, Statistics SierraLeone for providing relevant data and appropriate maps used during the survey and UNICEF Sierra Leone for funding the survey. The authors' appreciation also goes to the people of SierraLeone, without whose support and assistance this survey would not have been possible.

By Ernest Guevarra, Saul Guerrero, and Mark Myatt

In 2010, UNICEF approached VALID International Ltd. to designand conduct a national coverage survey of the government-runcommunity based management of acute malnutrition (CMAM)programme in Sierra Leone. Discussions with UNICEF and the

Sierra Leone Ministry of Health indicated that a spatially exhaustive setof SLEAC surveys (i.e. a SLEAC survey performed in every healthdistrict) augmented by one or two targeted SQUEAC investigationswould provide the information needed by both UNICEF and the SierraLeone Ministry of Health. The idea of using the two methods together inthis way is to use SLEAC to identify district programmes achieving lowand high coverage and to use SQUEAC to investigate the reasons for theobserved levels of coverage. Two variants of this model are outlined inFigure 1 and Figure 2.

This article describes how we used the SLEAC method to perform awide-area coverage survey of the national CMAM programme in SierraLeone. It also describes the SLEAC method in general terms.

The SLEAC method described: The simplified LQAS classifierThe SLEAC method classifies programme coverage for a service deliveryunit such as a health district. A SLEAC survey does not provide an esti-mate of overall coverage with a confidence interval for a single servicedelivery unit. Instead, a SLEAC survey identifies the category of cover-age (e.g. low, moderate, or high) that best describes the coverage of theservice delivery unit being assessed. The advantage of this approach isthat relatively small sample sizes (e.g. n = 40) are required in order tomake accurate and reliable classifications.

SLEAC uses the same simplified LQAS (Lot Quality AssuranceSampling) classification technique that is used in SQUEAC small-areasurveys. The differences between how the simplified LQAS classifica-tion technique is used in SQUEAC and SLEAC are:• The SLEAC survey sample is designed to represent an entire district

rather than a small area.• SLEAC surveys have no prior hypothesis regarding coverage. This

means that SLEAC surveys require larger sample sizes than SQUEACsmall area surveys.

• A target sample size for SLEAC surveys is decided in advance of data-collection. This is usually about n = 40 severe acute malnutrition(SAM) cases.

• SLEAC surveys may classify coverage into three (or more) classes.

Analysis of data using the simplified LQAS classification techniqueinvolves examining the number of cases found in the survey sample (n)and the number of covered cases found:• If the number of covered cases found exceeds a threshold value (d)

then coverage is classified as being satisfactory.• If the number of covered cases found does not exceed this threshold

value (d) then coverage is classified as being unsatisfactory.

Using SLEAC as a wide-areasurvey method

Figure 2: The SLEAC surveys identify low and high coverage programmes forinvestigation using SQUEAC

StartStart

Stop

SLEACsurveys

no

no

yesyes

SLEACsurveysCoverage

OK?

CoverageOK?SQUEAC

investigation(s)

SQUEACinvestigation(s)

SQUEACinvestigation(s)

Reformprogramme

Reformprogramme

ReformprogramCompare

& contrast

A survey team assigned to a 'seasegment' in one of the districts

1 Lot Quality Assurance Sampling

SLEAC stands for Simplified LQAS1 Evaluation of Access and Coverage. It is a quick andsimple method for assessing coverage in a programme area such as a health district.

Figure 4: Simplified LQAS nomogramfor finding appropriate valuesfor d1 and d2 given n, p1 and p2

The threshold value (d) depends on the number of cases found (n) andthe standard (p) against which coverage is being evaluated. A specificcombination of n and d is called a sampling plan. The following rule-of-thumb formula may be used to calculate a suitable threshold value (d) forany standard (p) and any sample size (n):

For example, with a sample size (n) of 40 and a standard (p) of 70% theappropriate value for d would be:

It is unlikely that a SLEAC survey will return the target sample size (n)exactly. If a survey does not return the target sample size (n) exactly thenthe classification threshold value (d) should be recalculated using theachieved sample size. For example:

Target sample size : 40Achieved sample size : 43

Standard : 70%

d :

Coverage is classified using the same technique as is used for SQUEACsmall-area surveys. For example:

n : 43d : 30

Covered cases found : 34Coverage classification : Satisfactory (since 34 > 30)

The simplified LQAS classification technique provides binary or two-tierclassifications. The method is usually extended to provide more granularclassifications in SLEAC surveys. Three classes are sufficient for mostSLEAC applications. Three-tier classifications require two samplingplans which are created using the rule-of-thumb formula presentedearlier.

For three-tier classifications there are two coverage proportions:

p1: The upper limit of the ‘low coverage’ tier or class

p2: The lower limit of the ‘high coverage’ tier or class

The ‘moderate coverage’ class runs from p1 to p2. For example:

Two classification thresholds (d1 and d2) are used and are calculated as:

Classifications are made using the algorithm illustrated in Figure 3.

This three-tier classification works well with small sample sizes (e.g. n= 40) provided that the difference between p1 and p2 is greater than orequal to about 20 percentage points.

Here is an example of the calculations required:

Sample size (n) : 40p1 : 30%p2 : 70%

d1 :

d2 :

Figure 4 shows a nomogram for finding appropriate values for d1 andd2 given n, p1 and p2 without the need for calculation.

Classifications are made using the algorithm illustrated in Figure 3.

40

Focus on coverage assessment

Figure 3: Algorithm for a three-class simplified LQAS classifier

Sample

Classify as high coverage

Classify as moderatecoverage

no no

yes yes

Number ofcovered cases

exceeds d2?

Number ofcovered cases

exceeds d1?

Classify as lowcoverage

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Programme coverage

p1 p2

Moderate coverage High coverageLow coverage

Figure 6: A coarse CSAS/quadrat sampleof villages suitable for a SLEAC survey

Figure 5: Mapping of the coverage classderived from the data shown in Table 1

50

40

300 5 10 15 20 25 30 35 40 45

p = 10%

p = 20%

p = 30%

p = 40%

p = 50%

p = 60%

p = 70%

p = 80%

p = 90%

d (round down to nearest whole number)Example showing d = 27 when n = 39 and p = 70%

Sam

ple s

ize (n

)

Low

Moderate

High

Legend

Intended catchment

Major road

Towns and villages

Sampling locations

10km

Figure 8: Villages selected using stratified systematic sampling

Figure 7: Selection of villages to besampled using CSAS sampling

A case-finding tally from a SLEACsurvey in one of the districts

Figure 9: Structure of samples in rural and peri-urban/urban districts

Legend

Intended catchment

Programme site

Clinic catchment

Major road

Towns and villages

Sampling locations

Centre of quadrat

Selected villages (5)

Quadratboundary

Villagenot selected

Sampling locations (villages) wereselected systematically from a completelist of villages sorted by clinic catchmentarea. This method can be performedusing village lists and does not require amap. Note that the sample is reasonablyevenly spread over the entire survey area.

District

Chiefdom

Chiefdom

Village

Village

Village

Village

Village

Village

Village

Village

Village

District

Section

Section

City Block

City Block

City Block

City Block

City Block

City Block

City Block

City Block

City Block

City Block

A : Rural districts B : Urban and peri-urban districts

Chiefdom Village Section

Table 1: Results of SLEAC surveys in eight clinic catchment areas using a targetsample size of forty (n = 40) cases for each catchment area and the classboundaries p1 = 30% and p2 = 70%

41

larger quadrats). The villages to be sampled are selected by theirproximity to the centre of each quadrat (Figure 7). The number/sizeof quadrats should be selected so as to spread the sample of villagesover the entire survey area. Another approach is to stratify by cliniccatchment area and systematically select villages from a complete listof villages sorted by clinic catchment area (Figure 8). This approachmay be used with any areas (e.g. administrative areas) for whichcomplete lists of villages are available. This first stage samplingmethod should be a spatial sampling method that yields a reason-ably even spatial sample from the survey area. Cluster samplingusing population proportional sampling (PPS), such as that used forSMART surveys, is not appropriate. The approaches outlined herecan provide a reasonably even spatial sample using village lists anddo not require the use of maps. It is important to note that samplingshould not stop when the survey has reached its required samplesize. Sampling only stops after you have sampled all of the selectedvillages

A within-community sampling method: This will usually be anactive and adaptive case-finding method or a house-to-house censussampling method (see Box 1). These methods find all, or nearly all,current and recovering SAM cases in a sampled village. Samplingshould be exhaustive. This means that you only stop sampling whenyou are sure that you have found all cases in the community.Sampling should not stop when you have met a quota or the widersurvey has reached its required sample size.

Focus on coverage assessment

If a survey does not return the target sample size (n) exactly then the classi-fication thresholds (d1 and d2) should be recalculated using the achievedsample size. For example, a set of SLEAC surveys classifying coverage inindividual clinic catchment areas using a target sample size of forty SAMcases (n = 40) for each catchment area and the class boundaries p1 = 30% andp2 = 70% returned the data shown in Table 1. The target sample size wasapplied to each of the clinic catchment areas separately. This allowed cover-age classifications to be made for individual clinic catchment areas. Thesecoverage classifications could be presented as a map as in Figure 5.

SLEAC can estimate coverage over several service delivery units. Coverageis classified using SLEAC surveys in individual service delivery units. Datafrom the individual service delivery units are then combined and coveragefor this wider area is estimated from the combined sample. The details ofcalculating a wide-area estimate from a set of SLEAC surveys are notcovered in this article.

The SLEAC method described: Sample Design

SLEAC uses a two-stage sample design:

First stage sampling method: This is the sampling method that is used toselect villages to be sampled. CSAS coverage assessments use the centricsystematic area sampling or quadrat method to select villages to be sampled(Figure 6). A similar method is often used to select villages to be sampled forSLEAC surveys. The number of quadrats drawn on the map may be muchsmaller than would be used for a CSAS assessment (this is the same as using

Box 1: Active and adaptive case-finding Box 2: Simple structured interview questionnaire to be applied to carers ofnon-covered cases

The within-community case-finding method used in both SQUEAC small-area surveys,SQUEAC likelihood surveys, SLEAC, and CSAS surveys is active and adaptive:

ACTIVE: The method actively searches for cases rather than just expecting cases to be found in a sample.

ADAPTIVE: The method uses information found during case-finding to inform and improve the search for cases.

Active and adaptive case-finding is sometimes called snowball sampling, optimally biasedsampling, or chain-referral sampling. The following method provides a useful starting point:

Ask community health workers, traditional birth attendants, traditional healers orother key informants to take you to see “children who are sick, thin, have swollen legs or feet, or have recently been sick and have not recovered fully, or are attending a feeding programme” and then ask mothers and neighbours of confirmed cases to helpyou find more cases using existing cases as exemplars.

The basic case-finding question (i.e. “children who are sick, thin, have swollen legs or feet,or have recently been sick and have not recovered fully, or are attending a feedingprogramme”) should be adapted to reflect community definitions / aetiologies of malnutri-tion and to use local terminology. Markers of risk (e.g. orphans, twins, single parents,neglected or abused children, households without land or livestock, &c.) may also beincluded in the case-finding question. It is important to avoid, if possible, highly stigma-tised terms (i.e. terms associated with poverty, child abuse or neglect, sexual libertinage,alcoholism, etc) as community members may be reluctant to slander their neighbours inorder to help you find SAM cases. It is important to ask about children attending a feedingprogramme (or specific feeding programmes). Failure to do this may result in bias towardslow coverage in your surveys.

It is important that the case-finding method you use finds all, or nearly all, cases in thesampled communities. Formal evaluations of the type of active and adaptive case-findingdescribed here have found that the method does find all, or nearly all, cases in the sampledcommunities provided that appropriate local terms and appropriate key-informants areused. Interviews such as those outlined in Box 2 are useful in designing the case-findingquestion and selecting the most useful key-informants. Sampling stops only when you aresure that you have found all SAM cases in the community. Sampling in a communityshould not stop because you have reached a quota or met the sample size required by thesurvey. Such early stopping is not allowed.

Care needs to be exercised in the choice of key-informant. Community leaders are a usefulpoint of entry but seldom make useful key-informants. They are most useful in helping youfind and recruit useful key-informants. You should avoid relying solely on communityhealth workers or volunteers who are attached to the programme as they may be unableor reluctant to take you to see children who are not in the programme.

It is important to realise that the active and adaptive case-finding method will fail in somesettings. The method has been found not to work well in some refugee and IDP (internallydisplaced persons) camp settings, in urban locations where there is a high populationturnover (e.g. around railway and bus stations, newly established or growing peri-urban‘shanties, etc), and in displaced and displacing populations. These settings are typified by alack or loss of strong extra-familial relationships, extended familial relationships, stronglocal kinship ties, collective loyalty, and simple (traditional) social structures. In thesesettings it may be very difficult to find useful key-informants or local guides, and snowballsampling will not work well for finding SAM cases when people do not know their neigh-bours well. In these settings it is sensible to search for cases by moving house-to-houseand door-to-door making sure that you measure all children by taking a verbal householdcensus before asking to measure children (this avoids sick or sleeping children being‘hidden’ to avoid them being disturbed by the survey team).

Questionnaire for carer of cases not in the programme

Village : ___________________________________________________________

Programme site :____________________________________________________

Name : ____________________________________________________________

1. Do you think that this child is malnourished? If YES ...

2. Do you know of a programme that can treat malnourished children?If YES ...

3. What is the name of this programme?_______________________________

4. Why is this child not attending this programme?

Do not prompt. Probe "Any other reason?"

Programme site is too far awayNo time / too busy to attend the programmeCarer cannot travel with more than one childCarer is ashamed to attend the programmeDifficulty with childcareThe child has been rejected by the programme

Record any other reasons ... ______________________________________

______________________________________________________________

5. Has this child ever been to the programme site or examined by programme staff? If YES ...

6. Why is this child not in the programme now?

Previously rejectedDefaultedDischarged as curedDischarged as not cured

Thank carer. Issue a referral slip. Inform carer of site and date to attend.

Clinic catchmentarea

Sample size d1* d2* Number ofcovered cases

Classification

Chawama 38 11 26 29 High

Matero 32 9 22 18 Moderate

Makeni 43 12 30 36 High

Chipata 35 10 24 15 Moderate

Ngombe 42 12 29 14 Moderate

Kalingalinga 37 11 25 10 Low

Mtendere 39 11 27 5 Low

Kanyama 42 12 29 23 Moderate

All 308 92 215 150 Moderate

* d1 and d2 calculated after data collection using achieved sample sizes

near even spatial spread of the selected villages across rural districts anda near even spatial spread of selected EAs across urban and per-urbandistricts. The structure of the district-level samples is shown in Figure 9.

A target sample size of n = 40 current SAM cases was used in both ruraland urban districts. This is the standard SLEAC sample size for largepopulations.

The number of PSUs (nPSU) needed to reach the target sample size in eachdistrict was calculated using estimates of average EA population andSAM prevalence using the following formula:

Average EA population was estimated as:

using data from the most recent (2004) Sierra Leone Population andHousing Census.

The percentage of the population aged between 6 and 59 months wasestimated as 17.7%. This is a national average taken from the SierraLeone 2004 Population and Housing Census. This estimate is used bySierra Leone government departments, UN agencies, and NGOs.

SAM prevalence rates were taken from reports of SMART surveys ofprevalence in each district that had been undertaken in the lean periodof the previous year. The prevalence of SAM using MUAC and oedemawas used since this matched programme admission criteria.

The Sierra Leone Central Statistics Bureau provided information on thetotal district populations and total number of EAs in each district. TheSierra Leone Central Statistics Bureau also provided lists of enumerationareas for the Western Area (urban and peri-urban) districts and large-scale maps (see Figure 10) of the EAs that were selected for sampling.

PSUs were selected using the following systematic sampling procedure:

Step 1: The lists of EAs were sorted by chiefdom for rural districts and by section for urban and peri-urban districts.

Step 2: A sampling interval was calculated using the following formula:

Step 3: A random starting PSU from the top of the list was selected using a random number between one and the sampling interval. The random number was generated by coin-tossing (see Box 3 for details).

42

This is a two-stage sample because a sample of villages in the surveyarea is taken first (stage one) and then a ‘census’ sample of current andrecovering SAM cases is taken from each and every one of the selectedvillages (stage two).

T he SLEAC meth od des crib ed: Sample si zeSLEAC uses a target sample size (n) which, together with prevalence andpopulation estimates, is used to decide the number of villages (nvillages)that should be sampled in order to achieve the target sample size. Atarget sample size of n = 40 cases from each service delivery unit inwhich coverage is to be classified is usually large enough for mostSLEAC applications.

The target sample size (n) together with estimates of the prevalence ofsevere acute malnutrition (SAM) in the survey area and population datais used to calculate the number of villages (nvillages) that will need to besampled in order to achieve the target sample size:

SAM prevalence refers to the average SAM prevalence in the catchmentarea of the service delivery unit. It is unlikely that SAM prevalence willbe known or known with good precision. SAM prevalence estimatesmay be available from (e.g.) previous SMART surveys. SAM prevalencevaries throughout the year (e.g. prevalence is usually higher beforeharvests than after harvests). This means that you should use the resultsfrom a nutritional anthropometry survey undertaken at the same time ofyear as the current SLEAC assessment.

Note that prevalence here is the estimated prevalence of theprogramme's admitting case-definition. This will usually not be theweight-for-height based ‘headline’ prevalence estimate reported by aSMART survey. The required estimate will usually be found in the needsassessment section of a SMART survey report.

If you do not have nutritional anthropometry survey results for the sametime of year as the current SLEAC assessment then you should useresults from the most recent nutritional anthropometry survey andadjust them using (e.g.) seasonal calendars of human disease, calendarsof food-availability, agricultural calendars, long term admissions datafrom nutrition programs, and long term returns from growth monitor-ing programmes.

The formula for the calculation of the minimum number of villages thatneed to be sampled in order to achieve the required sample size shownabove assumes that the case-finding method being used will find all, ornearly all, current and recovering SAM cases in sampled villages. If youare unsure of this then you should sample a larger number of villages.

Once these calculations have been made, sampling locations can beidentified and the survey undertaken (e.g. as shown in Figure 6 andFigure 8). A standard questionnaire, such as that shown in Box 2, shouldbe applied to carers of non-covered cases found by the survey. Datacollected using the standard questionnaire can be presented using aPareto chart (i.e. a bar-chart with bars ordered by and with lengthsproportional to the frequency of the reported barriers).

Background to the Sierra Leone national SLEAC survey

The CMAM approach to treating cases of severe acute malnutrition(SAM) in government health facilities was piloted in four districts ofSierra Leone in 2008. The programme was expanded to provide CMAMservices in selected health centres in all fourteen districts of the countryin 2010. This report describes the application of SLEAC to the assess-ment of the coverage of this national CMAM programme. The workreported here took place in March and April 2011.

SLEAC sample designSLEAC was used as a wide-area survey method to classify coverage atthe district level. The district was selected as the unit of classificationbecause service delivery of the national programme was managed andimplemented at the district level.

The primary sampling units (PSUs) used in the SLEAC surveys werecensus enumeration areas (EAs). In rural districts, EAs were individualvillages and hamlets. In urban and peri-urban districts, EAs were city-blocks or compounds. In rural districts, lists of potential PSUs weresorted by chiefdom. In urban and peri-urban districts, lists of potentialPSUs were sorted by electoral ward (sections). This approach ensured a

Focus on coverage assessment

The survey team

43

The PSUs selected by this procedure were sampled using a case-findingmethod tailored to the particular district:

• In rural districts, a district-specific case-finding question was developed from the base case-finding question:

Where can we find children who are sick, thin, have swollen legs orfeet, or have recently been sick and have not recovered fully, or are

attending a feeding programme?

This question was adapted and improved using information collectedfrom traditional birth attendants, female elders, traditional health prac-titioners, carers of children in the programme, and other key informantsto include local terms (in all local languages) and local aetiologicalbeliefs regarding wasting and oedema. This question was used as part ofan active and adaptive case finding method (see Box 1).

• In urban and peri-urban districts, house-to-house and door-to-door case-finding was used. This was done because it was felt that active and adaptive case-finding would not work well in these districts. Sampling was aided by the use of large-scale maps showing enumeration area (EA) boundaries (see Figure 10).

After all of the selected PSUs in a district had been sampled, the surveyteam met at the district headquarters for data collation and analysis. Thesimplified LQAS classification technique was applied to the collateddata. Coverage standards:

Low coverage: Below 20%.

Moderate coverage: Between 20% and 50%.

High coverage: Above 50%

were decided centrally by MoH and UNICEF staff before the start of thesurveys. These standards were used to create decision rules using therule-of-thumb formulae:

where n is the sample size achieved by the survey, p1 is the lower cover-age threshold (i.e. 20%), and p2 is the upper coverage threshold (i.e. 50%).

Coverage in each district was classified using the algorithm presented inFigure 3.

Table 2 presents the results of the surveys.

Figure 11 presents the same results as a map of per-district coverage.

A short questionnaire, similar to that shown in Box 2, asking about barri-ers to coverage was administered to carers of non-covered cases found.Data were tabulated from the questionnaires using a tally-sheet andpresented as a Pareto chart (see Figure 12A and Figure 12B).

The SLEAC implementation processThe process as described above was completed in eight weeks (44 work-ing days) staffed by fifteen mid-level health management staff and aprincipal surveyor provided by VALID International Ltd.

Box 3: Generating random numbers by tossing coins

Random numbers can be generated by tossing a coin. Tossing a coin has two outcomes(i.e. heads and tails) and the method of generating random numbers by tossing a coinworks by using powers of two. Here are some powers of two:

Each power of two is double the previous number so, for example, 212 = 2048 × 2 = 4096.To generate a random number between 1 and x by tossing a coin you must first find thesmallest power of two that is greater than or equal to x. If, for example, you need togenerate a random number between 1 and 28 you would use 25 (32) since this is thesmallest power of two that is greater than or equal to 28. This power of two, five in thiscase, is the number of coin-tosses (t) required to generate a random number between 1and 32. Write down powers of two starting at 20 and stopping at 2t - 1. For example:

Toss a coin t times and record the result of the tosses below each power of two. For example:

Replace each head result with its associated power of two and replace each tail resultwith zero. For example:

Add up these numbers and then add one. This is the random number. For example:

If a random number generated by this method is out of range (i.e. larger than youneed) then you should discard that number and start again.

