Constraints to achieving Sphere minimum standards for SFPs in West Darfur: a comparative analysis

The current conflict in Sudan's westernmost state of Darfur began in early 2003, although most humanitarian agencies only gained access to the area and began their response in mid-2004. The violence and displacement peaked in 2003. However, localised fighting has continued, as the underlying causes of the conflict (political and economic marginalisation of Darfur by the central government, proliferation of firearms from outside actors, competition for resources in a highly arid, resource poor area, among other factors) remain unresolved.

More than two million people, largely non- Arab agriculturalists, are believed to be displaced as a result of the conflict, with just over 700,000 of those living in camps or among host communities in West Darfur. The internally displaced people (IDPs) remain highly dependent on external aid, as their livelihoods are severely restricted due to insecurity and loss of assets during their flight. Although the Darfur Peace Agreement was signed by the central government and two of the three main rebel groups, little progress has been made towards lasting peace. The African Union (AU) deployed a small force in August 2004, but their numbers and mandate prevent them from truly protecting IDPs. A tentative agreement has been reached to deploy a combined UN/AU peacekeeping mission with a stronger protection mandate that could facilitate the return of some IDPs to their homes, but the deployment is yet to be realised.

Nutritional context

This protracted humanitarian crisis necessitates the continued operation of supplementary and therapeutic feeding programmes in many areas of the three Darfur states. The September 2006 Darfur-wide interagency nutrition survey found an estimated global acute malnutrition (GAM) ¹ of 10.6% in West Darfur, significantly higher than 2005's finding of 6.2% ². Although low relative to GAM rates in other regions and countries, malnutrition rates still fluctuate outside acceptable parameters and are likely even higher in localised pockets. For example, in the West Darfur state capital, El Geneina, the GAM was estimated at just over 12% in May/June 2005 and in 2006, despite the presence of General Food Distributions (GFDs) and selective feeding programmes ³.

Currently Concern is running a Communitybased Therapeutic Care (CTC) programme, comprising a Supplementary Feeding Programme (SFP) for moderate acutely malnourished individuals and an Outpatient Therapeutic Programme (OTP) for uncomplicated severe acutely malnourished individuals. Children with complicated severe acute malnutrition are referred to Therapeutic Feeding Centres (TFCs) for in-patient stabilisation/rehabilitation that are run by non-governmental organisations (NGOs) or Ministry of Health (MoH). The programme is facilitated by a strong community mobilisation component, targeting six programme areas through 18 distribution sites (see map in Figure 1). Programme caseloads at the end of November 2006 are outlined in Table 1. The CTC programme is complemented in Seleia, Kulbus and Mornei by Concern programmes in water and sanitation, hygiene promotion, health, and livelihood security.

Table 1 - SFP/OTP caseloads per site, end Nov 2006
Programme area	Sub location	SFP children under 5*	OTP	Total	June - November cumulative SFP cure rates	June - November cumulative OTP cure rates
Geneina	Durti Ardamata Riad Abuzar	50 80 207 240	18 11 26 11	68 91 233 251
Total		557	66	643	26.9%	66.4%
Krenig	Krenig	139	6	145
Total		139	6	145	41.7%	72.2%
Mornei	Mornei Romalaya	655 62	7 1	665 63
Total		717	8	725	60.5%	79.5%
Umshaleya	Umshaleya	324	10	334
Total		324	10	334	29.3%	59.4%
Kulbus	Wadi Bardi Batro Kulbus Helelat Dughush Mastura	38 16 36 39 31 0	5 6 7 4 6 0	43 22 43 43 37 0
Total		160	28	188	30.2%	87.6%
Seleia	Seleia hijilija Manjura Aro Sharow	99 39 16 40	15 4 11 16	111 43 27 56
Total		194	46	240	50.0%	53.3%
Total		2111	164	2276	40.9%	71.7%
* malnourished pregnant and lactating women and elderly individuals are also enrolled in the programme, numbers not presented here

Measuring SFP performance

SFP performance is measured principally against minimum standards for the proportion of exits that are cured, died and who have defaulted ⁴. The cure rates should meet or exceed 75%, death should be <3% and default <15% (Sphere minimum standards). These rates are calculated with total exits equal to the sum of cured, died, defaulted, transferred to OTP/hospital and failed. The widely used Medecins Sans Frontieres (MSF) nutrition guidelines interpret low cure and high death and non-responding (failure) rates as reflecting low programme quality, and high default as a sign of poor acceptability and/or accessibility ⁵.

