ODA BULTUM WOREDA, ETHIOPIA SEPTEMBER, 2013 BEATRIZ PÉREZ BERNABÉ AND LINDSEY PEXTON

Transcription

1 ODA BULTUM WOREDA, ETHIOPIA SEPTEMBER, 2013 BEATRIZ PÉREZ BERNABÉ AND LINDSEY PEXTON

2 ACKNOWLEDGEMENTS GOAL and the Coverage Monitoring Network (CMN) would like to express our great appreciation to all those who made possible the realization of this coverage assessment of the nutrition program in the Woreda (district) of Oda Bultum. In the first place, to ECHO and USAID for funding the CMN project which has contributed to creating capacity in the country and directly supporting this investigation. Thanks to all the participants of the SQUEAC training, including both MoH and GOAL staff for their effort, high level of engagement and for the quality of their work. Grateful thanks are extended to all at the GOAL office in Oda Bultum, for handling all the necessary preparations for the start-up and during the investigation in particular to Roza Dagne, Senior CMAM Programme Officer, for his organizational support and involvement throughout the study. The inputs in planning and carrying out the assessment given by Hailu Sitotaw, Senior Survey and Assessment Coordinator and Zeine Muzeiyn, Nutrition Programme Coordinator, have been greatly appreciated. Finally, the team address their most sincere gratefulness to the staff of the health facilities visited as well as the families and various community members for their hospitality, time and cooperation. Very special thanks to the mothers and children who took part in the investigation. ACRONYMS BBQ CDA CMAM ECHO HEW HC LQAS MAM MoH MUAC RUSF RUTF SAM SC SQUEAC USAID WHZ Barriers, Boosters and Questions Community Development Army Community-based Management of Acute Malnutrition European Office for Coordination of Humanitarian Affaires Health Extension Worker Health Centre Lot Quality Assurance Sampling Moderate Acute Malnutrition Ministry of Health Mid-Upper Arm Circumference Ready-to-Use Supplementary Food Ready-to-Use Therapeutic Food Severe Acute Malnutrition Stabilisation Centre Semi-Quantitative Evaluation of Access and Coverage United States Agency for International Development Weigh-for-Height Z-score 2

3 EXECUTIVE SUMMARY Oda Bultum woreda is one of the nineteen woredas in West Hararghe Zone, Oromiya Regional State. It is located in the eastern part of the country, 362km from Addis Ababa and 37km from Chiro [Asebe Tefere]. The woreda consists of 37 rural kebeles with an estimated population of 181,732, of whom 29,804 are children under five years 1. The population is predominantly ethnically Oromo and Muslim by religion, with an average family size of 5.0. Bedessa town serves as the main administrative center of Oda Bultum woreda. The CMAM program in Oda Bultum began in August 2008 under ownership of the Ministry of Health (MoH), with technical and logistical support provided by GOAL. There are currently 37 functioning health posts running Outpatient Therapeutic Programmes (OTPs) and 5 health centres (operating both OTPs and Stabilisation Centres). A coverage assessment based on the SQUEAC (Semi-Quantitative Evaluation of Access and Coverage) methodology took place in September 2013 to assess GOAL s CMAM project and to build the capacity of MoH and GOAL s staff in undertaking coverage assessments. Main barriers identified and recommendations to improve coverage are described in the table below. Barriers Recommendations Insufficient specific information about the program Increase efforts to sensitize the communities about the program, clarifying specific issues such as admission criteria (particularly marasmus), OTP days and that the treatment is free. Lack of screening in the community Strengthen existing outreach activities by increasing the number of CDA and revising the activity strategies Wrong admission criteria Lack of training of program staff/volunteers Weak referral system Promote adequate application of admission and discharge criteria Increase capacities of CDA and HEW - Train CDA screening with MUAC to improve case detection - Refreshment training for HEW on admission and discharge criteria Strengthen the referral system by developing a mechanism to monitor cases referred at community level, from OTP to SC (and vice versa) as well as for discharged children 1 CSA