Power of two Value20 121 222 423 824 1625 32

1 2 4 8 16

1 2 4 8 16

H T H H T

H T H H T

1 2 4 8 16

H T H H T

1 2 4 8 16

1 0 4 8 0

1 0 4 8 0

1 + 4 + 8 + 1 = 14

Power of two Value26 6427 12828 25629 512210 1024211 2048

Write down powers of 20 and stopping at 2t - 1.

Toss a coin t times and record the result of thetosses below each power of two.

Replace each head result with its associated powerof two and replace each tail result with zero.

Add up these number and thenadd one. This is the randomnumber.

Table 2: Coverage classification by district Figure 10: Example of a large-scale map showing enumeration area (EA) boundaries used when sampling in anurban districtProvince District SAM cases

found (n)CoveredSAM cases (c)

Lower decisionthreshold (d1)

Is c > d1? Upper decisionthreshold (d2)

Is c > d2? Coverageclassification

Northern Bombali 30 4 6 No 15 No LOW

Koinadugu 32 0 6 No 16 No LOW

Kambia 28 0 5 No 14 No LOW

Port Loko 30 2 6 No 15 No LOW

Tonkolili 28 1 5 No 14 No LOW

Eastern Kono 16 2 3 No 8 No LOW

Kenema 34 8 6 Yes 17 No MODERATE

Kailahun 34 4 6 No 17 No LOW

Southern Bonthe 41 7 8 No 20 No LOW

Pujehun 27 6 5 Yes 13 No MODERATE

Bo 22 6 4 Yes 11 No MODERATE

Moyamba 40 6 8 No 20 No LOW

Western Rural 46 6 9 No 23 No LOW

Urban 20 2 4 No 10 No LOW

National Total 428 54 85 No 214 No LOW

and

Focus on coverage assessment

Survey teams discussion how to organisea SLEAC survey in a new district

44

Three survey teams with five members each were used. Theteams were divided into two sub-teams. A survey team washeaded by a ‘captain’ who was in charge of managing the sub-teams, organising travel and survey logistics, and co-ordinatingsurvey activities with the principal surveyor.

Each district was divided into three segments. Segmentationwas informed by logistics with each segment being served by aroad (when possible).

Each survey team was assigned to one of the three segmentsand provided with:• A list of PSUs (sorted my chiefdom) to be sampled.• A list of the locations of CMAM programme sites.• A list of the names and home villages of chiefs and chief's

assistants for each PSU to be sampled.

Each survey team started case-finding in the farthest PSU andthen moved to the next-farthest PSU for case-finding and so-on.At the end of each day, the survey teams lodged in healthcentres, local guesthouses, or in villagers' homes. They restartedcase-finding on the following day. This continued until all thePSUs had been sampled. The survey teams then came togetherat the district headquarters for data collation and analysis andresults were shared with district-level health management staff.

Upon completion, the survey team was able to:• Classify coverage in each district (Table 2)• Map coverage by district for the whole country (Figure 11)• List barriers to coverage ranked by their relative

importance (Figure 12A and Figure 12B)

An overall coverage estimate was calculated but not reported.Figure 13 shows the calculation of the overall coverage estimateusing spreadsheet software.

A single SQUEAC investigation was carried out in the peri-urban‘Western Rural’ district. This district was chosen because a largenumber of cases were found in the SLEAC survey, coverage waslow, and it was conveniently close to where the survey team wasat the end of the SLEAC surveys. Figure 14 shows a concept mapsummarising the key findings of the SQUEAC investigation.

ConclusionThe work reported in this article supports the use of SLEAC forthe assessment of coverage of CMAM programmes over wide-areas up to national scale. The application of the SLEAC surveymethod in Sierra Leone proved to be easy to set-up and super-vise. The simplified LQAS technique for classifying coverage andthe use of tally-sheets to analyse and present ‘barriers’ dataallowed the survey team to analyse results of the survey for eachdistrict as soon as each SLEAC survey was completed. Feedbackwas immediately provided by the survey team to districtMinistry of Health officials on their programme's level of cover-age and the barriers to access and service uptake. Such ‘real-time’analysis and reporting of results is unique to SLEAC and has thepotential for real-time action and programme reform to be imple-mented. These findings demonstrate the usability of the methodby Ministry of Health staff and make SLEAC the coverage assess-ment method of choice when evaluating the coverage of CMAMprogrammes at a regional or national level.

For further information, contact: mark[at]brixtonhealth.com

Figure 12A: Barriers to service uptake and access(tallies from one SLEAC survey)

Figure 12B: Barriers to service uptake and access (national talliesfrom all SLEAC surveys)

Figure 13: Calculation of a wide-area coverage estimate using a spreadsheet

Figure 11: Map of per-district coverage

Low (< 20%)Moderate (20% to 50%)High (> 50%)

Previously rejected

Other reasons

Discharged then relapsed

No time / too busy

Distance

No RUTF

Lack of knowledge about SAM

Lack of knowledge about the programme

Number0 50 100 150 200 250 300

Figure 14: Concept map summarising the findings of the SQUEAC investigation

consistent with

Lessens(for some)

leads to mitigates

mitigates

leads to leads to

results in

results in

consistent with

results in

results in

leads to

leads to

reinforcesperceptions

leads to

leads to

Proximity ofcurrent cases to

OTP sites

HighOpportunity

Costs

Poor awareness and understandingof the programme

Effective coveragearea of OTP sites

is very small

LateTreatment

Seeking

LOWCOVERAGE

Low levelsof

defaulting

Perceptionthat the service

is not free

PoorStandardof Service

No RUTF

No activecase-finding or

community screening

Mothers & carersrecognise SAM

Focus on coverage assessment

45

Register now forcoverage assessmentworkshop in Oxford

Valid International together with Mark Myatt is planningtwo sets of 2-day coverage assessment workshops from27th to the 28th of March 2012 and 29th to the 30th of

March 2012 respectively. The workshop is an introductory courseto the various methods designed and used for assessing the cover-age of the outpatient therapeutic care programme (OTP) andsupplementary feeding programme (SFP) components of commu-nity-based therapeutic care (CTC) or community-basedmanagement of acute malnutrition (CMAM). These methods areSemi-Quantitative Evaluation of Access and Coverage(SQUEAC), Simplified Lot Quality Assurance SamplingEvaluation of Access and Coverage (SLEAC) and Simple SpatialSurvey Method (S3M).

The first workshop is aimed at mid-level managers and techni-cians particularly (but not limited to) those based at headquartersof international non-governmental organisations. The secondworkshop is intended for donors / funders and United Nationsorganisations. The workshop will be held at Jesus College,University of Oxford, UK. Each workshop is limited to 20-25participants.

At the end of the workshop, the participants would be able todifferentiate between the different coverage methods and deter-mine the programme contexts for which they are best suited. Thiswill enable more informed decisions on planning and budgetingfor appropriate coverage assessment methods for theirprogrammes in different country contexts. The workshop is notmeant to train participants on how to conduct coverage assess-ments.

Each 2-day workshop will start with an introductory sessionon the methods, their history and development and strengths andweaknesses. This will be followed by sessions that detail thefeatures of the individual methods. The sessions will be led by ateam of four highly-experienced practitioners headed by MarkMyatt, the lead developer of all the methods currently used forcoverage assessment. The workshop agenda is currently beingfinalised and will be made available to participants in February2012.

Participation in the two-day workshop costs £600. Thisincludes workshop fees, workshop materials and lunch andrefreshments for both days. Please note that participants will beresponsible for their own travel costs to and from Oxford.Participants who wish to stay in Oxford will be responsible fortheir own accommodation. Various options including discountedrate at Jesus College as well as travel information are availableonline.

Registration will start on the 2nd of January 2012. For furtherinformation and/or to signify your interest, visit the workshopwebpage at http://www.validinternational.org/coverage orcontact Basia Benda, email: [email protected]

The Technical Reference for Semi-Quantitative Evaluations ofAccess and Coverage (SQUEAC) and Simplified Lot QualityAssurance Sampling Evaluations of Access and Coverage

(SLEAC) Methods was developed by Action Against Hunger, BrixtonHealth, Concern Worldwide, Food and Nutrition Technical AssistanceII Project (FANTA-2/FANTA-2 Bridge), UNICEF, Valid International,and World Vision International, and will be published by FHI360/FANTA-2 Bridge with USAID funding in 2012.

The Technical Reference comprises dedicated sections on SQUEACand SLEAC methods and a series of case studies that address:• Assessing evidence and coverage in very high coverage

programmes• Assessing evidence and coverage in moderate coverage

programmes• Assessing evidence by wishful thinking (not a good idea)• Sampling without maps or lists• Using satellite imagery to assist sampling in urban settings• Active and adaptive case-finding in rural settings• Within-community sampling in an IDP camp• Within-community sampling in urban settings• The case of the ‘hidden defaulters’• Applying SLEAC: Sierra Leone national coverage survey (see field

article in this issue of Field Exchange)

The appendices include additional technical information on SQUEACand SLEAC, guidance on working with formulas, and a glossary ofSQUEAC and SLEAC terms.

When completed, the Technical Reference will be available for down-loading from the FANTA website: http://www.fantaproject.org

Technical Reference forSQUEAC and SLEAC Methods,2012

En-net is launching a new Coverage Assessment forum area fordiscussion of methods used in the assessment of programme cover-age of the outpatient therapeutic programme (OTP) andSupplementary Feeding Programme (SFP) components ofCTC/CMAM. This forum will also serve as a platform for discussingthe use of these methods in assessing coverage of other programmes.

The new forum area has five Technical Moderators on hand toprovide support and advice to challenging questions: ErnestGuevarra, Lio Fieschi and Allie Norris of Valid International, SaulGuerrero of ACF-UK and Mark Myatt (Independent). The forum areawill be overseen by the ENN Moderator, Tamsin Walters.

All questions welcome. Visit www.en-net.org.uk

To support the forum, a Coverage Assessment sub-section of the ENNResource Library has been established to locate key resources. Visit:http://www.ennonline.net/library and select Technical Resources,then Assessment.

Focus on coverage assessment

Coverage assessmentforum launched onen-net

News

Jan Komrska is a pharmacist working at UNICEF Supply Division leadingNutrition unit and responsible for procurement of products related tonutrition interventions of UNICEF.

The author acknowledges the work of UNICEF Supplies Division reflected in this article.

By Jan Komrska

Since 2007, the use of ready to use thera-peutic foods (RUTF) to treat severe acutemalnutrition (SAM) in young childrenhas been endorsed by the United

Nations and non-governmental organisations(NGOs), and received wide notice by themedia. With the proven success of RUTF, morecountries have adopted their use as part ofcommunity-based management of SAM(CMSAM), and demand for the products hassoared. Since 2006, UNICEF, the world’s majorpurchaser of RUTF1, has taken a series of stepsto shape the market and diversify the supplierbase. As a result of this effort, the market hasgrown from one qualified manufacturer in 2000to 19 today. This number is expected to increasein the coming years, especially in countrieswhere RUTF is used.

RUTF are high-energy foods fortified withvitamins and minerals, packed in individualportions providing energy intake of 500 kilo-calories. They can be in the form of a soft pasteor a crushable biscuit that is easy to swallow.Individual packaging allows easy handling andprevents contamination of the product betweenfeedings. UNICEF requires RUTF to be manu-factured by qualified suppliers in accordancewith stricter quality standards than normalfood products. The product most in demandand the subject of this article is RUTF in peanutpaste form.

The first peanut paste RUTF was developedjointly by the French Institute of Research forDevelopment and the manufacturer Nutriset in1996 as a fortified peanut spread, nowmarketed under the name Plumpy’Nut®.

UNICEF Supply Division is responsible forprocurement of specific products (includingRUTF) for UNICEF country programmes andexternal partners, assuring transparency inusing public funds and maintaining agreedproduct quality.

UNICEF procurement of RUTF in theperiod 2000-2010 UNICEF began to procure RUTF in 2000, whenNutriset was the sole qualified supplier andannual demand was below 100 metric tonnes(MT)2. The first long-term RUTF supplyarrangement (LTA) was established withNutriset on a sole-source basis in 2001.

Under an LTA, Supply Division placesorders with suppliers, based on requisitionsfrom UNICEF country offices. Suppliers areresponsible for manufacturing the product anddelivering it to the nearest seaport for shipmentby sea freight or, in urgent cases, to an agreedairport for air shipment. Further transportation

results of competitive bidding, for a period oftwo to three years with the supplier making thelowest acceptable offer, and eventually a back-up LTA with the supplier making the secondlowest acceptable offer. However this approachwas not applicable in the case of RUTF, becauseit would not encourage any further marketdevelopment and would leave UNICEF withone or two suppliers. Therefore, it was decidedto distribute total forecasted quantity among allcompanies meeting UNICEF technical require-ments for manufacturing facilities as well asproduct specifications.

Proposals were received from 13 companiesand seven proved to be able to meet definedrequirements for global supply of RUTF. LTAswere established subsequently with all sevensuppliers, expanding significantly the supplierbase (for more details see Table 1).

The second competitive bidding exercise forsupply of forecasted 54,000 MT of RUTF for theperiod 2011-2012 was issued by the end of 2010.The forecasted quantity was based on theassumption of continued expansion of CMSAMto new countries and scaling up of existingprogrammes. Proposals were received from 27companies out of which 12 met UNICEFrequirements for global supply of RUTF. LTAswere established subsequently with all 12suppliers listed in Table 2.

It is important to underline that RUTF prod-ucts manufactured by UNICEF-approvedmanufacturers comply with the Joint statementspecifications and they can be used by countryprogrammes interchangeably.

A key part of the procurement strategy wasto support the development of local productionin countries where RUTF is used, particularly inAfrica, in order to bring the supply closer to thebeneficiaries and reduce delivery lead times.

After successful audits of the manufacturingsites by Supply Division’s Quality Assurance

of RUTF to the beneficiary countries is assuredby UNICEF-contracted freight forwarders.Therefore, RUTF prices referred in this articleexclude shipping costs.

By 2004, demand began to rise as more coun-tries began piloting the use of RUTF, and itbecame increasingly urgent for UNICEF toidentify new sources of RUTF. During 2006,Supply Division began to work with manufac-turers in countries where the product could bemanufactured for local use, and approvedsuppliers in Niger and Ethiopia for localpurchase in 2006 and 2007 respectively.

With the publication of the UN JointStatement in 20073, demand increased dramati-cally, outpacing global production capacity. Thesituation became critical in 2008, when ahunger emergency in the Horn of Africa causeda spike in demand in the second half of the year.Even after a second global supplier (Vitaset,located in Dominican Republic) had beenapproved, the 11,000 MT ordered by UNICEF,still largely from Nutriset, did not meet peak indemand.

As a result of this experience, SupplyDivision made three key decisions in 2008: a) to initiate competitive bidding for RUTF in

order to open the market for new suppliersb) to begin conducting annual forecasting for

RUTF with individual country programmes, and

c) to conduct a study on RUTF supply chain performance in order to identify weaknessesand propose solutions.

These efforts were part of a larger procurementstrategy developed by Supply Division throughwhich UNICEF could leverage its buyingpower to influence the market, promoteincreased competition and ensure a diverse andsustainable supply base.

Implementation of the ProcurementStrategyIn 2008, in line with its procurement strategy,Supply Division launched the first competitivebidding exercise for the supply of forecasted20,000 metric tonnes (MT) of RUTF, for theperiod 2009-2010. This exercise was precededby a lengthy, multi-year process of advocacy forincreasing production capacity with existingsuppliers, identification of potential new globaland local suppliers (usually existing foodcompanies), and the development of manufac-turing standards coupled with inspection ofvarious manufacturing facilities.

For the majority of products, UNICEF typi-cally would establish a LTA, based on the

1 UNICEF is the largest, but not the only, purchaser of RUTF.

Other major purchasers include MSF, the Clinton

Foundation and different NGOs.2 1 MT contains 72 cartons of RUTF. It takes approximately

one carton to treat and save a child, so that each MT can

save the lives of 72 children.3 WHO/WFP/UNSCN/UNICEF. Community-Based Management

of Severe Acute Malnutrition. A Joint Statement by WHO,

WFP, UNSCN and UNICEF, May 2007

Increasing Access toReady-to-useTherapeutic Foods (RUTF)

2005 2006 2007 2008 2009 2010 2011

Global suppliers

1 1 1 2 6 7 12

Local suppliers 0 1 2 2 4 7 7

Total 1 2 3 4 10 14 19

Table 1: UNICEF-approved RUTF Suppliers, 2005-2011

46

Field Article

Test products

UN

ICEF

47

Nutriset proposed scaled prices for largerorders, and the price decreased gradually toabout €35.20 (48 USD) per carton in 2011.However, this drop in price was masked byfluctuations in the rate of exchange between theeuro and the United States dollar, because theprice is fixed in euros, as proposed by Nutriset.The RUTF prices offered to UNICEF by othersuppliers range from 48.00 USD to 56.50 USDper carton in 2010.

Prices of locally procured RUTFLocal producers import almost all of the ingre-dients and the packaging material, most ofwhich are subject to import duties. Often theprice offered initially is too high to justify localpurchase on a cost basis alone, even consider-ing the added cost of freight.

Supply Division carefully scrutinises theprices proposed by local companies in order toassure best value for money. They are comparedto the landed cost of globally procured RUTF(price at the factory plus freight cost to the finaldestination) and when found too high, UNICEFattempt to negotiate a lower price.

The prices of locally produced RUTF varysignificantly among the different suppliers. In2010, the local prices ranged between 57.00USD (in Madagascar) and 69.00 USD per carton(in Mozambique).

Freight costsRUTF is a heavy and bulky product, with acourse of treatment for one child (one carton)weighing about 15kg. The 16,000 MT of RUTFpurchased off shore in 2010, filled 830 40-footcontainers. Shipping the product by sea is themost economical option, but air shipmentmight be required when RUTF is needed foremergency response to natural disasters,drought or political instability. Air freight ofRUTF increases the landed cost by 100 per centwhereas sea freight only by 10 per cent.

In 2008, nearly 35% of RUTF purchased byUNICEF had to be transported by air, at a costof $8.5 million, in order to reach the beneficiar-ies on time. By 2010, less than 1% of the RUTFhad to be transported by air, costing less than$400,000. The decrease in shipment by air is aresult of RUTF supply chain analysis thatsuggested a number of tools to reduce freightcosts. These included introductions of forecast-ing of country programming needs,development of supplier base in locations closeto where the product is used and preposition-ing of stock in areas closer to emergency-pronecountries (e.g. Dubai for the Horn of Africa andGhana or Cameroon for the Sahel region).

Issues for future considerationAs more countries adopt and/or scale upCMSAM to treat SAM, the need for RUTF willcontinue to increase. Supply Division willcontinue to refine its procurement strategy andwork with UNICEF’s Programme Division and

external partners to address a number of issuesthat could affect RUTF availability and accessi-bility for the final beneficiaries.

UNICEF sees the continued diversificationof the supplier base, with more qualified manu-facturers in countries and regions closer to theend-users, as an important concern. This willfacilitate the eventual transfer of CMSAM anduse of RUTF from UNICEF and NGOs tonational authorities.

Building a strong network of smaller suppli-ers in countries with the highest incidence ofSAM therefore remains a priority. The work isongoing with potential manufacturers in SierraLeone, Uganda and Rwanda in order for themto become part of UNICEF-approved manufac-turers. However in the process of expansion ofUNICEF approved-manufacturers, the focus onproduct quality and safety shall remain central.

ConclusionsThe Supply Division strategy to increase theavailability of RUTF and assure a sustainableand diverse supplier base has accomplished itsmain goals. In 10 years, the supplier base hasgrown from one global supplier in Europe to 19global and local manufacturers in Africa,Europe and Americas. The seven local suppliersare in countries where RUTF is used, and morelocal manufacturers are expected to beapproved in the near future. Even with thissubstantial growth in production capacity,quality standards have been maintained.Forecasting of demand has been systematisedand orders can be filled promptly when theyare placed on time. Supply Division continuesto work with UNICEF country offices to refineforecasting of demand and to work with suppli-ers to manage global production. Air freightcosts have been reduced dramatically as a resultof these improvements, and stocks are beingprepositioned in key regions to allow for rapidresponse when necessary.

As outlined above, the major future challengeis the potential increased demand for RUTF ascountries adopt and expand CMSAM. SupplyDivision will continue to work with suppliers,country offices, UNICEF’s Programme Division,and external partners on development ofsustainable RUTF supply chain.

For more information, contact: Jan Komrska,email: [email protected]

staff in Niger and Ethiopia, further audits wereconducted at manufacturers in Malawi, theDemocratic Republic of the Congo,Mozambique, Madagascar and Tanzania andresulted in their approval for local purchases toUNICEF. The suppliers listed in Table 3 areauthorised to sell RUTF to UNICEF’s countryprogrammes locally. As of 2010, about 23% ofthe RUTF purchased by UNICEF was sourcedlocally.

As programme demand and productioncapacity increased, so did the volume ofUNICEF procurement, in terms of the numberof MT purchased and the number of countriesplacing orders. After a steady increase from2000 to 2008, procurement of RUTF decreasednot because of declining needs but as a result ofresource mobilisation challenges caused by theglobal economic crisis and availability of stocksin countries from 2008. However, orders haverebounded in 2010 and by mid-year hadsurpassed 2009 levels, reaching 20,690 MT as ofyear-end (2010) (for more details see Table 4).

ForecastingThe growing demand for RUTF, productioncapacity constraints and the volatility of pricesof raw materials made forecasting of demandincreasingly necessary. UNICEF first undertookglobal forecasting to collect information onneeds for RUTF and other nutrition products inJanuary 2009. This resulted not only in forecast-ing of global product needs (used for thebidding process) but also contributed to a betterunderstanding of the scale of implementationof CMSAM.