Concern's SFPs in West Darfur often fail to meet these key minimum performance indicators ⁶. The cure rates are depressed by high default and are reduced further by the necessity of transfers to OTP/TFC when children deteriorate, and by the discharge of children as 'failures to respond' after four months, if they have not yet reached discharge criteria ⁷. Poor SFP performance is a common problem facing most agencies working in nutrition in Darfur and several interagency meetings have been held for nutrition partners to discuss possible reasons ⁸. In the Concern programme, depressed cure rates can generally be seen across locations, but significant differences in performance can also be seen between programme areas. The most striking contrast is between Concern's two programmes with the highest caseloads: El Geneina and Mornei. While the more urban El Geneina SFP, which serves a mixed population of IDPs and residents, generally has the worst indicators, the Mornei SFP, based in West Darfur's largest internally displaced population (IDP) camp and serving a population of roughly 90% IDPs, has the best (see Figures 2 and 3).

Understanding factors that constrain performance

In order to better understand the reasons for the poor performance of Concern's West Darfur SFPs, the programme carried out several investigations in September and October 2006. The investigations attempted to find commonalities and differences between El Geneina and Mornei - the two sites with the highest number of admissions and the most dissimilar performance. The main investigations were: interviews with caregivers of children discharged as failures from both sites; a survey of El Geneina beneficiaries to establish registration status for General Food Distribution (GFD); Focus Group Discussions with beneficiary caretakers on their perceptions of malnutrition, the programme and their children's progress in it; observations at SFP distributions; discussions with the SFP team and some analysis of trends from programme admission and discharge statistics.

Differing SFP Performance at two sites: El Geneina and Mornei

The El Geneina SFP performance is the poorest of the Concern programmes. Cure rates were extremely low from June to November, 2006 (cumulative cure rate for the period: 26.9%), largely due to high default rates throughout (cumulative default: 47.5%), particularly in the months of cultivation (June/July) and harvest (October/November) (Table 1 and Figure 2). Failure rates have also been high and steady throughout, stabilising at roughly one fifth of all exits by the end of November (cumulative failure: 19.9%). Transfers to OTP/TFC represented the smallest exit category after death ⁹.

In contrast, Mornei SFP cure rates had nearly reached minimum standards by the end of the same six month period (cumulative cure rate: 60.5%) and default had reduced to within the minimum standard (cumulative default rate: 26.0%) (Table 1 and Figure 3). Failure rates were also lower overall in Mornei than in El Geneina (cumulative failure rate: 4.6%). The transfer rate to OTP/TFC, however, was higher in Mornei (cumulative transfer rate: 8.8%), which can be largely explained by a documented increase in the incidence of acute watery diarrhoea in the camp in June/July. The proximity of the Concern Mornei site to the MSF TFC/hospital may have also encouraged transfers during this time.

Possible contextual factors contributing to differing SFP performance

Before considering investigation results, it is important to note that the El Geneina programme's lower cure rates and higher default and failure rates may be partially explained by wider contextual differences between the two programme areas. In general, the El Geneina programme covers a larger, more diverse population of campbased IDPs, IDPs living with and among residents, and non-displaced residents (roughly 200,000 in total). In contrast, the Mornei programme covers mostly IDPs within the Mornei IDP camp (population of roughly 80,000). El Geneina is an urban, semi-sprawling town, while Mornei is a densely packed IDP camp with more limited trading and lesser access to neighbouring areas. These factors may affect beneficiary behaviour in several ways. First, El Geneina beneficiaries appear more mobile than those of Mornei because El Geneina is larger and more spread out, It also has a more developed cash economy with more opportunities for income generation, although the opportunities are generally irregular. The mixed IDP/resident population of El Geneina also seems to face less uniform security restriction on their movement within and outside the town, due to the diversity of relationships among the different groups (e.g. tribal, etc). Second, people in El Geneina seem more likely to spend time pursuing income generating opportunities because the opportunities exist and perhaps because a larger percentage of the population is forced to earn cash to buy food, since many are not registered as IDPs and therefore do not receive the GFD.