4 CONTENTS ACKNOWLEDGEMENTS... 2 ACRONYMS... 2 EXECUTIVE SUMMARY... 3 CONTENTS INTRODUCTION CONTEXT CMAM PROGRAM IN ODA BULTUM WOREDA OBJECTIVES GENERAL OBJECTIVE SPECIFIC OBJECTIVES METHODOLOGY GENERAL APPROACH STAGES ORGANISATION OF THE STUDY RESULTS STAGE 1: IDENTIFICATION OF AREAS OF LOW AND HIGH COVERAGE AND BARRIERS TO ACCESS STAGE 2: VERIFICATION OF HIGH AND LOW COVERAGE AREAS HYPOTHESIS SMALL AREA SURVEY DISCUSSION RECOMMENDATIONS ANNEX 1: EVALUATION TEAM ANNEX 2: CHRONOGRAME ANNEX 3: DATA COLLECTION FORM ANNEX 4: QUESTIONNAIRE FOR NON-COVERED CASES ANNEX 5: BARRIERS SOURCES & METHODS ANNEX 6: BOOSTERS SOURCES & METHODS

5 1. INTRODUCTION 1.1 CONTEXT Oda Bultum woreda is one of nineteen woredas in West Hararghe Zone, Oromiya Regional State. It is located in the eastern part of the country, 362km from Addis Ababa and 37km from Chiro [Asebe Tefere]. The woreda consists of 37 rural kebeles with an estimated population of 181,732, of whom 29,804 are children under five years 2. The population is predominantly ethnically Oromo and Muslim by religion, with an average family size of 5.0. Bedessa town serves as the main administrative center of Oda Bultum woreda. Livelihoods in the woreda mainly centre on rain fed agriculture, with mixed farming constituting 90% and agro-pastoralism estimated at 10%. Maize, sorghum, teff, wheat and barley are the major food crops while chat, coffee and pepper are the most important cash crops. The woreda has faced consecutive crop failure and/or below normal production over the past years, mainly due to the failure and/or poor performance of both kremt and belg rains. As a result of chronic food insecurity, Oda Bultum has been included in the Productive Safety Net Program (PSNP) since Currently a total of 20,501 beneficiaries across the woreda are targeted under either the public work scheme (16,910) or direct support (3,591). Cash payments are given for a period of six months, amounting to Ethiopian birr/person/month or birr/person/day. Those targeted in the public works programme are expected to work five days per month. Oda Bultum woreda is characterized as chronically food insecure, but according to the last nutrition survey conducted in the woreda, although the nutritional status of under 5 population has not improved significantly as compared with the year before, the malnutrition rate has significantly decreased compared with the base line survey (2009 and 2011) meaning that the current malnutrition rate can be considered normal. Based on national cut-offs, the prevalence of GAM (MUAC <120 mm and/or oedema) is 1.5% and SAM (MUAC <110 mm and/or oedema) 0.2%. The rate of GAM based on international standards (MUAC < 125 mm) goes up to 7.3% 3. 2 CSA Report on nutrition and retrospective mortality survey conducted in Oda Bultum Woreda, West Hararghe, Zone of Oromiya Region, conducted in March 2013, USAID and GOAL 5

6 1.2 CMAM PROGRAM IN ODA BULTUM WOREDA GOAL has a long history of emergency response in Ethiopia, and since 2005 has been USAID/OFDA s emergency nutrition response partner. GOAL Ethiopia s operational mandate in nutrition is to support the provision of a package of nutrition services which is targeted at the poorest of the poor and most vulnerable and also to strengthen the capacity of existing MoH structures in order to ensure sustainable services for local communities. The CMAM approach aims to build capacities within the local health services and the community to prevent and treat malnutrition. GOAL works in partnership with the local health bureaus to support them in developing their capacity to respond to nutrition emergencies. This approach aims to enable various program goals to be sustained and promote an acceptable exit when the situation has stabilized given contextual constrains. The capacity building of MoH staff and the training of health extension workers and Community Development Army emphasizes Nutrition Education. The MoH s policy is scaling-up delivery of essential health services to local communities. Their focus is on training of the health care worker and this initiative is supported by UNICEF. The CMAM program in Oda Bultum began in August 2008 under ownership of the Ministry of Health (MoH) with technical and logistical support provided by GOAL. There are currently 37 functioning health posts running Outpatient Therapeutic Programmes (OTPs) and 5 health centres (operating both OTPs and Stabilisation Centres). Despite the programme being operational for 5 years, this is the first time that a coverage assessment has been undertaken. The basis for the assessment was therefore the desire to gain a better understanding of programme access and coverage in order to drive further improvements. Developing capacity to undertake coverage assessments, particularly among GOAL s permanent Survey and Assessment Team was also paramount. Woreda and Zonal representatives were included in the exercise to create awareness of coverage monitoring, as well as to build capacity in coverage assessment tools and techniques. As such, 36 team members were drawn from GOAL s Survey and Assessment Team, GOAL s district level CMAM staff and woreda/zonal representatives. The process was coordinated by a Regional Adviser from the Coverage Monitoring Network (CMN), together with a Nutrition Adviser from GOAL. 6

