FOOD SECURITY AND NUTRITION ASSESSMENT IN REFUGEE SETTLEMENTS

Transcription

1 ithe FOOD SECURITY AND NUTRITION ASSESSMENT IN REFUGEE SETTLEMENTS REPORT November 2010

2 Food Security and Nutrition Assessment in Refugee Settlements Report

3 Acknowledgements The International Baby Food Action Network (IBFAN) Uganda is deeply grateful to the individuals, households and communities of refugees for their time and hospitality. This assessment report is the culmination of month s worth of effort of many people and organizations. The assessment was truly an integrated food security and nutritional assessment, and many individuals were involved in its design, the collection of data and the production of this report. We are deeply appreciative for the helpful comments of various organizations on the design of the survey, particularly the UN World Food Programme (WFP) and United Nations High Commissioner for Refugees (UNHCR). Special mention goes to Daniel Molla, Dorothy Nabiwemba Bushara, Kenneth Anyanzo, and Zachaeus Ndirima, Dr. Kevin Tsatsiyo and Dr. Julius Kasozi for their technical support in the planning process as well as during the implementation of the Assessment and finalization of the report. In the field, many organizations contributed staff, vehicles and other logistical support to ensure that everything went smoothly. In particular, we would like to thank OPM, WFP, UNHCR, NRC, AAH, GTZ and AHA. We also appreciate our gratitude to GTZ, AHA and OPM for providing full-time staff to the assessment. This report will be very useful in contributing to the further improvement of the food security, health and nutritional status of the population in the Refugee Settlements through providing information for proper planning and implementation of nutrition, food security and health related interventions. Dr. Gelasius Mukasa EXECUTIVE DIRECTOR - IBFAN UGANDA ii

4 Team Members Nakivale/Oruchinga Settlements (Isingiro District) John Jagwe Judith Awori Innocent Arinaitwe Louis Kasadha Kyangwali/Kyaka II (Hoima and Kyegegwa) Gerald Onyango Fiona Kuziga Lawrence Nyende Principal Investigator/Team Leader Barbara Nalubanga Tembo Co Investigators John Jagwe John Musinguzi John Musisi Data Manager Gerald Onyango Adjumani/Palorinya (Adjumani & Moyo) John Musisi Tim Mateeba Godfrey Idrifua Imvepi/Rhino/Kiryandongo (Arua & Kiryandongo Districts) John Musinguzi Gladys Njuba Grace Nambuusi Simon Opido iii

5 Table of Contents ACKNOWLEDGEMENTS... II TEAM MEMBERS... III ABBREVIATIONS AND ACRONYMS... XI EXECUTIVE SUMMARY... XII 1.1 Background Justification Overall Objective Specific Objectives The Food and Nutrition Conceptual Framework... 5 CHAPTER 2: METHODOLOGY Introduction Sample Size for Household Survey Sampling procedure for household survey Sampling Frame Survey Coordination Training Data collection Household survey Focus Group and Key Informant Interviews Anthropometric measurement Indicators, guidelines and formulae used Malnutrition Feeding Practices Food Security Data Management Data analysis Consent Limitations and potential biases CHAPTER 3: FINDINGS Demography iv

6 3.1.1 Study Area Age and Sex Distribution Child Health and Nutrition Child Malnutrition Main Causes of Child Malnutrition Child health Infant and Young Child Feeding Iron Deficiency Anaemia Iron Deficiency Anaemia among Children (6 to 59months) Main Causes of Iron Deficiency Anaemia Iron Deficiency Anaemia among Women (15 to 49years) Main Causes of Iron Deficiency Anaemia Health Interventions Treatment of Household Members Access and use of mosquito nets Coverage of Vitamin A supplementation Coverage of measles immunization Mortality Maternal Health Water and Sanitation Water Sanitation Food Availability: Crop and Animal Production Land Access and Agricultural activities Cultivation in the first season Total Yields cultivated in the previous season Livestock production Livestock Ownership by Type Relationship between Food Consumption and Animal ownership Food Access: Income, Expenditures and Coping Strategies Number and type of income sources Share of Food, Health and other expenditures Level of Monthly food expenditures and type of food purchased Coping strategies in the event of food shortages Main types of coping strategies according to the various household groups Coping Strategy Index Food Consumption Patterns Food consumption diversity and frequency Principles of the food consumption pattern analysis Main Food Sources Coverage of Food and Non Food Assistance v

7 Receipt of Food Aid Receipt of Food Aid by Sub Region Receipt of Food Aid by Programme Type of food aid commodities received Receipt of non-food items per sub region and per household group Household Food Insecurity and Risks to Lives and Livelihoods Prevalence of household food insecurity and short term risks to lives and livelihoods Child Education Household Priorities for Refugees Immediate Requirements/Priorities Longer Term Priorities CHAPTER 4: DISCUSSION, CONCLUSION AND RECOMMENDATIONS Nutrition Response options to address nutrition Food Security Populations at risk in need of assistance Response options to address food security Recommendations on programming and assessment follow-up ANNEXES Annex 1: Glossary of Terms Annex 2: Detailed Analysis by Settlements Annex 3: Comparison 2009 and 2010 Assessment Results Annex 4: Household Questionnaire Annex 5: Focus Group Questionnaire Annex 6: Key Informant Questionnaire Annex 7: NCHS Anthropometric Data Annex 8: Calendar of Events Annex 9: Uganda Refugee Statistics by Nationality and Settlement as of August 31, Annex 10: Quality Checks Annex 11: Survey Team Annex 12: List of Contributors Annex 13: Multivariate Logistic Regression Model for Key Indicators vi

8 List of Tables Table 2.1: Types of Data Collection... 7 Table 2.2: Assumptions and Estimated Sample Size per Sub Region... 8 Table 2.3: Proposed clusters by Sub Region... 9 Table 2.4: Classification of nutritional status using WFH expressed in Z-scores Table 2.5: Classification of children with chronic malnutrition by using Height For Age Table 2.6: Classification of Severity of Malnutrition in a Community by Prevalence of Acute and Chronic Malnutrition for Children Under 5 years of Age Table 2.7: BMI classification of Nutritional Status Table 2.8: Nutritional Status classified using MUAC Table 2.9: Cut off values for Crude Mortality and Under-5 Mortality Rates Table 2.10: Recommended Haemoglobin levels in g/dl Table 2.11: Cut offs for Anaemia prevalence in Communities Table 2.12: Copying Strategy Index calculation Table 2.13: Template for calculating the Food Consumption Score (FCS) Table 2.14: Food Consumption Groups (FCGs) and corresponding FCS thresholds Table 2.15: Typologies of the households according to their food consumption and access Table 3.1: Sampled sites by Sub Region and District Table 3.2: Percentage Distribution of Age and Sex among Sampled Children in South West sub region refugee settlements, October Table 3.3: Percentage Distribution of Age and Sex among Sampled Children in West Nile sub region refugee settlements, October Table 3.4: Acute malnutrition rates among children 6 to 59 months of age by Sub region, October, Table 3.5: Prevalence of Global Acute Malnutrition by Age category, October Table 3.6: Prevalence of Stunting and Underweight among children 6 to 59 months of Age, October, Table 3.7: Nutritional Status by MUAC in sampled children by Sub region, October Table 3.8: Morbidity among Children 6 to 59 months of Age, by Sub region, October Table 3.10 Prevalence of Anaemia among Children by Age and Sub Region Table 3.11: Vitamin A coverage among children 6 to 59 months, October, Table 3.12: Measles immunization coverage among children, October, Table 3.13: Crude and under-5 mortality rates for the among samples populations 2009 and vii

9 Table 3.14: Prevalence of Malnutrition among women (15 49 years) using MUAC and BMI, October Table 3.15: Main of Sources of Water by Sub Region, October 2010 (n = 2,070) Table 3.16: Amount of water used per person per day by sub region Table 3.17: Average Quantities of Crop Yields for the First Season of Table 3.18: Animal Ownership TLU Categories by Food Consumption Group, October Table 3.19: Income sources for households Table 3.20: Households owning Agricultural productive assets Table 3.21: Households owning Livestock assets Table 3.22: Households owning Domestic assets Table 3.23: Proportional expenditure on food, health, education and other items Table 3.24: Average expenditures of households on selected food items in a 30-day period Table 3.25: Average Monthly Food Expenditures per capita (UGX) Table 3.26: Coping strategies adopted by households by Sub region, October Table 3.27: Diet characteristics of the poor food consumption pattern Table 3.28: Diet characteristics of the borderline food consumption pattern Table 3.29: Diet characteristics of the acceptable food consumption pattern viii

10 List of Figures Figure 1.1: Refugee Settlements in Uganda... 1 Figure 1.2: The Food and Nutrition Security Conceptual Framework... 5 Figure 1.3: Enumerator interviewing a mother... 6 Figure 2.1: Structure of the survey team Figure 2.2: A child being tested for anaemia during assessment Figure Distribution of weight-for-height z-scores for overall sample compared to the reference population Figure 3.5: Prevalence of Anemia by Parity Figure 3.6: Sources of treatment for household members, October 2010 (n = 2,054) Figure 3.7: Comparison of mosquito net coverage in 2010 and 2009 by sub region Figure 3.8: Vaccination and Supplementation Coverage in the Sub regions, October Figure 3.9: Possible causes of death in the under and over fives, October Figure 3.10 Proportion of households able to cultivate some food in 1 st season of Figure 3.11: Main reasons for not cultivating any food in the 1 st season of 2010 in West Nile Region Figure 3.12: Main reasons for not cultivating any food in the 1 st season of 2010 in South-West Region Figure 3.13: Post Harvest handling practices among Refugees in South West Sub Region Figure 3.14: Average Quantities of crop yields for the first season 2009 compared to Figure 3.15: Livestock ownership by type by sub region Figure 3.16: Main constraints facing households engaged in livestock production Figure 3.17: Number of income sources per sub region Figure 3.18: Number of Income Sources by Food Security level in West Nile Figure 3.19: Number of Income Sources by Food Security level in South West Figure Wealth Quintiles for the West Nile and South West Refugee Settlements Figure 3.21: Coping Strategy Index by Sub Region Figure 3.22: Proportion of different food consumption groups by Sub region, October Figure 3.23: Food Consumption Pattern and Income Sources, Refugee Settlements Figure 3.24: Food consumption pattern and income source, October Figure3.25: Food consumption pattern and sex of Household Head, October Figure3.26: Food consumption pattern by income source, October ix

11 Figure 3.27: Households which received food aid in the period Jan-Aug Figure 3.28: Food Aid Beneficiaries Jan-August Figure 3.29: Non food items received by households Figure 3.30: Household Food Security by Sub region Figure 3.31: HH Food security status by sub region Figure 3.32: Household Food Security Status by Sex Figure 3.33: Ability to cultivate in Figure 3.34: Main constraints to crop cultivation and food security status Figure 3.35: Household food security by Animal Ownership Figure 3.36: Animal Ownership by food security groups Figure 3.37: Main constraints to crop cultivation by Food Security status Figure 3.38: Household Food security by income sources Figure 3.39: Household Food Security status by number of income sources Figure 3.40: Household Food security status by food expenditure Figure 3.41: Household Food security status by asset ownership Figure 3.42: Household Food security by coping strategy Figure 3.44: Percentages of boys and girls aged 6-12years not attending school Figure 3.43: School attendance of children 6-12 years by sub region Figure 3.46: Reasons why girls were not attending school Figure 3.45: Reasons why boys were not attending school Figure 3.47: Immediate priorities of households by sub region x

12 Abbreviations and Acronyms AAH Action Africa Help - International AHA Africa Humanitarian Action ARIs Acute Respiratory Infections CED Chronic Energy Deficiency CI Confidence Interval CMR Crude Mortality Rate DPT Diphtheria Pertusis Tetanus DRC Democratic Republic of Congo EFSA Emergency Food Security Assessment FAO Food and Agriculture Organization of the United Nations FGD Focus Group Discussion GAM Global Acute Malnutrition GTZ German Technical Assistance HFA Height For Age HH Household IBFAN International Baby Food Action Network IDA Iron Deficiency Anaemia MoH Ministry of Health TLU Tropical Livestock Unit MUAC Mid-Upper Arm Circumference NCHS National Centre of Health Statistics NGO Non-Governmental Organization NRC Norwegian Refugee Council OR Odds Ratio OPM Office of the Prime Minister OTP Outpatient Therapeutic Feeding Programme PCA Principal Component Analysis PPS Probability Proportionate to Size RDO Refugee Desk Officer SAM Severe Acute Malnutrition sd Standard Deviation SFC/P Supplementary Feeding Centre/ Programme SMART Standardized Monitoring and Assessment of Relief and Transition TFC/P Therapeutic Feeding Centre/Programme U5MR Under-5 Mortality Rate UDHS Uganda Demographic and Health Survey UGX Uganda Shillings UNHCR United Nations High Commissioner for Refugees UNICEF United Nations Children s Fund VAM Vulnerability Analysis and Mapping WES Water Environment and Sanitation WES Water and Environmental Sanitation WFA Weight For Age WFH Weight For Height WFP World Food Programme WHO World Health Organization xi

13 Executive Summary Background Several food security, health and nutrition assessments have been carried out in the Refugee Settlements since These have been conducted with technical and financial support from UNHCR, World Food Programme as well as the various implementing partners and Government. Until recently, there had not been a clear pattern of improvement or deterioration across the Refugee settlements in the sub-regions of Westnile and South West. The Government as well as the various partners have constantly put in place interventions to maintain the current levels of food security and nutritional status. While the levels of acute malnutrition and food insecurity in the assessment conducted in 2009 were not of major concern, it is important to re-assess the situation in order to find out how the situation may have evolved. This re-assessment will also be useful in monitoring the effectiveness of ongoing interventions as well as identify any gaps in their implementation. Based on this, a cross sectional population survey was conducted to assess the levels of health, nutrition and food security among the refugee populations in the settlements across west Nile and south Western sub regions in Uganda, October Objectives The overall objective is to assess indicators of health, nutrition, food security and retrospective mortality in the refugee settlements in order to generate information for improved programme and policy interventions. Methods The survey was designed to statistically represent the population in each of the two sub regions of South West and West Nile defined by the United Nations (UN) as Refugees. This population numbered approximately 109,317 people at the time of the survey and included persons living in the 9 refugee settlements. A two-stage cluster sampling method was employed. A total of 72 clusters were initially selected using a probability proportionate to size (PPS) method, however 5 additional clusters were assessed. Data were collected from 2,072 households with 11,510 individuals, and 2,035 children were assessed. Findings Nutrition: The prevalence of Global Acute Malnutrition among the population in Refugee Settlements was 2.2 percent which according to the National Guidelines on Nutrition Assessment was within the acceptable levels of below 5percent. The prevalence was about the same in West Nile at 2.2pecent and South West at 2.1 percent. However, prevalence of stunting of 26.7 percent was rated poor according to the guidelines in the 2 sub regions and specifically serious for South West sub region at 35.8percent. The prevalence of severe acute malnutrition xii

14 (SAM) among children 6 to 59 months of age fell to 0.4 percent in 2010 [95% CI: %] from 0.5 percent [95percent CI: %] in The crude mortality rate (CMR) was 0.37 deaths per 10,000 per day [95% CI: ], which was below the emergency threshold of one death per 10,000 per day. The under-five mortality rate (U5MR) was 0.33 deaths per 10,000 per day [95 % CI: ] and was also below the emergency benchmark of two deaths per 10,000 per day. While breastfeeding practices were quite good among refugees (99.4 percent of children less than two years of age had been breastfed at some point during their life), the rate of exclusive breastfeeding (24.4percent) among children less than six months of age indicated a need for improvement. The promotion of breast feeding and complementary feeding practices also needed improvement only 8.1 percent of children aged 6 to 24 months received four to five meals consisting of mainly cereals per day. Acute malnutrition among children is a very sensitive indicator to changing conditions. While this survey found that 2.2 percent of children were moderately or severely malnourished, an additional 5.9 percent of children were found to be mildly malnourished, using MUAC as an indicator. These children could rapidly fall below this threshold if conditions deteriorate. Until positive and sustainable changes to livelihoods that lead to improved food security among the population are attained, it is essential that support be continued comparable to what is provided at the current levels. As the situation improves, and refugees are able to engage in agricultural livelihoods, gradual reduction of such support will be imperative. Anaemia: The 2009 assessment identified IDA as a public Health problem in a severe form in the vulnerable population groups and there has not been any change in Of the 1,093 children tested for anaemia, 53.6percent (95% CI ) were anaemic (<11g/dl). Anaemia was more prevalent in children from West Nile at 61.1percent (95% CI ) than those from South West at 45.3percent (95% CI ). The prevalence of Iron Deficiency Anaemia was 32.5percent among pregnant women contributing the highest proportion compared to non pregnant at 17.5percent, giving an overall prevalence of IDA among women at 27.4percent. The main causes of IDA especially among children were significantly related to child illness, method of faecal disposal with households using bushes and community latrines having the high proportion of children with anaemia; use of unsafe water for drinking and the treatment methods for drinking water with boiling of drinking water causing high prevalence of anaemia compared to those using chlorination (e.g. aquatab). Nutrition, health and food-related programmes: Fever was highly prevalent among children aged 6 to 59 months of age and reportedly affected 35.3 percent [95% CI: %] over the two weeks prior to the survey. The availability of mosquito nets at the household level improved from 48.9 percent in 2009 to 52.3 percent [95% CI: %] in Mothers/caretakers sought treatment for 99.1 percent of children reportedly suffering from any type of illness in the previous two weeks. Almost all child illnesses (94.2percent) were reportedly treated at health centres, followed by 3.4 percent treated at hospitals. Health services were mostly provided by Non-Governmental Organizations (NGOs) in existing government facilities in the settlements. Coverage of measles vaccination among children 9 to 59 months of xiii