As a result, it seems that carers of El Geneina beneficiaries may perceive a greater opportunity cost for attending SFP distributions where a relatively small amount of Corn Soya Blend (CSB)/oil/sugar porridge mix is distributed, when they could use the time and effort to pursue income generating opportunities. For Mornei beneficiaries, with fewer income generating activities, however, it seems the cost benefit ratio weighs in favour of regular SFP attendance, resulting in more favourable performance indicators. It is interesting to note that the cost benefit of attending OTP distributions, where Plumpy'nut and more focused medical care are provided to a more obviously malnourished child, seems equally high in both Mornei and El Geneina. In both settings, OTP default rates are uniformly low and cure rates consequently higher (Table 1).

Awider livelihoods analysis (including cultivation practices) and further investigation in to differences in clinic access, quality of health care and factors affecting uptake of health care in each location would also be beneficial in understanding the observed performance differences.

Table 2 - Results of 'failure' exit interviews
	EL GENEINA (n=53)	MORNEI (n=27)
WFH% on discharge	Mean: 78.5%+/-3.3%, Median 79% (range 72%-84.5%) Averages are <80% because the majority of children are being discharged from the programme malnourished	Mean 79.2% (+/-3.3%) / Median 80% (range 72%-84%) Slightly skewed distribution because nearly half of the children actually had WFH% ?80% when finally discharged
Length of stay in the programme	Mean 107 days (+/-19 days)	Mean 212 days (+/- 119 days) Very long because some failures were children who had been dis charged failed and readmitted multiple times
Sex	71% girls, 28% boys Expected given the lower caseload of boys (see text)	59% girls, 41% boys Expected given the lower caseload of boys Length of stay in the programme
Age	Median 24 months - younger, complementary feeding age children more likely to fail	Median 16 months - younger, complementary feeding age children more likely to fail
Household head	26% (n=13/50) female headed	Only one child came from a female headed household
Household size	Mean 7 (+/-3)	Mean 5 (+/-1.9)
Residence	Most failures from sites with highest caseloads 30 % reported to be residents 70 % said they were IDPs	All but one child came from an IDP household
GFD	25 % (n=14/50) said they did not have a GFD card	Every child came from a house in receipt of GFD
Main food source	69% stated a reliance on GFD 8% stated a reliance on market purchase Remaining used own produce/received gifts/received payment in kind	All caretakers reported reliance on GFD as their primary food source
Income	83% households had an income Median daily income was equivalent to US$2 The main sources were: Daily labour - 59% Family member with salary - 28% selling agricultural produce - 10% Petty trade - 3%	Every household had an income (although not stated and sources were poorly specified, but petty trade appears to be important)
Water source	Hand pump - 58% Shallow well or wadi (seasonal river) - 42%	All households reported using piped water
Clinic visits	Data was not collected in the El Geneina interviews	Clinic use was high: 26 of the 27 failed children were taken at least two times in the month prior to failure, of which 10 were taken more than two times
Supervisor perceptions of cause of failures at end of interview	40% of failures were perceived to be the result of unresolved or repeated sickness/underlying medical problems 36% to be household food insufficiency and probable sharing Other suspected primary reasons were regular absence (11%) and child dislike of the Corn Soya Blend (CSB) or lack of appetite (6%)	56% were perceived related to poor child care: mother reportedly busy, frequently absent, child not fed regularly. In two cases, abrupt weaning was cited as initial cause not yet resolved. Sharing as a problem was not mentioned Unresolved medical problems, lack of appetite and a problem cooking the CSB mentioned as secondary reasons

Investigation results

Follow up of beneficiaries failing to recover

Follow up of failures highlights different contextual factors that limit individual progress. Interviews were carried out with the caretakers of nearly all the children exited from the SFP as 'failures' in September and October; 53 in El Geneina and 27 in Mornei ¹⁰. Although the sam-ple size was small and selection purposive (only caretakers of failures were interviewed), key findings highlight constraints to improved programme performance in each location (Table 2).

In both programmes, young girls comprise a larger proportion of the failures. This is probably due to a skew in the demographics of the caseload with increased admissions of female children during the complementary feeding period (6 months - 24 months) (data not presented here). Nutrition surveys in El Geneina and Mornei show no higher risk of girls becoming malnourished than boys, indicating that girls may only be more likely to be brought for enrolment. It is possible that they are more likely to fail than boys once enrolled, but a more complex statistical test would be required to assess this. In addition, it appears, unsurprisingly, that children from female-headed households and from larger families are more likely to fail.