The 2010 forecast indicated that UNICEFcurrently is implementing CMSAM in about 55countries, where there are 6.1 million childrenwith SAM. UNICEF country programmesintended to reach 1.8 million children (about 30per cent), revealing a large gap in coverage,especially given the 20 million children esti-mated to be suffering from SAM worldwide.

RUTF PricingPeanut-based RUTF consists of milk powder(30%), sugar (28%), peanut butter or paste(25%), vegetable oil (15%) and vitamin andmineral premix (1.6%).4 The product is packedin standard foil sachets. The milk and packag-ing material (aluminum foil) are the mostexpensive components, followed by the premix,peanut butter, sugar and oil.

Somewhat surprisingly, the entry of newsuppliers into the market has not resulted inany dramatic drops in price, for a number ofreasons. One possible reason might be that newsuppliers do not have the large productionvolumes that usually result in lower prices, andare also faced with high-start-up costs.

Prices of globally purchased RUTFThe pricing structure in the LTAs is complexbecause of volume discounts, payment terms,currency used and other factors. The initialprice paid by UNICEF for RUTF in 2001 wasabout €41.50 (56.80 USD) per carton. In 2006, asthe average size of orders began to increase,

4 Mark J. Manary. Local production and provision of ready-to-

use therapeutic food (RUTF) spread for the treatment of

severe childhood malnutrition. Food and Nutrition Bulletin,

vol. 27, no. 3 (supplement) 2006, The United Nations

University

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Countries orderingRUTF

1 4 6 7 8 17 26 41 48 45 52

Total 3 84 57 109 344 1,793 2,697 4,552 10,741 8,129 20,690

Table 4: UNICEF Global Purchases of RUTF, 2000-10 (in MT)

Global supplier Product Name

1. Nutriset (France) Plumpy Nut®

2. Vitaset (Dominican Republic) Plumpy Nut®

3. Diva Nutritional Products (South Africa) Generic name***

4. Insta EPZ (Kenya) Generic name

5. Challenge Dairy (United States)* Generic name

6. Tabatchnick Fine Foods (United States) Nutty Butta

7. Compact (India) EeZee PasteTM

8. Compact (Norway) EeZee PasteTM

9. Edesia (United States) Plumpy Nut®

10. Nutrivita (India) Plumpy Nut®

11. JB/Tanjaka Foods (Madagascar)** Plumpy Nut®

12. Mana Nutritive Aid Products (UnitedStates)

Generic name

*Dairy-based, not peanut-based, RUTF. **The first company located in programmatic country capable of exporting RUTF***Supplier agreed to remove branded name Imunut from the labels

Table 2: UNICEF-approved global RUTF suppliers

Global supplier Product Name

1. STA (Niger) Plumpy Nut®

2. Hilina (Ethiopia) Plumpy Nut®

3. Project Peanut Butter ( Malawi) Plumpy Nut®

4. Valid Nutrition (Malawi) Plumpy Nut®

5. Amwili (DR Congo) Plumpy Nut®

6. JAM (Mozambique) Plumpy Nut®

7. Power Foods (Tanzania) Plumpy Nut®

Table 3: UNICEF-approved local RUTF suppliers

Field Article

News

UNHCR Technical Workshopon the Operational Guidanceon the use of SpecialNutritional Products

Sarah Style is an ENN consultant working onthe UNHCR/ENNAnaemia Control,Prevention andReduction Project.

Andy Seal is a lecturer and researcherat the UCL Centre for InternationalHealth and Development and asenior technical advisor to the ENNon the UNHCR/ENN Anaemia Control,Prevention and Reduction Project.

By Sarah Style and Andrew Seal

In July 2011, UNHCR, in collaboration withENN, held a five-day technical trainingworkshop to introduce the recently finalised

‘Operational Guidance on the Use of SpecialNutritional Products to Reduce MicronutrientDeficiencies and Malnutrition in RefugeePopulations’.

The Operational Guidance (describedfurther in Field Exchange 41) deals with certainfood supplementation products (FSP) (includ-ing micronutrient powders (MNP) and certainlipid-based nutrient supplements (LNS)) thatare currently being used, or being consideredfor use, in blanket supplementary feedingprogrammes within UNHCR operations. Theseproducts are designed for use in programmesaiming to prevent malnutrition, with a particu-lar focus on micronutrient deficiencies. Theguidance has been developed to help countrystaff deal with the challenges and confusion indesigning, implementing, monitoring and eval-uating programmes using these new FSPs.

The workshop was designed to provide bothtraining on the use of the Operational Guidanceand the opportunity for sharing of informationand experiences of using these products inrefugee camps across UNHCR operations. Theoverall aim was to enhance best practice andcontribute to participants being better able tomeet the needs of the refugee populations withwhom they work.

The specific objectives of this technical work-shop were to:a) Provide standardised guidance on best

practice in planning, implementing, monitoring and evaluating programmes using special nutritional products

b) Improve the design, implementation and monitoring & evaluation of programmes using special nutritional products

c) Build the capacity of technical health and nutrition staff from UNHCR and WFP in effectively managing programmes using these products in refugee situations.

Twenty health and nutrition staff fromUNHCR and WFP attended the workshop,which was held in Mombasa, Kenya.Participants represented operations in 12 coun-tries: Algeria, Bangladesh, Chad, Djibouti,Ethiopia, Kenya, Nepal, Rwanda, Sudan,Tanzania, Uganda and Yemen.

Workshop activities and agendaThe workshop followed the general structure ofthe Operational Guidance beginning with Stage1 on the first day (define the nutritional prob-lem) through to Stage 6 (monitoring andevaluation)1. The workshop covered the keyoperational components relating toprogrammes using specific FSP intended forpreventative purposes. See Table 1 for furtherproduct information.

A variety of different learning methods wereincorporated throughout the workshop, includ-ing presentations, case studies, individualexercises, group work, testing of various toolsin the Operational Guidance, and optionalevening sessions. Participants were invited topresent their experiences, learning, and thechallenges on the ground concerningprogrammes using FSPs. At the end of theworkshop, participants were allocated timewith facilitators to discuss any questions orchallenges that they were facing relating totheir country of operation and the way forward,and how to incorporate the principals and bestpractice outlined during the workshop.

Follow upWorkshop evaluation findings (daily feedbackand final evaluation) were positive and construc-tive. The main challenge thatparticipants identified for practice arebudgetary constraints to purchase andmanage products due in part to othercompeting operational needs. Themain actions identified by participantsto improve product related practicewas to review their camp behaviourchange and communication (BCC)strategy and to improve M&E systemsincluding use of the M&E toolsprovided in the Operational Guidance.A clear expectation of the workshopwas that all participants wouldconduct follow up training on theirreturn to help build institutionalknowledge.

Feedback from participants have been incor-porated in the final version of the OperationalGuidance which will shortly be available forpublic dissemination and download viahttp://info.refugee-nutrition.net

For further information, contact CarolineWilkinson at UNHCR, Geneva, email: [email protected]

FSP / FBF Targetagegroup

Product descriptions

Micronutrientpowder

6-59m MNPs provide no energy (kcal) in thediet. They are usually packaged inindividual sachets to provide a doseof selected vitamins and minerals inpowder form, to be added to foodsdirectly after cooking. MNPs havebeen shown to be efficacious intreating and preventing anaemia.Branded products include Sprinkles™and MixMe

Low quantity LNS*

6-24m Highly fortified peanut-based pastethat contains vitamins and mineralsin addition to providing energy. It isusually packaged in individual dailysachets and is to be eaten eitherdirectly from the sachet or added tocomplementary food. It has beenshown to improve linear growth inyoung children. Product brandsinclude Nutributter®2

Mediumquantity LNS

6-36m Highly fortified peanut-based pastethat contains vitamins and mineralsin addition to providing energy. It isusually packaged in individualweekly pots. However it will also beavailable in the form of daily sachets,which is the preferred form for distri-bution. It has been used inprogrammes to prevent increases inGAM in young children during peri-ods of food insecurity. Productbrands include Plumpy’doz®

FBF+ / ++ 6-59m /6-24m

FBF+ e.g. Corn-Soy Blend (CSB) is afood for young children and othervulnerable groups, as well as thegeneral population. Its content ofvitamins and minerals has beenmodified compared to previousformulations. It is recommended as apartial replacement for nutritionallyinadequate local diets.FBF++ is a newly developed FBF forinfants over 6 months and youngchildren. It contains milk powder andhas a higher energy density thanother types of FBF.

Table 1: Summary of Food Supplementation Products(FSP) and Fortified Blended Food (FBF) for use in children aged 6-59 months that are covered by theOperational Guidance

1 See news piece in Field Exchange 41 for

schematic.2 This is the only product currently approved for

use by UNHCR. Products are approved on a

case by case basis. Similarly, only Plumpy’doz

is approved as a medium quantity LNS.

Cre

dit:

/ U

NH

CR /

F. C

ourb

et,

Eth

iopia

,

2008

Eritrean refugees in Shimelbarefugee camp, Ethiopia

48

49

infoasaid: communication in emergencies

When an emergency strikes, be it natu-ral disaster or man-made,organisations working on the

ground, together with the national authorities,rush to respond. Their actions are orientatedtowards getting in supplies, arranging logistics,and ensuring programmes such as Water,Sanitation and Hygiene (WASH), Health,Shelter and Nutrition can deliver on theircommitments. Two-way communication, i.e.the sharing and receiving of information, withdisaster-affected communities should and canbe a cross-cutting part of humanitarianresponse and in itself, a form of aid.

Communication is essential to successfulemergency programming in many ways. It canalert the population about the disaster and raiseawareness on potential threats and thus miti-gate risk. If it is two-way, it can improveprogramming by being more needs-based; ifdone well and effectively, it can improve cover-age overall and at the same time, reach morevulnerable people. It can support the coordina-tion effort by creating the space and the meansfor organisations and communities to work andtalk together. It can also be an important andeffective tool in addressing the psycho-socialneeds of a population who have experienced adisaster first hand.

infoasaid is a 2 year, DFID-funded project thatis being implemented by two media develop-ment organisations - Internews and the BBCWorld Service Trust. The overall goal of the

project is to improve the quality of humanitar-ian responses by maximising the amount ofaccurate and timely information available toboth humanitarian agencies and affected popu-lations through enhanced informationexchange between them in the critical first fewdays and weeks of an emergency.

The project has two main objectives:• To strengthen the capacity and preparedness

of aid agencies to respond to the informationand communication needs of crisis affected populations.

• To partner with a number of aid agencies tohelp inform and support their communica-tions response in a variety of emergency contexts.

In order to achieve objective one, infoasaid isdeveloping a range of tools that will be avail-able to the entire humanitarian community.These include: • Media and Telecoms Landscape Guides for

22 of the world's most disaster and conflict-prone countries. These guides provide a comprehensive picture of the media and telecommunications landscape, information on media consumption patterns and a useful contact directory of media and telecoms oper-ators. The guides are a practical tool for aid agencies to refer to when deciding which channels of communication to use in order to access different populations.

• A library of generic messages and accompa-nying guide. This has been developed in collaboration with a number of different clusters/sectors in humanitarian response, including WASH, Health, Nutrition, Food Security Protection and Education. The messages aim to provide information to affected populations about the scale and nature of the emergency, risks and threats and how to mitigate them and information about programme interventions. The accompanying guide explains the importanceof context, culture, and delivery methods ineach emergency situation. The messages aredesigned to be used as a reference tool and ideally, should be translated, piloted and adapted to suit the local context and to ensure comprehension before an emergencyhas occurred. The message library will soonbe available for use in the form of a web based tool.

• Generic questions on information needs andaccess which need to be adapted and then integrated in agencies’ needs assessment frameworks.

• An interactive e-learning course to raise awareness of, and provide basic skills on, communicating with disaster affected communities. The two hour module takes the learner through a number of different emergency scenarios.

In order to achieve objective two, infoasaid ispartnering with a number of aid agencies tohelp inform and support the integration ofcommunications into their emergencyprogrammes. Partnerships have been signed

with Save the Children, Merlin, World Vision,ActionAid and the International Federation ofthe Red Cross and Red Crescent Societies(IFRC). infoasaid is currently piloting an inter-vention with Action Aid in Kenya around foodsecurity, using two new open source softwareplatforms – Frontline sms and Freedom Fone.Scoping missions are also taking place in northeast Kenya and Puntland with Save theChildren to see how integrating communicationinterventions can improve their emergencyresponse.

infoasaid also provides additional support inthe following areas:• Context analysis (analysis of the media land

scape)• Undertaking information needs and access

assessments• Development of communications plans and

strategies• Development and dissemination of key

messages to affected populations• Linking aid agencies with media develop-

ment organisations, the tech-community and the local media

• Piloting new technology innovations• Design and roll out of small scale pilot

projects• Undertaking learning reviews post-response,

in order to learn lessons and document good practice

infoasaid informs its advocacy efforts through itsfield based experiences and the learninggleaned through the reviews undertaken postresponse. infoasaid hopes to consolidate itslearning in the form of a Humanitarian PracticeNetwork (HPN) Paper that it will develop andlaunch in collaboration with the OverseasDevelopment Institute (ODI) in mid-2012. infoasaid is also providing technical and finan-cial support to the ‘Communicating withDisaster Affected Communities (CDAC)Network’ with the view to strengthen theNetwork and enable it to take forward the workthat infoasaid is currently doing.

For more information on infoasaid visit:www.infoasaid.org or contact Anita Shah, Headof Project, email: [email protected]

News

Info

asa

idIn

foasa

id

Info

asa

id

50

News

Experiences of theNutrition inEmergencies RegionalTraining Initiative

Abigail Perry is a nutritionist withextensive experience in developmentand emergency work. She worked as aResearch Associate at UCL and was thecoordinator and lead trainer for theNutrition in Emergencies Regional

Training Initiative (NIERTI) project between 2009 and2011

Jessica Meeker is a graduate of theLSHTM Public Health Nutritioncourse, coming from a previouscareer in IT consultancy manage-ment. Whilst working as a ResearchAssistant at UCL, she conducted

much of the research on Nutrition in Emergenciescompetencies

The NIERTI project started in the Autumn of 2009 andhas just concluded. The project was funded by anOFDA award to the Emergency Nutrition Network(Agreement No. DFD-G-00-09-00289-00) and imple-mented by UCL-CIHD.

Andrew Seal is a Lecturer inInternational Nutrition at UCL wherehe conducts research and training onvarious aspects of nutrition in emer-gencies. He was the project managerfor the work described here and is

continuing the work with NIERTI partners to takeforward this initiative

By Abigail Perry, Jessica Meeker, and Andrew Seal,UCL Centre for International Health andDevelopment (UCL-CIHD)

The upward trend in the frequencyand magnitude of humanitariandisasters is set to continue and

ongoing economic instability, escalatingfood and fuel prices, climate change andurbanisation are predicted to amplifyhumanitarian needs1 . This projection ofincreased vulnerability has implicationsfor development; losses associated withhumanitarian disasters are thwartingprogress towards meeting developmentgoals2. Central to efforts to protectcommunities and reduce vulnerability isthe development of capacity to preparefor and respond to crises effectively3.The 2005 Humanitarian ResponseReview noted that the lack of capacity inthe humanitarian sector hampers thequality and appropriateness of emer-gency response. The review provokedmore detailed examination of humani-tarian capacity, including a 2007 reportfocusing on ‘nutrition in emergencies’(NIE)4. This review, completed on behalfof the IASC Global Nutrition Cluster(GNC), identified two key NIE capacitygaps: (i) inadequate ‘resident’ capacityand (ii) lack of nutritionists able tohandle the demands of emergencies.

A number of projects have sinceemerged that have attempted to addressthis gap. In 2008, a standard set of tech-nical documents, known as theHarmonised Training Package (HTP)were developed5. More recently, theGNC has initiated a capacity develop-ment project for cluster coordinators. Inaddition, two projects have been devel-oped with support from the Office forDisaster Assistance (OFDA) to theENN6; one focusing on in-service andpre-service training in NIE (imple-mented by NutritionWorks) and theother on the development of profes-sional short courses. In this report wediscuss the key findings from the secondof these projects, the Nutrition inEmergencies Regional TrainingInitiative (NIERTI).

NIERTI objectivesThe overall goal of the NIERTI was toincrease the availability of regular, highquality training in emergency nutrition.Two principles underpinned the devel-opment of the training model. First, weviewed capacity development in thebroadest sense and thus as a process thatrequires investment in three levels ofcapacity: (i) the individual, (ii) theorganisation and (iii) the enabling envi-ronment7. Keeping this in mind, weelected to focus on strengthening thecapacity of relatively senior national andinternational individual practitioners.This group represents the ‘tip’ of thehuman resource capacity pyramid.Although investment is needed in NIEcapacity at every level, building thiscadre will improve organisationalcapacity and create an enabling environ-ment for effective capacity developmentamong the other cadres. Second, thetraining would be implemented in part-nership with academic institutions. Ouraim was to develop a model that has

financial and market sustainability andthat can be owned and maintained bypartner institutions, rather than beingdependent on (unpredictable) externalfunding. The three partner institutionswere the American University of Beirutin Lebanon, the University of Makererein Uganda and the Asian DisasterPreparedness Centre in Thailand. Allthree institutions were already runningnutrition courses as well as the PublicHealth in Complex Emergencies shortcourse8. The NIERTI sought to work inpartnership and strengthen and developtheir training capacity in NIE.

Development of the trainingmodelCostFinancial sustainability was identifiedfrom the outset as a key factor in thesuccess of this initiative. As such, weworked closely with our partners todevelop realistic budgets that ensuredall the costs associated with running thecourses would be covered by course feesalone. The fee level for each course wasbased on estimated cost and antici-pated participant numbers. Thebudgets were reviewed and revisedfollowing each course to ensure that onone hand, course fees are kept at aminimum and access is maximised, andon the other, that the courses are finan-cially sustainable.

ContentThree courses of varying duration (6, 10and 12 days) were developed and pilottested. This process enabled us to deter-mine the optimal duration and balanceof topics. A set of learning objectiveswas developed for each course module.This helped to shape the content and toselect appropriate course exercises. Thetechnical content was based on the HTP,supplemented where necessary withexercises developed specifically for theinitiative. Each module consisted of atechnical reference document (compiledin a course manual for participants),generic PowerPoint presentations andpractical exercises.

Previous reviews of NIE capacitydevelopment and interviews with NIEpractitioners highlighted the impor-tance of including activities that enableparticipants to apply knowledge. As a

1 UN General Assembly Economic and Social

Council (2011) Strengthening the coordinaton of

emergency humanitarian assistance of the United

Nations. Report of the Secretary General,

Geneva, July.2 UNDP (2004). Reducing disaster risk: a challenge

for development. UNDP, New York. 3 Capacity for Disaster Risk Reduction Initiative

(2011) Basics of capacity development for

disaster risk reduction. 4 Gostelow L (2007). Capacity development for

nutrition in emergencies: beginning to synthesise

experiences and insights. NutritionWorks / IASC

Global Nutrition Cluster.5 NutritionWorks, ENN, GNC (2011). Harmonised

Training Package revised modules, v2.

http://www.ennonline.net/htpversion26 See ENNs 5 year strategy that reflects capacity

development on NiE as one of ENNs priority

activities, http://www.ennonline.net/pool/files/

reports/enn-5-year-strategy-2010-to-2015-final.pdf7 See footnotes 3 and 48 http://www.phcetraining.org/

Part of the group work inthe Uganda training

A P

err

y/N

IETN

, U

ganda,

2011

Part of the field work inthe Uganda training

A P

err

y/N

IETN

, U

ganda,

2011

51

result, a 1-day emergency simulation exercise was developedin consultation with a range of NIE experts. The simulationwas reviewed and revised following each course. Field-basedexercises relating to the management of acute malnutritionand nutrition surveys were adapted from the HTP and testedduring the course in Uganda.

EvaluationModules were delivered by experienced facilitators and feed-back was obtained using a variety of evaluation approaches.All materials were piloted at least twice and reviewed andrevised from one course to the next based on comments fromcourse participants, our partners and course facilitators.Module-specific and end-of-course evaluation forms wereadministered to participants. Feedback from each partnerorganisation and the course facilitators was obtained verballyand via email. Follow-up of course participants was doneapproximately 3 months after each course ended.

AssessmentA multiple-choice questionnaire was developed and tested toassess the progress of participants. Each of the 40 questionswas designed to correspond to one of the three levels of learn-ing, based on a modified version of Bloom’s taxonomy(knowledge, combined comprehension and application, andproblem solving)9. The MCQ was pilot tested before beingadministered to course participants. The test was given at thestart and end of the course and scores were reviewed andanalysed by participant and topic.

ResultsCourse uptakeOverall, 131 applications were received for the three coursesrun during the pilot phase. The first course (held in Lebanon)was restricted to practitioners from the Middle East and only19 applications were submitted. Fifty-six applications werereceived for each of the two other courses (held in Ugandaand Thailand).

A total of 67 people attended the courses. The majority ofparticipants worked either for an international non-govern-metnal organisation (INGO) (28%) or a UN agency (34%). Therest worked for government ministries (9%), academic insti-tutions (8%), local NGOs (6%), donor organszations (6%) orwere self-employed (9%). Sixty-six per cent of participantsworked in nutrition, 15% in health, 9% in general emergencyprogramme management and 8% in food security. One partic-ipant worked in logistics.

One third (33%) of participants were working in WesternAsia (including Middle Eastern countries), 18% in NorthernAfrica (predominately Sudan), 13% in Eastern Africa, 12% inSouthern Asia, and 6% each in Western Africa, South-eastAsia and Australia. The rest had a regional/global remit.Overall, 73% of participants were national staff (i.e. workingin the country where they are from).

An overview of each of the coursesrun during the pilot phase of the NIERTIis given in Table 1.