15 age was 57.9percent [95% CI: percent] with evidence of a card and vitamin A supplementation coverage was 49.5percent [95% CI: percent]) Overall, 87.8 percent [95% CI: %] of households were using a safe water source for drinking water. This does not however include households that had access to treated water from a potentially unsafe source since in most times households also treated water from the clean source and the latrine coverage was noted to be at 70.6 percent [95% CI: percent] with South West having a higher coverage of 73.9percent [95% CI: percent] than West Nile 67.4percent [95% CI: percent]. Compared to the 2009 report, there was an improvement in latrine coverage from 61.7 per cent to 67.4percent in West Nile while a drop was noted to be from 79.6percent to 73.9percent in South West. Overall, approximately 75.9percent of the households admitted to having received food aid in the period January to August 2010 with the highest proportion of households reported in South West sub region at 91.6percemt than West Nile at 63.1percent. This coverage steadily dropped from the month of January through to August 2010 from 94percent in the first quarter to 84.4percent in the third. Food security: In 2010 food consumption in households declined from that of Overall the proportion of households with acceptable food consumption decreased from 57.5 percent to 38 percent, while households with poor food consumption decreased from 12 percent to 7.7 percent. The highest proportion of households that had acceptable food consumption was in South West at 45.3percent compared to West Nile sub region at 31percent, a decline from 63percent and 52percent respectively. Household food sources and a variety of other data (e.g. expenditure, asset ownership, income sources and household demographics) were analyzed to establish whether the observed household food consumption levels were sustainable and based on self-reliant means. This analysis revealed that the decline in food consumption could have been a result of fewer sources of income to purchase food, over reliance on unskilled and agricultural labour, female headed household and illiteracy. Availability of food aid as the major source of food could not have played a role given that the proportion of households reported obtaining food as the major source remained the same as that of 2009 at 20percent. The total number of people in Refugee Settlements in South West and West Nile sub regions was estimated at 115,885 as of August 2010, and of this figure, 17,924 were in West Nile and 97,961 in South West. The scale of the population at risk was calculated by multiplying the survey findings, i.e. the proportion of households for each food security level by these population data. Approximately 80percent of all households required assistance to ensure a minimum degree of adequacy in their access to food with about 44.6percent (55,985 people) to be considered for food aid. Therefore, targeting modalities for this population should be strengthened as high priority assistance needs exist for such households identified as being foodinsecure. The group of vulnerable households accounts for just more than one third (35.9percent) of the refugee population in the South West and West Nile settlement (i.e. approximately 38,042 people) and should, wherever possible, be reached through modalities other than general relief rations, i.e. Food for Work, school feeding, and Food for Education. xiv

16 Other forms of income and development support (e.g. agricultural inputs and extension) should also be a priority Recommendations Nutrition and health: Coverage of measles immunization, de-worming and vitamin A capsules was observed to still be below the national targets. Where coverage was not sufficient, mop-up campaigns should be conducted. Assessments are needed to determine the true coverage rates following immunization campaigns, and to identify the reasons for low coverage. Working to boost coverage of immunization and other programmes (e.g. vitamin A supplementation, bednets, de-worming and selective feeding) through integrated child health day programming is recommended; Health services, particularly those provided by NGOs, were utilized by the population to treat child illness. For the longer term, there is a clear need to strengthen government health services (particularly through outreach); Community mobilisation for disease prevention should be focused on in the next implementation strategy. Whereas overall household access to mosquito nets has improved from 48.9 percent in 2009 to 52.3 percent in 2010, there is clear need to expand both the coverage and usage of nets among children less than five years of age as well as pregnant women; malnutrition among pregnant and lactating women was a notable problem overall, with 8.3percent of these women being underweight. Rather than pursuing a blanket approach to supplementary feeding for pregnant and lactating women, the need for supplementary feeding should be based on local assessments to identify pockets of high maternal malnutrition. Zonal teams should be utilised to identify the malnourished women and managed appropriately; and several partners reported to support health and nutrition related interventions therefore the collaboration and coordination between such partners and the district local government should be strengthened by putting in place a strong coordination mechanism. Iron Deficiency Anaemia: Targeted interventions to provide iron supplements to especially pregnant women need to be strengthened further especially monitoring their consumption of such supplements; Food-based approaches to increase iron intake through food fortification and dietary diversification are important sustainable strategies for preventing iron deficiency among refugees; given that the assessment found out that the current water, sanitation and hygiene practices were strongly associated with the anaemia prevalence in the settlements; national guidance on de-worming should strongly be followed and enforced; utilization of ITNs should be promoted through community campaigns and/or other available opportunities and channels; improving uptake of family planning services as well as involvement of men in preventing and controlling anaemia in women is critical; promotion of exclusive breastfeeding for 6 months, followed by breastfeeding with complementary feeding into the second year of life will contribute to the control of iron deficiency anaemia in children; Strengthening the Health Management Information System to be able to provide reliable information on prevalence and the effectiveness of interventions should be prioritised; and finally, to be effective and sustainable, strategies must be led with firm technical as well as political commitment and strong partnerships involving all relevant sectors/line ministries xv

17 Food: Continued support in terms of regular and timely delivery of food aid is needed to sustain gains; in order to facilitate future surveys, improve targeting and programming, efforts should also be continued in providing detailed information (village level population data) by the Office of the Prime Minister (OPM); In some areas, more effort needs to be made to familiarize the population with the importance and how to prepare fortified blended food (CSB). Mechanisms should be established in problematic areas to ensure the dissemination of key nutrition messages to communities (perhaps alongside food distribution), and that the population gets maximum nutritional benefit from CSB and other food commodities. Since maize has been used as both food and cashcrop, seedlings/seeds of other crops should be provided, such that emphasis is put on food consumption than selling; and different partners provide support in the area of food security in terms of availability and access, however, little emphasis has been put on utilization and therefore building the capacity of key players in the area of nutrition such as CHWs and VHTs is critical for this purpose. Selective feeding: The levels of Global Acute Malnutrition of below 5percent do not warrant selective feeding; therefore strengthening government capacity to prevent and/or manage malnutrition within the Health Facility settings and communities is recommended. Water and sanitation: There is particular need to further improve access to clean and safe water and sanitation for all refugees to at least 95percent coverage. The amount of water used per person per day should be increased especially in the South West sub region from about 12l/p/d to at least 15l/p/d; efforts should be made for almost all households to own latrines of standard size (4x2x15feet) and 30metres away from a house and any water source. The latrines should be well built with a wall and roof as well as availing hand washing facilities (water and soap) nearby; latrines based in trading centres, health facilities, schools and/or other institutions should also benefit from the WATSAN related support and/or interventions. During the land distribution process, efforts should be made to ensure that it is 100metres away from a water source to avoid contamination of the source and efforts should be made to strengthen the community based maintenance system to avoid breakdown; the design of houses to allow adequate ventilation and separation of people from animals should be emphasized and provision of durable roofing materials such as iron sheets should be done. Design of standard shelters should be provided to refugees on arrival; promote tree planting for housing and shelter; and strong behaviour change communication interventions should be put in place such as handwashing through use of pulley jerican as washing facilities and soap; effective boiling of drinking water as well as storage; and household chlorination of drinking water. Food Security: For the next year, general relief food rations will continue to be required for about 80percent of refugees that have been identified as food-insecure and vulnerable (at risk). However, ongoing monitoring and, as appropriate, adjustments in the general food distribution will be essential, including verification of the actual crop quantities harvested, changes in food availability, prices and income earning opportunities. Enhanced coverage of agricultural assistance will be critical to reduce the food aid reliance of settled households and enable the successful resettlement of refugees. It is recommended to keep the food basket under regular review and explore resourcing of locally available and acceptable food staffs wherever possible. xvi

18 Chapter 1: Introduction 1.1 Background Uganda enjoys a fair degree of political stability, especially after having undertaken relatively successful military operations against the Lord s Resistance Army (LRA). However, the seasonal fighting of various groups and militias in the neighbouring Democratic Republic of the Congo (DRC) has forced thousands of Congolese to flee to Uganda. More than 50,000 have entered the Kisoro and Kanungu districts since November Nearly 30,000 Sudanese repatriated in 2009, bringing the total number of Sudanese returnees from Uganda to some 100, Currently, there are nine (9) refugees settlements of Imvepi and Rhino Camp (Arua District); Palorinya (Moyo district); Adjumani (Adjumani district); and Kiryandongo 2 (Kiryandongo district) in West Nile region while Kyaka (Kyegegwa district); Kyangwali (Hoima district); and Nakivale and Oruchinga (Isingiro district) are located in the South Western sub region as shown in Figure 1.1. As of August 2010, the refugee statistics indicated about 150,597 refugees were living in Uganda (refer to Annex 9) and mainly from the DRC (55.8percent), Rwanda (10.3percent), Somalia (10.5percent), Sudan (13.1percent), Eritrea (3.7percent), Burundi Figure 1.1: Refugee Settlements in Uganda 1 UNHCR Global Appeal 2011 update Uganda, 2 for the purpose of this Assessment, Kiryandongo was grouped with the Westnile region, 1

19 (4percent) and other countries such as Ethiopia, Kenya, Pakistan, Chad, and Tanzania contributing 2.6percent to the total refugee population. The refugee settlements of Nakivale, Kyangwali and Kyaka II hosted the highest number of refuges as indicated in Figure 1.1 above Many refugees, particularly those newly arriving from the DRC, continue to lack adequate access to water, social and basic health services. Many also have little or no access to post-primary education, and have no legal remedies in cases of sexual and gender-based violence during flight. The provision of adequate food assistance has also been a challenge, which contributes to insecurity in the refugee settlements. Livelihood opportunities and agricultural production need to be augmented to complement the food assistance and to prevent long-term dependency on humanitarian assistance3. In 2009, the Global Acute Malnutrition rate for West Nile and Southwest sub-regions was found to be at 3.4percent and 2.8percent respectively with no clear pattern of improvement or deterioration from the previous assessments. Levels of GAM were highest in the settlements of Adjumani, Imvepi/Rhino Camp and Kyangwali at 4.7percent, 4.5percent and 4.2percent respectively while Kyaka, Palorinya and Nakivale had the lowest of 1.5percent, 2.3percent and 2.1 percent respectively. The survey was undertaken in August during the major harvest and just after the hunger gap that falls between March to July each year. The food intake in refugee settlements was found to be adequate at 88percent, but in most cases, depended on food assistance. In West Nile and Southwest, depending upon the circumstances in the settlement, the food assistance comprised between 40 and 60 percent of their recommended daily allowances (RDA)4. New arrivals and extremely vulnerable individuals received 100 percent. One of the key constraints was that there was limited access to cultivable land. Under the Development Assistance to Refugee-Hosting Areas (DAR) programme, the Government of Uganda provided land to refugees, but the pieces were of limited size. The most prevalent micronutrient deficiency disorder among refugees was anaemia with almost 54.4 percent of children and 32.7percent of mothers being anaemic 5. Vitamin A supplementation with evidence of a card was noted to be low at 33.2percent increasing the risk of refugees to Vitamin A deficiency. UNHCR is working together with the Government of Uganda in support of the Peace Recovery and Development Plan which aims to facilitate the transition from relief to recovery and development in northern Uganda. Activities include among others strengthening the capacities of local communities and authorities through a three-year 3 UNHCR Global Appeal 2011 update Uganda, 4 WFP, UNHCR, IBFAN Uganda. Food Security and Nutrition Assessment in Refugee Settlements, August

20 Post Repatriation Rehabilitation and Livelihood Programme designed to benefit some 50,000 people in the settlements 5. UNHCR will also continue to assist refugees who live in the settlements and in urban areas in Kampala. This will mainly be done by supporting government operationalise the new Refugee Act as well as institutionalize protection structures; responding to the continuous influx of Congolese asylum-seekers in a timely and effective manner; through provision of favourable protection environment, fair protection processes and basics needs and services through ensuring that all people of concern are able to avail themselves of their protection, social and economic rights; while promoting durable solutions for refugees in the settlements including voluntary repatriation for those who wish to return. All these are expected to be achieved by the end of 2010 through strengthened and improved protection delivery; protected natural resources and the shared environment; strengthened partnerships with local actors; increased access by refugees to primary education and basic health services. Furthermore, basic domestic and hygiene items will be available to all refugees; improved shelter and infrastructure in settlements; satisfactory sanitary conditions put in place; and increased livelihood opportunities. 1.2 Justification Several food security, health and nutrition assessments have been carried out in the Refugee Settlements since These have been conducted with technical and financial support from UNHCR, World Food Programme as well as the various implementing partners and Government. Until recently, there has not a clear pattern of improvement or deterioration across the Refugee settlements in the sub-regions of West Nile and South West. The Government as well as the various partners have constantly put in place interventions to maintain the current levels of food security and nutritional status. Following the 2005 Assessment, the levels of anaemia were high in both women of the reproductive age group and children and a number of various interventions were put in place by WFP and UNHCR. While the levels of acute malnutrition and food insecurity in the assessment conducted in 2009 were not of major concern, it is important to re-assess the situation in order to find out how the situation may have evolved. This re-assessment will also be useful in monitoring the effectiveness of ongoing interventions as well as identify any gaps in their implementation. Based on this, a cross sectional population survey was conducted to assess the levels of health, nutrition and food security among the refugee populations in the settlements across West Nile and South Western sub regions in Uganda. 5 UNHCR Global Appeal 2011 update Uganda, 3

21 1.3 Overall Objective The overall objective is to assess indicators of health, nutrition, food security and retrospective mortality in the refugee settlements in order to generate information for improved programme and policy interventions Specific Objectives i. To determine the levels of retrospective crude mortality rates and age specific mortality rates for under-5s in a specified time period ii. iii. iv. To assess the current feeding practices of the children in the target group, children 6-59 months; Determine prevalence of malnutrition (wasting, stunting and underweight) among children aged 6-59 months (and/or measuring cm in length or height); Determine the prevalence of anaemia among children 6-59 months and pregnant and non pregnant women years; v. Determine the coverage of health interventions (e.g. routine immunization coverage, DPT, Measles, polio and de-worming) and Vitamin A supplementation among children under five; vi. vii. viii. ix. Assess access to and coverage of safe water and sanitation facilities; Assess mosquito net coverage; Determine incidence of common diseases (fever, diarrhoea, measles and ARI) among the target population, two weeks prior to the assessment and access to/uptake of health services for treatment; Undertake factors associated with malnutrition; x. Assess the current food security status of households, including food consumption and dietary diversity (using 7-day dietary recall methods) and use of coping strategies; xi. xii. xiii. Analyze factors that determine household food security status; Determine the coverage of food security assistance provided to households since January 2010; and Recommend appropriate immediate as well as medium to long term courses of action by the government, UNHCR, WFP and other stakeholders based on the findings of the assessment. 4

22 E X P O S U R E T O S H O C K S A N D H A Z A R D S 1.4 The Food and Nutrition Conceptual Framework The food and nutrition security analysis was based on an understanding of food and nutrition security as well as vulnerability. The Food and Nutrition Security Conceptual Framework shown in Figure informed not only the selection of indicators for analysis, but also the design of field assessment instruments. The household food security conceptual framework adopted by EFSAs considers food availability, food access and food utilization as core determinants of food security, and links these to households asset endowments, livelihood strategies, and political, social, institutional and economic environment. During the food and nutrition security assessment, the conceptual framework served to Context / Framework Food Availability / Markets Basic Services and Infrastructure Political, Economical, Institutional, Security, Social, Cultural, Gender, Environment Agro- Ecological Conditions / Climate Individual Food intake Household Food Access Nutrition Status / Mortality Care / Health Practices Health Status / Disease HH food Production, Gifts, Exchange, Cash, Earnings, Loan, Savings, Transfers Natural Physical Human Economic Social Capital / Assets Health and Hygiene Conditions Figure 1.2: The Food and Nutrition Security Conceptual Framework provide a way of visualizing the relationships among factors that affect food and nutrition security, which was helpful during data collection and analysis as well as report writing. Individual Level HH Level Livelihood Outcomes Livelihood Strategies Community / HH Level Livelihood Assets The food security status of any household or individual is typically determined by the interaction of a broad range of agro-environmental, socio-economic and biological factors. As with the concepts of health or social welfare, there is no single, direct measure of food security. However, the complexity of the food security problem can be simplified by focusing on three distinct but interrelated dimensions: (i) aggregate food availability, (ii) household food access, and (iii) individual food utilization. Vulnerability is a forward-looking concept for assessing community and household exposure and sensitivity to future shocks. Ultimately, the vulnerability of a household or community depends on its ability to cope with exposure to the risks associated with shocks such as drought, flood, crop blight or infestation, economic fluctuation and conflict. The ability to manage these risks is determined largely by the characteristics of a household or community, particularly its asset base and the livelihood and food security strategies it pursues. 6 Source: Emergency Food and Nutrition handbook 5

23 The framework shows that exposure to risk is determined by the frequency and severity of natural and human-induced hazards, and by their socio-economic and geographical scope. The determinants of coping capacity include the levels of a household s natural, physical, economic, human, social and political assets, the levels of its production, income and consumption, and its ability to diversify its income sources and consumption to mitigate the effects of the risks it may face at any moment. The details are in Figure 1.2. Figure 1.3: Enumerator interviewing a mother 6

24 Chapter 2: Methodology 2.1 Introduction Both quantitative and qualitative methods were employed for data collection. This assessment was conducted from 10 th to 24 th October Although the timing of this assessment was planned to coincide with the same timing as that of August 2009 in order to maximize the comparability of the two surveys, there were noted delays that were beyond control. Furthermore, the assessment did not fall within the hunger gap that is between March and July. To estimate the comprehensive status of food security and nutrition in the refugee settlements, the assessment was composed of three components. Table 2.1 presents these components and types of data collection for each component. In particular, the assessment was designed to provide statistically representative estimates of key indicators for the overall refugee population in Uganda and in particular, the 2 sub regions in West Nile and South Western sub regions. Table 2.1: Types of Data Collection Primary data Secondary data Assessment Component Quantitative Methods Qualitative methods Household survey Observation Focus Group Discussion Nutrition X X X Health X X X X Food Security X X X X Nutrition component included anthropometric measurements among children 6 to 59months of age and mothers 2.2 Sample Size for Household Survey For quantitative data collection methods, sample size calculations were made to ensure that key indicators were statistically representative at the sub regional level and/or of the overall refugee population. Sample size was calculated with 0.05 of statistical significance (that is, 95% CI was used) for key indicators (Table 2.2). Based on last year s assessment results, assumptions were made that each household would have an average of one child 6 to 59 months of age and a household size of five. Prevalence estimates for each of the key indicators were based on the previous survey carried out in 2009 in the refugee settlements. An estimate for acute malnutrition of 3.3percent was therefore used. 7