Two main contextual differences were indicated in the failure interviews. First, the difference in GFD ration coverage (explained below) was reflected among the non-responders: 30% of El Geneina failures were reportedly residents and 25% came from households not in receipt of GFD, whereas in Mornei only one failure was a resident and all the failures were receiving the GFD. This suggests that sharing may be contributing more to failure in El Geneina, while other factors may be contributing in Mornei. Secondly, access to/use of protected water sources appears poorer among failed children in El Geneina than in Mornei (where a central tap system is in place) ¹¹, suggesting that water borne diseases may constrain individual children's progress more in El Geneina during the time period of investigation.

Information on income was hard to collect and the data from the two sites are not comparable. Nevertheless, answers in El Geneina suggest beneficiary households might, in general, have low and irregular income. As outlined above, the casual nature of income generating opportunities may have a negative effect on attendance and child progress.

Further investigations are now required, involving:

• Interviews with caretakers of cured children for comparison in the form of a beneficiary profile.
• In El Geneina, further investigation of the nature of the SFP children's illnesses, care taker health seeking behaviour and uptake/outcome of clinic referrals from SFP.
• In El Geneina, investigation of the adequacy of protected water sources (access/availabil ity) and reasons for preferential use of unprotected sources.
• Further investigation of the possible bias in bringing girls over boys to SFP and/or greater likelihood of failure among girls.

Observations from Beneficiary Focus Group Discussions

Focus Group Discussions in El Geneina distribution sites confirmed and further explained some of the findings of the failure interviews. They also highlighted other possible factors contributing to high rates of failure, absence and default (Boxes 1 and 2). Chief among them was the indication that the SFP ration was being shared and not lasting the 14 days between distributions and that the targeted utilisation of the pre-mixed CSB (with oil and sugar) for the SFP child, compared to the CSB provided in the GFD, was not understood.

Observations of the programmes in action suggest that a degree of programme fatigue is likely to blame for poor performance indicators. This is most significant among the staff who may not be communicating to mothers clearly why their child is admitted, the aims of the programme (including how long the child is meant to be enrolled and when they should come) and indicators of the child's progress towards meeting that aim. This is in need of improvement.

It is also apparent that the stand alone nature of Concern's El Geneina OTP/may not have as much impact as the integrated programming approach of Mornei, where Concern is the camp manager and the SFP has close links to Concern livelihoods support, health and hygiene promotion and water and sanitation activities that address underlying causes of malnutrition. Although other actors are working in these sectors in El Geneina, coordination across agencies and the larger and more diverse population and area make success in this context more of a challenge.

These observations suggest that communication with caretakers needs to be improved, and that better inter-sectoral programming might be a key to optimising SFP performance.

Difference in GFD coverage of SFP beneficiaries at the two sites

GFD coverage differences by site might be a key reason for the performance differential. The World Food Programme's (WFPs) GFD strategy in Mornei, a town of originally 7,000 and now swollen to approximately 78,000 with IDPs, has been to register all current Mornei residents for an 'IDP scale ration' ¹². As such, it is widely accepted that GFD coverage in Mornei is nearly 100% (note all failed children came from households in receipt of GFD). In El Geneina however, the original pre-conflict resident population is much larger than the current IDP population, and only IDPs are targeted for GFD ¹³. Of the 793 SFP and OTP El-Geneina beneficiary carers asked if they were registered for GFD in October, only 59% (n=468) said they had cards ¹⁴. It seems likely that the lower GFD coverage among El-Geneina's SFP catchment households is a major cause for the lower performance observed ¹⁵.

Recommendations

The investigations shared here were not planned as a research project. However, they have assisted in identifying means to try and improve programme performance, as well as guiding what further investigations need to be undertaken. As a result, Concern is undertaking the following activities:

Providing names of El Geneina SFP beneficiaries (residents, camp-dwelling IDPs and noncamp- dwelling IDPs) lacking GFD to WFP's implementing partner, Save the Children-US, for inclusion in the GFD, which is currently limited to camp-dwelling IDPs only.