CostThe first of the three courses (held inLebanon) was subsidised since it was thefirst ‘pilot’ course. The fee was US$500for 6 days (excluding accommodation)and the cost of course facilitators wascovered by the project funding. Thiscourse has not been included in thefollowing analyses because it is notrepresentative of the actual costs ofrunning NIE training. The next twocourses had a fee of US$ 2,500 includingmeals and accommodation. The coursein Uganda was 12 days and the Thailandcourse lasted 10 days. This equates toUS$208 per person per day for Ugandaand US$250 per person per day forThailand. As a comparison, the 5-daycourse in NIE run by the University ofWestminster (London, UK) costs theequivalent of US$ 1,100 (based oncurrent exchange rates), excludingaccommodation and meals other thanlunch. This equates to US$ 220 per dayor US$ 2,640 for 12 days.

Of the 48 people who attended thefull-price courses in Uganda andThailand, 83.3% were funded by theiremployer, 10.4% were sponsored byanother organisation and 6.3% were selffunded. Feedback from participantsindicated that the course was viewed asgood value for money. Of the 222enquires that had been received aboutthe initiative by the end of August 2011,23 were requests for funding support.None of these individuals subsequently

attended a course. Sixteen people whowere offered a place on a course reportedthat they were unable to attend becauseof a lack of funding.

The most expensive components ofthe courses were the facilitators followedby accommodation/meals for partici-pants. Expenditure on facilitators wasUS$ 1,227 per participant in Uganda andUS$ 661 per participant in Thailand. Inboth cases this does not include a fee forthe lead facilitator or for an additionalfacilitator who taught several modules,both of whom work for UCL CIHD. Thecost per participant in Thailand waslower because we had more participantsand were able to secure facilitators whowere based locally and who were not ina position to accept a fee (which cannotalways be guaranteed). Even so, thecourse in Thailand only just covered allcosts using fees and in Uganda weincurred a loss that was buffered usingproject funding. The cost of accommoda-tion, meals and the training venue was$72 per participant per day in Ugandaand $110 per participant per day inThailand.

EvaluationThe response to each course was posi-tive. Overall, 56% of participantsstrongly agreed and 42% agreed that thecourse they attended met the goal toprepare them to respond to the nutri-tional needs of people affected byemergencies. The improvement in

News

Location Lebanon Uganda Thailand

Partner American University of Beirut University of Makerere Asian Disaster PreparednessCentre

Duration 6 days 12 days 10 days

Number ofparticipants

17 (+2 students) 20 28

Modulescovered

Introduction to NIE Introduction to NIE Introduction to NIE

Causes of malnutrition Causes of malnutrition Causes of malnutrition

Humanitarian system Humanitarian system Humanitarian system

Humanitarian standards and M&E Humanitarian standards and M&E Humanitarian standards and M&E

Rapid assessments Rapid assessments Rapid assessments

Nutrition surveys Nutrition surveys Nutrition surveys

- Surveillance and early warning Surveillance and early warning

- Food security and livelihoods Food security and livelihoods

Food assistance Food assistance Food assistance

Micronutrient interventions Micronutrient interventions Micronutrient interventions

- Advocacy Advocacy

Management of severe acutemalnutrition

Management of severe acutemalnutrition

Management of severe acutemalnutrition

Management of moderate acutemalnutrition

Management of moderate acutemalnutrition

Management of moderate acutemalnutrition

Infant feeding in emergencies Infant feeding in emergencies Infant feeding in emergencies

Emergency preparedness Emergency preparedness Emergency preparedness

Emergency simulation Emergency simulation Emergency simulation

Professional development Professional development Professional development

- Pre-/post-test MCQ Pre-/post-test MCQ

- Field training (nutrition surveys) -

- Field training (management ofSAM)

-

- Field training (management ofMAM)

-

Table 1: Overview of the NIERTI courses run to date

9 Anderson LW et al. (eds.) (2001). A Taxonomy for

Learning, Teaching, and Assessing: A Revision of

Bloom's Taxonomy of Educational Objectives. New

York: Longman.

Simulation exercise: Inter-agencycoordination, Thailand course

A P

err

y/N

IETN

, U

ganda,

2011

52

News

facilities in Thailand were excellent value formoney and it is difficult to achieve similar at alower cost.

The other main contributor to the cost of thecourses is also challenging to solve. The mainreason why NIE training is expensive isbecause of the reliance on relatively few NIEpractitioners who need to be flown in to teachon courses. We were not able to run the NIEcourses using only staff from the partner organ-izations. None of the available facilitators hadexperience of working in emergency contextsand all felt that practical experience was anessential prerequisite. As such, external facilita-tors had to be employed. Significant effort wentin to identifying experienced facilitators basedin each region. However, the NIE community isstill relatively small, thework commitments ofpotential facilitators areunpredictable, experi-enced staff are in highdemand and expect to bepaid a fee that is commen-surate with theirexperience, and relativelyfew NIE practitioners havesufficient teaching experi-ence. The latter point isparticularly important;although course partici-pants appreciated theextensive experience of thefacilitators while the abil-ity to transmit informationand to manage teachingsessions effectively wasalso valued highly.

The most obvious solu-tion to this is to build thecapacity of national practitioners and academicstaff. This approach would reduce costs andcould promote sustainability. Development ofthe capacity of academic staff will require expo-sure to emergency nutrition programmes. Arecent piece of research looking at pre-servicetraining in NIE found that academic staff can beisolated from the work of NGOs and UN agen-cies10. This gap needs to be bridged in order toenable academics to gain hands-on experience.Training in the sector is still more commonlyundertaken by agencies rather than highereducation institutions. Although this approachcan improve capacity within the humanitariansector, it has the potential to further weaken theacademic system that could provide a sustain-able approach to strengthening capacity in thelonger term (8).

The other constraint to strengtheningnational academic capacity is the continuingissue of low salaries, high staff attrition andconflicting demands among academic staff indeveloping countries11. In order to establishacademic institutions as providers of high qual-ity nutrition training, this problem will need tobe addressed. This might also help to attractexperienced NIE practitioners in to theacademic sector.

Providing practical training in NIETwo key themes emerged from the evaluationsand course participant follow-up. First, partici-pants emphasised the importance of theproblem-solving course exercises, field-basedtraining and the emergency simulation forenabling them to put into practice the theory

One solution to this is to include classroom-based teaching as part of an integratedwork-based training programme that includesfield exposure. In 2010, the Consortium ofBritish Humanitarian Agencies (CBHA)launched a competency-based trainingprogramme for humanitarian staff, whichincludes field placements. More recently, Savethe Children UK and Concern Worldwide initi-ated 1 and 2-year emergency nutritioninternships, which include placements at headquarters as well as at field sites. Another optionwould be to develop partnerships betweenacademic institutions that can provide theorytraining and humanitarian agencies operatingin relevant countries. Practitioners could attendtraining and then opt for a shorter and moreaffordable work placement in their country oforigin. This arrangement would also help tobridge the gap between the academic and oper-ational agencies in these countries. Forpractitioners already working within thehumanitarian system, a relevant work place-ment could be one of the pre-requisites forattending the theory training.

Standardisation of training and the need forprofessional competenciesA recurrent issue that emerged during thedevelopment of the NIERTI was the lack ofcommon understanding of the competenciesrequired by NIE staff. Although experiencedpractitioners have a reasonable sense of what

performance and ability of participants is diffi-cult to quantify; however there were some signsthat the courses had a positive impact. Theaverage score for the MCQ increased from52.0% at the start of the course in Uganda to63.5% at the end. In Thailand the average scoreincreased from 52.5% to 65.5%. Only 29 of the67 participants responded to the follow-upemail. Among these 29, there were numerousexamples of positive actions resulting fromattendance on the courses. The majority ofresponders had provided training to their ownstaff or to others. A group from Lebanon haddeveloped a country-specific emergencypreparedness plan for nutrition. Others hadbeen identifying ways to strengthen themanagement of moderate acute malnutrition aspart of CMAM.

Based on feedback from the course evalua-tions, an 11-day ‘standard’ NIE course has nowbeen developed and this will be the model thatwill be implemented by each of the trainingpartners. This course includes the emergencysimulation exercise but the field-based trainingis designed as an optional add-on to the stan-dard course, depending on the opportunitiesavailable and the logistic feasibility at eachtraining site.

DiscussionOn one level, the overall goal of the NIERTI toincrease the availability of high quality trainingin NIE has been met. There are also indicationsthat the training model will be sustained by ourpartner organisations. The courses have gaineda positive reputation among the NIE commu-nity and, at the time of writing, all threeinstitutions have taken steps to implement theirnext course. Each institution has started toinvestigate ways to strengthen their own capac-ity and to institutionalise the training. Thisincludes engaging regional practitioners todevelop a network of known NIE facilitatorsand incorporating the course (or elements of thecourse) into post-graduate training. There alsoappears to be a reasonably steady market forthe courses. Despite increased focus on thedevelopment of NIE capacity, we are onlyaware of one other specific course in NIE that isrun on a regular basis. Other training initiativesthat have emerged over the past few years arestill predominately ad-hoc or one-off eventshosted by international agencies. In light of theextent of the capacity gap in nutrition and thehigh turnover of staff within the humanitariansector, investment in occasional training, whilebeing very useful, will not be sufficient.

The process of setting up the NIERTI hashighlighted a number of issues that will need tobe considered if we are to make furtherprogress in addressing the lack of capacity inthe sector. These have been grouped into threemajor themes: the cost of capacity develop-ment, providing practical training in NIE andstandardisation of training.

The cost of capacity developmentThe NIERTI course fee is undoubtedly prohibi-tively high for some individuals and agencies.However, the fees provide an indication, andperhaps an under-estimation of the real cost ofrunning a course of this nature in the currentclimate. The cost of accommodation/mealsseems quite high but in fact it is extremely diffi-cult to find alternatives at a lower cost inUganda, other than in hostel-type settings. The 10 NutritionWorks (2011). In-service and pre-service training

in NIE. http://www.ennonline.net/meetings/servicetraining

covered. Second, participants felt that post-course internships, work placements andmentoring would have enabled them to consol-idate their learning further. The role of practicaltraining in NIE was emphasized in the 2007capacity review by Gostelow. The field-basedtraining included in the course held in Ugandawas well received but this is a difficult activityto maintain. Running field training in an emer-gency programme can be disruptive and thenature of emergency response does mean thatfrom one year to the next it is difficult to guaran-tee that relevant programmes exist or areaccessible in the country/region where trainingis being held. Field training also lengthens theduration of courses, which increases both thetime and financial commitment for participants.

Training session on infant and young childfeeding in emergencies, Beirut, 2010

A P

err

y/N

IETN

, Beirut,

2010

53

News

In order to evaluate the quality and impact ofnutrition interventions, agencies need effec-tive monitoring and reporting systems in

place for programmes. A retrospective analysisof emergency supplementary feedingprogrammes (SFPs) published in 2008 by theEmergency Nutrition Network (ENN)1 high-lighted inadequate reporting standards andraised concerns over the quality of the inter-ventions themselves. Based on the study’srecommendation and funded by the US Officefor Disaster Assistance (OFDA), the ENNtogether with Save the Children UK (SC UK), ledthe development of a monitoring and report-ing guidelines and software, guided by asteering group of interested agencies. ThisMinimum Reporting Package (MRP) for SFPswas subsequently piloted in four countries2. Anew phase of rollout has now begun, fundedby ECHO and led by SC UK. Initiated in thesecond half of 2011, this 18 month project willsee MRP rollout with approximately ten ECHOimplementing partners and involving up to 30SFPs. An additional module to integrate report-ing for severe acute malnutrition (SAM)programmes is also being developed to facili-tate reporting where both SFP and SAMtreatment are managed under one interven-tion.

The overall goal of the MRP is to improvemonitoring of the performance of SFPs imple-mented by NGOs and WFP (and WFP partnerswhere relevant), to enhance programmemanagement and agency accountability.Given the different capacities of implementingagencies and contexts within which they work,the MRP has been configured for two levels ofusage: essential reporting (adhering to SphereStandards) for basic users and optionaladvanced reporting for agencies requiringmore information. The MRP also allows report-ing on programme characteristics (e.g. specificcountry/area/ programme context) and popu-lation-based statistics (e.g. coverage andmalnutrition rates) to aid interpretation ofprogramme performance indicators. Fieldimplementation can be tailored to each organ-isation’s capacity, adapting to context-specificchallenges in communication and staff train-ing, including allowing for MRP use in systemsthat are mainly paper-based at field level.

A further goal of the ECHO MRP project is toimprove understanding of the overall efficacyand effectiveness of SFPs through conductinga final analysis of the performance data of allprogrammes utilising the package during theproject period.

MinimumReportingPackage (MRP) onSupplementaryFeedingProgrammes

NIE personnel need to know and be able to do,there is a great deal of subjectivity in this. As aresult, the NIE courses that exist (includingthose provided by the NIERTI) feature differentcombinations of topics, are of varying durationand involve either no assessment or assess-ments that tend to measure knowledge gainedrather than the ability of participants to func-tion effectively. This makes it extremelydifficult to ascertain whether participants ofNIE courses have the necessary competencies,and whether the courses themselves areproviding effective training.

One way to address this would be to adopt amore systematic, competency-based approach.Other sectors in the humanitarian communityhave recently moved towards competency-based training, assessment and recruitment. Aset of core humanitarian competencies is nowbeing used by the CBHA and the child protec-tion and logistics sectors use standardcompetency frameworks to measure the abilityand performance of staff.

Preliminary work on developing a competencyframework for the NIE sectorAs an addition to the NIERTI, we have takeninitial steps to developing a competency frame-work for the sector using the Child Protectionin Emergencies (CPIE) framework as a guide.The CPIE competency framework is now usedas the basis for recruitment, training and staffdevelopment by a number of humanitarianagencies. The benefits of this are perceived to bevery positive; one staff member commentedthat it had improved recruitment processes andtraining and has increased staff motivation andcareer progression (K. Bisaro, personal commu-nication).

Research into the identification of competen-cies for NIE was designed to be ascomprehensive as possible, and was conductedby reviewing existing competency frameworks,NIE-related job descriptions, academic andtraining course content, and by conductinginterviews with key stakeholders and NIERTIcourse participants.

The competencies regarded as essential foran international emergency nutritionist havebeen extracted and compiled into a basic frame-work, which we will be publishing shortly. Toincrease the usability of the framework, specific

behavioural indicators have been developed foreach competency at three progressive levels.These facilitate its use for creating job profiles,conducting assessment during recruitment andtraining and for staff to identify areas forprofessional development.

We now have the foundations for a compe-tency framework for NIE. However, there arestill a number of actions required before thiscould serve as a functional tool.1. More detailed mapping of career paths and

job profiling is required to clarify which competencies are required at each level

2. The framework needs to be reviewed by thesector to identify gaps and to agree each of the indicators.

3. Finally, the framework will need to be approved and adopted by GNC member agencies.

What next for the NIERTI?The NIERTI will continue to support partnerorganisations to run NIE training courses basedon the materials developed and tested duringthe pilot phase. We would also like to incorpo-rate competency-based training and assessmentas part of broader efforts to professionalise thehumanitarian sector. Having a standard frame-work for measuring the competencies attainedby NIE practitioners would help to improvequality and performance in the sector. In themeantime, we are keen to maintain the qualityof training provided under the NIERTIumbrella, which will in turn ensure continueddemand and financial sustainability. This willprovide the opportunity for our partner organ-isations to continue to strengthen their owncapacity to provide NIE training. Any deterio-ration in the reputation of NIERTI courseswould have a negative impact on existingcourse providers and would jeopardise thefuture of the initiative. To try and ensure thatquality is maintained, we will continue tocontrol the use of NIERTI training materials.Any organisation that wishes to run a coursebased on these materials will be free to do so, aslong as they can guarantee they will be deliv-ered to a high standard.

For further information, and to request accessto the materials, please email: [email protected].

A P

err

y/N

IETN

, U

ganda,

2011

Simulation exercise: Clustercoordination, Thailand course

54

Specific MRP objectives are:• to provide guidance for reporting

monitoring information from SFPs by providing standardised criteria and a standardised reporting system

• to facilitate the process of reporting by providing standard data collection tools and a user-friendly database

• to promote quality improvement and lessons learning in SFPs through real-time reporting of programme statisticsand data comparisons within agency programmes, as well as across agenciesas appropriate

The MRP includes:• MRP guidelines with standard

definitions of variables and indicators (includes SAM for practical use by agencies required to report on both SAM and moderate acute malnutrition(MAM))

• Software/database (the ‘eMRP’) • eMRP database user manual (step-by-

step guide)• Data collection forms

Key expected benefits of the MRP are• Increased timely monitoring and

reporting capacity of implementing partners and thus higher quality performance and impact

• Enhanced accountability to beneficiariesand donors

• Facilitation of programme supervisionand cross programme/agency comparisons

• Lessons learned through an end-of-project analysis

The MRP project will continue to besupported throughout 2012 by a team offour SCUK staff along with advisors andconsultants. Moving forward, the teamwill provide training, mentoring, and real-time support as needed to agencies usingthe MRP, including a helpdesk andmonthly report feedback

Although this 18-month project neces-sarily has short-term goals, the MRP couldhave longer term, more sustainable appli-cations. It is foreseen that MRP use andapplication by agencies will continueafter the ECHO project ends in December2012.

For further information, please contactJennifer Martin, email: [email protected].

1 Measuring the Effectiveness of Supplementary

Feeding Programmes in Emergencies, Carlos

Navarro-Colorado, Frances Mason and Jeremy

Shoham, Humanitarian Practice Network Paper

63, September 2008. ODI2 Download report at: http://www.ennonline.net/

pool/files/research/mrp-report-final.pdf

We are pleased to include with this editionof Field Exchange, a complementarycopy of the ‘MOYO chart’. Named after

the MOYO1 Nutrition Ward, Blantyre, Malawiwhere it was originally developed, this is a lowcost job aid that helps health workers correctlyassess and interpret a child’s weight-for-height.

Background to developmentField Exchange readers will be familiar with thechallenges that inspired the chart: a busy nutritioncentre, overworked but often underexperiencedfront line staff, the need to ensure consistent, highquality patient assessment and referral to appro-priate treatment. In such settings, optimising eachstep of the patient care pathway is critical.

The MOYO chart came about as we realised thattraditional weight-for-height lookup charts weresometimes part of the problem rather than part ofthe solution. Adapting and repackaging them intoa more user-friendly slide chart format seemed tomake a difference2. Following some further refine-ment, we went on to formally test our new designin a randomised controlled trial in Ethiopia. Thisshowed that the MOYO chart was preferred overtraditional charts and enabled significantly(p=0.011) more accurate assessment of nutrionalstatus3. In settings where weight-for-height is stillused, it acheives these benefits by guiding health-caare workers through the process of:• determining whether to measure standing

height or recumbent length• rounding the length/height measurement

appropriately• allowing easy identification of weight-for-

height z-score and correct diagnosis of SAM (Severe Acute Malnutrition) or MAM (Moderate Acute Malnutrition).

• determining an appropriate target weight for discharge (either -2 WHZ or -1 WHZ).

Insert in Field Exchange 42Two versions of the MOYO chart are currentlyavailable, both using 2006 WHO growth standards:

i) Boy/Girl split sex chart (as recommended by WHO4 and included with Field Exchange 42)

ii) Joint sex chart (responding to field demand butawaiting formal testing)

Plentiful white space is available on the chart sothat bulk buyers have scope to customise thechart to include local protocols or other locallyimportant text/graphics.

AvailabiltyFollowing this initial distribution via FieldExchange - made possible thanks to a grant fromUCL (University College London) Futures Fund -further copies of the MOYO chart are available tobuy via the health education charity Teaching Aidsat Low Cost (TALC). Unit costs will depend onorder size. One hundred per cent of profits fromsales of the chart are retained by TALC to supportits wider educational objectives. Chart customisa-tion (e.g. with local protocols) is available onrequest for bulk buyers – contact TALC to discussyour needs: web: www.talcuk.org, email:[email protected], tel (UK): +44(0) 1727 853 869

We hope you find the chart useful and thank themany people whose comments and suggestionshave been critical to its development. Any furtherfeedback to help with future versions is alwaysvery welcome: contact Marko Kerac, email:[email protected] or Andy Seal, email:[email protected]

1 ’Life’/’Health’ in local language, Chichewa2 Kerac, M, A. Seal, H. Blencowe, and J. Bunn. Improved

assessment of child nutritional status using target weights

and a novel, low-cost, weight-for-height slide chart. Trop

Doct, 2009. 39(1): p. 23-6.3 Sikorski, C., M. Kerac, M. Fikremariam, and A. Seal, Preliminary

evaluation of the Moyo chart—a novel, low-cost, weight-for-

height slide chart for the improved assessment of nutritional

status in children. Transactions of the Royal Society of

Tropical Medicine and Hygiene, 2010. 104(11): p. 743-7454 WHO child growth standards and the identification of severe

acute malnutrition in infants and children. A joint statement by

the World Health Organization and the United Nations

Children's Fund. May. 2009 (accessed 19 Sept 2010)];

Available from: http://www.who.int/nutrition/publications/

severe malnutrition/9789241598163/en/index.html.

By Marko Kerac and Andrew Seal, UCL Centre forInternational Health & Development, UK

Improving patient assessment: The ‘MOYO’Weight-for-Height Chart

E-learning course on Social Safety Nets

The UN Food and Agricultural Organisation(FAO) and the World Bank have just releaseda new e-learning course on Social Safety

Nets to meet decision makers’ need for under-standing the role safety nets play in reducingpoverty and building food security.

The course is aimed at decision makers whomay not be technical experts, but need to under-stand the best options for implementing socialsafety nets in their specific context. The coursehighlights key issues to be considered for makingthe right decisions. It is also useful for anyonewishing to gain a solid overview of Social SafetyNets. Case studies provide examples of good

practices. Nutrition is referred to in a few exam-ples and case studies.

The course is available for free at: http://www.foodsec.org

This website also provides access to many otherresources provided by the EC-FAO Programme onLinking Information and Decision Making toImprove Food Security. These include standards,tools and methodologies, e-learning courses andtraining materials, and food security country briefs.The programme is based at the FAO and funded bythe European Union’s Food Security ThematicProgramme.