25 Because a two-stage cluster sampling was used, it was necessary to increase the sample size by a factor which would allow for the design effect. Design effects (shown in Table 2.2) were estimated using the Food Security and Nutrition Assessment in the Refugee Settlements in 2009, and previous surveys conducted in the settlements in The desired precision was based on the estimated prevalence, as well as a consideration of relevant cut-offs for programmatic action. Table 2.2: Assumptions and Estimated Sample Size per Sub Region Nutrition Indicator Target group (Statistical Unit) Estimated prevalence/ proportion: Po Design Effect; DEFF Precision : d Individual Required sample size: n + 10% non response rate Household/ Individuals Acute Malnutrition Children 6-59 months 3.3% 2% 1.4% 1, ,744 BMI Lactating women 3.7 2% 1.4% Anaemia Women Pregnant/lactating women 32.7% 2% Children Children 6-59 months 65.7% 2% Mortality Crude Mortality rate All household members % 0.5 5,122 5,634 1,127 7 Food Security 8 Household Based on a CMR of 0.21 per 10,000 per day, a CI of 95%, a precision of 0.5, a design effect of 2 and a recall period of 90 days, a total of 1,127 households per sub region were required (Table 2.2). To determine the number of households required, an assumption of five persons per household was made. The CMR estimate for sample size calculations was based upon the 2009 Food Security and Nutrition Assessment. Specific to food security, information was collected in households where CMR and nutrition data were gathered for comparison purposes. 2.3 Sampling procedure for household survey Sampling Frame The sampling frame for this survey consisted of approximately 115,885 people residing in 9 refugee settlements in West Nile and South West sub regions in Uganda. This frame consisted of the population identified as refugees by UNHCR and the Office of the Prime Minister as of August 2010, additional WFP beneficiary figures were utilized. It is important to emphasize that the survey did not represent the host communities in the settlements. 7 Assumption that each household would have an average of 5 people 8 Food Security information was collected in households where mortality and nutrition data were gathered for comparison purposes 8

26 A multi stage cluster sampling methodology based on the SMART Nutrition Survey Guidelines was used. The first entailed the selection of 36 clusters per sub region using the probability proportional to population size technique, while the second stage entailed the selection of households. (1) Selection of primary sampling units (clusters) Due to the fewer number of persons in the settlements compared to the previous years, it was decided to include 36 clusters from each of the two sub regions in the sample (to reach a total of 72 clusters for the overall estimate). Based on 2009 experience, it was expected that some of the clusters that were drawn would have fewer households. To compensate for this, an additional four clusters were drawn for each sub regions, leading to a total of 40 clusters from West Nile, and 40 clusters from South Western (Table 2.3). To prevent potential bias, the survey protocol required that the teams visit the maximum number of accessible clusters of those selected. Clusters were chosen using Probability Proportionate to Size (PPS) methods from the list of settlements annexed in 9. Table 2.3: Proposed clusters by Sub Region Sub region Settlement Population No of clusters No of proposed replacement clusters West Nile Adjumani 7, Palorinya 2, Imvepi 1, Rhino Camp 1, Kiryandongo 4, Total 17, South West Kyaka II 18, Nakivale 56, Oruchinga 1, Kyangwali 22, Total 97, Overall 115, (2) Second stage: selection of sampling (cluster location within chosen community) Once survey teams arrived in each district, they met with WFP and UNHCR, NGOs and other local officials to try to find additional information about the populations included in the sample this often resulted in population lists, which were used for further sampling. Upon arrival at the actual cluster location, the teams held discussions with Camp Commandants and other key informants to further refine this information, following reassurances that the data would not be linked to the delivery of food or nonfood items. To determine the actual location of clusters within selected locations, the PPS method was used. The goal was to reach a population size of 100 to 200 households from which to choose the final 30 households. In large settlements, several stages of sampling were 9

27 required. Three main methods were used to achieve PPS sampling, depending on the situation. Geographical segmentation method: This method was used most frequently to narrow down the sample. This method involved the creation and use of a map of the area where the population lived, and the division of that population into multiple segments. A cumulative population list by zones was compiled, and a random number table used to select the cluster location. If each sector was of equal size, one sector was chosen using a random number table. Population lists method: Population listing was also used as a technique to narrow the sample. In the early stages of cluster selection, a list of villages and the population belonging to them was compiled, and PPS was used to select one or more villages. This was based on the assumption that all households within the community belonged to only one village, and no household belonged to more than one village (this assumption was true in all of the communities that were encountered). (3) Selection of the basic sampling unit (household) To ensure that the required number of children was met, a decision was made to select 30 households in each of the (36) clusters. Two sampling methods were utilized based on the situation in the settlement: (i) (ii) Using the geographical segmentation method, once the cluster location was identified, the team leader walked the boundary of the cluster with a community leader. The total number of households was divided by 30 (the required number of households) to provide a sampling interval, which was usually between 5 and 20 depending on the size of the cluster location. The team leader then identified each selected household, and after getting initial consent from a household member, marked the household with chalk. With the household listing method, the selected local leaders were asked to list every household belonging to their village using the definition of household in section Each household was allocated a unique number on that list, and a random number table was used to select the first household. The total number of households was then divided by 30 to get the sampling interval for selection of subsequent households. All chosen households were selected, whether or not they contained a child 6 to 59 months of age. If household members were not present, community leaders were asked to bring them to the house to be interviewed. Households were visited at least two times in an effort to identify household members, unless time or logistic constraints 10

28 limited the amount of time spent in a cluster. Basic demographic information was taken from an adult household member (usually brother or sister of the head of household), if available. If the members had departed permanently, or were not expected to return before the survey team had to leave the village, the household was skipped and not replaced. Where possible, survey teams visited the cluster location on two successive days. 2.4 Survey Coordination In collaboration with the Office of the Prime Minister - Refugee department, IBFAN Uganda held stakeholders meetings both at the national and district levels where the inception report was presented and input from the stakeholders made as well as their roles and responsibilities identified and emphasized for ownership of the report. The stakeholders among others included RDO, UNHCR, WFP and NGOs working in the Refugee Settlements (NRC, AHA, GTZ, DRC and AAH). Other persons met included District Health Officers and District Production coordinators from the respective districts given the health and food security components in the Assessment. The survey team comprised of 1 Principal Investigator, 4 co-investigators, 8 supervisors drawn from national and district levels and 36 enumerators from the 2 sub regions. These were teamed into 4 groups as shown in Figure2.1. Each of the groups had 3 teams of 1 supervisor and 3 enumerators giving a total of 4 people per team. In total, there were 12 teams for the whole assessment. Annex 11 shows the list of individuals who composed the teams. Principal Investigator Team Leader- Oruchinga/Nakivale Team Leader- Kyaka/Kyangwali Team Leader- Kiryandongo/Imvepi/Rhino Team Leader- Adjumani/Palorinya Data Manager 2 Supervisors 2 Supervisors 2 Supervisors 2 Supervisors 9 Enumerators 1 Data Entrant 9 Enumerators 1 Data Entrant 9 Enumerators 1 Data Entrant 9 Enumerators 1 Data Entrant 4 Data Entrants Figure 2.1: Structure of the survey team 11

29 2.5 Training Training on and piloting of the questionnaire was conducted jointly for all the 4 groups at the national level who later travelled to the districts to train the district based enumerators and supervisors. A training guide adapted from the national guidelines on health and nutrition assessments was adapted to include the food security component and used in order to harmonize the training for the 4 groups. All enumerators received a three days of classroom training, including a one-day practical orientation in one of the villages in the settlements not included in the survey. This was done under close supervision. Team members were trained jointly on the rationale for the survey, sampling, consent, questionnaire administration, interviewing, anthropometric measurements and referral. All the teams underwent three standardization sessions until they had acceptable inter-observer variation. 2.6 Data collection A single household structured questionnaire with all the described variables was used to collect the data from all the households (Annex 4). Qualitative data was obtained through conducting Focus group discussions and Key informant interviews (Annexes 5 and 6) Household survey Household: A group of people who routinely ate out of the same pot, and slept in the same structure or family compound (or physical location). It is possible that they may have lived in different structures. Members of a household were not necessarily relatives by blood or marriage. If several separate families were living in the same compound, they were regarded as separate households. If a polygamous family lived and ate together, they were considered one household. Mortality was assessed using the retrospective household census method. Respondents were requested to list all members living in the household at the time of the bomb that went off in Kampala on the 11th July 2010, which they all accepted were familiar with. First, all members living in the household at that time were listed by age and sex, with the head of the household listed first. The respondent was then asked where each person was at the time of the interview. Possible choices were: alive and living in the household, alive and living elsewhere, missing or dead. Deaths occurring in each household in the period were recorded along with the date of occurrence. Individual sub region-based local calendar of events were developed and used to determine ages of household members and dates of death (Annex 8). Cause of death was collected from the head of the household or the respondent, if not the head of the household. Each questionnaire had descriptions of the causes of death listed; and codes included 12

30 diarrhoea, fever, measles, difficulty breathing, malnutrition, violence/conflict-related deaths and other. Enumerators asked questions of each mother with a child 6 to 59 months of age in the household regarding pregnancy, parity and breastfeeding. For mothers with children 6 to 24 months of age, questions were asked regarding breastfeeding initiation and duration as well as infant and young-child feeding practices. Information was also gathered on each child 6 to 59 months of age from an adult household member (preferably the mother). Questions were asked regarding vitamin A supplementation, de-worming, DPT 3 and measles vaccination and recent illness. Vaccination and supplementation records were reviewed where available. However, mothers reports were also taken as evidence of vaccination against measles and receipt of vitamin A supplementation. On food security, questions on the frequency and diversity of food consumed in definite recall time (7days) were asked. To allow the assessment to analyze household food security in relation to the socio-economic situation of households, data on livelihood, coping strategies, expenditure on food, land access and agricultural activities, ownership of agriculture and livestock assets and preparedness for harvest were included in the household questionnaire. On household health, questions on the sources of water used for drinking and other purposes were asked. Other questions were related to treatment of drinking water, amount of water used in the household, disposal of faeces and rubbish as well as treatment of household members when sick, the use of mosquito nets and the type of fuel used for cooking in the households Focus Group and Key Informant Interviews Focus Group discussions were conducted in all the 9 settlements with the aim of collecting additional information on issues related to the health, nutrition, food security and the general living conditions. After a team arrived in a community, the purpose of the assessment was explained to the community leaders. They were then asked to nominate 10 to 12 persons, representing both males and females and other groups within the community, to participate in the FGDs. A questionnaire (Annex 5) was used to guide the discussions. This questionnaire included questions concerning food security, health situation and status of health care services, and major problems in the community and coping strategies. The FGD questionnaire was designed so that the interview could be completed in approximately one to two hours. The Key Informant interviews took the same process although the target audience was mainly the settlement commandants, Implementing Partners for Health and Food Security, 13

31 District Production Coordinators and Health Officers making a total of 36 Key Informants (Annex 6) Anthropometric measurement Enumerators measured children s weight, height/length and assessed the presence of oedema. Children were weighed to the nearest 100 grams with a UNICEF UNISCALE. For children younger than two years of age or less than 85 cm, length was measured to the nearest 0.1 centimetre in the recumbent position using a standard height board. Children 85 to 110 cm were measured in a standing position. Oedema was assessed by applying thumb pressure to the feet for approximately three seconds, and then examining for the presence of a shallow print or pit. Whereas MUAC was measured on all mothers in the survey using a MUAC measuring tape, heights and weights were taken for only non pregnant women 15 49years. MUAC was also taken for most children apart from those with nutritional oedema. Haemoglobin concentration and haematocrit using a heamocue system was done to determine the prevalence of anaemia among women and children in the settlements. Where facilities existed, malnourished children and women were referred to therapeutic feeding centres for treatment of severe malnutrition (MUAC <11.5cm) or to supplementary feeding programmes for treatment of moderate malnutrition (>11.5cm to <12.5cm). 2.7 Indicators, guidelines and formulae used Malnutrition Acute malnutrition rates or wasting (indicating recent food shortages and morbidity or acute illness) was estimated using weight for height (WFH) index values combined with the presence of oedema. The WFH indices were compared with NCHS and WHO 2006 reference population. WFH indices were expressed in Z-scores both for NCHS and WHO Children were categorised according to their nutritional status i.e. Global, Moderate and Severe Acute malnutrition, using the guidelines in Table 2.4. The report used WHO 2006 indices however NCHS are in Annex 7. Table 2.4: Classification of nutritional status using WFH expressed in Z-scores Nutritional Status Severe Acute Malnutrition Moderate Acute Malnutrition Global acute malnutrition Expressed in Z-score WFH < -3 SD* and/or existing bilateral oedema WFH < -2SD to -3SD, and the absence of oedema WFH <-2 SD *SD Standard Deviation 14

32 Stunting: Chronic malnutrition was obtained using height for age index and was expressed using Z scores. Children were classified as severe, moderate and global stunting as shown in Table 2.5. Table 2.5: Classification of children with chronic malnutrition by using Height For Age Nutritional Status Severe Stunting Moderate Stunting Global Stunting Expressed in Z-Score HFA Below -3 SD HFA Between -3 and SD HFA Below -2 SD The severity of the malnutrition situation in the sub regions was interpreted with the help of the bench marks provided in Table 2.6. Table 2.6: Classification of Severity of Malnutrition in a Community by Prevalence of Acute and Chronic Malnutrition for Children Under 5 years of Age Severity of Acute WFH <-2 z-scores Chronic malnutrition HFA malnutrition (%) <-2 z-scores (%) Acceptable <5 <20 Poor Serious Critical Chronic Energy Deficiency (CED): The Body Mass Index (BMI) is used to determine the CED in non pregnant women. BMI is the result of dividing weight (in Kg) by height squared (in meters). Cut offs used are as indicated in Table 2.7. Table 2.7: BMI classification of Nutritional Status BMI Interpretation < 16 Severe Malnutrition 16 and 16.9 Moderate Malnutrition Mild Malnutrition (At risk) Normal Over weight Above 30 Obese Mid Upper Arm Circumference: Mid upper arm circumference measurements for children between 6 59 months and all women were assessed. Bench marks for defining nutritional status of both children and women using MUAC is as indicated in Table

33 Table 2.8: Nutritional Status classified using MUAC MUAC for children (6-59 months) Interpretation < 11.5 cm Severe Malnutrition 11.5 cm and < 12.5 cm Moderate Malnutrition 12.5 cm - < 13.5 cm Mild Malnutrition (At risk) 13.5 cm Good Nutritional Status MUAC for Women (15-49 years) <21 cm Severe Malnutrition < 21.5 cm Moderate <22 cm Mild cm Normal cm Overweight > 30 cm Obese Mortality: The mortality rate (MR) was determined for the whole population (CMR) and children under 5 years (U-5MR). The severity of mortality in the sub regions was determined using the bench marks in Table 2.9 for both Crude mortality rate (CMR) and under five mortality rate (U-5MR). Table 2.9: Cut off values for Crude Mortality and Under-5 Mortality Rates Indicator CMR U-5MR Alert Level (Relief program under control) 1/10,000/day 2/10,000/day Critical Emergency Level (Emergency out of control) 2/10,000/day 4/10,000/day Famine/major catastrophe 5/10,000/day 10/10,000/day Morbidity: This was estimated from the number of reported cases of illnesses over the past two weeks prior to the survey. Children reporting any illness were compared to the total number of children in the survey and the prevalence obtained. Given the National Health Service guidelines, any prevalence above 40percent is of public health significance. Anaemia: Peripheral blood was taken from finger sticks and used for the analysis of nutritional status and risk factors in children and women. Anaemia was tested using the Haemocue Photometer, which measures the amount of haemoglobin in a blood drop placed into a specialized disposable cuvette. Children and women were classified as severe, moderate and normal as shown in Table Figure 2.2: A child being tested for anaemia during assessment 16

34 Table 2.10: Recommended Haemoglobin levels in g/dl Population group Haemoglobin g/dl Overall Moderate Severe Children (6-59 months) <7 Non-pregnant women - >15 years <7 Pregnant women <7 Table 2.11 presents the classification of iron deficiency anaemia for communities. Table 2.11: Cut offs for Anaemia prevalence in Communities Public health significance (Category) Prevalence of anaemia (%) Severe > or = to 40 Moderate Mild Feeding Practices Early initiation of breastfeeding: Proportion of children born in the last 24 months who were put to the breast within one hour of birth and calculated as children born in the last 24 months who were put to the breast within one hour of birth/children born in the last 24 months. Exclusive breastfeeding under 6 months: Proportion of infants 0 5 months of age who were fed exclusively with breastmilk and calculated as Infants 0 5 months of age who received only breastmilk during the previous day/ Infants 0 5 months of age Continued breastfeeding at 1 year: Proportion of children months of age who were fed breastmilk and defined as Children months of age who received breastmilk during the previous day/ Children months of age Introduction of solid, semi-solid or soft foods: Proportion of infants 6 8 months of age who received solid, semi-solid or soft foods and defined as Infants 6 8 months of age who received solid, semi-solid or soft foods during the previous day/infants 6 8 months of age. Children ever breastfed: Proportion of children born in the last 24 months who were ever breastfed and defined as Children born in the last 24 months who were ever breastfed/ Children born in the last 24 months Continued breastfeeding at 2 years: Proportion of children months of age who are fed breastmilk and was calculated as Children months of age who received breastmilk during the previous day/ Children months of age 17