Discussing with WFP the feasibility of registering El Geneina residents and non-camp dwelling IDPs for the GFD, and the insufficiency of the proposed further reduced 'resident ration scale' for resident households containing vulnerable children.¹⁶

Investigating probable causes of malnutrition in children from El Geneina resident households through more in-depth interviews and home visits.

Renewing programme focus on addressing the wider causes of malnutrition. As a start, Concern will conduct a Positive Deviant Inquiry to determine factors contributing to malnutrition vs. positive nutritional status among the GFD-reliant population in Mornei and will use results to promote positive behaviours among SFP beneficiary caretakers.

Increasing efforts to coordinate better with other service providers to address underlying causes and link El Geneina SFP beneficiaries to them.

Planning for reassessment of SFP performance following improved GFD coverage and the possibility of initiating livelihoods activities.

Re-initiating regular training for nutrition and outreach staff/volunteers on infant and young child feeding, causes of malnutrition, personal and environmental hygiene, communication skills, community mobilisation.

Distributing BP100 donated by MSF-France as a supplement to the supplementary ration in El Geneina.

Distributing extra non-food items in addition to the premix and soap (starting with cooking pots for those cured and discharged) to increase the mother's cost benefit of attendance and child recovery.

Renewing programme focus on community mobilisation and engagement.

Concern CTC supervisors have been asked to ensure that:

• Communication with mothers on programme aims and benefits are improved and that staff take a problem solving approach with caretakers of children showing slow weight gain.
• Children who meet criteria for 'slow weight gain' are visited at home to determine causes prior to discharge.
• SFP teams improve education on appropriate infant and young child feeding practices and the importance of use of protected water sources.
• Children are properly screened and referred to clinic, referrals are followed through, and results of clinic visits are reviewed and incorporated into analysis of poor recovery.

Outstanding questions

The investigation raised several other questions related to the appropriateness of the current model:

• minimum standards for SFP performance indicate that the general ration needs to be increased?
• How does inclusion of CSB in the general ration affect its perceived value among caretakers? Could education on use and targeting of CSB included in the general ration for children, be an effective alternative to SFP?
• To what degree does illness contribute to malnutrition in relation to household food insecurity in these contexts? Would cash transfers or social protection schemes for IDPs be more appropriate if Sudan, overall, has a surplus of grain production?
• Should cure, death and non-response (failure) rates be used to judge programme quality alone? Or should other indicators of the quality of the supporting services (health, water and sanitation, food security, GFD etc), and the appropriateness of a food-based SFP be developed/considered?
• Are admission criteria appropriate? It seems that many mothers of children with a Weight for Height (WFH)% at the upper end of the eligibility criteria do not perceive their children to be malnourished. If they do not perceive their children to be malnourished, those children will not be given extra food; i.e. the ration will be shared.

Given the widespread failure of Darfur SFPs to reach minimum Sphere performance standards, a number of main points were raised for further discussion with MoH/UNICEF/WFP including: a review of the adequacy/impact of the current GFD ration and its targeting strategy in El Geneina and a review of the current size of the SFP ration as an increase may be required to account for uncontrollable sharing and familial clustering of malnutrition ¹⁷. Alternatively, the GFD ration could be increased, and, due to the culturally engrained practice of food sharing, this may be more cost-effective and have a greater impact than increasing the SFP ration ¹⁸. Finally, improved inter-sectoral programming and coordination among agencies, particularly for targeting and referral, is needed to address effectively the causes of malnutrition, to be led by the coordinating UN agencies.

For further information, contact Victoria Sibson, email: victoria.sibson@concern.net and Kate Golden, email: kate.golden@concern.net.

About the authors

Victoria Sibson started work as a nutritionist with GOAL in Ethiopia in January 2004, from where she moved to Darfur in October 2004. She became MSF-H's Darfur nutrition coordinator in June 2005 and has been a Nutrition Programme Manager for Concern's West Darfur programme since June 2006.

Kate Golden is one of Concern's Dublin-based Nutrition Advisors. Prior to starting this post, she was the Programe Manager for the West Darfur Nutrition Programme (2006) and worked as nutritionist in Concern's South Sudan (2004/5) and Ethiopia (2003/4) programmes.