AConsortium of universities offering the EDAMUS Masters degree on‘Sustainable Management of Food Quality’ has been granted theErasmus Mundus label by the European Union. This means a number of

attractive scholarships are now available to students coming from both Europeanand non-European countries. The programme is coordinated by the University ofMontpellier 1 (France), in collaboration with a large international partnership ofacademic institutions.

Students from developing countries can apply for a 2-year scholarship to theMasters course (24,000 euros per year). Within the second year of this MasterCourse, there is a specialisation in ‘Nutrition in Developing Countries’.

There are 10 scholarships offered in 2012 for the course beginning inSeptember 2012. It is anticipated there should be 10 scholarships in 2013 and eachyear following, for 5 years.

The deadline for applications is January 30th, 2012 (receiving date) forstudents demanding an Erasmus Mundus scholarship and April 1st, 2012 for theothers.

For further information, visit: http://www.master-edamus.eu/

En-net updateBy Tamsin Walters, en-netmoderator

55

News

Attractive scholarshipfor EDAMUS Mastersprogramme

Government of SudanCMAM Training Course onInpatient Management ofSevere Acute Malnutrition:Training Materials (2011)

In 1999, the World Health Organisation (WHO) published the Management ofsevere malnutrition: A manual for physicians and other senior health workersand in 2002 the Training course on the management of severe malnutrition.

FANTA-2, in collaboration with national partners in Sudan, adapted and built onthe WHO documents, the 2009 Government of Sudan Community-BasedManagement of Severe Acute Malnutrition manual, and other materials todevelop training materials for inpatient management of SAM designed for physi-cians, nurses, and nutritionists in hospitals in Sudan. While the training coursefocuses on inpatient care, the training materials are compatible with theCommunity-Based Management of Acute Malnutrition (CMAM) approach andthe Sudan context.Course materials include a set of training modules, three training guides, a set ofjob aids, forms, and checklists, a book of photographs and several videos used inthe training, related slide presentations and several documents to help with plan-ning and additional reading. All files can be downloaded from:http://www.fantaproject.org/publications/sudan_CMAM_IC_2011.shtml.

Support for development of the training materials was provided by the UnitedStates Agency for International Development (USAID) Bureau for Democracy,Conflict, and Humanitarian Assistance, Office of U.S. Foreign DisasterAssistance, and the Bureau for Global Health, Office of Health, InfectiousDiseases, and Nutrition.

Twenty-seven questions were posted on en-net in thethree months up to Christmas, eliciting 102 replies. Inaddition, 17 job vacancies have been posted.

Recent discussions have included: how to deal with unso-licited donations of breastmilk substitutes (BMS) during theHorn of Africa crisis, measuring undernutrition and vulnerabil-ity in older people, the use of different cut-offs for assessingundernutrition in different contexts and countries, how toreport uncured registered cases at the end of a community-based management of acute malnutrition (CMAM)programme within indicators of programme performance, andtarget weight setting for HIV-positive moderately acutelymalnourished children in outpatient therapeutic care.

Two recent questions have sought evidence and guidanceon how mid upper arm circumference (MUAC) changes duringthe treatment of acute malnutrition. These build on previousdiscussions concerning appropriate exit criteria for CMAMprogrammes that predominantly use MUAC for admission. Arecent study, published in the Nutrition Journal, wasconducted in Kenya to quantify MUAC changes among dehy-drated children some of whom were malnourished,http://www.nutritionj.com/content/10/1/92. The authorswere able to quantify that a one percent (1%) change inweight, was associated with a 0.40 mm change in MUAC.

Examination of data from programme record cards fromCommunity therapeutic care (CTC) programmes during theCTC research programme and in the initial roll-out of CTC /CMAM show a range of median MUAC gains (0.25 mm per dayto 0.50 mm per day) . However, these are averages of averagesover an entire treatment episode. The rate of change is notconsistent over the entire treatment episode and an uncompli-cated case of severe acute malnutrition (SAM), typically followsa growth curve with rapid gains early in the treatment episode.

A MUAC study currently being undertaken in Malawi byValid International is also looking at MUAC and weightchanges during treatment of SAM, with results expected laterthis year. A body of data is building around these issues thatcould lead to improved future guidance for programmemanagement.

To join the discussion, go to http://www.en-net.org.uk/ques-tion/602.aspx.

In the Cross-cutting Issues forum area a call for contributionshas been posted for the recently launched ALNAP survey toinform the upcoming ‘State of the Humanitarian System’Report. A link to the survey can be found at http://www.en-net.org.uk/question/604.aspx. Don’t miss this chance toinclude your views and ensure that the voices of the emer-gency nutrition community are well-represented. The surveyonly takes 7-10 minutes to complete.

Two new forum areas have recently been launched on en-net:Coverage Assessment, http://www.en-net.org.uk/forum/16.aspx (see news piece in this issue of Field Exchange), andUpcoming Trainings, http://www.en-net.org.uk/forum/15.aspx. ENN has received several requests for training onnutrition in emergencies, particularly from within Africa andwe hope this forum area will provide a useful place for adver-tising courses and bringing people together with an interest indeveloping capacity to respond.

To join a discussion and share your experience or to post aquestion, visit www.en-net.org.uk

1 Mwangome, MK, Fegan G, Prentice AM and Berkley JA. Are diagnostic

criteria for acute malnutrition affected by hydration status in hospitalized

children? A repeated measures study Nutrition Journal 2011, 10:922 Mark Myatt analyses

56

News

UNHCR standardised nutritionsurvey guidelines and training

Melody Tondeur is anENN Consultant and oneof the team membersimplementing theAnaemia Control,Prevention and

Reduction Project, a collaborative workbetween the UNHCR and ENN.

Andrew Seal is Lecturerat UCL Centre forInternational Healthand Nutrition and asenior technical advisorto the ENN on the

UNHCR/ENN Anaemia Control,Prevention and Reduction Project.

By Melody Tondeur and Andrew Seal

Nutritional outcomes continue to be ofconcern in most refugee contexts.UNHCR recommends the measure-

ment of anthropometric status, anaemia andother associated indicators (including water,sanitation and hygiene (WASH) and mosquitonet coverage) on an annual basis in order tomonitor situations and to react in a timelymanner to any deterioration.

A 2009 review of UNHCR nutrition surveysworldwide highlighted a lack of standardisa-tion in methodologies, in the type ofinformation collected and in data presentation.

In order to measure trends over time in thesame region, as well as across regions andpopulations, surveys need to be replicable andthe same definitions, geographic boundaries,and methodologies need to be used from yearto year. To help overcome these challenges,UNHCR encourages the adoption of the inter-nationally recognised SMART1 methods forsurvey design and anthropometric measure-ments.

However, surveys conducted by UNHCRand its partners typically collect data on a muchwider set of indicators to allow for the monitor-

Training on survey guidelines

UNHCR and ENN in collaboration with the Centres forDisease Control and Prevention (CDC) implementedtwo 8-day long regional training workshops in Kenya(Naivasha, October 2010) and Hungary (Budapest,UNHCR training centre, May 2011) for key nutritiontechnical staff and UNHCR’s implementing partners.

The workshop objectives were to:• Improve the quality and reliability of nutrition

survey data collected in refugee operations.• Establish standardised data collection, analysis and

reporting for refugee operations.• Build capacity of technical staff from UNHCR and

implementing partners for conducting standardised nutritional surveys.

The expected training outcomes were that partici-pants know how to use the SMART tools appropriatelyand how to assess anaemia, WASH and mosquito netcoverage indicators, are able to use standard methodsto implement nutritional surveys, gather reliable data,and analyse and report on results, and can identifyareas of improvement for nutritional survey imple-mentation among refugee populations.

Candidates were prioritised based on criteria using aquestionnaire developed by Action Contre la FaimCanada (ACF-Ca) and adapted by ENN and UNHCRthat included experience in nutrition surveys, training,data collation, report writing and working in refugeepopulations. A total of 29 participants from 13 coun-tries from UNHCR offices, six non-governmentalorganisations (NGOs) and one government agencywere selected.

The training covered SMART (SMART StandardisedTraining Package), Epi Info (and the ENA/Epi Info hybridsoftware) and the UNHCR Standardised NutritionSurvey Guidelines (SNSG). Evening sessions wereoffered on a daily basis and covered the day’s topics.

Participants were evaluated on the SMART compo-nent of the training only, using the SMART

Standardised Training Package tests. The post-test wasmore complex than the pre-test (more information onthe SMART methodology concepts and the ENA soft-ware) that limits interpretation. Results suggest theSMART knowledge of the group either stayed at asimilar level or increased slightly. A formal evaluationof the rest of the training was not conducted. Insteadinformal evaluations were conducted at variouspoints during the training. Daily evaluationscompleted by participants helped to refine theagenda and review the training experience. Dailyquizzes helped to assess individual and group lessonlearning and knowledge retention, and allowedparticipants to identify areas of weakness to work on.

Based on daily evaluations and an end of trainingfocus group discussion, the level of overall satisfactionwith the training was found to be high. Participantsfelt they had acquired knowledge, but would haveliked more time and practical exercises on thosesessions which were considered technically complex,e.g. on sample size and sampling, plausibility check,report writing, analysis with ENA and Epi-Info, themortality survey and the standardisation test. In theBudapest training, participants felt that infant andyoung child feeding (IYCF) should be added as a topicin future similar training events. A survey module onIYCF is currently being finalised (Dec 2011). A post-training assignment was sent to participants from theBudapest training and is currently being analysed(Dec 2011).

Follow-up is needed to ensure that trainees applywhat they have learned. This has since been pursuedin refugee camps in Algeria, Ethiopia and Kenya in2011 where the ENN has provided technical supportand capacity building for the implementation of thestandardised nutrition surveys.

For more information, contact Caroline Wilkinson orAllison Oman, email: [email protected]

ing of programme performance and key riskfactors for malnutrition. UNHCR has also beenintroducing a number of innovative interven-tions to control and reduce anaemia in refugeepopulations, including the use of micronutrientpowders and lipid-based nutrient supplements,and strengthening existing efforts on malariacontrol and deworming. The need to monitorprogress with these interventions, as well ascollect data for other key indicators, resulted inan initiative to develop the UNHCRStandardised Nutrition Survey Guidelines(SNSG). This work was undertaken in collabo-ration with ENN and UCL, with modules onthe measurement of anaemia, WASH, anti-malarial bed nets, food security, and IYCF beingdeveloped over a period of two years.

The UNHCR survey guidelines are dividedinto two main sections. Section 1 is a quickreference guide in the form of a Fact Sheet,focusing on the key practical steps involved ina standardised nutrition survey. Section 2contains the standardised survey modulesoutlining the information to follow for training,data collection, analysis and reporting, andfocuses on the recommended core indicators.The guidelines can be used as a reference docu-ment for designing and implementing anutrition survey or as a tool for training surveyworkers on conducting the standardised nutri-tion survey.

The guidelines are designed to cover mostemergencies and all stable, protracted campsituations (except urban settings). In the future,these guidelines will be updated and improvedas lessons are learnt on their application andusefulness, and the field of survey methodol-ogy evolves.

A number of tools are available to assist ateach step of the survey process. For example,there are spreadsheets for supplies planningand producing trend graphs. All the tools andguideline documents can be downloaded from: http://www.unhcr.org or fromhttp://info.refugee-nutrition.net

These guidelines are designed specificallyfor camp settings, however they can be adaptedto other contexts. Comments and questions canbe directed to: [email protected]

1 Standardised Monitoring and Assessment of Relief and

Transitions. http://www.smartmethodology.org

Sara

h S

tyle

, EN

N/U

NH

CR,

Kenya,

2011

Dadaab nutrition surveys,August-September 2011

57

News

Putting nutritionproducts in theirplace: ACF positionpaperSummary of position paper1

Action Against Hunger | ACF Internationalhas recently produced a position paperon the role of products in the treatment

and prevention of global acute malnutrition(GAM). The paper focuses on the treatment ofsevere acute malnutrition (SAM), looks at theprevention and treatment of moderate acutemalnutrition (MAM) and considers the role ofnutritional products as part of broader strate-gies to build resilience to, and prevent,under-nutrition. The paper is written to beaccessible to field staff requiring guidance onwhether or not to use a certain type of product,and also to provide clear guidance on ACF’sposition for policymakers. The following is asummary of the key points in this paper.

The role of management within protocols addressing SAMAction Against Hunger | ACF Internationalrecognises that there are multiple factors thatcontribute towards recovery of patients fromSAM. Whilst therapeutic products, includingReady to Use Therapeutic Foods (RUTF), havebeen a proven key element of success, it isimportant to note at the outset that the supplyof a product is accompanied by other compo-nents. A medical protocol is used to tackleunderlying infections and to respond to new orworsening symptoms appropriately. Withinoutpatient care, weekly follow-up is essential,with criteria of how to respond to patients whoare not recovering properly. Inpatient carerequires continuous observations by qualifiedmedical personnel. Continual emphasis isplaced on the role of appropriate infant andyoung child feeding (IYCF) within sustainablerecovery. A number of ACF missions havefound it helpful to introduce focus groups, indi-vidual counselling and home follow-up ofnon-responders and defaulters as ways ofimproving recovery rates and reducing theproportion of defaulters. In ACF’s experience,even if the supply chain of RUTF is disruptedtemporarily, the continuation of good manage-ment procedures that emphasise these otherprogramme elements can help ensure success-ful outcomes.

RUTF local production and validationACF procures RUTF directly from a variety ofsuppliers. However, many projects receive in-kind contributions from UNICEF, which maybe sourced from further suppliers. ACF is in theprocess of validating several suppliers and isdesigning its own approval protocol followingUNICEF and Médecins Sans Frontières (MSF)validation mechanisms. ACF would supportthe establishment of an independent approval

board so that it is not only the main RUTFcustomers who are the quality regulators.

ACF see the potential advantages of procur-ing RUTF from local suppliers based indeveloping countries using local food productswherever possible, and is therefore fullysupportive of this approach. ACF will be pursu-ing a validation process with local supplierswhere product efficacy, safety and quality havebeen demonstrated.

RUTF and patentsACF acknowledges that the global demand forRUTF will continue to increase, particularly asRUTF usage becomes integrated into existinghealth systems. No single producer can be reliedupon, due to constraints in production capacityand in the global supply chain. ACF thereforewelcomes the steps Nutriset have taken recentlyto make the patent agreement more accessible tolocal producers. ACF does not consider patentsas inherently wrong, recognising the role theycan play in protecting RUTF quality and localproducer viability, so long as measures continueto be put in place to ensure a sustainable globalsupply chain of RUTF.

RUTF acceptanceACF encourages governments to evaluate thewealth of existing data that supports use ofRUTF in the treatment for SAM. ACF supportswidespread uptake of the community basedmanagement of acute malnutrition (CMAM)approach by health facilities and communitiesthrough government management, and thelocal production of RUTF to meet the demand.The scale-up of CMAM should be alongside(and not replace) initiatives that look at allforms of malnutrition.

Cost of RUTFSince the cost of many RUTF products is highlylinked to fluctuating milk powder prices, ACFwelcomes continued research into lower costalternatives using locally available products.ACF also recognise that locally-made RUTFproducts may not be inexpensive, due to lack ofsubsidies for key ingredients as found in somedeveloped countries. However, ACF feels theadded benefit to the local economy broughtabout by local factories should be factored intoany cost-benefit analysis when choosing prod-ucts, and suggests that quantitative researchinto such local benefits is continued andpublished. ACF believes that for now, RUTF isthe most effective treatment for SAM whenused with proper management and medicalprotocols, and should be continued despite thehigher product costs in comparison to other

nutrition interventions. ACF advocate for thescale-up of the treatment of SAM.

Coverage and funding of SAM treatmentACF fully supports the scaling up of CMAMactivities for the treatment of SAM. RUTFproduction and the CMAM approach needs tobe dramatically increased if all children withSAM are to be reached. ACF believe that theonly way to achieve substantial coverage isthrough supporting governments to integratethe CMAM approach into existing health facili-ties where possible.

RUTF and impact on breastfeedingACF fully supports and agrees that exclusivebreastfeeding for infants less than 6 months ofage is essential for optimum child health. ACFactively promotes this best practice, as well asadvocating for sustained breastfeeding for chil-dren aged 6-24 months and beyond. As atreatment for SAM, ACF does not consider thatthe use of RUTF undermines breastfeeding, butacknowledges that careful follow-up shouldmonitor this risk. If international protocols arefollowed, then no RUTF should be given toinfants below 6 months. For children aged 6-24months, breastfeeding is actively encouragedbefore the child is offered RUTF.

Potential negative impacts of rapidweight gain following RUTF treatmentThere is research indicating an associationbetween rapid childhood weight gain of thinchildren and later chronic disease. It is not clearwhether the short period of weight gain seen inSAM children would have any negative conse-quences later in life, particularly if the childreturned to a normal weight after recovery.Rapid early weight gain and then continuedobesity for childhood would seem to be a greatercause for concern, although more research isneeded to verify this. ACF will continue to treatSAM children with RUTF as part of the manage-ment protocol due to the elevated risk ofimmediate mortality, but acknowledges the clearneed for further research.

Summary of ACF position on productsdesigned for the treatment of MAMRegarding the use of products for treatingMAM (which is not the only availableapproach):• ACF acknowledges that programmes

involving traditional corn soya blend (CSB) have not been highly effective,

1 AAH/ACF International (2011). Products are not enough:

Putting nutrition products in their proper place in the treat-

ment and prevention of global acute malnutrition.

Summary Paper, November 2011

María P

ére

z N

egro

, ACF-S

pain

, Leso

tho

News

Conference ongovernmentexperiences ofCMAM scale up,Ethiopia, 2011

The Emergency Nutrition Network (ENN) incollaboration with the Government of Ethiopialed a conference to capture government expe-

riences of scale-up of Community-basedManagement of Acute Malnutrition (CMAM) in AddisAbaba, 14-17th November, 2011. The Conferencewas co-funded by Irish Aid, UK Aid (DFID) and CIDA.

The purpose of the conference was to provide a‘live’ forum for the discussion of issues relating tonational scale-up of CMAM from a governmentperspective. Senior government representativesfrom 22 African and Asian countries participated inthe conference, in addition to representatives fromUN agencies, non-governmental organisations,academia, bilateral donors, foundations and individ-ual experts. A total of 144 delegates attended.

The first three days focused on CMAM scale up.Nine countries with experience of scale up (Ethiopia,Pakistan, Niger, Somalia, Kenya, Ghana, Sierra Leone,Malawi, Mozambique) shared experiences for discus-sion based on carefully constructed case studies. TheIndia delegation shared particular considerationsaround CMAM scale-up and an additional 12 coun-tries shared their experiences. Day 4 considered thefindings of the CMAM experiences in the context ofthe Scaling Up Nutrition (SUN) movement.

Detailed country case study write up, a synthesispaper highlighting common themes and findingsfrom the country case studies and conferenceproceedings, and a meeting report are beingfinalised and will feature in a special edition of FieldExchange 43. Meanwhile, film footage of many ofthe speakers, presentation summaries and back-ground information are available at:www.cmamconference2011.org

and that the general trend coming out of recent research shows ready to use supplementary foods (RUSF) may be a

viable and potentially more effective alternative.

• Much of the research involving RUSF has come out of Malawi and Niger, and ACF welcomes further research in different settings to help validate RUSF for international use. ACF will also continue with its own research on the topic.

• CSB++ is a promising product that has been re-designed to take account of many of the nutritional limitations of traditional CSB. Lack of field-based evidence surrounding CSB++ makes it difficult to take a position on it at this stage, but ACF welcomes further research on it.

• Fortified blended foods with the additionof oil can be effective if efforts are made to minimise defaulting. ACF recommendscontinuing use of such products due to their lower cost and often local availability.If CSB++ is not available, and fortified oil is not available to accompany other fortified blended products, then effec-tiveness will be compromised. In this scenario, ACF recommends the use of a RUSF.

• ACF notes the potential effectiveness of RUTF in treating MAM but does not support this practice. If given a choice between RUTF and RUSF for the treatment of MAM, ACF recommend using RUSF due to the importance of keeping a sustainable pipeline of RUTF for those who most need it and its sometimes lower cost. It is important to avoid any potential confusion in the community regarding the use of RUTF as part of treatment and not as a food which should be shared or sold.

• Good management of programmes, including close follow-up of children, and counselling in relation to caring andfeeding practices is just as important as the choice of product. Further research is needed on this topic.

• ACF stresses the need to remember that products given for the treatment of MAM should be given in conjunction with the international medical protocol and close follow-up.

Summary of ACF position on productsdesigned for the prevention of acutemalnutritionACF regards IYCF best practices as crucialin the effort to prevent malnutrition. Everyeffort should be made to ensure that theintroduction of products does not under-mine breastfeeding, and any focus on aproduct intervention should always beaccompanied by adequate IYCF work. Thisshould include the promotion of, andsupport to, appropriate caring practices andunderstanding what the barriers are to goodIYCF and caring practices.

There are many approaches to preventingacute malnutrition, of which the provision ofproducts is just one option. Preventingmalnutrition requires a multi-sectoralapproach and therefore, even if products areused, they should only form one part of theresponse.

Where it has been determined that adirect nutritional intervention is required, afood-based approach should be used wherepossible to prevent acute malnutritionbecause of its lower cost, better sustainabil-ity and cultural appropriateness. However,in contexts of displacement or natural disas-ters where food supplies have been cut off,products may be the only viable option toprevent a high caseload of patients withacute malnutrition. In this scenario, Readyto Use Complementary Foods (RUCF) couldbe used. ACF consider that RUTF or RUSFshould not be used in prevention due totheir overlap with treatment of SAM andMAM.

ACF acknowledges that there is notenough evidence in the field of productsand preventative malnutrition and iscontinuing to conduct its own research inthe field.