35 Duration of breastfeeding: Median duration of breastfeeding among children less than 36 months of age - The age in months when 50percent of children 0 35 months did not receive breastmilk during the previous day Food Security A household's ability/possibility to obtain food from its own production, stocks, purchase and gathering or through transfers (gifts from relatives, members of the community, government, or external assistance). Household Asset Diversity: Common household assets were profiled to include both productive and non productive assets. Household asset diversity was then derived based on the frequency of households owning at least an asset based on a derived scale. Households were profiled in terms of those owning at least one or more of the listed assets in increasing order of scale. Households with higher scores indicated more asset diversity in terms of ownership of more different items rather than more of individual items. Livestock production analysis: This was based on household responses on ownership of small animals (shoats) and large animals and poultry. Analysis was in terms of households owning at least one or more of the different livestock, and owning much more in numbers of three categories; cattle, shoats and poultry. Access to livestock products was analyzed in terms of production months by categorizing months into a scale of 1-3 where households access products in fewer months were assigned a lower scale and the more months, a higher scale. Land Access and Agriculture Activities: Based on the responses from the households, households that were able to cultivate or not able to cultivate were presented in percentages by running frequencies of the responses. Reasons for not being able to cultivate were ranked according to frequencies of occurrences. This was then related to the Food Consumption Scores and hence significant predictors of food security. Main income Source: Members of the households were asked what kind of activities they did to earn a living and the main source of income. For each income source, the highest frequency of occurrence for a particular percentage reported was recorded. This was then deduced by categorizing the percentages into 3-4 categories of very low, low, moderate or high, depending on how the responses were presented. This was then compared with the level of food security in a particular household. Expenditures: Expenditures were analyzed by recording the 30 day expenditure on each food item in Uganda shillings into appropriate categories of expenditure; e.g. 0-10,000, more than 10,000 to 100,000 and more than 100,000 depending on the expenses and categorized as food insecure, vulnerable and food secure respectively. 18

36 Shocks and coping: Respondents were asked if they had in the 7 days experienced a situation when they could not provide enough food for their households as well as adopting particular behaviours caused by the situation. The Coping Strategy Index (CSI) was then used for the analysis of these strategies in terms of their severity and frequency (Refer to Table 12 for details). The higher the score, the more severe/frequent coping mechanisms were employed and were then compared between households and groups as well as changes overtime. Table 2.12: Copying Strategy Index calculation In the past 30 days, as a result of not having enough food, how often has your household had to: All the time/every day Fairly often/3-6 times per week Occasionally/ 1-2 times per week Rarely/less than once a week Never Raw Score Severity Weight Score=relative frequency x weight Relative Frequency Score a. Rely on less preferred and less expensive foods? 2 b. Borrow food, or rely on help from a friend or relative? 4 c. Purchase food on credit? d. Gather wild food, hunt or harvest immature crops? 4 8 e. Consume seed stock held for next season? 6 f. Send household members to eat elsewhere? g. Send household members to beg? h. Limit portion sizes at meal times? i. Restrict adults' consumption so that children can eat? j. Feed working household members at the expense of non-working members? k. Ration the money available and buy prepared food? l. Reduce number of meals eaten in a day? m. Pass entire days without eating? NA 2 8 Total Household Score Food Assistance: This was based on frequencies of those household that either received or not received food assistance between January to August Further analysis was done on the frequencies of households that received food aid at least any one of the listed six categories and the frequency by months. 19

37 Non Food Assistance: The non food assistance items were categorized into appropriate groups including production, health, extension services, loans, and ran frequencies of households that received any assistance among the grouped items. Child Education: Frequencies of boys or girls reported to be staying out of school were carried out. The main reasons for not attending school were ranked from the most frequently reported to the least. Household Priorities: This was analyzed by ranking the most frequently reported priority by the respondents for the short term and the long term. Food Consumption Score: This was calculated based on the types of foods and frequencies with which they were consumed during a seven-day period. Food items usually consumed by households were grouped into food groups according to their relative nutritional value. Households were then be asked to recall foods consumed in the previous seven days and how many days each food item in a food group was consumed. Each item was given a frequency score 0-7 depending on the number of days it was consumed in the recall week period. Each food group was then assigned a weight reflecting the nutritional density. The household FCS was then calculated by multiplying each food group by each food group weight and then summing these scores into one composite score and categorizing them into poor, borderline and acceptable food consumption groups. These are summarized in Tables 2.13 and Table 2.13: Template for calculating the Food Consumption Score (FCS) Food Item (examples) Food Group Weight (A) Maize, rice, sorghum, millet, bread and other cereals Cereals and Cassava, potatoes and sweet tubers 2 potatoes Beans, peas, groundnuts and cashew nuts Pulses 3 Vegetables, relish and leaves Vegetables 1 Fruits Fruit 1 Beef, goat, poultry, pork, eggs and fish Meat and fish 4 Milk, yoghurt and other dairy products Milk 4 Sugar and sugar products Sugar 0.5 Oils, fats and butter Oil 0.5 Condiments Condiments 0 Composite score Days Eaten in past 7 days(b) Score (A X B) 20

38 Table 2.14: Food Consumption Groups (FCGs) and corresponding FCS thresholds Food consumption Group Cut off Poor 21 Borderline > Acceptable > Data Management Data analysis Nutrition and health Data was entered in Microsoft access. The calculation and analysis of anthropometry indices were conducted in ENA. ADDWIN (Addati 6.0) was used to conduct both PCA and cluster analysis. Analysis of all other variables was carried out in Microsoft excel and SPSS 15. Household Food Security Profiling To define the food security of the households, four indicators were considered: food consumption; food aid dependency; proportion of the weekly food expenditure per capita compared with the total weekly expenditure per capita; and absolute value of the weekly expenditure per capita. First, the number of meals eaten during the last 24 hours was asked. Second, to analyse food consumption patterns, households were classified according to dietary diversity and frequency of food consumption, i.e. the number of days on which a household consumed each of the food items in the last seven days prior to the interview. A total of 23 food groups were used including: maize, rice, sorghum, millet, wheat, cassava, potatoes, yams, bread, matooke, beans and peas, vegetables, groundnuts, fresh fruits, fish, meat, blood, eggs, sugar, oil/fats, milk, beer residue and condiments. The steps taken for this analysis were to: apply the number of days on which each food item was consumed to a Principal Component Analysis (PCA) in order to identify key principal components (PCs) that more clearly explain diversity and frequency of consumption of major food items; of all the 23 PCs created, whose total number is equivalent to the number of food items, select the PCs with greater variance of PC scores (loadings) so that the cumulative variance would account for percent; 21

39 through cluster analysis, classify all the households into several food consumption patterns according to the scores (loadings) of PCs selected; and further classify identified food consumption patterns into three groups: (i) acceptable food consumption; (ii) borderline food consumption; and (iii) poor food consumption. Dependency on food was computed basing on the food items supplied by WFP. Food items considered for the analysis were maize, cooking oil, beans. Households were then categorised into those where more than 50% of food consumed is food aid, those where less than 50% of food consumed is food aid, and those that did not receive any food aid. A second set of clusters were produced through PCA and cluster analysis based on food and non-food expenditure. The number of principal components saved for the subsequent analysis was the number required to explain approximately 90% of the total variance. Average expenditure shares were then calculated for each cluster with the sum of all shares (all expenditures) equal to 100. Households were then categorised into those spending less than 50percent on food and those spending more than 50percent on food. Weekly Per capita food expenditures were computed and aggregated into per-capita quintiles. Through this process, each household was assigned to its respective quintile for per capita food expenditures. Households falling within the bottom two quintiles were taken as households having the least per capita expenditure on food. Households were then grouped into those spending less than 275 shillings/capita/week (average for Refugee Settlements for the bottom two food expenditures quintiles), and those spending more than 275 shillings/capita/week. To establish the household food security profile, household groups were categorized on the basis of a combination of food consumption patterns, food aid as a source of food, and food expenditures, using the following criteria: Criteria for food aid: - more than 50% of food consumed - less than 50% - 0% Criteria for food expenditures: - less than 50% and less than 275 shillings/capita/week - more than 50% and less than 275 shillings/capita/week The three groups are composed of various typologies of households, as illustrated in the Table 2.15: 22

40 Table 2.15: Typologies of the households according to their food consumption and access Food Consumption Pattern Share and amount of food expenditures / capita / week Fa: Proportion of food aid in primary source of foods consumed at households Food Aid = 0% Food Aid > 50% 50% Food Aid > 0% (No food aid) Poor Borderline Acceptable > 50% and 275 shillings (economic insecurity) < 50% or 275 shillings (economic security) > 50% and 275 shillings (economic insecurity) < 50% or 275 shillings (economic security) Groups No.1 to 4 (in red): very unsatisfactory food consumption and access at the time of the survey, with high risks to lives or severe economic insecurity 1) Unsatisfactory food consumption pattern: severe risks to lives (negative effects expected on health and nutrition); 2) Borderline food consumption pattern with high or medium reliance on food aid: moderate risk for health and nutrition, high dependence on external assistance; 3) Borderline food consumption pattern with no reliance on food aid but large share of food expenditures and low amounts of food expenditures: moderate risk for health and nutrition, severe economic insecurity; 4) Acceptable food consumption pattern with high reliance on food aid and large share of food expenditures and low amounts of food expenditures: high dependence on external assistance, severe economic insecurity. This whole group of households is considered to be severely food insecure and at high risk to lives and livelihoods on the short term. Groups No.5 to 7 (in yellow): unsatisfactory food consumption and access at the time of the survey, with medium risks to lives and severe or moderate economic insecurity 5) Borderline food consumption pattern with no reliance on food aid: moderate risk for health and nutrition; 6) Acceptable food consumption pattern with high reliance on food aid but small share of food expenditures or high amounts of food expenditures: high dependence on external assistance; 7) Acceptable food consumption pattern with medium reliance on food aid and high share of food expenditures and low amounts of food expenditures: moderate dependence on external assistance, severe economic insecurity. 23

41 This whole group of household is considered to be moderately food insecure and at medium risk to lives and livelihoods on the short term. Groups No.8 and 9 (in green): acceptable food consumption, with low risk to lives and moderate or low economic insecurity 8) Acceptable food consumption pattern with medium reliance on food aid and low share of food expenditures or high amounts of food expenditures: moderate dependence on external assistance, economic security; 9) Acceptable food consumption pattern with no dependence on food aid. This whole group is considered food secure and at low risk to lives and livelihoods on the short-term. 2.9 Consent All household members received a verbal explanation of the survey including both the household questionnaire and anthropometry. Enumerators were provided with a paragraph in their guide requesting consent from the interviewee. The consent or refusal was recorded on the form by the interviewer. Households were informed that the survey was confidential, and that their responses would not affect food distributions. Participation was voluntary and household members had the right to refuse to answer any or all questions as well as anthropometric assessments Limitations and potential biases All population-based sample surveys have potential limitations and biases. In emergency contexts, uncertainty usually exists about estimates of population figures. This was the case in this assessment as some settlements were still updating their population figures. Findings of this survey can only be generalized to the overall population of approximately 115,855 people categorized as Refugees by the UN and OPM that the survey was designed to represent. While the survey used the most up-to-date population estimates from lists constructed by UNHCR, 31 st August 2010, survey findings cannot be extrapolated to other potentially vulnerable groups not included on those lists, such as urban refugees. Potential interviewer and measurement bias are also important to consider in surveys, particularly those with a large number of survey team members. To minimize this potential bias, interviewers and anthropometrists underwent three days of classroom training and a field test. In addition, anthropometrists underwent three standardization activities to ensure quality measurements; both at the national and districts levels. Each question was reviewed by the group to ensure that the content and meaning was understood and would make sense in a local context. Each team was supervised by 24

42 national and district supervisors with wide experiences in conducting assessment. Questionnaires were checked by these supervisors for completeness and accuracy before departing the cluster. Recall bias is also important to consider in any retrospective survey of mortality. A remarkable event of the bomb blast that took place in Kampala better remembered in all settlements was used. A shorter recall period of three months, similar to that used last year was utilized, and helped minimize the potential recall bias, and also helped respondents remember when deaths occurred, specific local calendars were developed for each of the 2 sub regions prior to the survey (Annex 8). The same calendars were also used by enumerators when asking other questions requiring recall, such as child age, and whether or not they had been immunized in the past six months. Additionally, because only households in which members were present on the day of the survey were included in the sample, households in which all members had died would not have been selected, potentially leading to an under-estimation of mortality. Assurances were made during the introduction of the survey at both the community and household levels that information provided regarding household composition and recent deaths would not affect rations or receipt of non-food items. However, one cannot discount the possibility of information bias (respondents purposefully giving misinformation). At the time of the survey, some settlements were undergoing registration or repatriation activities. Sensitivity to registration and the presence of WFP and UNHCR staff and vehicles may have influenced respondents answers for fear of reduction of household rations or repatriation. Most likely, this bias would lead to an underestimation of mortality, with fewer deaths reported. Causes of death were collected through self report. There may be inaccuracies in the causes reported by respondents, which may have been influenced by the local perception of disease or the perception of health workers. 25

43 Chapter 3: Findings 3.1 Demography Study Area The survey sample included 2,072 households. Of the total number of households, 1,053 were from West Nile and 1,019, in South West. A total of 77 clusters were selected, 36 from West Nile and 41 from South West. The details are shown in Table 3.1. Table 3.1: Sampled sites by Sub Region and District Sub region District Settlement Number of clusters sampled Number of Households West Nile Adjumani Adjumani Moyo Palorinya Kiryandongo Kiryandongo Arua Imvepi/Rhino Camp Total South West Kyegegwa Kyaka Isingiro Nakivale Isingiro Oruchinga 1 30 Hoima Kyangwali Total Overall Within the sub regions, West Nile comprised of Adjumani, Imvepi/Rhino Camp, Kiryandongo 9 ; and Moyo while South West were Kyaka, Oruchinga, Kyangwali and Nakivale refugee settlements Age and Sex Distribution The results in Tables 3.2 and 3.3 show the number of children surveyed disaggregated by sex and age in the 2 sub regions. 9 Kiryandongo was included in the Westnile cluster for the purpose of this assessment 26

44 Table 3.2: Percentage Distribution of Age and Sex among Sampled Children in South West sub region refugee settlements, October 2010 Age Group (months) Boys Girls Total Ratio no. % no. % no. % Boy:Girl Total The demographic sex ratio comprised of equal proportion of males and females with the overall ratio of 1:1 female to male in both sub regions which indicates that there was no sample bias in the survey. Table 3.3: Percentage Distribution of Age and Sex among Sampled Children in West Nile sub region refugee settlements, October 2010 Age Group (months) Boys Girls Total Ratio n % n % N % Boy:Girl Total Generally, in both sub regions, the number of children assessed from the different age groups was almost the same. The survey collected more detailed data on 2,072 children (6 to 59 months of age). Among these children, 1,022 (49.4percent) were boys and 1,050 (50.6percent) were girls. The excluded subjects and quality check results are presented in Annex Child Health and Nutrition Child Malnutrition Table 3.4 shows that anthropometric measurements were taken on 2,055 children across all sub regions (1,061 West Nile and 994 South West). The results presented are based upon weight-for-height z-scores and the presence of nutritional oedema. For the overall sampled population, the prevalence of GAM was 2.2 percent [95% CI: ] and SAM was 0.4 percent [95% CI: ]. These results are lower than the previous levels found in the 2009 survey i.e. 3.3 percent [95% CI: ] of GAM in 2009 and 0.5 percent [95% CI: ] of SAM in These figures were also lower than the national prevalence of wasting of 6percent (UDHS, 2006). 27

45 Table 3.4: Acute malnutrition rates among children 6 to 59 months of age by Sub region, October, 2010 West Nile (n=1061) South West (n=994) Total (n=2055) n(%) 95% CI n(%) 95%CI (%) 95% CI Global Acute Malnutrition 23 (2.2) (2.1) (2.2) Moderate Acute Malnutrition 22 (2.1) (1.4) (1.8) Severe Acute Malnutrition 1 (0.1) (0.7) (0.4) Table 3.4 further shows that the prevalence of GAM was almost the same for both sub regions at 2.2 percent [95% CI: percent], and 2.1 percent [95% CI: ] for West Nile and South West respectively) with SAM of 0.1 percent [95% CI: percent] and 0.7 percent [95% CI: percent] with 6 cases of oedema identified respectively. Figure 3.2 shows the distribution of weight-for-height z-scores for the child population aged 6 to 59 months relative to the reference population. The figure shows that the entire distribution of weightfor-height z-scores of children in the overall sample was shifted to the right in comparison to the reference population indicating that malnutrition was not of public health concern in the settlements. Figure Distribution of weight-for-height z-scores for overall sample compared to the reference For children 6 to 59 months of age, the prevalence of acute malnutrition was statistically higher among boys (2.9 percent) than among girls (1.5 percent). When malnutrition was disaggregated by age, the prevalence of acute malnutrition among children under 24 months of age was higher than those above 24 months at 3.6 percent [95% CI: percent] and 1.4 percent [95% CI: percent], respectively. The details are as shown in Table 3.5. Table 3.5: Prevalence of Global Acute Malnutrition by Age category, October 2010 Prevalence of Global Acute Malnutrition Number (%) 95% CI 6 to 11 months to 23 months to 35 months to 47 months to 59 months Total