Many thanks go to Assistant Nutrition Programme Manager, Bahraddin Hassan, for assistance in interpreting the data and to Concern's Nutrition Advisor for the Horn of Africa, Nicky Dent, for assistance editing. Most of all, thanks to Concern's West Darfur nutrition staff and volunteers, many of whom have been working tirelessly in the emergency nutrition programmes since mid-2004.

---

¹ <-2 weight for height Z-scores in the 6-59 month old age group

² Nutrition in Crisis Situations, Report no.11 UN SCN, November 2006

³ Concern nutrition surveys of el Geneina town

⁴ Concern's current protocols state that SFP children who are absent at three consecutive distributions are discharged as defaulters on the third occasion. (This default criteria is endorsed by UNICEF W. Darfur and used by most agencies).

⁵ MSF Nutrition Guidelines, Paris, 1995

⁶ Other measures of performance include average length of stay, rate of weight gain, and coverage but are not calculated regularly and are not a focus of this article.

⁷ Concern's current SFP protocols state that children not reaching a weight for height of ?85% at two consecutive weighings before a maximum of nine consecutive fortnightly distributions, should be discharged as failed, if they have not already been exited in another category. Failures who meet admission criteria on the day of discharge are readmitted the same day.

⁸ e.g. 19/03/06, West Darfur state Nutrition working group meeting summary: Current Problems relating to the Poor Performance of the SFP in West Darfur, UNICEF

⁹ Importantly, the death rate represents only those deaths reported by community sources after follow-up by team. Due to SFP caseload, some but not all defaults are followedup to rule out death as a cause; therefore, the number of SFP deaths could be slightly higher than represented here.

¹⁰ Failures in Mornei were much less frequent than in El Geneina during this period, accounting for only 3% of exits in October 2006 vs. 18% in El Geneina.

¹¹ That said, a significant 'crisis' in provision of water through the tap system was reported in June/July, with many households reportedly reverting to shallow wells, resulting in increased diarrhoea.

¹² The West Darfur ration scale differentiates residents from IDPs. The 'IDP ration scale' applied in El Geneina and in Mornei is 13.5kg cereal (100% full standard ration), 0.375kg pulse (25% full standard ration), 1.12kg CSB (75% full standard ration), 0.675kg oil (75% full standard ration), 0.075kg salt (25% full standard ration), and 0.225kg sugar (25% full standard ration). This provides about 90% of population requirement for energy (recommended 2100kcal/person/day); 112-123% protein and 70-79% fat (depending on cereal and pulse included). Amounts of calcium, iodine, vitamin A, riboflavin and vitamin C are deficient (data analysed using NutVal 2004).

¹³ An estimated 197,507 reside in the feeding centre catchment area (data from 2006 National Immunisation Day), while October distribution monitoring data from Save the Children US indicated that only 65,935 people (33% of total) in the catchment area received a ration.

¹⁴ This is lower than the percent having GFD cards found in September (525/724 or 73%), which reflects active case finding and admission among the resident population in late Sep/early Oct.

¹⁵ A 2005 MSF-H internal investigation of SFP performance, including South and West Darfur, supports this suggestion: "that SFPs have the best impact on child recovery where access and food aid is relatively stable" (Evaluation of Supplementary Feeding Programmes in Darfur, Uganda and Somalia, MSFH October 2005)

¹⁶ The proposed resident ration scale for Mornei or El Geneina comprises: 6.75kg cereal (75% full standard ration); 0.375kg pulse (25% full standard ration); 0.75kg CSB (50% full standard ration); 0.45kg oil (50% full standard ration); 0.075kg salt (25% full standard ration) and 0.225kg sugar (25% full standard ration). This provides about 50% of the population requirement for energy and half of the recommended ration of 2100 kcal/person/day

¹⁷ It is recommended a take-home supplementary ration should provide 1000-1200 kcal/person/day (Management of Nutrition in Major Emergencies, WHO Geneva, 2000). The current WFP ration is on the lower end of that scale, providing 1017 kcals/person/day from 200g CSB, 20g oil, 20g sugar/person/day, providing 1017kcals (14% energy from protein and 28% from fat)

¹⁸ It must be noted that other agencies in Darfur have distributed much larger rations to SFP beneficiaries with limited impact on performance indicators, e.g. MSF-H 2005 South and South-West Darfur (personal communication).