Overall summaryThe solution to the management of acutemalnutrition does not lie with productsalone. Nutrition causal analysis should beused to determine which integratedapproach should be used. Products can playa supporting role, but need to be consideredas part of a wider package of support.Indeed, periods when there are temporarilyreduced or interrupted supplies of thera-peutic products emphasize the importanceof continued medical monitoring andmanagement to contribute to successfuloutcomes for the affected individuals andtheir families.

Treatment of SAM requires the use oftherapeutic milks and RUTF in accordancewith the CMAM approach.

The management of MAM is less clear-cut. There is often an artificial dividebetween treatment and prevention, andapproaches designed for one can have animpact on the other.

If a direct nutrition intervention involv-ing supplementation of diets is required, thechoice between products and a food-basedapproach depends on the context and takeinto consideration nutrition requirements,nutritional quality, time frame, sustainabil-ity, programme setting, available fundingand cultural acceptability.

The field of nutrition products is rapidlyevolving. ACF will continue to conduct rele-vant research to help inform programmepolicy. A continual review of ACF positionswill be needed as more products becomeavailable and as more research findings areshared.

For more information on ACFs position,contact: Sandra Mutuma, email:[email protected]

ACF are not alone in having to consider how andwhen to use various products in programmesthat prevent or manage MAM in differentcontexts. We would welcome your thoughts andexperiences on this topic and have opened an en-net discussion around this athttp://www.en-net.org.uk under ‘Preventionand treatment of moderate acute malnutrition’.Letters to Field Exchange on this topic are alsowelcome. Send to: [email protected]

58

59

Improving blanket supplementaryfeedingprogramme(BSFP) efficiencyin Sudan

Pushpa Acharya is currently work-ing as Head of Nutrition for theWorld Food Programme in Sudan.She has a PhD in HumanNutrition from the University ofMassachusetts. She has over 21

years of professional experiences with nationalgovernments and UN agencies.

Eric Kenefick is current Head ofProgramme for WFP in Sudan. Hehas spent the past 15 years work-ing or WFP and UNICEF invulnerability analysis and M&E.He is a graduate of Tulane

University’s School of Public Health and TropicalMedicine.

The project is implemented by the State Ministryof Health. The commitments of the Director ofNutrition Ms Sit Eldat Ahmed Al, and Ms KhaldaKhalafalla – WFP Kassala, were critical in thesmooth functioning and success of the programme.Dr Amal Abdalla – WFP Sudan – coordinated theproject and provided the technical guidance without which the intended results would nothave been achieved.

By Pushpa Acharya and Eric Kenefick

Map 1: Prevalence of acute malnutrition in Sudan

Prevalence of acute malnutritionacross Sudan is high and ranges from11 to 29%1. Specific causes of acutemalnutrition are largely unknown.

High rates are observed during both non-lean and lean seasons. Major efforts are beingexerted by the Ministry of Health andhumanitarian aid agencies to treat malnour-ished children with therapeutic programmesimplemented through approaches involvingin particular community based managementof acute malnutrition (CMAM), in-patientcare, and targeted supplementary feedingprogrammes. Additionally, in areas withhigher acute malnutrition rates, blanketsupplementary feeding programmes (BSFP)are implemented during lean seasons as apreventive approach. In spite of all theseprogrammes, repeated survey results showthat acute malnutrition rates remainunabated (see map).

Pilot to improve BSFP performanceIn 2010, WFP’s targeted supplementaryfeeding programme (SFP) aimed at treat-ing moderately malnourished childrenreached over 200,000 children. In addition,415,000 children aged 6-59 months werereached through a BSFP aimed at prevent-ing the usual peak of acute malnutritionobserved during lean seasons in Darfur.While the targeted SFP met the SPHEREstandard for all performance indicators

across Sudan, 2009 programme monitoringdata of the BSFP led to questions about itsefficiency in reducing rates of acute malnu-trition usually observed during lean seasons.

In order to improve efficiency of theBSFP in Sudan, means of improvingperformance of the programme wereexplored. A pilot was designed and imple-mented in one area in Kassala State.Kassala was selected because of WFP’s pre-existing SFP programme, presence of aWFP nutritionist and the relative safetyand accessibility of the area compared toDarfur and other areas of conflicts. Thepilot programme began in March 2010 andis continuing until end of 2011.

Nutritional and programming contextPrior to the pilot, GAM prevalence rates inKassala were usually high and similar tothat seen in Darfur (19.2% SHHS 2006, 15%SMoH 2009, 16.7% SHHS 2010). Mukaram,one of the shanty towns on the outskirts ofKassala town, was selected for the pilot bythe State Ministry of Health (SMoH). Thearea is one of the poor neighbourhoods ofKassala and is situated relatively near tothe main town and hence easier for SMoHstaff to monitor. Prior to the pilot study,mid upper arm circumference (MUAC)measurements of all children in the catch-

1 Sudan Household Health Survey 2010

Field Article

Table 1: Number ofchildren under 5 yearsenrolled in SFP in twoPHCs, Mukaram,Kassala, 2009Month Number of

children

January 137

February 125

March 20

April 208

May 227

June 200

July 156

August 106

September 129

October 145

November 55

December 90

Role play as part of the programme in Mukaram

P A

chary

a/W

FP,

Sudan

Boundaries shown on this map do not imply official endorsement or acceptance by theUnited Nations World Food Programme. This map is for planning purpose only. WFP cannot guarantee that this map product is error free and therefore we accept no liability orconsequential and indirect damages arising from use of this map product.

Data Sources:VAM Unit: WFP Sudan

Sudan Boundaries: Sudan Interagency Mapping (SIM)

meetings/education sessions, facilitatingdiscussion on topics related to food, feeding,food hygiene and food safety, cooking demon-stration, and mixing of the’ super’ cereals(CSB+), oil and sugar prior to the distribution,etc. All traditional practices that posed a threatto the children’s and women’s nutritional statuswas strongly discouraged. Example of suchpractices included ceasing breastfeeding assoon as the mother is pregnant even when thebreastfed child is too young to do so. The‘change agents’ were also responsible formaking home visits – once a week for eachhome – in order to increase awareness of otherhousehold members of issues discussed in theclub meetings. This facilitated changing riskytraditional feeding practices such as early cessa-tion of breastfeeding, eating from the sameplate/dish even for very young children, lowfeeding frequencies and poor food hygiene, etc.

A Knowledge Attitudes Practice (KAP)survey was conducted to understand the localfood and feeding habits. Existing educationmaterials were then adapted to address issuesidentified by the KAP survey. The ‘changeagents’ were trained for three days by the MoHon these topics and on facilitation techniques.

Target groups and enrolmentAll pregnant and lactating women and childrenunder five years of age were targeted with theBSFP, providing approximately 500 kcal. Theration consisted of super cereals 100g, 10g oiland 10g sugar mixed prior to distribution on abi-monthly basis. All children were measuredevery month and their growth monitored.Children who were identified as moderatelymalnourished were referred to the targeted SFPand children identified as severely acutelymalnourished children were referred to theoutpatient therapeutic programme (OTP).

The enrolment in the programme was suchthat within four months of the start of theprogramme, 100% of all children under fiveyears in the catchment areas of these PHCswere registered in the programme (expectedunder-five children 1500 – 15% of the totalpopulation). This blanket enrolment also deter-mined that there was an increase in theidentification of acutely malnourished children.At the start of the enrolment, almost 22% of thechildren in the programme were identified to besuffering from acute malnutrition.

ResultsThe nutritional status of all the children in thepilot programme was monitored on a monthlybasis (see Table 2). Children identified as moder-ately malnourished at enrolment were referred tothe SFP centres located in the same health facilitywhere they received the regular SFP ration (1200

60

Field Article

Figure 1: Trend in admission in eight SFP centres in Kassala during lean andnon-lean seasons, Jan 2009-Dec 2010 (as percentage of under five catchmentpopulation by centre)

25.0

20.0

15.0

10.0

5.0

0.0

Jan

–20

09

May

–20

09

Jun

–20

09

Jul –

2009

Aug

–20

09

Nov

–20

09

Dec

–20

09

Jan

–20

10

May

–20

10

Jun

–20

10

Jul –

2010

Aug

–20

10

Nov

–20

10

Dec

–20

10

Beryai

Wau Nour

Mukram

Althower

Garb Algash

Alangas north

Ansar Alsana

Alshapy

ment areas of these two Primary Heath CareCentres (PHC) confirmed a high prevalence ofacute malnutrition where 15.5% of the childrenhad MUAC below 125 mm2.

There were and remain two SupplementaryFeeding Programme (SFP) centres located in thearea within the PHC. The catchment area of thetwo centres was estimated to have a populationof 10,000 people with 1,500 children under fiveyears of age. These centres usually saw highnumbers of children with moderate acutemalnutrition (MAM) enrolled in the SFPprogramme throughout the year. The admissionnumbers for MAM cases in 2009 for the two SFPcentres in Mukuram can be seen in Table 1. Veryhigh numbers of admission were seen at the SFPcentres located in Mukaram in March 2010 at thestart of the pilot. This may have been due tointensive communication with the communityprior to the start of the pilot combined with theprovision of a blanket food ration, therebyattracting greater numbers of MAM childreninto the Mukaram catchment area.

Figure 1 shows the trend in admission ofmalnourished children from January 2009 toDecember 2010 in ten SFP centres in Kassala,including those in Mukaram. This is presentedas a percentage of MAM cases enrolled in theSFP relative to the estimated number of chil-dren under five in the catchment area of each ofthe centres (15% of the catchment population).The the coverage of SFP in the catchment areasis unknown. Seasonal trends are observed withhigher prevalence rates during the lean season(May – Sept)3 and lower rates during thepostharvest season (Oct-Dec). The coverage ofthe SFP in Kassala is unknown4.

Pilot intervention designThe design of the pilot study included intensivecommunity engagement and sensitisation. Acommunity club was established in each of thehealth centres. The clubs were equipped withcooking facilities for recipe development/trials,toys to entertain children while their mothersparticipated in discussions, and other essentialresources. These facilitated participation ofwomen in the discussion/awareness sessionsheld in the clubs twice a week. Prior to theestablishment of the clubs, meetings were heldwith the community elders to sensitise them tothe objective of the programme and also toidentify and select community change agents.

The tasks of the ‘community change agents’(the frontline workers of the intervention)included support to the MoH staff in growthmonitoring of the children enrolled in the BSFP,keeping children entertained during the club

kcal per day per child as a take home ration).Children who were not malnourished at enrol-ment received half the ration of the targeted SFP.

The recovery of the children who weremalnourished was rapid. A significant proportion(68%) of malnourished children enrolled in the SFPgained sufficient weight within 4 weeks to recover.This recovery was sustained on the lower BSFPration over the 12 months period following recov-ery. Additionally, children who were notmalnourished at enrolment remained healthythroughout the year, even during the lean/hungerseason. A survey was conducted in July 2011 on281 randomly selected children in Mukaram5. Theproportion of children with a weight for height zscore (WHZ) below – 2 SD was less than 1%. MeanWHZ was found to be 0.40±0.43.

Replication of the modelWith impressive results from the model pilotedin Mukaram, the SMoH requested WFP toexpand the integrated blanket supplementaryfeeding programme (IBSFP) into North Delta,where the latest survey had indicated that theacute malnutrition rates was 16.5%6. There wasno existing SPF centre in North Delta. Hence,the MoH with support from WFP establishednew SFP centres in four PHCs. The expansionalso entailed establishment of community clubsin the PHCs. The Mukaram model was dupli-cated in all aspects. Table 3 provides thepreliminary data from monthly monitoring ofthe nutrition status of the children enrolled inthe programme. While the results are not asimpressive as Mukaram, they reflect the successof the overall programming approach.

CostThe cost of the ration/child including Food-for-Work provided for the community mobilisers(at a ratio of 50 children/community mobiliser)is 0.09 USD per child/day7. The additional costfor the printing of registers, education materi-als, training of community mobilisers, toys,mats, and sun shelter for the clubs for Mukaramwas 0.81 cents per child. The latter cost is a one-off fixed cost at the start of the programme. Thetotal cost per child per year in the blanket SFPwas 33.66 dollars. The cost of the targeted SFPration ranged from 12.4 -14.9 USD per child ifthey recovered from MAM within 10 to 12weeks of enrolment in the programme. TheBSFP cost was therefore at least twice as high asthe targeted SFP. However given that under theBSFP children don’t succumb to malnutritionyear after year, the overall programme cost ismuch lower as fewer children present fortargeted supplementary feeding.

Lessons learnedThe community involvement in the project fromthe design stage onwards played a significantTable 2: Nutrition status of children enrolled in the

BSFP pilot

Month Total number ofchildren <5years registered

Number ofmalnourished

Global acutemalnutritionprevalence

March ‘10 1782 390 21.8%

April ‘10 1851 134 7.2%

May ‘10 1821 85 4.6%

June ‘10 1866 90 4.8%

July ‘10 1882 68 3.60%

August ‘10 1790 73 4.07%

September ‘10 1861 45 2.40%

October ‘10 1856 44 2.37%

November ‘10 1841 32 1.37%

December ‘10 1856 19 1.02%

January ‘11 1901 15 0.78%

February ‘11 1901 12 0.63%

2 Prevalence in acute malnutrition

with MUAC criteria is found to

be much lower than with WHZ

criteria in Sudan.3 For some centre records were

not available for all months; for

this reason the lines are not

continuous.4 Coverage survey of CMAM is

ongoing - UNICEF5 Records of 5 children were

flagged and 19 children did not

have complete data.6 Report of Nutrition and

Mortality Survey in North Delta

March 2011, SMoH7 At commodity prices as of 14

October 2011

61

Field Article

Evaluation of Concern’sresponse to the HaitiEarthquake

By Andy Featherstone

Andy Featherstone is an independent humanitarian research and policy consultant.With significant experience working for NGOs and interagency networks, hespecialises in helping agencies learn from their work and in bridging the gapbetween policy and practice.

The earthquake that struck Haiti on January 12, 2010 did so with devastatingconsequences. More than 200,000 people were killed, 300,000 were injuredand over one million were left homeless. The cocktail of extreme vulnerabil-ity coupled with the huge loss of life and massive destruction wrought on

Haiti’s largest urban area and political and commercial hub effectively decapitatedthe state. It left hundreds of thousands of people traumatised and without the meansnecessary to sustain life and livelihood. It was this that precipitated the tremendousgenerosity that saw Concern raise 28 million euro and embark on its largest single-country humanitarian programme since it was established in 1968.

An evaluation was undertaken eight months after the earthquake1 and followedan unprecedented expansion of the programme and staff. The purpose of the exer-cise was to review the appropriateness, timeliness, efficiency and effectiveness ofboth the interventions carried out and operational support systems with an impor-tant focus being placed on documenting lessons learnt.

The evaluation findings were as follows.

TimelinessThe response was timely, particularly the early support to water, sanitation andhygiene (WASH). An area where Concern performed particularly strongly was inquickly moving beyond a focus on support to urban areas to meeting the needs ofdisplaced and host communities in rural areas.

While Concern was successful in providing timely assistance in important sectorsof its response, the organisation lacks consensus over the use of its surge capacitymechanisms, the Emergency Response Team (ERT) and Rapid Deployment Unit(RDU). It is urgent that agreement is reached over how to manage and deploy theseassets to most effect in the future.

1 Featherstone. A (2010). Evaluation of Concern’s Response to the Haiti Earthquake. October 2010

Signs at the water stand encourage people to drink chlorinated water– prior to the painted signboards being erected, people were notconvinced that the water was safe to drink

Rois

in G

allagher/

Conce

rn,

Haiti 2011

role in ensuring successful implementation andoutreach of the programme. Initiation of the projectthrough the community leaders created strong linkbetween the targeted community and the SMoH.

The community change agents took ownership ofthe project and felt a sense of responsibility towardsthe community members. Use of the change agentseased the task of convincing the community about theneed for behaviour change and also facilitated the taskof the health staff in the health centres. Food for workplayed an important role in motivating change agents

The SFP centre attracted children and women frombeyond the usual catchment area of the health centres.

The toys made available at the health and socialclub assisted the nutrition educators to entertain chil-dren while women were discussing and listening tothe nutrition education and take accurate measure-ments of children by making them relaxed during themeasurements and consequently increasing the accu-racy of the measurements.

Cooking demonstrations of various recipes ofcomplementary foods from locally availablecommodities and CSB at the health club gave thechance for women to learn proper food preparationand hygiene practices while also keeping womeninterested while health and nutrition messages weredelivered.

Community club meetings provided opportunitiesfor women to discuss a wide variety of topics beyondfood, feeding, food safety and food hygiene.

Intensive monitoring by the MoH and WFP as wellas the community leaders was important for the over-all outcome of the pilot.

ConclusionsChanging harmful infant and young child feedingpractices requires active participation of the commu-nity in the learning process. When food availabilityand quality is enhanced through the provision ofsmall quantities of highly fortified food combinedwith the intensive engagement of the communityaround harmful feeding practices, the impact of foodaid is significantly increased. The size of theprogramme allowed intensive monitoring by SMoHand WFP. The challenge lies in taking the pilot to scale.

For more information, contact: Pushpa Acharya,email: [email protected]

Table 3: Monthly beneficiary number and proportion ofmalnourished children enrolled in IBSFP in North Delta area,May – July 2011

Month IBSFP centres Total < 5childrenRegistered

TotalMAMCases

TotalSAMCases

GAMrate

May

Umalguraa 402 47 22 17%

Britani 352 40 14 15%

Hadalia residents 391 50 10 15%

Hadalia IDPs 301 35 15 17%

Total/average 1446 172 61 16%

June

Umalguraa 498 34 10 9%

Britani 356 31 7 11%

Hadalia residents 200 25 5 15%

Hadalia IDPs 305 28 9 12%

Total/average 1359 118 31 11%

July

Umalguraa 442 22 7 7%

Britani 428 21 8 7%

Hadalia residents 361 28 12 11%

Hadalia IDPs 305 16 6 7%

Total/average 1536 87 33 8%

August

Umalguraa 442 13 6 4%

Britani 428 9 7 4%

Hadalia residents 361 21 6 7%

Hadalia IDPs 305 11 12 8%

Total/average 1536 54 31 6%

62

Evaluation

There is a need for greater fluency within theorganisation in applying streamlined humanresources (HR) systems that better supportemergency recruitment (like the use of genericterms of reference and the quicker adoption ofrecruitment systems that can fast-track staffappointments, inductions and entry to theprogramme).

While the organisation has a comprehensivefinancial management system, the experienceof the team in trying to link programme outputswith donor contracts to facilitate contractmanagement has been a challenging one andwould benefit from review.

CoverageConcern has performed well against the cover-age criterion, choosing to launch a multi-sectoralresponse which targeted both urban and ruralcommunities. Given the extraordinary nature ofthe crisis, an extraordinary response wasrequired. While it means that Concern has had towork hard to deliver against such a largeprogramme, it has done so successfully.

RelevanceThe programme has maintained its relevancethroughout the response by providing a mix oftargeted assistance (through cash for work andlivelihoods programmes) and blanket distribu-tion of basic commodities. While the context ofthe crowded urban environment has made itdifficult at times to attain minimum standards,interviews with project participants suggestthat the services being provided by Concern aremeeting their needs.

The use of SPHERE standards and indicatorsin Concern’s programmes has provided animportant yardstick for success. Given the clar-ity that the WASH team now has aboutprogress that needs to be made against clusterstandards, it will be important that swiftprogress is made to achieve these.

ConnectednessConcern’s established presence in the countryand the strong links it has with communitiesprovided an important platform for the earth-quake response. This has served both toprovide much needed assistance and furthercement Concern’s relationship with projectparticipants. Since the organisation is alsoworking in sectors it has established a compe-tence in, interventions in these areas oftenbenefit from strong analysis.

they demonstrate significant innovation,achieve a level of excellence in response, orshow proficiency in a particular area ofresponse.

Responding to the needs of rural andurban communitiesConcern has built on its established presence inthe country to extend both relief and recoveryactivities to those living outside the immediateenvirons of Port-au-Prince. Its operations in LaGonave and Saut D’eau have been backed upby a robust analysis of the numbers ofdisplaced and the impact this displacement hashad on the local infrastructure and economy. Inthe first three months of response, Concernprovided cash, tents to meet emergency shelterneeds and non-food items to targeted benefici-aries in these areas. Given the propensity foraid to be targeted at the most visible and mostnumerous claimants, the targeting of rural areasby Concern is noteworthy. Not only does it havethe potential to ensure that those affected by theearthquake and subsequent displacement aresupported, but it goes some way to slowing theinevitable return to Port-au Prince which thelimited services available in the city would havestruggled to accommodate.

Transitional shelter design and deliveryThe roll-out of the T-shelter programme hasbeen considered by many (including the shel-ter cluster coordinator) to be exemplary. Whileit took some time to conduct the baselinesurvey, to assess needs and to procure thematerials, the programme has benefited fromsound targeting and excellent organisation ofwork processes which has allowed for swiftproduction of the shelters. The design is inno-vative and has taken account of the needs forearthquake- and hurricane-proofing. Theapproach taken towards the use of contractorsto manage shelter construction using laboursourced from the camp has allowed the workto progress quickly, while fostering ownershipand transferring important skills to membersof the camp population.

Humanitarian leadership & coordinationThroughout the response, Concern staff haveshown a commitment to participating inhumanitarian leadership and coordinationforums. While the Country Director is part ofthe Humanitarian Country Team (HCT), sectorstaff members have also played prominentroles in clusters and sub-cluster groups.

While considerableprogress has been made inbuilding a coherent Concernteam in Haiti, it will beimportant to continue tostrengthen ways of workingto ensure strong integrationbetween all parts ofConcern’s mandate, whetherlong-term development orhumanitarian response.

AccountabilityGood progress has beenmade in cascading keyaccountability principlessuch as the provision ofinformation, consultationand participation of commu-nities throughout theConcern programme. Animportant area for Concern and the broaderhumanitarian community will be to ensurethat camp committees are consistently workingin the best interests of the people they repre-sent.

There is some urgency in establishingaccountable and representative camp commit-tees and Concern should continue to workwith the cluster to find workable solutions.