46 The prevalence of stunting and underweight rates were as provided in Table 3.6. Table 3.6: Prevalence of Stunting and Underweight among children 6 to 59 months of Age, October, 2010 Malnutrition West Nile South West Total N (%) 95%CI n (%) 95%CI n (%) 95% CI Global stunting (<-2 z-score) Moderate stunting(<-2 z-score and >=-3 z-score) Severe Stunting (<-3 z-score) Global underweight (<-2 z-score) Moderate under weight (<-2 z- score and >=-3 z-score) Severe underweight The prevalence of stunting in the South West region of 35.8percent [95% CI: percent] was higher than that in the West Nile region of 17.8percent [95% CI: percent] giving an overall prevalence of 25.7percent [95% CI: percent]. Based on the benchmarks of chronic malnutrition, this situation was noted to be poor according to the national survey guidelines, There was a reduction by 3.3percent from that of 2009 of 30percent. The levels of stunting were however slightly lower than those reported in the UDHS 2006 of 38percent nationally. The prevalence of underweight was also noted to be higher in South West than West Nile sub regions at 4.8percent [95% CI: percent] and 6.2percent [95% CI: percent] respectively with the overall prevalence of 5.5 percent [95% CI: percent]. The stunting and underweight levels by settlement are shown in Annex 2. Table 3.7 indicates that only 5.9 percent [95% CI: percent] of the children in West Nile and South West sub regions were at risk of malnutrition. Almost all children had good nutritional status at 92percent [95% CI: percent]. Table 3.7: Nutritional Status by MUAC in sampled children by Sub region, October 2010 South West West Nile Total n % 95% CI n % 95% CI N % 95% CI Good Mild (At risk) Moderate Severe Grand Total ,

47 Furthermore, WHO urges serious intensified action in countries where the prevalence of malnutrition for children 0-59 exceeds 20percent for stunting, 5 percent for wasting and 10percent for under-weight. For this matter, the high levels of stunting calls for urgent and bold interventions in the Refugee Settlements Main Causes of Child Malnutrition Relationship between child malnutrition, disease and the health environment (a) Relationship between nutritional status and illness Malnourished children (wasted) were significantly more likely to have been sick in the previous two weeks compared to the non-malnourished (p = 0.035). There was a strong relationship between children suffering from fever in the two weeks prior to the survey and being stunted at 20.2percent than those wasted and underweight at 2.4percent and 6.3percent respectively. However, almost one third of the children in the survey reported fever, so this association may be a reflection of a relatively high prevalence. The means weight for height and weight for age z-score were significantly associated with having suffered from diarrhoea in the two weeks prior to the survey (p>0.05). About 3.6percent of children who had had diarrhoea were acutely malnourished, compared to 2.1percent of those who had not had diarrhoea. The same was observed for cough, with wasting prevalence of 2.4percent among children who had suffered from cough. Chronic malnutrition is a long-term process and therefore this recent bout of diarrhoeal illness is not directly related to stunting. To identify the determinants of child nutrition by MUAC, multivariate analysis (Cox s Logistic regression model) was performed. Annex 13 shows that lack of measles immunization was highly significant and had a positive relationship with child malnutrition by MUAC. A child who had not been immunized against measles was more likely to be malnourished than the counterpart who had been immunized against measles Relationship between nutritional status and source of water Wasting, stunting and underweight were significantly associated with the type of drinking water source (p<0.05). Prevalence of stunting was higher in households drinking unsafe water than safe water at 41.3percent compared to 23.7percent; underweight was 10.5percent compared to 5.2percent; and wasting at 3.5percent compared to 2.4percent. The interaction between poor food consumptions (inadequate dietary intake including micronutrients coupled with diseases and infections contributed by unsafe water) greatly contributed to the rates of under-nutrition. 30

48 Relationship between nutritional status and type of latrine There was no statistical significant difference between malnutrition and faecal disposal by households (p>0.05). However, by percentage reporting, the prevalence of stunting was higher in households disposing faeces in bushes at 29.5percent, followed by those using community latrines at 28.8percent, private latrines at 24.2percent and other places at 21.6percent. In addition, a similar trend was observed with underweight among households disposing faeces in bushes at 4.7percent, followed by those using private latrines at 2.5percent, and community latrines at 0.7percent. Furthermore, the prevalence of wasting was higher in households disposing faeces in bushes at 10percent, followed by those using community latrines at 7.3percent, other places at 5.7percent and private latrines at 5.1percent according to percentage reporting. This therefore reveals that poor faecal disposal was a predisposing factor as far as under-nutrition was concerned bearing in mind that there is no single causal factor but an interaction between various factors Relationship between nutritional status and household food security (a) Household food consumption patterns In general, the food frequency and dietary diversity revealed that there were relatively a big number of stunted children in households with all the 3 categories with the highest proportion being observed in acceptable food consumption group, followed by poor and borderline at 25.9percent, 23.9percent and 22.9percent respectively. In reference to wasting, a high number of wasted children were identified to be among households in the borderline consumption group, closely followed by poor and acceptable although there was no significant difference in this category at 7percent, 6percent and 5percent respectively. It is therefore possible that intra-household food distribution could have been a major factor although in this assessment; data was not collected on it and therefore recommend that future assessments look into it Overall household food security situation Household food security (defined as a combination of food consumption, food expenditures and level of dependency on food aid) was not significantly associated with child malnutrition, whether acute (wasting) or chronic (stunting). This does not mean that no malnourished children were found in severely or moderately food insecure households, but it indicates that other factors than food security seemed to play a stronger role as determinants of malnutrition. Most of the food insecure households were beneficiaries of food aid, whereas many of the food secure households (in terms of access to food) were found to have malnourished children. While it cannot be concluded that food aid is protecting against malnutrition, the findings indicate that problems of access to food (as identified by low food expenditures and high dependence on food aid). Therefore, the assessment 31

49 indicated that food consumption was not the main determinant of malnutrition in the refugee settlements at the time of the survey Food access, proxied by the number and type of income sources The number and type of income sources influence food access, one of the components of food security. The highest prevalence of wasting was found in households relying on begging at 6.5percent, sale of handcrafts at 4.2percent and brewing at 3.2percent as their main income sources, while the lowest prevalence was observed among those relying on sale of non food cash crops, animal products, fishing, hunting and gathering as well as salaries and sale of non food animal products, all below 2percent. Household wealth was negatively related to child stunting. A child from a wealthy (rich) household was less likely to be stunted compared to the counterpart from a poor household. The differences in the mean Z-scores for wasting were significant between children living in households relying mainly on the sales of livestock/animal products compared to those in households relying on waged labour (p< 0.05) or on the sales of non food cash crops. The mean wasting Z-score of children living in households relying on food aid was also significantly better than those in households depending on wage labour (p<0.05). On the other hand, stunting was significantly higher among children living in households relying mainly on fishing, hunting and gathering at 42.9percent; food production, 26.8percent; begging, 29percent; commercial activity, 26percent; and unskilled labour, 28percent compared to those living in households relying on the waged salary and sales of livestock/animal products (based on the mean Z-score, p<0.05 and p<0.01 respectively). It is important to note that linkages between the income source type and acute malnutrition may be compounded by other factors. In particular, sales of food crops or livestock/animal products were mostly performed by more settled refugees, who were also more likely to obtain their water from unsafe sources and use bush for faecal disposal. Similarly, just settling in refugees were more likely to rely on the sale of food aid for their income, and the majority were living in transit camps where they had better access to safe water sources and improved latrine facilities, as well as to nutrition programmes. The results indicate that a high dependence on selling food aid for income generation was not associated with higher risks of acute malnutrition. However chronic malnutrition tended to be more widespread among children of these households. Acute malnutrition was higher in households with the characteristics of settled residents higher number of animals, reliance on selling crops and commercial activity. 32

50 Food availability, proxied by cultivation and animal raising Crop cultivation: The mean weight for height Z-score was significantly higher for children in households having cultivated more than 2 ha of cereals in the first season of 2010 compared to those having planted less or not at all (p<0.001 in both cases). There were no significant differences in wasting (mean Z-score) between children living in households who had not planted cereals and those in households who had planted less than 2 ha. The lower wasting prevalence associated with high cereal acreage is interesting to note as it does not seem influenced by other factors such as the poorer access to safe water sources and disposal of faecal matter in bushes. Stunting (mean Z- score) was also higher among children of households who had not planted cereals compared to children of households having cultivated more than 2 ha (p<0.05). Animal raising: The mean weight for height Z-score was lower for children living in households with a large number of animals compared to those in households with no or with small numbers of animals. The difference was almost significant when considering TLUs and significant when considering the ownership and numbers of poultry/goats (p<0.05). The mean Z-score for stunting was higher for children living in households with a large number of animals (3 TLUs) compared to those in households with 1 or 2 TLUs (p<0.05). When looking at the animal species, better mean Z-score for stunting was noted for children in households with 1-5 poultry/goats compared to those in households with no poultry/goats (p<0.05). These results are consistent with the association found between the reliance on the sale of livestock/animal products as a main source of income and higher wasting and lower stunting prevalence compared to some other income sources. They may illustrate an effect of the consumption of animal products on linear growth but other factors must also be taken into account Relationship between malnutrition, sex and literacy of the head of household The household data did not indicate associations between the sex of the head of household and the prevalence of child malnutrition. Stunting was significantly associated with the literacy level of the head of household, with a slightly higher prevalence of stunted children in households with illiterate heads (53.4percent) compared to literate heads (46.6percent). Multivariate analysis further confirms that mother/caregiver s education was marginally negatively related with child malnutrition (under-weight and wasting). A child born to or given care by a literate mother or caregiver was less likely to be malnourished than the counterpart under an illiterate caregiver/mother. The findings are in agreement with those of the UDHS 2006 results that revealed that largely literate household heads were men. 33

51 3.2.3 Child health Mothers/caregivers were asked whether each child aged 6 to 59 months in the household had been sick in the previous two weeks. Overall, mothers reported that 86.5 percent [95% CI: percent] of all children had been sick during the recall period. Survey team specifically asked about diarrhoea, fever and cough (Table 3.8) in the two weeks prior to the survey. Overall, fever was the most frequently reported symptom included in the recall and affected 35.3 percent of children [95% CI: percent], followed by potential respiratory infections reported in 22.9percent of children [95% CI: percent]. Diarrhoea, defined as three or more loose or watery stools in the past 24 hours, was reported in 11.2 percent [95% CI: percent] of children. Table 3.8: Morbidity among Children 6 to 59 months of Age, by Sub region, October 2010 West Nile South West Total Type of symptom No (%) 95% CI No (%) 95% CI No (%) 95%CI Fever , Measles Diarrhoea ARI/Cough Skin diseases Eye diseases Other No Illness , , , Note: Fever was self-reported by the caregiver and not clinically verified; the survey did not record cases of diagnosed malaria separately. Fever and ARI/Cough were more pronounced in West Nile (43.8percent [95% CI: percent] and 24.8percent [95% CI: percent]) than South West (24.6percent [95% CI: percent] and 20.5percent [95% CI: percent]) respectively. Key informants and Focus Group Discussion with people in the communities confirmed that the most common diseases in the settlements were malaria (fever), RTI/Cough, worms, diarrhoea, malnutrition, intestinal infection, skin infection, trachoma, eye infections Infant and Young Child Feeding Breastfeeding status was based on maternal recall of the child feeding practices (children aged 6 to 24 months). Exclusive breastfeeding refers to children who receive only breast milk, or breast milk with only vitamin/mineral supplements or medicine. 34

52 Complementary feeding is defined as those children who receive breast milk and solid or semi-solid food from 6 months of age to 2 years. About half, 53.7 percent of mothers put their child to the breast within the first hour of birth, while 45.7 percent waited more than hour to commence breastfeeding. Overall, 99.1 percent of children less than two years of age had been breastfed at some point during the first two years of life. At the time of the survey, 75.3 percent of children aged 6 to 24 months were still breastfeeding. The rate of exclusive breastfeeding among children at six months of age in the sample was 31.2 percent. Breastfeeding was stopped at the mean age of 16 months, however, 52.4percent of mothers introduced solid, semi solid and soft-foods before 6months of age, which was earlier than the recommended six months. The main foods introduced before 6 months included cereal porridge at 44.8percent, followed by water at 22.3percent with South West contributing a higher proportion 47.9percent for cereal porridge while West Nile was water at 25.5 percent. Just over 50 percent of the children receiving complementary food ate less than three meals per day compared to the recommended five meals a day. Only 8.1 percent of children received four to five meals per day. The main foods fed to children at complementary foods included cereal meal at 41.7percent and legumes at 23.4percent with no significant difference between the two sub regions. It should be noted that meat and meat products, fish and milk were hardly given to the children. At the sub region level, there was no difference in the percentage of mothers who reported having breastfed their child at some point during the first two years of life. Initiation of breastfeeding within the first hour of birth was suboptimal, with 60percent and 47.6percent of children commencing breastfeeding more than one hour after birth and not receiving colostrum in West Nile and South West respectively. In all sub regions, approximately 58percent of children were receiving less than two meals per day, which was remarkably lower than the recommended five times a day for this age group. 3.3 Iron Deficiency Anaemia Iron Deficiency Anaemia among Children (6 to 59months) Of the 1,093 children tested for anaemia, 53.6percent (95% CI ) were anaemic (<11g/dl). Anaemia was more prevalent in children from West Nile at 61.1percent (95% CI ) than those from South West at 45.3percent (95% CI ). The mean haemoglobin level in children was 10.7g/dl with a range of g/dl. Table 3.9 summarizes the prevalence and severity of anaemia in children and women. 35

53 Table 3.9 Population Group Children (6-59months) NP Women (15-49yrs) P Women (15-49yrs) N Haemoglobin Levels and the Prevalence of Anaemia among Children and Women Mean Hb (g/dl) Range (g/dl) Total Anaemia n(%) Mild Anaemia n(%) Moderate Anaemia n(%) Severe Anaemia n(%) 1, (53.6) 433 (39.6) 124 (11.3) 29 (2.7) (17.5) 0 (0) 104 (15.8) 11 (1.7) (32.5) 7 (7.2) 19 (22.9) 2 (2.4) NP=Non Pregnant; P=Pregnant. Cut offs for anaemia in the different population groups are based on the Management of Nutrition in Major Emergencies, WHO 2000 The National Survey guidelines indicate that the prevalence level of below 5percent, 5-20percent, 20-40percent and above 40percent define the level of public health significance within a population as normal, mild, moderate, severe respectively. The anaemia prevalence among children (6 to 59months) was observed to be severe. Compared to the prevalence of anaemia in children of 54.4percent in 2009, there was a slight decrease to 53.6percent in 2010 and specific to Westnile and South West sub regions, the prevalence increased from 38.9percent to 45.3percent for South West and Westnile reduced from 69.8percent to 61.1percent in 2009 and 2010 respectively Main Causes of Iron Deficiency Anaemia There are large number of risk factors in combination that may have contributed to the development of anaemia. These included nutritional deficiencies, parasitic and other infections as well as physiological status. Age: The child's age was associated significantly (p<0.05) with the occurrence of anaemia; the risk was higher among children 6 to 24 months than those within the 25 to 59months age bracket. This age range, significantly associated with anaemia, is in line with the most important changes that occur in child-feeding regimens, namely complementary feeding and exposure to the households eating patterns. The short median duration of exclusive and total breastfeeding, along with the low iron content in complementary foods in children under-two years of age as indicated in this assessment may partially explain these increased risks. 36

54 Table 3.10 Prevalence of Anaemia among Children by Age and Sub Region West Nile South West Child's Age Category Total Child's Age Category Total 6-24 months months 6-59 months 6-24 months months 6-59 months n % For these children, breastfeeding and especially prolonged breastfeeding is an important strategy for preventing anaemia, given the high bioavailability of iron in human breast milk. Furthermore, the diets that these children were exposed to may be reflected in the high anaemia rates, particularly because of the low iron content, and poor in iron absorption activators fruits and vegetables. Such eating practices are known to facilitate the development of anaemia. Hygiene and sanitation practices: Households accessing borehole water had more anaemic children than those accessing other sources at 99.1percent in West Nile than 54.3percent in South West (p<0.05). Households treating their drinking water in South West had more anaemic children than their counter parts in West Nile at 52.3percent and 20.7percent respectively. When closely examined, households boiling their drinking water had more anaemic children (54.2percent) than those treating it by either using water guard or aquatab (Chlorination) at 44.4percent according to percentage reporting. However, the degree of boiling the drinking water was not ascertained and hence it is possible the subjects were not boiling the water properly. Based on the observations coupled with the Focus Group discussions, borehole water was not boiled or even treated for drinking and this led to some of the household members having diarrhea. This easily leads to infections that are closely linked to IDA. Methods used for treating drinking water in West Nile sub region could be more effective than those in South West. Food Consumption patterns: The food consumption patterns at the household levels were greatly associated (p<0.05) with Iron Deficiency Anaemia among children aged 6 to 59months. There were more anaemic children households with acceptable food consumption than borderline and poor at 68percent; 27.4percent and 4.7percent (Refer to Figure 3.4). 37

55 There is a possibility that children in the borderline and poor category were mainly fed on green leafy vegetables than their counterparts. Surprisingly, children fed on foods rich in iron were more anaemic at 63.9percent than those not given iron rich foods (54.7percent) although this was not statistically significant. This clearly indicates that the foods given to children had less iron and/or other factors affected iron absorption from such foods. Nutrition status of children: About 8.2percent of the underweight children were anaemic; 25.2percent stunted and 3.4percent wasted. This indicates that the nutritional status of the children had some contribution to the anaemia in such children. This is because there are studies which show that stunting can be used to determine the severity between malaria and anaemia. This can be revealed by the number of children found to have fever in our findings. Household Food Security: Households that were food secure had 18.8percent anaemic children compared to those that were food insecure and vulnerable at 41.2percent and 40 percent respectively. Food security status of the households had a lot to do with the prevention and control of anaemia among children (6-59months). Method of faecal disposal: There was a more pronounced difference between anaemic children and disposal faecal of matter (p<0.05). The prevalence of anaemia was influenced by the way faecal disposal was done there were more anaemic children in households disposing faecal matter in bushes (65percent), community latrine (69.9percent) and neighbour s latrine (52.1percent) and others (75percent) than households with children who were not anaemic at 35percent, 30.1percent, 47.9percent and 25percent respectively. This further confirms the earlier finding on the risks community latrines and bushes had on the general wellbeing of the community. Identifying strategies on how to improve faecal matter disposal within the settlements is highly recommended. Hand-washing: The inadequate hand washing contributed to the prevalence of anaemia in the settlements especially among households where members of households had even to be reminded about washing their hands at 58.3percent Iron Deficiency Anaemia among Women (15 to 49years) Of the overall 1,498 women who participated in the survey, only 739 women were assessed for anaemia. Of these, 83 were pregnant and 656, non pregnant. The prevalence of malnutrition among women assessed is presented in Table The prevalence of Iron Deficiency Anaemia was 32.5percent among pregnant women contributing the highest proportion compared to non pregnant at 17.5percent, giving an overall prevalence of IDA among women at 27.4percent. 38

56 Specific to the sub regions, women in West Nile (36.1percent [95percent CI: percent]) were more anaemic than their counterparts in South West at 26percent [95% CI: percent]). There was a significant drop in the prevalence of IDA in West Nile from 51.3percent in 2009 to 36.1percent in 2010; and an increase in South west from 14.1 percent to 26percent. The prevalence of anaemia by settlement is as indicated in Annex 2. According to the public health significance, overall, the prevalence of anaemia among all women irrespective of their physiological status was found to be in the moderate category ( percent) refer to Table 2.10 for detailed thresholds Main Causes of Iron Deficiency Anaemia Parity: This refers to the number of deliveries. The prevalence of anaemia among women decreased with the increase in parity that is from 24percent for those having given birth to one child through to 15percent with those with 6 children; however this remained stagnant for those having given birth to more than 6 children ranging between 19percent with those with 5 children to 20percent with 9 children. The details are as shown in Figure Prevalence of anaemia Prevalence of anaemia Figure 3.5: Prevalence of Anemia by Parity Pregnancy creates a large demand for iron, which is needed to develop the foetus and placenta and to expand a woman s blood volume. Additional iron is lost with blood lost at delivery. When the iron demands of pregnancy are combined with the iron demands of adolescent growth, girls enter adulthood at great risk of iron deficiency. The postpartum period is a time of recuperation of iron status, as iron in additional red blood cells made during pregnancy becomes available. This is especially true during the period of full breastfeeding and lactational amenorrhea, because the iron cost of breastfeeding is typically less than the iron cost of regular menstruation. Hygiene and Sanitation practices: There was a significant difference between anaemia among women and source of drinking water (p=0.01) with more households accessing piped water (32percent), protected spring (33percent) and open well/spring (26.7percent) having anaemic women than those accessing water from boreholes at 16.5percent. There was however more households boiling drinking water having anaemic women than those treating it by chlorination at 24percent compared to 14.3percent. Food Consumption patterns: When compared to the 3 food consumption groups, there were more anaemic woman and among the poor (21.4percent) than borderline (20.1percent) and acceptable (18.5percent). There was however no significant difference among the 3 consumption groups (p> 0.05). 39

57 Household Food Security: By Household Food Security, 19.9percent of food insecure households were anaemic; 21.4percent vulnerable and 15.8percent food secure. 3.4 Health Interventions Treatment of Household Members For household members with reported illness in the past two weeks, mothers/caregivers were asked whether they had sought treatment. Overall, almost all members of the households 99.2 percent [95% CI: percent] had been brought to a health service provider for treatment. The proportion of members who had been brought for treatment did not vary much by sub region with 99.9 percent [95% CI: percent] in West Nile and 98.3 percent [95% CI: 96.2 Figure 3.6: Sources of treatment for household members, October 2010 (n = 100.4percent] in South West. As shown in Figure 3.6, Health Centres were the main source of treatment for household members, followed by hospitals and government clinics. The Key informants also reported that there was at least one health facility in a refugee settlement with some settlements having one HC III, while the others had two or more health facilities, in most cases HC IIs. However, there were concerns from KIs that due to the inadequacy of medicines in the health facilities supported by the District Health Services, clients had to move from one facility to another with records/books to access the medicines. This calls standardisation of health medical forms to avoid getting more medicines at a given time. The main services reported to have been family planning, VCT, OPD, IPD, MCH, PMTCT, Referral, TFP, delivery, maternity, family planning, HIV/AIDS services and PHC services. Many people accessed health services from the nearest health facility for treatment when they fell sick and the distance to the HF was reported to be 5 to 10 km. The Key informants also revealed that in some facilities, the staffing level was up to 90percent while in other health facilities the staffing level was as low as 30percent. The Focus Group Discussants confirmed that there were qualified health workers at the health facilities in the settlements but also revealed that there was shortage of medicines, resulting from irregular/inadequate supply and at most times there was only malarial treatment. Furthermore, it was reported that Health Workers 40

58 were absent during weekends and on public holidays, which affected access and utilisation of services during those times. It was also reported that the refugee settlements had two VHTs/ CHWs (Community Health Workers) per village and the qualifications ranged from having the ability to read and write; having acquired an O level certificate, being an enrolled nurse or nursing assistant. In Adjumani specifically, the key informant interviews, revealed that extension workers at sub-county level were used as community contact points Access and use of mosquito nets Overall, 52.3 percent [95% CI: percent] of the households reported all household members having mosquito nets, which was a slight improvement in mosquito net availability from the 2009 survey at 48.9 percent (Figure 3.7). Mosquito net coverage was highest in West Nile at 59.8 percent [95% CI: percent], and a slightly lower coverage in South West at 44.6 percent [95% CI: percent] in South West. Mosquito net usage for the previous night among children 6 to 59 months of age was reported at 78.4 percent [95% CI: percent] West Nile South West Total Figure 3.7: Comparison of mosquito net coverage in 2010 and 2009 by sub region It was further confirmed through the focus group discussions and key informant interviews that ITNS had been adequately distributed to all households in the settlements with support from UNHCR, AHA, UN foundation, GTZ, and AAH. It should be noted that the utilization of the ITNs was not assessed however; the KIs reported that they were used for building/housing, fishing and sold in markets Coverage of Vitamin A supplementation Overall, about half the children aged 6 to 59months, (49.5percent [95% CI: percent]) had received a vitamin A supplement with evidence of a card, in the six months prior to the survey (Table 3.11). This is an improvement from that reported in 2009 of 44.4 percent. It should be noted that the coverage is far below the Vitamin A supplementation national target of 85percent. 41

59 Table 3.11: Vitamin A coverage among children 6 to 59 months, October, 2010 West Nile South West Total No (%) 95% CI No (%) 95%CI No (% 95%CI Vitamin A Coverage Yes with card , Yes without card No with card No without card Total 1, , , At the sub regional level, the coverage of vitamin A supplementation with evidence of a card among children 6 to 59 months of age was higher in West Nile at 56.9percent [95% CI: percent] compared in South West at 41.8percent [95% CI: percent]. The main source of vitamin A supplementation in both sub regions was through routine immunization and CHILD days Coverage of measles immunization The measles vaccination rate for children aged 9 to 59 months with evidence of a card was 57.9percent [95% CI: percent] (Table 3.12). There was a great improvement from that reported in 2009 of 44.4percent. Table 3.12: Measles immunization coverage among children, October, 2010 West Nile South West Total No (%) 95% CI No (%) 95% CI No (%) 95% CI Measles Vaccination Coverage Yes with card , Yes without card No with card No without card Total 1, , , At the sub regional level, the measles coverage with evidence of a card was higher in West Nile at 72.8 percent [95%CI: percent] compared to 42.1 percent [95% CI: percent] in South West. This coverage is mainly attributed to the CHILD days exercises conducted a few days to the survey in the different settlements. This coverage is still below that national measles coverage target is 90percent. 42

60 Figure3.8 shows the different vaccination and supplementation coverage in the sub regions of West Nile and South West with West Nile showing a slightly higher coverage than South West Measles DPT 3 Vitamin A De-worming West Nile South West Overall Figure 3.8: Vaccination and Supplementation Coverage in the Sub regions, October 2010 The KIs and FGDs confirmed that CHILD days/ Immunization campaigns had just taken place in October Mortality A total of 11,510 individuals were reportedly living in the sampled households at the beginning of the recall period on 11 th July Of this total, 10,986 (95.4 percent [95% CI: percent]) were reported alive and still living in their household at the end of the recall period, 481 (4.2 percent [95% CI: percent]) were reported alive and living elsewhere, 38 (0.3 percent [95% CI: percent]) were reported to have died since that time and 5 (0.0 percent) were reported missing or their status was not recorded. In the results that follow, missing persons were not classified as dead. The recall period between 11 July 2010 and the date of the surveys varied between 104 and 115 days across the sub regions. The point estimate for the CMR during this period was 0.37 deaths per 10,000 per day [95% CI: ], which was below the emergency threshold of one death per 10,000 per day (Table 3.13). The U5MR of 0.33 [95percent CI: ] for the approximate three-month recall period was below the emergency benchmark of two deaths per 10,000 per day and slightly below the CMR. 43

61 Table 3.13: Crude and under-5 mortality rates for the among samples populations 2009 and Crude Mortality Rate 0.21 (95% CI: ) 0.37 (95% CI: ) Under 5 Mortality Rate 0.77 (95% CI: ) 0.33 (95% CI: ) CMR = Crude Mortality Rate as deaths per 10,000 persons per day U5MR = Mortality Rate for children less than five years of age per 10,000 per day Whereas the CMR had slightly increased in 2010 from 0.21 to 0.37, the under 5 mortality rate had reduced from 0.77 in 2009 to 0.33 in Of the 384 individuals who died during the recall period, the cause of death was not reported for 31.1 percent of the cases. Figure 3.9 shows the reported causes of death with fever being the highest cause of death in under 5 years. Figure 3.9: Possible causes of death in the under and over fives, October Maternal Health Overall, 173 (10.5 percent [95% CI: percent]) mothers of children 6 to 59 months of age in the sample were pregnant at the time of the survey. Approximately half of all mothers (54.2 percent [95% CI: percent]) were breastfeeding a child at the time of the survey; while only 17 (1percentpercent [95% CI: percent]) were pregnant and breastfeeding at the same time. A total of 1,671 women (15 49 years) were assessed. Slightly less than half 43.8percent [95% CI: percent]) had secondary and tertiary education. Most of the women interviewed 94.2percent [95% CI: percent]) were mothers to children assessed. The prevalence of anaemia among women is fully presented and discussed under section 3.3. The prevalence of malnutrition among women (15 49years) is as shown in Table

62 Table 3.14: Prevalence of Malnutrition among women (15 49 years) using MUAC and BMI, October 2010 Malnutrition West Nile South West Total No % 95%CI No % 95% CI No % 95%CI MUAC < 21.5cm BMI < The prevalence of underweight using BMI among women (non pregnant) in South West and West Nile was 3.8percent [95% CI: percent] and 13.9percent [95% CI: percent]. It should be noted that under weight among women in West Nile is about 3.4 times more than that in South west. Overall, underweight among women was 8.3percent [95% CI: percent] a significant increase from that reported in 2009 of 3.7percent. Cases of overweight and obese women was 12.3percent [95% CI: percent] and 3percent [95% CI: percent] respectively. MUAC was taken from 1,563 women, of these, 0.8percent [95% CI: percent] were severely malnourished and 1.4percent [95% CI: percent) were moderately malnourished. Women at risk of under-nutrition were 3.2percent [95% CI: percent]. 3.8 Water and Sanitation Water Using the water, environment and sanitation (WES) sector partners and UNICEF s classification of safe water source (i.e., piped water/standpipe, protected well/spring, borehole with hand-pump, water bladder or tanker truck), overall, 91.1 percent [95% CI: percent] of the population was using a safe water source as their main source of drinking water. Variations existed across sub regions: 99.5 percent [95% CI: percent] in West Nile and 75.8 percent [95% CI: percent] in South West. Households were also asked about treating their drinking water. Overall, 28percent [95% CI: percent] of households reported using treated water. West Nile had the lowest proportion of households using treated water at 22 percent [95% CI: percent] compared to South West at 34.3percent [95% CI: percent]. Treatment of drinking water was mainly by boiling at 61percent followed by chlorination 10. at 37.8percent in all settlements. Whereas boiling was mainly reported in South West at 91.8percent [95% CI: percent] and West Nile at 13.5 percent CI: percent]; chlorination was mainly in West Nile at 86.5percent [95% CI: percent] and South West at only 6.2percent CI: percent]. 10 For the purpose of this survey, chlorination was household water treatment for drinking such as adding water guard, aquatab, etc 45

63 Additionally, water, environment and sanitation (WES) sector partners and UNICEF defined safe water as either using a safe water source or treated water. Overall, 87.8 percent [95% CI: percent] of households were defined as having access to safe water. West Nile had the highest proportion of households with access to safe water at 99.5 percent [95% CI: percent], while South West had only 75.8 percent [95% CI: percent]. It is important to note that the storage conditions of water were not assessed, which may have adversely affected the safety of the water. The details are shown in Table Table 3.15: Main of Sources of Water by Sub Region, October 2010 (n = 2,070) Drinking Water Source West Nile South West Total No % 95%CI No % 95% CI No % 95%CI Piped water through tap Water from open well/spring Water from protected well/spring Water from borehole 1, , Surface water (River, dam, run off) Rain water Other Total 1, , , The borehole water was mainly brought up by hand pump or pulley at 96.6percent [95% CI: percent]). Adult women were responsible for collecting water at 55.2percent [95% CI: percent]), followed by everybody, young girls and children at 15.5 percent [95% CI: percent]), 8.1percent [95% CI: percent]) and 7.9percent [95% CI: percent]) respectively. FGDs and KIs confirmed that the main source of water was boreholes, followed by wells and water tanks in the West Nile sub region. Most households in the sub region had access to borehole and other water sources within a reach of 1-4km radius. The average amount of water used in the Refugee Settlements in the two regions is presented in Table The mean amount of water used per household in the refugee settlements was 86 litres, an increase from 76 in 2009; while the average amount of water used per person per day was 15.3 litres, with no much difference from that of 2009 of 15.6l/p/d which is equivalent to the recommended of 15l/person/day. However the refugee settlements in South West region used less amount of water recommended per person per day of only 11.9 l/p/d. Following the FGDs and KIs, it was observed that three to six 20litre jerry-cans were used in a day by a household. 46

64 Table 3.16: Amount of water used per person per day by sub region Settlements Amount water (litres) No. of households Amount of water per household *Water amount per person South West 60,624 1, West Nile 117,400 1, Overall 178,024 2, *Mean number of persons per household was taken at 5 persons per household Sanitation (a) Latrine Coverage: Overall, the latrine coverage was noted to be 70.6 percent [95% CI: percent] with South West having a higher coverage of 73.9percent [95% CI: percent] than West Nile 67.4percent [95% CI: percent]. Compared to the 2009 report, there was an improvement in latrine coverage from 61.7 per cent to 67.4percent in West Nile while a drop was noted to be from 79.6percent to 73.9percent in South West. The latrines were of different standards and types from privately to communally owned latrines however, for the purpose of this assessment, only the privately owned latrines were used to calculate the latrine coverage. The respondents reported that that the main source of faecal waste disposal was through private and community latrines followed by bushes and in some areas, holes were dug and mud put on top. Privately owned latrines were locally constructed using grass, poles, mad and polythene sheets however, the materials for constructing and cleaning the latrines were not readily available and the latrines were too near the houses where rain water easily washed dirt and flies into the houses, contaminating their food. Community latrines especially in Arua were constructed by individuals with help from IPS who gave tools for maintaining boreholes and constructing latrines. It was also revealed that a few areas had by-laws for the construction and maintenance of latrines which included all households having pit latrines and hand washing facilities. Whoever would not comply with the laws was either fined a fee between UGX 15,000 and UGX 50,000 or subjected to labour depending on the sub region. (b) Hand-washing: This was so pronounced in the settlement with 90.4percent of households reporting washing their hands at any given time. There were more households in West Nile reporting washing their hands at critical times, after easing oneself, before and after serving and eating meals, including feeding the children at 84.3percent [95% CI: percent] compared to those in South West at 66.1percent[95% CI: percent]. There were however no hand-washing facilities observed in the different households at the time of the assessment. (c) Rubbish Disposal: The most common method for rubbish disposal was composite pit at 65percent [95% CI: percent] followed by the garden and bushes at

65 % households percent [95% CI: percent] and 14.3percent [95% CI: percent] respectively. This was an increase from 62.7percent in There was however no significant difference in the 2 sub regions with 69.3percent [95% CI: percent] in West Nile and 60.4percent South West [95% CI: percent]. (d) Shelter: Through general observations, FGDs and KIs, it was observed that almost all refugees had temporary shelters that were not conducive for human habitation. The shelters leaked, had inadequate aeration and animal stayed together with humans. The respondents reported that they could not obtain poles and grass for thatching the houses given the policy of not cutting trees. UNHCR provided plastic sheetings on arrival but there was follow up to find out whether they were still viable. In conclusion, the existing water and sanitation interventions in the settlements included: borehole rehabilitation through communal contribution, support from government and agencies dealing in WATSAN and self reliance communal approach. However the discussions revealed that though these interventions existed, many areas did not have clean water and no specific containers for storing drinking water. For example, water from Lake Nakivale was noted to be very dirty, even when purifiers were used and also cutting down of trees to boil water was a challenge given the policy that prevented tree cutting by government. 3.9 Food Availability: Crop and Animal Production Land Access and Agricultural activities Cultivation in the first season In principle, each of the households in the settlement was entitled to a 0.5 hectare piece of land to engage in agricultural activities. However, the ability of the members of the households to cultivate food in the first season of the year (March to July) was established as shown in Figure No Yes 0.0 West Nile (n=1057) South-West (n=980) General (n=2037) Regions Figure 3.10 Proportion of households able to cultivate some food in 1 st season of 2010 Key informants at community level indicated that refugees from West Nile sub region were more likely to have cultivated this year than in South West sub region given that 48