EffectivenessConcern has been effective in mounting a largemulti-sectoral humanitarian programme inHaiti. In particular, the breadth of theprogramme, the timeliness of many of its earlyinterventions and the prioritisation of meetingrural in addition to urban needs has beenimpressive in such a complex context.

Protection and peace building are nowbeing mainstreamed across the humanitarianprogramme. In the absence of governmentpolicy on durable settlement solutions andwith elections planned for November 2011, it islikely that these cross-cutting areas willbecome ever more relevant. It will be impor-tant that there is sufficient capacity and thatactivities are fully integrated across all ofConcern’s humanitarian work in Haiti.

Two resources are highlighted - a review ofthe excellent organisation-level meta-evalua-tion conducted in 2009 and the Preparing forEffective Emergency Response (PEER) docu-

ment which summariseorganisational knowledgeand learning. Condensingthese into a set of succinct(1-2 pages) documents,highlighting key lessons forprogramme design anddelivery and organisationalsystems and ways of work-ing, would be a wiseinvestment for the future.

Identifying best practiceThe evaluation highlightedthe significant contributionwhich Concern has madein meeting the needs ofe a r t h q u a k e - a f f e c t e dcommunities in a timelyand effective manner. Anumber of these are worthyof particular mention as

Allis

on S

helley/C

once

rn,

Haïti, 2

010

Out-patient nutrition clinic atPlace de la Paix, Port-au-Prince

Eighty per cent of schools in Port auPrince were damaged or destroyed.

Concern has hired ‘animators’ tofacilitate play and learning.

Allis

on S

helley/C

once

rn,

Haïti, 2

010

relevant to the urban environment. While‘camps’ are often administratively more easyto support, the lack of space and the infiltra-tion of powerful gangs into the over-crowdedurban environments has created a significantchallenge to agencies who are more used toworking in peri-urban or rural environments.The lesson here may be to stretch humanitar-ian comfort zones and look at methodologiesto support smaller and more decentralisedsettlements that focus on the importance ofcommunity, that benefit from strong links withlocal authorities and which have strong linkswith the private sector to ease the process andsustainability of handing over services. Inseeking to address these issues directly, theapproach taken by Concern in Tabarre Issacamps has much to offer. In working with bothresettled and host communities, and in tryingto plan the settlement less as a regimentedcamp, Concern’s programme has taken a‘neighbourhood’ or ‘community’ approachwhich is considered as best practice by many inthe cluster.

Innovative approaches to addressingvulnerability: The Baby Tent ProgrammeHaiti is the first humanitarian response wherethe concept of ‘baby tents’ has been deliveredto scale, in a context where women hadsuffered significant trauma and where the useof infant formula was often prioritised overbreastfeeding practices. The Baby Tents were aspace which offered privacy, care and coun-selling and which could advocate for, educateand support women on breastfeeding. Whereinfants were not breastfed, the baby tentsmonitored the use of infant formula. Thus theyprovided potentially life-saving services forboth breastfed and non-breastfed infants.

A high value placed on an independentprocurement capacityAn important lesson has come from triallingthe United Nations Humanitarian ResponseDepot (HRD) which is available for both pre-positioning of stock items and procurement.Experience from the earthquake responsestrongly suggests the need for Concern toretain an independent procurement capacityas quotes from the HRD procurement agencywere found to be uncompetitive and leadtimes were considered to be lengthy. While theinitial reliance on air freight has a significantcost attached to it, it did ensure that theprogramme could scale-up swiftly andensured that minimum quality standards forprocured items were met.

63

Interviews showed they had an astute under-standing of both the resource requirementsthat coordination commitments place on theorganisation but also the opportunity thisprovides to influence the humanitarianresponse far beyond what Concern couldachieve through other means. While somemembers of the humanitarian communitymay dismiss the value of coordination asbeing too time-consuming, it is the way ofdoing humanitarian business. In contexts suchas Haiti, it is essential for prioritising assis-tance and avoiding duplication. The team hasused the forum that the clusters offer for rais-ing issues of concern to the widerhumanitarian community such as protection,shelter design issues and the threat of forcedevictions. Interviews with ClusterCoordinators highlighted the value they placedon Concern’s participation, as well as theimportant contribution that staff members havemade to the work of the clusters.

Lessons learnedConcern’s response in Haiti has necessitatedinterventions that have either been unprece-dented or provide the potential to lead theorganisation in new directions or challengecurrent ways of working.

Responding to urban disastersHumanitarian organisations in Haiti havestruggled to make their earthquake responses

Review ofIntegratedFood SecurityProgrammein MalawiSummary of published review1

Areview of an Integrated FoodSecurity Programme (IFSP),implemented by GTZ2 inMalawi from 1997 to 2004, has

recently been published by TuftsUniversity. The IFSP in Malawi was acomplex, multi-sector activity that soughtto improve food security and nutrition inone of the country’s most vulnerable,least-performing regions. Theprogramme was implemented by GTZ(now GIZ)3 on behalf of the GermanFederal Ministry for EconomicCooperation and Development (BMZ)between 1996 and 2003 (with a 12-monthextension supported by the EuropeanUnion).

The IFSP’s end-line evaluationreported that the intervention hadachieved its objectives. A subsequentreview was undertaken to considerwhether gains made in the past had beensustained and to draw lessons from thisexample that may contribute to newthinking on models of integrated, multi-sectoral programming. The reviewhighlights that “food security approachesto nutrition require systemic, multidisci-plinary and inter-sectoral approaches”(UN Standing Committee on Nutrition,2009, p. 1). It goes on to argue that theempirical evidence remains limited ofwhat actually works on the ground,where attempts are made to introducepackages of interventions that addressmultiple sectors at once.

This review was conducted over aperiod of five months (November 2010through March 2011) and involved twofield trips. The findings presented in the

Evaluation

Rois

in G

allagher/

Conce

rn,

Haiti 2011

Cleaning drains as partof Cash for Work

Temporary shelters nearingcompletion at Tabarre

Rois

in G

allagher/

Conce

rn,

Haiti 2011

64

review rest on three kinds of information:documented evidence (project, consultant,and published reports), insights shared byvarious experts and stakeholders, and directexperiences of the reviewers from the villagevisits.

Villages were visited based on meeting one offour criteria: • Early adopters versus later adopters (villages included from the outset of the IFSP in 1997 versus those included later).• Villages that recorded positive nutrition gains over the course of the IFSP versus those that were less positive.• Villages that served as ‘controls’ in Phalombe District (prior to that district’s separation from Mulanje).• Villages with specific examples of ‘success stories’ versus ‘failures’.

The review authors emphasised the limits toattributing findings of the review in any statis-tically significant manner. The IFSP could notmaintain ‘pure’ control groups due to admin-istrative re-districting in the late 1990s thatremoved original control villages fromMulanje District. Furthermore, there weremany other agents of change both acrossMulanje and beyond. While the review doesnot ascribe causality to the IFSP versus anyother influences, an attempt is made to drawinferences about the role of the interventionwhere expected outcomes were achieved (asdocumented in the end-line evaluation). Thisis achieved by careful post-hoc interviewswith those directly involved (beneficiaries aswell as implementers), an assessment of theplausible links between inputs and outcomes,and triangulation across multiple sources ofdata with Mulanje and neighbouring districts.

The intervention which started in 1997encompassed 185 villages (roughly 40,000households). Mulanje District was selectedbecause it represented “an area which haschronically suffered the greatest food deficitover the last 10 years, compared to other areasin Malawi”. An obvious manifestation of theseverity of local problems showed up in thecauses of paediatric mortality in MulanjeDistrict Hospital, which in 1993 were reportedas malnutrition (21% of cases), malaria (19%),and anaemia (11%). In 1997, malnutrition andanaemia were still among the top three causesof death locally (at 18% and 14%, respec-tively), with HIV/AIDS having taken overfirst place.

The initial roll-out villages were chosenlargely because at that time they had “nointerventions by other donor agencies”, inother words, they were seen as “pristine” andmore likely to demonstrate changes moreclearly in the absence of other donor activities.

Key findings of the review The IFSP in Mulanje was successful in most ofits aims. Not only were conclusions of the2004 Final Evaluation Report confirmed, butmany of the gains identified then have beensustained. This represents an important“proof of concept” of this particular approachto integrated programming. Successes can beidentified as helping bring positives changesin food security (measured by outcomes

across multiple sectors), changed thinkingand behaviours at community level (thatpersist a decade later where ‘early adopters’have continued to innovate), and newapproaches to tackling food insecurity thathave been adopted by the public sector locallyand nationally. The IFSP has influencedgovernment thinking on food and nutritionsecurity more broadly, leading to many of itsprinciples being embedded in current nationalpolicies.

In terms of specific successes, child nutri-tion was improved (reaching the target set of a10% reduction in the prevalence of stunting)and most sectoral targets were also achieved.Gains that can be attributed by varyingdegrees to the IFSP include enhanced agricul-tural productivity and output in several staplecrops, the cultivation of an enhanced range ofcrops (as a result of introduction and promo-tion of new and improved seeds), andreduced losses to crop and livestock diseases -all leading to higher levels of farm output.This in turn contributed to reduced periodswhen farm households have no food in theirstores, higher household incomes, andincreased local investments in productiveassets, including in the natural resource base.

Beyond agriculture, the IFSP promotednon-farm income diversification activities thathave since proliferated, allowing for morediversified livelihoods and disposableincome. Access to market (for sale of crops,purchase of food, and engagement in cottageindustries) was improved through access roadand bridge construction, still well maintainedin most instances. The supply of clean waterhas improved significantly, and maintenanceof water points has been good, largelysupported through village committees. Accessto food for-work represented an importantsafety net for food-insecure households whocould not immediately benefit from enhancedfarm productivity and market access.Improved supply of food and income hassupported enhanced diet diversity and qual-ity. A wider range of foods is consumed todaythan prior to the IFSP, and also compared withmost other parts of the country. Food preser-vation activities have enhanced diet choicesand reduced post-harvest losses.

The process of community engagementwas valuable and valued. Community andgovernment training in problem-solvingprocesses are still in use today. Many villagecommittees are still functional, and thepromotion of ‘demand responsive’ models ofservice delivery had durable impact on theway that public servants conduct their busi-ness. The IFSP model was widely promoted inMalawi and its lessons have been incorpo-rated into training and policy agendas sincethe end of the intervention.

Conclusions of the reviewA number of broad conclusions emerge. TheIFSP represents a model of integratedprogramming, carefully designed around acore conceptual framework, which achievedits targets. But it is not the only possiblemodel, either for achieving such targets or forapproaching integrated programming as a

Evaluation

1 Webb. P (2011). Achieving Food and Nutrition Security:

Lessons Learned from the Integrated Food Security

Programme (IFSP), Mulanje, Malawi. June 2011.

Download at: http://www.donorplatform.org/load/111882 German Technical Cooperation. This is now part of GIZ

since 1 January 2011 (see footnote 3).3 Deutsche Gesellschaft für Internationale Zusammenarbeit

(GIZ) GmbH. The German Society for International

Cooperation (GIC) Ltd is a federal enterprise that supporst

the German Government in achieving its objectives in the

field of international cooperation for sustainable

development.

process. The Mulanje example should be care-fully analysed against other potentially viableapproaches in seeking to understand how bestto leverage actions across multiple sectors toachieve gains in agriculture, nutrition, andhealth simultaneously.

The IFSP model appears to have been rela-tively cost-effective. At roughly US$59(around €40) per household, or US$11 person(€8) per year, the package of IFSP interven-tions compares well with a range of otherintegrated programmes in Malawi and else-where. That said, not every element of thepackage worked equally well, with homegardens, some health interventions, and somecrops performing weakly compared with othercomponents of the programming.

The successful (versus weak) aspects of thisactivity shone a spotlight on the importance ofcultivating leadership for change.Engagement of community leaders as stake-holders and the intensive training of villagersin leadership roles and committee processeswas critical. So too was establishing appropri-ate incentives and buy-in across district- andnational-level ministries so that ‘ownership ofleadership’ was cultivated and service deliv-ery and programme implementation allbenefitted. Identification and support for earlyadopters (leaders in innovation) matteredimmensely to ‘start-up’ activities in the realmof livelihood diversification. Attention to thisprocess aspect of programming was critical.

Questions raised by the review that shouldframe debate on future integrated program-ming include: i) Could the same outcomes have been

achieved for less cost? ii) If so, what is the minimum versus

desirable menu of interventions that would (together) generate the best possible outcomes for least cost?

iii) Would the unit cost of the package introduced in Mulanje rise or fall if taken up at scale across the country?

iv) Should such packaged interventions seek to promote absolute change or accelerate relative change (to bring ‘lagging’ regionsor communities up to par with the rest of their country)?

v) Can integrated programmes be designed to buffer future shocks, not just resolve pre-existing vulnerability to food insecurity, and what would that add to the cost of a package of integrated services and inputs?

Many such questions can only be answeredthrough operations research on a next genera-tion of multi-sectoral integrated programme,which this review concludes is a reasonabledevelopment policy priority.

65

Multi-prongedapproach to themanagement ofmoderate acutemalnutrition inGuineaBy Dr Jean-Pierre Papart andDr Abimbola Lagunju

Dr. Jean-Pierre Papart MD,MPH, is Health advisor,Fondation Terre deshommes, Lausanne,Switzerland

Dr Abimbola Lagunju MD, isRegional Health Advisor,Fondation Terre des hommes(West Africa)

The authors acknowledge the support of theteam in Guinea, in particular Mamadi Kaba(MD), Mariama Ba, Sonia Panzani (FondationTerre des hommes), Marie-Jeannne Haubois(former Fondation Terre des hommes).

Fondation Terre des hommes(Fondation Tdh) is a Swiss ChildRights advocacy organisation basedin Lausanne, Switzerland and

founded in 1960. The Foundation employsthe UN Convention on the Rights of theChild as its guiding principle in its two prin-cipal domains of action, maternal and childhealth (MCH) and child protection. Its inter-vention strategy in these two domains ispredicated on empowerment of beneficiar-ies, system reinforcement and advocacy.Fondation Tdh has nutrition-focused MCHand child protection projects in five countriesin the West Africa sub-region – Benin,Burkina Faso, Mauritania, Togo and Senegal

Management of acute malnutrition inGuineaIn Guinea, the management of acute malnu-trition is undertaken at three different levelsof facility, depending on the severity of thecase:• Therapeutic nutrition centres (CNT)

manage phase 1 of complicated severe acute malnutrition (cSAM)

• Outpatient nutrition centres (CNA) treat severe acute malnutrition without complications (sSAM) and also phase II of cSAM

• Supplementary nutrition centres (CNS) manage patients with moderate acute malnutrition (MAM).

This management strategy alongside themodel for interventions (see Figure 1) wasdeveloped by the Ministry of Public Health(MSPH) and published in The NationalGuidelines of Management of AcuteMalnutrition in May 2008. National guide-lines admission criteria at health facility levelare shown in Table 1.

Fondation Tdh in GuineaFondation Tdh has been supporting govern-ment health facilities in Conakry, Guinea in

This article describes the experience of Fondation Terre deshommes in the management ofmoderate acute malnutrition insupplementary nutrition centressupported by the organisation inGuinea.

the management of MAM since 2005. This deci-sion was informed by the reported increase in theglobal acute malnutrition (GAM) rates between1999 and 2003 in the city of Conakry. The 1999Demographic and Health Survey reported 10.9%GAM prevalence in children aged less than 5years. The QUIBB (Enquete sur le Questionnaire desIndicateurs de Base du Bien-etre) survey of 2003showed that global acute malnutrition was 14.4%among the same age group. Conakry had thesecond highest rate of global acute malnutritionamong the eight regions in the country in 2003.

In December 2007, Fondation Tdh supportedtwo communal medical centres (CMCs) (Ratomaand Flamboyant), both located in the Communeof Ratoma and a private medical facility (St.Gabriel Dispensary) in the adjoining commune ofMatoto to adopt and put into practice the newnational guidelines on the management of acutemalnutrition in the treatment of sSAM. Thisinvolved use of RUTFs and health facility basedmanagement of MAM. In December 2008, twoadditional health centres (CS) in Ratomacommune (Lambandji and Wanindara) also intro-duced these activities using the nationalmanagement guidelines. Fondation Tdhsupported these facilities by putting in place amonitoring system to follow up on performanceand provide necessary technical advice. This arti-cle is based on the monitoring of the performanceof these four health facilities between 2008 and2010. See Figure 2 for map reflecting Conakrycommunes.

Health facility network in RatomaCommuneThe Commune of Ratoma has 20 quarters (admin-istrative units). The public health system consistsof sixteen health facilities: two CMCs and 14 CS.The CMCs, which have the same facilities asdistrict hospitals (surgery, paediatric and internalmedicine units, hospitalisation), serve as referralunits for the health centres. Of the 16 health facil-ities in the commune, none has a CNT, five have aCNA and all 16 have a CNS (five facilities haveboth a CNA and a CNS). All these health facilitieshave a nutrition unit manned by trained govern-ment staff. All identified cSAM cases are referredto the Institute National de Santé de l’Enfant(INSE) for stabilisation. After stabilisation, theyare referred back to the health centres for ambula-tory management of phase 2. Fondation Tdhpresently provides technical support to all thehealth facilities in Ratoma Commune.

Fondation Tdh support to health facilitiesAll the activities of the nutrition programme(anthropometric assessment using measurements ofweight, mid-upper arm circumference (MUAC)),cooking demonstrations, counselling) are carriedout by the government staff of the nutrition units ofthe health facilities. Fondation Tdh provides techni-

Field Article

Guide to acronyms used (equivalent in genericCMAM terminology)

cSAM complicated severe acute malnutritionsSAM uncomplicated SAMMAM moderate acute malnutritionCNT Stabilisation Therapeutic centre

(manage Phase I cSAM) CNA Outpatient Therapeutic centre (treat

sSAM and phase II cSAM)CNS Supplementary Feeding centres

(manage MAM) CMC Communal Medical CentresMSPH Ministry of Public HealthCS Health Centre

Preparation of enriched porridge aspart of the programme to managemoderate acute malnutrition

Fondation T

dh,

Guin

ea,

2011

Table 1: National guidelines admission criteria andhealth facility level

Admission criteria* Health facilitylevel for care

ModerateAcuteMalnutrition

• W/H between 70% and79.9% of the median

• MUAC between 11 and 12cm (for length > 65 cm)

• Absence of oedema

CNS

SevereAcuteMalnutrition

• W/H < 70% of the median • MUAC < 11 cm (for length > 65 cm)• Absence of medical

complications)

CAN

• W/H < 70% of the median• MUAC < 11 cm (for length > 65 cm)• Presence of medical

complications

CNT

W/H: weight for height *Based on the NCHS references

66

Field Article

1 The management of sSAM has also been the subject of

several monitoring reports, which may be consulted.

Contact the authors for details.

Communityscreening

< 5years AMSuspected

AMReferred

SAM

MAM

MAMRecovery

MAMTreated

Child

cSAM cSAMTreated

sSAMTreated

cSAMStabilised

sSAMRecovery

sSAM CMin CNA

Transferg CNA

Referralg CNS

MAM CM Activity

Transfer g CNA

sSAM

Referalg CS

Screeningin CS

cSAM CMin CNT

Transferg CNT

cal, material and equipment support to the nutri-tion units of the health facilities.

The staff of Fondation Tdh comprises twomedical doctors and a nutritionist. Technicalsupport includes training and on the job super-vision. Fondation Tdh also invests in quarterlynutrition programme supervision of all thehealth facilities by the higher authorities of theConakry City Health Directorate. FondationTdh staff also assist the health facilities incollating and analysing data generated fromtheir nutrition activities. Through these analy-ses, weaknesses are jointly identified anddecisions are reached on corrections. Further,on the request of the health authorities,Fondation Tdh acts as an active interfacebetween the health authorities and agencieslike UNICEF and World Food Programme(WFP) for supplies to the health facilities. Thisis a temporary arrangement pending the timethe authorities study the reporting mechanismsof these agencies and identify a liaison personfor this activity.

In addition to the technical support,Fondation Tdh also provides material supportto the nutrition units, such as stationery, IEC(information, education, communication)materials, cooking materials, MUAC tapes andweighing scales.

Urban Community Health Workers inRatoma communeIn six of the 20 quarters of Ratoma, FondationTdh in collaboration with the communal healthauthorities, recruited and trained 32 urbancommunity health workers (UCHW). TheUCHWs participate in the community screen-ing of children for acute malnutrition in theirneighbourhoods, follow up on defaulting casesand engage in the promotion of healthy nutri-tion practices through cooking demonstrations,counselling on breastfeeding and hygiene (SeeFigure 3).

Screening of children for acute malnutritionScreening of children for acute malnutritiontakes place at two levels – in the neighbour-hoods (active screening) and at the healthfacility level (passive screening). See Figure 3.

Active screeningEach of the UCHWs is assigned an area withintheir neighbourhoods. The UCHWs compile alist of all the children within their area and visit

Figure 1: Model of intervention for the management of acute malnutrition in Guinea

their homes on a monthly basis. In the course ofthe visits, MUAC of children 6-59 months ismeasured and recorded. Children with MUACof <125mm are referred and accompanied bythe UCHW to the health facility in their area.The nutritional status of referred children isfurther assessed (weight and height are meas-ured, MUAC re-measured) by trainedgovernment health workers and then childrenassigned an appropriate treatment regime asdirected by the National Guidelines. TheNational Guidelines stipulate that children withcSAM are immediately referred and accompa-nied to the INSE. Children with sSAM are put onthe ready to use therapeutic food (RUTF) regime(see below). Children with MAM are givenrations (when available) and their mothersadvised to come for weekly cooking demonstra-tion and training on appointed days.

Passive screeningThis is conducted in the health facilities byhealth workers (nurses and doctors) for all thechildren (0-5 years) who have come for consul-tations due to an illness or to well-baby clinics(0-11 months). The weights and heights of allthese children are measured and recorded andin the case of children between 6-59 months (orwhose lengths are >65 cm), MUAC is also meas-ured. Children between 6-59 months presentingwith acute malnutrition are assigned to theappropriate treatment as directed by theNational Guidelines.