66 Reasons the Rwandese refugees were denied access to land by the Ugandan government while the Somalis were are not traditionally inclined to crop production. This was further confirmed with the household survey that showed that 72.5percent of households in West Nile had cultivated land compared to 58.1percent in South West sub regions giving the overall of 65.9percent for both sub regions. Figures 3.11 and 3.12 show reasons given by households for not having been able to cultivate any food in the first season in West Nile and South West sub-regions respectively. Specific to the West Nile sub region, for those not cultivating, the main reasons given were limited access to land (44.8percent), sickness and disabilities (34.3percent) and poor weather (11.9percent). No access to land Poor weather Land conflicts We are not agriculturalists Insufficient family/household labour insufficient seeds and tools Sickness or physical inability Prohibited by the government The land is infertile/marginal Prohibited by the clan Insecurity % Households in West Nile Region (n=268) Figure 3.11: Main reasons for not cultivating any food in the 1 st season of 2010 in West Nile Region The reasons given for the South Western sub region were almost the same as those from the West Nile sub region and included limited access to land (36.3percent) as well as sickness and disabilities (28percent). However, reason given such as not traditional inclined to crop production was listed as the third main reason in South West at 28percent. It should also be noted that the proportion of households reporting no access to land and sickness/disabilities were more pronounced in West Nile than South West sub regions. 49

67 Reasons No access to land Poor weather Land conflicts We are not agriculturalists Insufficient family/household labour insufficient seeds and tools Sickness or physical inability Prohibited by the government The land is infertile/marginal Prohibited by the clan Insecurity % Households in South-West Region (n=339) Figure 3.12: Main reasons for not cultivating any food in the 1 st season of 2010 in South-West Region In general, approximately 40percent of the households indicated that limited access to land was the main reason for not cultivating any food on the first season of 2010 with approximately 30.8percent of the households indicating sickness and disabilities while 7.9percent was due to poor weather. Compared to the Assessment report of 2009, there was a reported increase in the number of households that cultivated land in the first season of 2010 compared to that of 2009; from 60percent in 2009 to 65.9percent in However, although there was a remarkable increase in West Nile sub region from 55percent in 2009 to 72.5percent in 2010; that was not the case in South West, where there was a decline from 65percent in 2009 to 58.9percent in Total Yields cultivated in the previous season The crops cultivated by most households were cassava, sweet potatoes, maize and sorghum. An overall picture indicates that there were more households that cultivated crops in West Nile compared to South West which was different from the findings of the 2009 survey. 50

68 Table 3.17: Average Quantities of Crop Yields for the First Season of 2010 Region Yield West Nile South-West Overall n Yield Mean n Yield Mean n Yield Mean Maize , , , Millet Sorghum , , , Potatoes , , , Cassava , , , Bananas , , Per Capita Yield 1, ,132 1, , , ,910 1,124 Maize , , , Sorghum 309 8, , , Millet Potatoes , , , Cassava , , , Bananas , ,578 86, , ,534 65, For the households which were engaged in crop production, the yields obtained from the crops they produced in the first season of 2010 were generally quite low compared to the national averages. This is as shown in Table Whereas cassava (157,274kg) recorded the highest yields in West Nile sub region followed by sweet potatoes (123,810kg) and maize (110,982kg), South West sub region, maize (98,064kg) recorded the highest yields, followed by cassava (42,979kg) and sorghum (7,330kg). The reasons for the low yields included: Low input usage that is mininal usage of improved seed and fertilisers, poor agronomic practices e.g. mulching, water conservation, pest, disease and weed maganement. According to the FGDs and KIs, there were concerns about the soil fertility in certain areas in South West especially near Lake Nakivale where there were signs of soil salinity, which is unfavourable for crop production. Furthermore, it was observed that the post harvest and handling practices in the communities were very inadequate coupled with the serious soil erosion around Lake Nakivale. This was mainly attributed to poor farming practices and land fragmentation caused mainly by polygamous families. Therefore involvement of the production coordination offices in the districts at all levels of planning, implementation, monitoring and evaluation of interventions in the refugee settlementrs was crucial. Specifically to access to land, there was disgruntlement about land allocation whereby some refugees complaied of being allocated plots of land far away from their homesteads yet the land near them was left vacant or even worse leased to nationals. Communities which were able to engage in crop production ought to be accorded more land than those which are not. 51

69 Average Quantities of Crop Yields (Kg per Ha) Figure 3.13: Post Harvest handling practices among Refugees in South West Sub Region The comparison of the crop yields in 2009 with that of 2010 is as shown in Figure 3.14.There was a general improvement in yields in the first season of 2010 compared with that of 2009 with a remarkable increase in the yield of maize from 52kg/ha in 2009 to 367kg/ha in FGDs indicated that although they were able to cultivate food, it took only 2 months on average to be completed Maize Millet 0 39 Sorghum 7 6 S.Potatoes 12 Cassava 19 Bananas Figure 3.14: Average Quantities of crop yields for the first season 2009 compared to

70 % Households Livestock production Livestock Ownership by Type The household survey assessed the ownership of livestock and reasons for not owning were also established. The survey findings are depicted in Figures 3.15 and Cattle Sheep Goat Pig Rabbit Poultry Donkey Livestock types owned West Nile South West General Figure 3.15: Livestock ownership by type by sub region In general, the majority of households owned poultry and goats at 47.1 percent and 30.1percent respectively with West Nile sub region recording the highest proportion of households owning poultry (60.2percent) and goats (44.1percent). The South West sub region recorded higher proportion of households owning poultry (32.2percent), goats (15percent) and pigs (8.8percent). This implies that the comunities were moving towards diversified income and food sources. The least reared livestock was donkey, rabbit and sheep in both sub regions. The main constraints facing livestock production in the refugee settlements included pests and diseases (66.4percent) and followed by lack of veterinary services(10.5percent) and limited availability of pastures and animal feed (10.2percent). There was however no much difference in the constraints faced among the 2 sub regions, a possible indication that there were limited livestock services provided to the community in the sub regions. This further confirms that refugee communities were excluded from extension service provision from the Key informants. 53

71 Reasons Other Theft Insecurity Lack of veterinary services Shortage of water Shortage of pasture/feed Inadequate labour Parasites/diseases Poor breed % Households West Nile (n=639) South-West (n=263) General (n=902) Figure 3.16: Main constraints facing households engaged in livestock production This assessment could not compare the livestock production of 2009 with that of 2010 due to the fact that the one of 2009 did not assess this component Relationship between Food Consumption and Animal ownership Based on the FAO guidance, a common unit to describe livestock numbers of various species as a single figure that expresses the total amount of livestock present was used. The concept of an "Exchange Ratio", whereby different species of different average size could be compared and described in relation to a common unit - 1 Tropical Livestock Unit (TLU) was utilised. Table 3.18 shows animal ownership (in Tropical Livestock Unit - TLU) by food consumption category. It is evident that the communities falling into the poor food consumption category recorded the highest proportion of households not owning any animals at 61.8 percent. 54

72 Table 3.18: Animal Ownership TLU Categories by Food Consumption Group, October 2010 Region West Nile South- West Animal Ownership TLU categories Acceptable Borderline Poor Total n % n % n % n % No Animals Less than 1 TLU TLUs More than 2 TLUs Total No Animals Less than 1 TLU TLUs More than 2 TLUs Total Food Access: Income, Expenditures and Coping Strategies Income sources, Expenditures and Assets Number and type of income sources (a) Number of income sources Generally, diversity of income sources reduces the risk of becoming food insecure. Figure 3.17 shows that the majority of households in the 2 sub regions had at least one main income source at 41.9percent. The proportion of households relying on three or more sources of income was higher in West Nile possibly indicative of a more diversified livelihood option in the sub region. In general, incomes were much higher in West Nile than South West. This could possibly be attributed to the difference in proximity to markets e.g. West Nile sub region was linked to Southern Sudan and the trade route to Kampala. Figure 3.17: Number of income sources per sub region 55

73 There was however a slight difference between households with 2 income sources in the 2 sub regions with 33.4percent in South West and 35.5percent in West Nile. Figures 3.18 and 3.19 show the number of income sources by food security group for West Nile and South West respectively: Figure 3.18: Number of Income Sources by Food Security level in West Nile Figure 3.19: Number of Income Sources by Food Security level in South West It should be noted that: The less the income sources, the more food insecure the households were. There were more households with one income source at the same time food insecure with 70.7percent in South West and 56.4percent in West Nile; The higher the income sources, the more food secure households were. There were more households in West Nile with 3 or more sources of income and also more households being food secure in West Nile (36.3percent) than South West at 17.8percent. 56

74 Generally, for both West Nile and South West Nile sub regions, it was observed that households with more diversified sources of income tended to be more food secure than those with one or two. (b) Type of income source Table 3.19 shows the main sources of income for the households which included unskilled wage labour (57.6percent), commercial activity (54.6percent), agricultural labour (53.5percent), food crop production (53.1percent) and trading in food crop or non food crops (52percent). The least sources of income for the households were rental of property (0percent), non food cash crops sales (13.8percent) and fishing, hunting and gathering (17.5percent). More than half (57.6percent) of households were engaged in unskilled wage labour, an indication of an unstable source of income which makes them vulnerable to food insecurity. The main types of income when compared by sub regions indicate that most were higher in South West than West Nile sub region: unskilled wage labour at 67.6percent and 45.2percent; commercial activity at 55.5percent and 52.7percent; agricultural labour at 62.9percent and 43.5percent; and food crop production at 57percent and 49.5percent respectively. However, trading in food or non food crops was higher in West Nile than South West at 62percent compared to 37percent. This confirms the earlier observation in relation to proximity to market and trade routes. Table 3.19: Income sources for households Income Source West Nile South-West Total n % n % n % Food crop production Non-food cash crops sales Livestock production Animal products sales Trading in food crop or non food crops Commercial activity Petty trading Unskilled wage labour Agricultural labour Handicrafts Brewing Sale of natural resources (charcoal, wood) Remittances or kinship Salaries, wages Rental of property (parcels, buildings) Government allowances Fishing, hunting & gathering Gifts from neighbours, relatives Begging Food aid Others Source: Survey data,

75 Household productive and domestic assets Ownership of productive and domestic assets (a) Ownership of Agricultural Productive Assets Table 3.20 shows the various productive assets owned by households in the 2 sub regions of West Nile and South West, with the majority of households owning at least a hoe (83.8percent) followed by a bicycle (34.6percent). There was however, no much difference in the ownership of the productive assets in the 2 sub regions. Table 3.20: Households owning Agricultural productive assets Asset West Nile South West Overall n % n % n % Hoe Plough Tractor Fishing net Bicycle This is a possible indicator that households were able to cultivate land. (b) Ownership of Livestock Assets Table 3.21 shows the proportion of households owning livestock in the 2 sub regions of West Nile and South West, with the majority owning poultry (47.1percent) and goats (30.1percent). Specific to the sub regions, there were more households owning poultry (60.2percent) and goats (44.1percent) in Westnile than South West (32.2percent and 15percent respectively) sub regions. Table 3.21: Households owning Livestock assets Asset West Nile South West Overall n % n % n % Rabbit Poultry Donkey Cattle Sheep Goat Pig Owning of livestock assets such as goats and chicken can easily converted into cash. This implies that households in West Nile sub region were able to access food other than what was provided through consumption and/or selling of the livestock assets. (c) Ownership of Domestic Assets Table 3.22 shows the proportion of households owning domestic articles in the 2 sub regions with most households owning at least a bed (75.1percent), followed by chairs 58

76 (71.1percent), mattress (69.2percent) and table (63.2percent). Generally, households in West Nile sub region owned more assets than those in South West sub region. Table 3.22: Households owning Domestic assets Asset West Nile South West Overall n % n % n % Radio/tape Cellophone Sewing machine Automobile Boat/Motor Canoe Motorcycle Television Bed Table Chairs Mattress Generator In conclusion, households owned more of non productive assets than productive assets with households in West Nile sub region owning more assets than those in South West sub region. This also shows that refugees in West Nile were more settled than their counterparts in South West Asset Ownership by Food Security Group Data was further analysed and indicated that households with assets were more food insecure than those without assets for example households owning poultry, cattle and goats were more food insecure than those owning at 61.5, 96.6, 79.8 percent compared to 38.8, 3.45 and 20.2 percent respectively. In conclusion, ownership of assets does not necessarily translate into households being food secure; however, assets could be easily converted to cash to purchase food as a coping mechanism Wealth Index The Wealth Index is a composite measure of the cumulative living standard of a household. It was calculated using data on a household s ownership of beds, tables, chairs, mattresses, generator, radio/tape, cell phone, sewing machine, bicycle, automobile, boat and motor, canoe, fishing net, hoe, ox-plough, motorcycle, television, tractor, cattle, sheep, goats, pigs, rabbits, poultry, donkeys, types of water source for drinking water, type of fuel used for cooking, and facilities for fecal disposal. Weights (factor scores) for each of the assets was generated through principal component analysis and summed by household. The resulting asset scores were standardized in relation to a standard normal distribution with a mean of zero and a standard deviation of one. These scores were summed by household, and individual households ranked according to the total score of the household. The households were then divided into 59

77 wealth quintiles as Lowest/Poorest, Second/Poor, Middle/Moderate, Fourth/Rich, and Highest/Richest. More than half of the households in the West Nile region (56.3%) were rich (among the fourth and fifth quintiles) compared to 22.9% in the South West region, as shown in Figure 20 The majority of households in the South West region (53.5%) were poor (among the first and second quintiles), while 26.7% of households in the West Nile region were poor. Figure Wealth Quintiles for the West Nile and South West Refugee Settlements Major household expenditures Share of Food, Health and other expenditures Table 3.23 depicts the average expenditures of households in each sub region on selected food items recorded for a 30-day period prior to the survey. Food expenditures represented on average 27.2percent of total monthly expenditures, while health expenditures represented 26.1percent. There was however no much difference between the share of food and health expenditures in the 2 sub regions of West Nile and South West. Table 3.23: Proportional expenditure on food, health, education and other items West Nile South-West Total Item n % n % n % Expenditure on Food 1, Expenditure on Health Expenditure on Education Expenditure on the rest

78 Level of Monthly food expenditures and type of food purchased Table 3.24 shows that the highest expenditure was on cereals (17,933 UGX), followed by meat, eggs and fish (6,851 UGX), sugar (5,991 UGX) and pulses (5,822 UGX). The least expenditures were mainly on cooked or processed foods (241 UGX) and drinking water (444 UGX). Table 3.24: Average expenditures of households on selected food items in a 30-day period West Nile South-West Total Expenditure on the following items n Mean (UGX) n Mean (UGX) n Mean (UGX) Cereals (sorghum, millet, maize, rice, etc) , , ,933 Cooking oil 894 5, , ,323 Meat, eggs, fish 826 5, , ,851 Pulses, (G.nuts, beans ) 831 5, , ,822 Sugar 814 5, , ,991 Milk, yoghurt, cheese 671 1, , ,333 Fruits & vegetables 707 2, , ,995 Cooked or processed food Expenditure on drinking water Other foods 740 1, , ,283 The mean monthly food expenditures were slightly higher amongst households in South West (54,199 UGX) compared to West Nile (42,290 UGX) sub regions giving an overall average expenditure among households of 48,215 UGX per month as shown in Table This may be related to the fact that having higher incomes does not necessarily translate into expenditures on food. West Nile had a higher capita expenditure than South West sub region, reflecting the higher limitations on incomeearning activities for these households. Table 3.25: Average Monthly Food Expenditures per capita (UGX) Past 30-day food expenditures Average monthly food expenditures (UGX) Average Monthly Food Expenditures per capita West Nile 42,290 1,055 South West 54,199 1,015 Total 96,489 2, Share of food expenditures out of total monthly expenditures On average, households dedicated 27percent of their monthly expenditure to food purchases. This apportioning was fairly the same in West Nile and South-West (27.1percent and 27.3percent respectively). This situation reflects a relatively fair economic precarity of the refugee populations of refugees, other factors of deprivation assumed normal. Food expenditure by poor, borderline and acceptable categorization had a different picture: Households in poor category tended to spend 61

79 more on food than acceptable and borderline food consumption groups (West-Nile: poor 60percent; borderline 53percent and acceptable 55percent, while South-West: poor 59percent; borderline 38; and acceptable 4percent). This could be more of an indicator of from-hand-to-mouth way of living than good economic capacity. Increase in the average amount of monthly food expenditure per capita did not show any marked corresponding increase in the proportion of the total expenditure of the households. The households expenditure on food in this situation, of low income earners/living standards, may not be relied on to deduce a specific fact because the reasons are many fold. Households could have been spending more on food because they were not producing enough of their own food, were getting insufficient food, and/or they had been made aware of the importance of diets, through sensitization by service providers (like NGOs) in the settlement. On the other hand households may not have been allocating much share of their expenditure on food because they were producing enough food of their own, getting sufficient food aid or getting food handouts from other households. Even this may not mean that their economic situation that would impact on food access was good Coping strategies in the event of food shortages Main types of coping strategies according to the various household groups Overall, 79.7percent of households experienced food shortages 1 week prior to the assessment. The main types of coping strategies and the percentage of households adopting these strategies in the 7-day period prior to the survey are shown in Table Overall, the two main strategies used in the first instance to cope with food shortages were relying on less preferred or less expensive food (74.5percent), eating less amounts of food at meal time (62.7percent) and reduce number of meals eaten in a day (65.2percent). Feeding working members of the family (4.6percent), consume seed stock for next season (16.3percent) and sending household members to eat elsewhere (18.3percent) were much less frequent as a first response. There were little differences in the household coping strategies between the two sub regions, with the exception of purchase of food on credit with 40.1percent and 50.1percent and consumption of seed stock held for the next season at 21.9percent and 10percent for West Nile and South West sub regions respectively. 62

80 Table 3.26: Coping strategies adopted by households by Sub region, October 2010 Coping strategy West Nile South-West Overall n % n % n % Rely on less preferred or less expensive food? Borrow food or rely on help from a friend or neighbour? Purchase food on credit? Gather wild food, hunt or harvest immature crops? Consume seed stock held for next season? Send household members to eat elsewhere? Limit portion size at meal time? Restrict consumption by adults in order for small children to eat? Feed working members of the household at the expense of nonworking members? Reduce number of meals eaten in a day? Skip entire days without eating? Source: Survey data, 2010 Compared to 2009 where 49percent of households in South West and 86percent in West Nile reported having experienced food shortage in the week prior to the assessment, the 2010 assessment indicates that only 75.2percent of households in South West and 84.1percent in West Nile had experienced food shortage an increase from that of Coping Strategy Index Coping Strategy Indices (CSI) have been constructed as a continuous variable taking into account the frequency and severity of the diet-related coping strategies whose answers were elicited from respondents. The higher the score, the more frequent a household uses severe coping strategies. From the index, quintiles were computed. Thus, high quintile represents higher coping strategy index. For all the refugee settlement regions, the mean CSI was The South West region refugee camps had a mean CSI (33.7) than those in the West Nile region (29.8). The high CSI in the South West region refugee camps would mean that refugee households in that region used severe diet related strategies more frequently than their counterparts in the West Nile region. Among all the refugee settlements, 37.8percent of the households fell within the high to very high coping strategy quintiles and 42.5percent fell within the very low to low CSI quintiles (See Figure 3.21). The South West region has a higher proportion of refugee households falling within the fourth and fifth quintile (41.5%) as compared to the West Nile region with 34.3% households falling within the same quintiles. 63

81 Figure 3.21: Coping Strategy Index by Sub Region 3.11 Food Consumption Patterns Food consumption diversity and frequency Principles of the food consumption pattern analysis In order to enable comparisons with the Food and Nutrition Security Assessment conducted in 2009, the same methodology was used to determine three food consumption groups: poor ; borderline ; and acceptable ; based on the diversity and frequency of consumption of food items during the 7 days preceding the survey. A total of 8 food groups were used that included: (i) cereals and tubers, (ii) pulses, (iii) vegetables, (iv) fruits, (v) meats, (vi) milk, (vii) oils and (viii) condiments Characteristics of the food consumption patterns in each group a) Poor dietary diversity and frequency of food consumption Table 3.27 shows the food scores for average dietary diversity and food consumption frequency. The diet for the poor consumption group was characterized by: a low consumption of maize, cassava, beans, peas, vegetables and oils; rare/no consumption of rice, sorghum, millet, wheat, potatoes, yams, bread, matooke, groundnuts, fresh fruits, fish, meat, blood, eggs, sugar, milk, beer residue and condiments. Consumption of this diet is expected to contribute to malnutrition in the short-term (such as wasting) and if sustained, to longer term damage such as stunting in children, low birth weight and micronutrient deficiencies, and to limit the physical capacity (and productivity) of individuals engaged in physically demanding activities. 64

82 Table 3.27: Diet characteristics of the poor food consumption pattern Past week food Consumption Maize Rice Sorghum Millet Wheat Cassava Potatoes Yams Bread, Mandazi, Chapati Matooke Beans, Peas Vegetables Groundnuts, Simsim, sunflower Fresh fruits Fish Meat Blood Eggs Oil Sugar Milk Beer Residue Condiments Average score 25 Always (6 7 days) Number of days of consumption in the past week Often Sometimes (4 5 days) (2 3 days) Rarely/Never (0 1 day) b) Borderline dietary diversity and frequency of food consumption The diet of the borderline food consumption group was characterized by: always consuming maize (6 7 days); frequent consumption of beans/peas and vegetables (4 to 5 times per week); low consumption of consuming cassava and oils/fats (2 3 days); rare/ no consumption of rice, sorghum, millet, wheat, potatoes, yams, bread, matooke, groundnuts, fresh fruits, fish, meat, blood, eggs, sugar, milk, beer residue and condiments (once per week or none) such a diet can be expected to contribute to malnutrition (such as stunting in children), low birth weights and micronutrient deficiencies (especially anaemia) if such consumption continues for long periods. Details are shown in Table

83 Table 3.28: Diet characteristics of the borderline food consumption pattern Past week food Consumption Maize Rice Sorghum Millet Wheat Cassava Potatoes Yams Bread, Mandazi, Chapati Matooke Beans, Peas Vegetables Groundnuts, Simsim, sunflower Fresh fruits Fish Meat Blood Eggs Oil Sugar Milk Beer Residue Condiments Average score 36 Always (6 7 days) Number of days of consumption in the past week Often Sometimes (4 5 days) (2 3 days) Rarely/Never (0 1 day) c) Acceptable dietary diversity and frequency of food consumption The diet of the acceptable food consumption group was characterized by: frequent consumption of maize, beans/peas, vegetables and oil (4 to 5 times per week); low consumption of cassava, sugar and condiments (2-3 times per week); rice, sorghum, millet, wheat, potatoes, yams, bread, matooke, groundnuts, fresh fruits, fish, meat, blood, eggs, milk, and beer residue (once per week or none). Refer to Table Despite being labelled acceptable, it is worth noting that some households and specific individuals within this food consumption group will continue to be at risk of malnutrition in the longer term (especially stunting), depending on the amounts of food items that they effectively consume, the share of food within the household, and individual characteristics. 66

84 Table 3.29: Diet characteristics of the acceptable food consumption pattern Past week food Consumption Maize Rice Sorghum Millet Wheat Cassava Potatoes Yams Bread, Mandazi, Chapati Matooke Beans, Peas Vegetables Groundnuts, Simsim, sunflower Fresh fruits Fish Meat Blood Eggs Oil Sugar Milk Beer Residue Condiments Average score 51 Always (6 7 days) Number of days of consumption in the past week Often Sometimes (4 5 days) (2 3 days) Rarely/Never (0 1 day) Proportions of the various household food consumption groups by sub regions Using the above classification, 7.7percent of the households in Refugee settlements had a poor food consumption pattern, 54.3percent borderline and 38percent acceptable. This shows that there were more households at risk of being food insecure. Figure 3.22 shows that the South West sub region had the highest proportion of households with poor food consumption pattern while West Nile sub region had higher proportion of households with a borderline consumption pattern (63.6 percent) compared to those in the South West sub region (44.7percent). The refugee settlements in the South West sub region had a higher proportion of households with an acceptable food consumption pattern (45.3percent) when compared with those in the West Nile sub region (31percent). 67

85 Figure 3.22: Proportion of different food consumption groups by Sub region, October Relationship between food consumption and the number and type of income sources The number of income sources (p = 0.01) and type of income generating activities were related to the food consumption pattern. The highest proportion of households among those with a poor consumption pattern had one source of income (Figure 3.23). A poor food consumption pattern was associated with a reliance on unskilled labour and agricultural labour. Conversely, acceptable food consumption was more frequent among households relying on petty trade, salaries, and livestock production (Refer to Figure 3.24) Figure 3.23: Food Consumption Pattern and Income Sources, Refugee Settlements

86 % Households Others Food Aid Begging Gifts Fishing Government Allowance Rental Of Property Salaries Remittances Sale Natural Resources Brewing Handicrafts Agricultural Labour Unskilled Labour Petty Trade Commercial Activity Trade in Food Crop and Non FoodCrops SaleAnimalProducts Livestock Production NonFoodCashcrops Food Crop Production 0 Poor Borderline Acceptable Figure 3.24: Food consumption pattern and income source, October Relationship between food consumption and other household characteristics There were no clear relations between the size of the households and their food consumption patterns. The proportion of households with poor food consumption pattern was slightly higher among male- than female-headed households (9.0percent versus 6.3percent), and the proportion of households with acceptable food consumption higher (41.1percent versus 34.7percent). Similarly, the proportion of households with poor food consumption pattern was slightly higher among households whose head was illiterate (12percent) than those whose head was literate (8percent), and the proportion of households with acceptable food consumption was slightly higher in literate (57percent) than illiterate at 51percent). 69

87 % Households Acceptable Consumption Borderline Consumption Poor Consumption Female Male Figure3.25: Food consumption pattern and sex of Household Head, October 2010 Figure3.26: Food consumption pattern by income source, October Main Food Sources Table 3.26 shows that the main source of food was by purchase from the market at 51.6percent, followed by own production (21.5percent) and Food Aid (20.5percent). There were households purchasing food from South West than West Nile at 57.5percent and 44.9percent respectively. Food Aid was more pronounced in South West than West Nile sub regions at 27.6percent and 14.1percent respectively. There were more households reporting own production in West Nile than in South West at 25.2percent and 17.3percent respectively. Compared to the 2009 report: 70

88 % Households West Nile sub region: The proportion of households purchasing food increased from 51percent in 2009 to 57.5percent in 2010 while the number of households reporting Food Aid as their main source of food fell from 21percent in 2009 to 14.1percent in Furthermore, the proportion of household reporting own production as the main source of food increased from 17percent in 2009 to 25.2percent in South West sub region: The proportion of households purchasing food decreased from 58percent in 2009 to 44.9percent in 2010 while the number of households reporting Food Aid as their main source of food increased from 18percent in 2009 to 27.6percent in Furthermore, the proportion of household reporting own production as the main source of food increased from 15percent in 2009 to 17.3percent in Overall Refugee Settlements: The proportion of households purchasing food decreased from 54.5percent in 2009 to 51.7percent in 2010 while the number of households reporting Food Aid as their main source of food slightly increased from 19.5percent in 2009 to 20.5percent in Furthermore, the proportion of household reporting own production as the main source of food increased from 16percent in 2009 to 21.5percent in Coverage of Food and Non Food Assistance Receipt of Food Aid Receipt of Food Aid by Sub Region Overall, approximately 75.9percent of the households admitted to having received food aid in the period January to August The highest percentage was observed from the South West region. The details are as shown in Figure Yes No West Nile South-West General Regions Figure 3.27: Households which received food aid in the period Jan-Aug 2010 Overall, the proportion of households having received food aid gradually decreased from the month of January through to August 2010 with the coverage dropping from 94percent in the first quarter to 84.4percent in the third quarter (Figure 3.28). 71

89 Figure 3.28: Food Aid Beneficiaries Jan-August 2010 The proportion of households receiving food aid since the beginning of the year through to the third quarter was steady in West Nile Sub region compared to South West which kept of dropping from 93percent in January through to 67.7percent in the third quarter. This is could be linked to the ever increasing number of refugees influx and subsequent resettlement in South West sub region Receipt of Food Aid by Programme There were a number of programmes in which food aid was tagged to and these included Food for Education, Supplementary feeding Programme, Therapeutic Feeding Programme, General Food distribution and Food for work. Respondents indicated that food aid was mainly received from GFD at 70percent (South West-38.7 percent; West Nile 31.3 percent, followed by Food for Education, 1 percent; Food for Work, 0.7 percent; SFP, 0.5 percent and TFP at 0.4 percent. The FGDs had a concern about the back feeding policy of WFP because food items were borrowed for the months WFP was not providing food and they needed to be paid back Type of food aid commodities received Focus Group Discussions indicated that they had received the various commodities such as cereals, pulses, oil, CSB, sugar, and salt. The completeness of the food ration (in terms of content, but with no indication of the amounts) was slightly better in West Nile than South West. There were generally concerns that the type of foods being given to the Refugees in monotonous and did not in some insistences take care of the ethnics groups such as Somalis whose main food is rice and wheat. Specific to the CSB which was received and sold off, this was attributed to the inadequate knowledge on the benefits of the blend. 72

90 % Receipt of Non-Food Items Receipt of non-food items per sub region and per household group Figure 3.29 depicts some of the non-food items which were received by the households from the government or humanitarian agencies. In general, the highest proportion of households reported having received mosquito nets and soap. According to the FGDs and KIs, the soap was provided to all women of the reproductive age group on a quarterly basis. The other items are also depicted but it is important to note that the durability and longevity of the items was not captured. Some 66.4percent of the households mentioned receipt of mosquito nets, 52.4percent blankets, 16.8percent hoes, 15percent pangas and 14.8percent blankets. The proportions of non-food aid beneficiaries were higher in West Nile than South West sub regions. These differences were quite large and could be partly explained by variations of the proportions of refugees in the settlements in the 2 sub regions and/or programming decisions may also have come into play. FGDs as well as observations while in the field indicated that most of the shelter for the refugees was leaking and this partly contributed to the poor health situations in the settlements West Nile South-West General Hoes Pangas Seed or seed voucher Eating utensils Jerry can Blankets Soap Plastic sheet Sleeping mat Mosquito net Items Figure 3.29: Non food items received by households. 73

91 3.13 Household Food Insecurity and Risks to Lives and Livelihoods Prevalence of household food insecurity and short term risks to lives and livelihoods Food consumption frequency and diversity during the 7 days preceding the survey can be taken as a short-term proxy of food insecurity and risks to lives, given that the food consumption pattern affects directly health and nutrition. The degree of reliance on food aid as a source of food, and the share as well as level of food expenditures in the 7 days preceding the survey can give an indication on risks to livelihoods: reliance on food aid reflects dependence on external assistance, and the share and level of food expenditures reflect the degree of economic security, given the well-established relation between food expenditures and poverty. However, because this information is based on a very limited timeframe (a few days before the survey), solid conclusions on risks to livelihoods must take into account other factors able to capture longer-term perspectives. The methodology in section on food security profiling was applied to categorize household groups on the basis of a combination of food consumption patterns, food aid as a source of food, and food expenditures Proportion of food insecure households at short-term risk to lives and livelihoods Figure 3.30 shows that almost half of the households (45.2percent) in the refugee camps were found to be at high risk to lives and livelihoods in the short term, 34.8percent at moderate risk and 19.9percent at low risk. The West Nile Sub region presented the highest proportion of households at high short-term risk to lives and livelihoods while the South West refugee settlements had the lowest. West Nile sub region: 49.5percent severely food insecure and at high short-term risk to lives and livelihoods, 32.7percent moderately food insecure and at medium risk, and 21.9percent food secure and at low risk; South West sub region: 41.1percent severely food insecure and at high risk, 36.9percent moderately food insecure and at medium risk, and 21.9percent food secure and at low risk. 74

92 Figure 3.30: Household Food Security by Sub region Another way at looking at these results is to examine the distribution of the households among the refugee settlements with more than half of the severely food insecure/high risk households and moderately food insecure households found in the West Nile refugee settlements. The details are as shown in Figure Figure 3.31: HH Food security status by sub region Food insecurity/risks to lives and livelihoods, and household characteristics Food insecurity/ risks to lives and livelihoods, and characteristics of the head of household a) Sex and literacy level of the head of household A slightly higher proportion of households severely food insecure and at high risk to lives and livelihoods were headed by a woman compared to food secure/low risk households. Figure 3.32 shows that among the female-headed households, almost half (46.4percent) were severely food insecure and 18.9 percent were food secure (24percent), compared to 44.1percent and 20.9percent respectively of male-headed households. 75

93 Figure 3.32: Household Food Security Status by Sex Food insecurity/risks to lives and livelihoods, and food availability Food insecurity/risks to lives and livelihoods, and crop cultivation a) Cultivation practices and ownership of land Households severely food insecure and at high risk to lives and livelihoods were less likely to cultivate than moderately food insecure/medium risk and food secure/low risk households. The importance of cultivation was also shown by the fact that 50.5percent of those not cultivating were severely food insecure and only 16.6percent were food secure, compared to 42.4percent and 21.7percent respectively of those cultivating. Similar trends were observed with regard to the ownership of land for cultivation. Figure 3.33 provides the details. b) Main constraints to crop cultivation Figure 3.33: Ability to cultivate in 2010 Figure 3.34 indicates that no access to land was the main reason mentioned by most households, followed by sickness or physical inability. Food secure/low risk households were more likely to report not being agriculturalists and no access to land as the reasons for not being able to cultivate. 76

94 Figure 3.34: Main constraints to crop cultivation and food security status Food insecurity/risks to lives and livelihoods, and animal ownership a) Number and type of animals owned Figures 3.35 and 3.36 show that households severely or moderately food insecure and at high or medium risk to lives and livelihoods were much less likely to raise any animals and those that had animals were more food secure than those without. Figure 3.35: Household food security by Animal Ownership Figure 3.36: Animal Ownership by food security groups 77

95 b) Main constraints with animal raising Figure 3.37 shows that most households mentioned parasites/disease and lack of veterinary services as the main constraints to livestock production. Figure 3.37: Main constraints to crop cultivation by Food Security status Food insecurity/risks to lives and livelihoods, and food access Food insecurity/risks to lives and livelihoods, and income sources a) Number and type of income sources Figure 3.38 and 3.39 show that the number and type of income sources differed between households of different food security and risk status. Households food insecure/at high or medium risk to lives and livelihoods were more likely to rely on agricultural labour, begging or food aid, and less likely to rely on commercial activities, salaries and trade in food and non food crops. Figure 3.40: Household Food security status by number of income sources Figure 3.38: Household Food security by income sources 78

96 Figure 3.39: Household Food Security status by number of income sources Food insecurity/risks to lives and livelihoods, and food expenditures Figure 3.40 shows that all the households dedicated on average a large share of their expenditures to food purchases, whatever their food security situation. However, the proportion of households dedicating more than 80percent of their expenditures to food was much higher among households severely food insecure compared to the other households. Figure 3.40: Household Food security status by food expenditure Food insecurity/risks to lives and livelihoods, and ownership of assets Figure 3.41 shows that households that were food insecure and at high or medium risk to lives and livelihoods owned less assets than households that were food secure and at low risk. Figure 3.41: Household Food security status by asset ownership 79

97 Food insecurity/risks to lives and livelihoods, and coping strategies in the event of food shortages Figure 3.42 shows that coping strategies employed by households when faced with food shortages during the month preceding the survey did not differ much between households according to their food security and risk to lives and livelihoods. Households moderately food insecure and at medium risk were slightly more likely to borrow food, purchase food on credit or send household members to eat elsewhere and less likely to feed working members of the household at the expense of the nonworking members but the reasons for this pattern are unclear. Figure 3.42: Household Food security by coping strategy Households that were food insecure and at high risk were more likely to eat less expensive food, limit portion size and reduce the number of meals eaten Regression analysis: causal factors of food insecurity A regression analysis was conducted to identify the main causes of household food insecurity and distinguish independent factors and factors which are correlated between them (colinearity). The variables included in the regressions took into account the cross-tabulations carried out previously. The independent factors influencing food security were: The status of the household: female-headed households were more likely to be food insecure than the male headed households and households headed by illiterate heads were more likely to be food insecure than those with literate heads; Cultivation practices: Households not cultivating were more likely to be food insecure than those cultivating and those owning land were more food secure than those without; 80