The total case load of malnutrition for RatomaComune and St Gabriel is shown in Table 2.

Management of SAMThe child with sSAM is prescribed Ready to usetherapeutic foods (RUTF)1 and routine drugslike Vitamin A, antibiotics and anti-helminthics.RUTF is given on a weekly basis to the childrenand at the end of each week, the child presentsat the health facility for a check-up until thechild attains and maintains the target weight(85% of the median % weight-for-height) at twoconsecutive weekly checkups (option 1) oroption 2 (discharge on reaching 85%) withouttwo consecutive check-ups.

Monitoring reports indicate that St GabrielDispensary has had major problems withdefaulting in the treatment of sSAM. In 2010, ofa total number of 1,133 recorded exits, 35.6%were cases of defaulting. The reason for this isthat, unlike the four other CNAs supported by

Fondation Tdh, St. Gabriel Dispensary has areputation for inexpensive and quality medicaltreatment which extends far beyond Ratomacommune. The dispensary charges a flat ratewhich includes costs for consultation, labora-tory tests and medicines. Many people travelgreat distances, sometimes up to 200 km, toseek treatment is this dispensary. Normally,parents do not recognise signs of acute malnu-trition in their children and come to thehealthcare centre expecting treatment forillnesses rather than for malnutrition. Thehealthcare centre staff identify malnourishedcases through routine measurement of the chil-dren. When a child is diagnosed as sufferingfrom acute malnutrition and the parents aretold that treatment is necessary and that thiswill involve several weekly check ups (a total of5 to 10 visits), many of them fail to return to thecentre after one or two follow-up visits. This isdue to the distance they have to travel eachtime. These cases present a problem for theDispensary, because there are no CNAs or CNSin the villages of origin of these children towhich they can be referred.

All identified cases of cSAM are referred toINSE. Fondation Tdh supports the patientsthrough the payment of transport costs fromthe referring health facility to INSE. Further,Fondation Tdh pays the treatment costs ofcSAM cases referred from any of its interven-tion centres. cSAM cases are managed with F75formula until stabilisation and then referredback to referring centre for Phase 2 manage-ment as prescribed by the National Guidelines.

Management of MAMBreastfed children <6 months, identified byweight for age, suffering from MAM are treatedfor any underlying illness and their mothers arecounselled on appropriate breastfeeding tech-niques and practices. The children are dischargedfrom the programme when consistent weightgain is established through breastfeeding.

Children between 6-59 months sufferingfrom MAM (identified using W/H) are treatedwith enriched porridge, prepared twice a weekat the health facility and distributed to children.The porridge is constituted to deliver100kcal/100mls. The children are fed there.

RATOMA

MATOTO

DIXIN

MATAMKALOUM

Figure 2: Map of Guinea focused on Conakry communes

Key to figure: AM: acute malnutrition; SAM: Severe acute malnutrition; MAM: moderate acute malnutrition;CS: health centre; cSAM: complicated SAM; sSAM: Uncomplicated SAM; CNA: ambulatory nutrition centres;CNS: supplementary nutrition centres; CM: case management

67

Mothers are advised and shown how toprepare the enriched porridge for their chil-dren at home and to give them the porridgetwice a day in addition to family foods.Depending on availability (through the WorldFood Programme supplies), there is a weeklydistribution of premixed food (Corn Soy Blend(CSB), sugar, oil, salt).

Anthropometric measurements of these chil-dren are taken and recorded weekly and a childis discharged from the programme in accor-dance with either option 1 or 2 (outlined aboveunder SAM management). Although option 2is less reliable, most facilities supported byFondation Tdh adopt this approach.

Programme monitoringThis monitoring report is based on datacollected over a period of 36 months (1stJanuary 2008 to 31st December 2010) fromCMC Ratoma and CMC Flamboyant.Additional data were collected over a 12-month period (1st January 2010 to 31stDecember 2010) from two CS in RatomaCommune – CS Wanindara and CSLambandji. Data from St Gabriel (Matotocommune) were not available for analysis.

A total of 7,033 cases of malnutrition inchildren 0-59 months were treated by thefour health facilities during the period underreview. Of this number, 2,343 (33.3%)presented with SAM, while 4,690 (66.7%)were treated for MAM. Of the total numberof children treated for MAM, 162 (3.5%) were<6 months of age, 4,207 (89.7%) were aged 6-23 months and 321 were aged 24 – 59 months.

Table 3 shows the relevant centre, casenumbers and year of data collection for theSAM and MAM cases.

Anthropometric profile and admissioncriteria of the children (0-59 months)treated for MAMThe criteria for identification of MAM usedwere weight for age for children <6 monthsand 70% - 80% of the median W/H index forchildren aged 6-59 months Of the 4,690 casesidentified as MAM at these centres during thereference period, 35 (0.7%) were false positivediagnoses and five (0.1%) were SAM casesthat were eventually referred to CNT.

The average weight for height (W/H)percentage of the median (NCHS unisexcurve) at admission was 77% for boys andgirls combined (p=0.509). Applying thegender specific NCHS W/H references, theaverage W/H z score on admission was -2.8for boys and -2.5 for girls (p<0.001). The samedifferential is found using the 2006 WHOgrowth standards: W/H z score -3.3 for boysand -2.8 for girls (p<0.001).

The programme criteria are based onNCHS references. For comparative purposes,the 2006 WHO growth standards were usedas a test reference for analysis. Of thoseadmitted to the programme, 5.2% werejudged not acutely malnourished (W/H zscore >-2), half (50.7%) of admissions weremoderately acutely malnourished and 44.4%were severely malnourished based on 2006WHO standards. With the NCHS bi-sexcurves, the figures are 12.6%, 70.5% and16.9% respectively. Thus use of the WHOGrowth Standards would greatly increase thenumbers of children considered severelymalnourished, and reduce the number ofthose classified as moderately malnourished.

Performance record of MAM casemanagementIn the management of MAM, the nationalcase management guidelines of malnutritionstipulate that children should be dischargedon attainment of 85% of the median of theW/H index. However, some mothers aban-doned treatment before their children couldattain the discharge weight. Table 4 showsthe status of children on exit from the MAMmanagement programme that reflects 5.8% ofmothers of MAM children defaulted fromtreatment. The default rate from treatmentvaried little between Ratoma andFlamboyant (p=0.867). Reasons for defaultvaried from change of location of residence(within the city) by the mother, lack of time,and return to their village of origin. The vastmajority (94.1%) attained the anthropometriccriterion for recovery. Five cases werereferred to CNT for treatment for SAM.

Time needed to attain target weightThe mean time needed for a child to attain85% of the median was 3.5 weeks. The length

of stay varied from year to year and from healthfacility to health facility as shown in Table 5.

The longest period of stay recorded was nineweeks in CMC Ratoma in 2010. This may be partiallyexplained by the fact that some of the childrenadmitted into the programme were borderline cases(MUAC 110mm or slightly above or 70% of themedian % weight-for-height or slightly above).Further, some mothers did not regularly attend thetwice-weekly cooking demonstration and training.The duration of stay in the programme also dependson the ability of the mother to source and prepareappropriate food for the child at home in the periodbetween the cooking demonstration and food distri-bution in the health facilities. The lowest duration ofstay recorded was 1 week – these were also border-line children with MUAC at 123 mm or above or 79%of the median % weight for height or slightly below.

Average weight gainThe average weight gain for children treated forMAM in CMC Flamboyant was 7.7g/kg/day,whereas in Ratoma, it was 7.6g/kg/day over thethree year period. There were fluctuations in thisaverage weight gain depending on the CMC and theyear in question. Table 6 shows the average weightgain in the two CMC and the two CS over a periodof three years

Weight gain and length of stay in theprogrammeNormally the length of stay should be inverselyproportional to weight gain. However, this correla-tion could not be established in this study because

Field Article

Figure 3: Communal (active) screening / health facility based (passive) screening

Active community screening

Age < 6 monthsBreastfeeding

problems

Poorgeneral

condition

≥ 1 sign ofserious condition

Poor sucklingability

Bilateraloedemas

Referral �CS

MUAC <12.5 cm

11.0 cm 12.5 cm

MUAC

Testing for bilateraloedema

Weight for Height (WH)

12.5 cm > MUAC > 11.0 cmor

80% (-2 SD)>WH>70% (-3SD)and

No bilateral oedema

MUAV < 11.0 cmor

WH<70% (-3SD)or

Bilateral oedema

Testing for complications

Referral�CNS

Referral�CNA

Referral

Cases screened in CS(active screening)

Cases referred �CS(passive screening)

Testing for bilateral oedema

MUACchild > 65 cm

Table 2: Screening1 of children and cases of malnutrition inRatoma Commune (+ St. Gabriel)

Children screened MAM2 cases MAM3 cases

2008 87,739 9,316 (10.6%) 1,953 (2.2%)

2009 101,197 9,371 (9.3%) 2,259 (2.2%)

2010 100,047 6,793 (6.8%) 1,393 (1.4%)

1 This includes children screened at community level UCHW as well as children identified through passive screening. St. Gabriel does not have a UCHW network for active screening.

2 MUAC 110 – 120 mm or 70% - 80% of the median % weight-for-height

3 MUAC <110mm or <70% of the median weight-for-height

Table 3: Sources, case numbers and year of data for SAM andMAM cases

2008 2009 2010 Total

CMCRatoma

1314 1047 851 3212

CSWanindara

0 0 509 509

CSLambandji

0 0 302 302

CMCFlamboyant

858 1187 965 3010

TOTAL 2172 (30.9%) 2234 (31.7%) 2627 (37.4%) 7033 (100%)

Table 5: Duration of treatment to attain target weight in MAM childrenCMC Ratoma CMC Flamboyant CS Wanindara CS Lambandji

Max Min Mean Max Min Mean Max Min Mean Max Min Mean

2008 6 1 4.1 7 2 5.1 n/a n/a n/a n/a n/a n/a

2009 6 0 3.2 7 1 3.2 n/a n/a n/a n/a n/a n/a

2010 9 1 3.7 5 2 3.3 8 2 3.6 6 2 3.2

Total 9 0 3.6 7 1 3.7 8 2 3.6 6 2 3.2

Table 4: Status of MAM children on exit from the programme

YEAR Status at discharge (according to National Guidelines) Total

Recovery Default Transfer to CNT

2008 1320 (92.6%) 106 (7.4%) 0 1426

2009 1435 (95.3%) 65 (4.3%) 5 (0.4%) 1505

2010 1660 (94.4%) 99 (5.6%) 0 1759

Total 4415 (94.1%) 270 (5.8%) 5 (0.1%) 4690

Table 6: Average weight gain in two CMCs and two health centres

2008 2009 2010

CMC Flamboyant 8g/kg/day 9.7g/kg/day 7g/kg/day

CMC Ratoma 10.7g/kg/day 5.7g/kg/day 5.9g/kg/day

CS Wanindara - - 6.5g/kg/day

CS Lambandji - - 5.8g/kg/day

there were external factors beyond the control ofthe programme managers despite their efforts tocontain them. These factors included the regu-larity of mothers at the twice-weekly cookingdemonstration sessions, quality and quantity offood which they gave to their children in-between the twice-weekly visits to the healthfacilities and episodes of illness like malaria,diarrhoea or respiratory tract infections. Effortsto contain these external factors included:• Home visits for defaulting mothers by CHWs• Support to mothers of children during an

episode of illness• Distribution of premixes (CSB, oil, sugar

and salt) to mothers of MAM children after each weekly cooking demonstration in the health facility when available.

The role of mothers in the management ofmoderate acute malnutritionThe compliance of mothers to advice on thepreparation of enriched porridge, feeding prac-tices at home and their attendance record at thetwice-weekly food preparation demonstrationsplayed an important role in the favourable exitoutcome of moderate acute malnutrition. Despitethe challenges of increases in the prices of basicfoodstuffs, many mothers were able to mobiliseresources to purchase ingredients and prepare

enriched porridge for their children irrespectiveof supplies of premix via the programme.

ConclusionsGovernment health facilities when given theappropriate technical, material and equipmentsupport can deliver good results in the manage-ment of MAM cases. On the job training andsupervision, feedback on performance andregular higher level supervision of nutritionactivities played an important role in the qual-ity of results posted by the four health facilitiesthat were considered in this study.

To many mothers, MAM is not an illness andcompliance with treatment, particularly when itrequires weekly presence in a health facility, is achallenge. The multi-pronged approach ofcommunity sensitisation, information andeducation, home visits and facility-basedmanagement of MAM children through weeklyweighing and cooking demonstration adoptedby Fondation Tdh and the authorities of thehealth facilities of the project, improved theknowledge of mothers about acute malnutri-tion in general, and MAM in particular.

Cooking demonstrations that included feed-ing MAM children in health facilities andeducation and counselling on good household

infant and young feeding practices were accept-able services to mothers. Compliance withtreatment and the recovery rate was high andthe rate of default was low. Lack of knowledgeamongst mothers on appropriate feeding prac-tices is likely to have been a significant factor incausing malnutrition in those children admittedto the MAM programme.

The role of UCHWs is pertinent even wherethere is a wide network of private and publichealth facilities. Through the active screening ofchildren in their homes in the quarters, manychildren who otherwise would not have beenbrought to the health facility because they werenot perceived as being ill by their mothers,were identified and sent to the health facility fortreatment for MAM.

Finally the management of MAM in an urbansetting demands a multi-pronged approach thatinvolves trained health workers, communityhealth workers, information education andcounselling of mothers, good supervision and anadequate level of food security.

For more information, contact: Dr. AbimbolaLagunju, email: [email protected] and Dr. Jean Pierre Papart, email: [email protected]

68

Field Article

As we went to print, we heard the sadnews of the premature death of KariEgge, whom many of you will have had

the privilege of knowing and working with.

Kari graduated from Mahtomedi HighSchool in 1985 and George WashingtonUniversity in 1989. Kari earned a PhD in PublicHealth from Tulane University where shecombined studies with HIV/AIDS- related proj-ects. In the following years, Kari dedicatedherself to international causes. Her career inhumanitarian aid work spanned over 20 yearswith Peace Corps, Catholic Relief Services (CRS)

and the International Division of American RedCross (AmRC). As part of CRS’ EmergencyResponse Team, she played a key role in numer-ous emergencies, including the Goma volcanodisaster, Afghanistan, Southern Africa droughtand the tsunami in Indonesia. In addition, Karihelped train hundreds of emergency respon-ders. Her last posting was to Thailand, workingwith those in countries affected by the 2004tsunami. During her career, she made lastingfriendships around the world. Kari was also aloving mother to Dylan (6) and Isabelle (8),taking every opportunity to introduce them tonew cultures and environments.

There has been a huge outpouring of trib-utes on Facebook from a global community ofcolleagues and friends who supported herthrough her illness and are now celebrating herlife. Here are just a few:

“Kari will be greatly missed. Her enthusiasm andpassion for improving the humanitarian causewill continue to inspire all of those who the pleas-ure to work with her.”

“She was articulate, strong, and sensible andclearly had the experience to back up all thecontributions she made”.

“A mom who managed to become an amazingmother and still contribute to the humanitarianfield for a long time. This affected me profoundlyand I know I made decisions differently because ofher example. It is amazing how much impactKari’s life had, not just on her immediate familywho were blessed by her love and mourn herdeparture, but also on those of us who weremerely in her presence and felt changed by her lifeand the way she chose to live it.”

Some lovely memories are also shared athttp://thegeographyofsoul.wordpress.com/2012/01/31/kari-noel-egge/

There will be gatherings for her and celebra-tions of her life happening around the world(Minnesota, Washington DC, Kenya andBangkok). This speaks to the incredible impactthat she has had on the world.

We express our sincere condolences to herchildren, Dylan and Isabelle, on their loss, aswell as to their father, Graham Eastmond, herparents, Robert and Deanna Egge, brother Kirkand family.

Mary Lung’aho, Jennifer Rosenzweig and all of the ENN Team

Dedication to Kari Noel Egge

Dedication to Mr AbdikarimHashi Kadiye

Shared by two of her close friends andcolleagues, Mary Lung’aho and JenniferRosenzweig

Shared by Leo Matunga, Nutrition Cluster

Coordinator, Somalia

We extend condolences to the familyand friends of Mr Abdikarim HashiKadiye and Mr Duale, who died on the

12th of January 2011 in Somalia. Mr Kadiye thehead of Tofiiq Umbrella Development organisa-tions (TUOS) and also, the nutrition cluster focalpoint for Hiran and Galgaduud, had justrecently organised and chaired the Hiran-Galgaduud cluster meeting on the 10th ofJanuary 2011. Mr Kadiye was killed togetherwith his driver, Mr Duale, in an ambush byunknown gunmen between Dhusamareb andEl Dhere. This is a senseless loss of life and a

huge loss to the work of the cluster in theregion. Mr Kadiye had been the nutrition focalperson for Galgaduud for almost a year now,was a hard working member of the cluster, andhad tried under difficult conditions to continueto organise partners meetings in the area andensure smooth operations of the nutritionprogramme.

We extend our heartfelt condolences to MrKadiye and Mr Duale’s families, friends and theTUOS staff.

Leo Matunga and the ENN Team

69

People in aid

Participants in the UNHCR Operational Guidance workshop (see news piece this issue).

UNHCR standardised nutrition survey guidelines and trainingregional training in Budapest, May 2011 (see news piece).

Invite to submit material to Field Exchange

Many people underestimate the value of their individual field experiences andhow sharing them can benefit others working in the field. At ENN, we are keento broaden the scope of individuals and agencies that contribute material forpublication and to continue to reflect current field activities and experiencesin emergency nutrition.

Many of the articles you see in Field Exchange begin as a few lines in an emailor an idea shared with us. Sometimes they exist as an internal report thathasn’t been shared outside an agency. The editorial team at Field Exchangecan support you in write-up and help shape your article for publication.

To get started, just drop us a line. Ideally, send us (in less than 500 words) yourideas for an article for Field Exchange, and any supporting material, e.g. anagency report. Tell us why you think your field article would be of particularinterest to Field Exchange readers. If you know of others who you think should

contribute, pass this on – especially to government staff and local NGOs whoare underrepresented in our coverage.

Send this and your contact details to:Marie McGrath, Sub-editor/Field Exchange, email: [email protected] to: ENN, 32 Leopold Street, Oxford, OX4 1TW, UK. Tel: +44 (0)1865 324996 Fax: +44 (0)1865 324997

Visit www.ennonline.net to update your mailing details, to make sure you getyour copy of Field Exchange.

If you are not the named recipient of this Field Exchange copy, keep it or passit on to someone who you think will use it. We’d appreciate if you could let usknow of the failed delivery by email: [email protected] or by phone/post atthe address above.

70

Editorial teamJeremy ShohamMarie McGrathDeirdre Handy

DesignOrna O’Reilly/BigCheese Design.com

The Emergency Nutrition Network (ENN) is a registered charity in the UK

(charity registration no: 1115156) and a company limited by guarantee and

not having a share capital in the UK (company registration no: 4889844).

Registered address: 32, Leopold Street, Oxford, OX4 1TW, UK. ENN

Directors/Trustees: Marie McGrath, Jeremy Shoham, Bruce Laurence, Nigel

Milway, Victoria Lack, Arabella Duffield

Jeremy Shoham (Editor),Marie McGrath (Sub-editor)and Carmel Dolan are ENNTechnical Directors.

Orna O’ Reillydesigns andproduces all ofENN’s publications.

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

Katherine Kaye isthe part-timeadministrationassistant at theENN.

Chloe Angood is a nutritionist working part-timewith ENN on a number of projects and supportingHuman Resources.

Thom Banks is theENN's DeskOperations Officerand provides logisticaland project supportto the ENN team.

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

The opinions reflected in Field Exchangearticles are those of the authors and donot necessarily reflect those of theiragency (where applicable).

The Team

Field Exchangesupported by:

The Emergency Nutrition Network (ENN)

grew out of a series of interagency meetings focusing on food andnutritional aspects of emergencies. The meetings were hosted byUNHCR and attended by a number of UN agencies, NGOs, 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. ENNs five year strategy (2010-2015) is available atwww.ennonline.net

Contributors for this issue

Thanks for the pictures to:

Cover

Office SupportKatherine KayeMatt ToddThom Banks

WebsitePhil Wilks

Nune MangasaryanBernadette CichonPushpa AcharyaJose Luis Alvarez MoranAllison ShelleyWFP photo libraryUNICEF photo libraryFAO photo librarySave the Children UKConcern WorldwideJenny AkerPatrick WebbSandra MutumaRachel Bezner KerrMelody Tondeur

María Pérez NegroBruno Spada/MDSAraceli ReyAbigail PerryJan KomrskaKate SadlerPatrick WebbJean-Pierre PapartNaomi CosgroveGwyneth CoatesKate A. GreenawayLeisel TalleyCarlos Navarro-ColoradoBrid KennedyJennifer Martin

Front: Goma Issune, a recipient at the mobile cash transfer programme. © Concern Worldwide, Niger, 2010Back: Amaram Moussa, mobile cash transfer recipient,© Concern Worldwide, Niger, 2010

Jeff FeldmesserHedwig DeconinckMark MyattErnest GuevarraSaul GuerreroAnne WalshSandra MutumaPhilip JamesAndrew SealSarah StyleAbigail PerryJessica MeekerMelody TondeurJose Luis AlvarezMoranBrian Mac DomhnaillSaul GuerreroBernardette CichonJan KomrskaNaomi CosgroveJane EarlandAurélie Rozet

Mathias Grossiord Cecile SalpeteurJean-Pierre PapartAbimbola LagunjuPushpa AcharyaEric KenefickKate A. GreenawayElizabeth C. JereMilika E. ZimbaCassim MasiBeatrice MazinzaKawanaLola GostelowTamsin WaltersLeisel TalleyCarlos Navarro-ColoradoAndy FeatherstoneMara NwayoJennifer MartinMarko KeracHilary Heine

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

Tel: +44 (0)1865 324996

Fax: +44 (0)1865 324997

Email: [email protected]

www.ennonline.net