7 2. OBJECTIVES 2.1 GENERAL OBJECTIVE To assess the coverage of GOAL s CMAM program and to understand the barriers to accessing health care in the areas of intervention within the Woreda of Oda Bultum for children aged between 6 to 59 months, based on the Semi-Quantitative Evaluation of Access and Coverage (SQUEAC) methodology. 2.2 SPECIFIC OBJECTIVES To develop capacity of GOAL (GOAL Ethiopia Survey and Assessment team as well as those directly involved in the program) and Ministry of Health staff to undertake CMAM program coverage assessments using SQUEAC methodology. Assess the global estimation of coverage in the target areas of the program. Identify high and low coverage areas within the intervention area. Identify barriers to access to treatment of severe acute malnutrition based on information collected from mothers/caretakers of children with severe acute malnutrition identified during the investigation and who are not enrolled in the program. Make recommendations based on the results of the evaluation to improve the access to treatment of severe acute malnutrition and increase the level of coverage in the program intervention area. Write a report presenting the results of the evaluation and taking into account the differences identified. 7

8 3. METHODOLOGY 3.1 GENERAL APPROACH The coverage assessment tool, Semi-Quantitative Evaluation of Access and Coverage (SQUEAC) 4, was developed by Valid International, FANTA, Brixton Health, Concern Worldwide, ACF and World Vision in order to provide an efficient and accurate method of identifying barriers to service access and to estimate the coverage of nutrition programs. This is a relatively time efficient method and gathers large amounts of relevant information; promotes the collection, use and analysis of data; and provides information on program activities and possible reforms. The need for human, financial and logistical resources is relatively small. Furthermore, it is easily reproducible and ensures program monitoring at low cost. SQUEAC is an interactive, informal and intelligent investigation that collects a large amount of data from different sources (i.e. using routine data as well as additional data collected in the field), using a wide variety of methods and providing the means to organise the data. It is a semi-quantitative assessment as it combines both quantitative and qualitative data. The analysis of these data is guided by the two fundamental principles of exhaustiveness (of information up to the point of saturation) and triangulation (information is collected from different sources using alternative methods, crossing checking data until findings become redundant before being validated). By focusing on the collection and intelligent analysis of data during the field phase, the investigation sheds light on the operation of the service whilst simultaneously providing an educated guess on coverage which allows for a smaller sample size to be used in the final stage. 3.2 STAGES SQUEAC allows for the regular monitoring of programs at low cost, helps identify areas of high or low coverage and provides explanations for such situations. All of this information allows the planning for specific and concrete actions in order to improve the coverage of programs. The SQUEAC methodology consists of three main stages: 4 Myatt, Mark et al Semi-Quantitative Evaluation of Access and Coverage (SQUEAC)/Simplified Lot Quality Assurance Sampling Evaluation of Access and Coverage (SLEAC) Technical Reference. Washington, DC: FHI 360/FANTA. 8

9 3.2.1 STAGE 1: Identification of high and low coverage areas and barriers to access This stage is based on the analysis of both quantitative data and qualitative information (already available and collected during the investigation) in order to understand the various factors influencing coverage, some of which have a positive effect and some a negative effect on coverage. The SQUEAC approach helps to identify and understand these relevant factors and their effects. The evaluation of these factors helps to develop a trend in the coverage rates prior to conducting a field investigation in well-defined areas. Analysis of quantitative program data: routine data (monthly reports) and records of individual monitoring (register book and individual cards). The analysis of routine data is used to assess the overall quality of services, to identify trends in admissions and performance, and to determine if the program meets needs. This stage also helps to identify potential problems related to the identification and admission of beneficiaries as well as problems related to their treatment. Information such as MUAC measurements at admission and numbers of defaulters can be used to assess early detection, recruitment and effective communication channels. It also provides information on differences in raw performance between different health facilities. Collection and analysis of qualitative data through meetings in the community and health facilities with those involved directly or indirectly in the program 5. This phase of the investigation is twofold: it serves to better inform and explain the results of the analysis of routine data and it also helps to understand the knowledge, opinions and experiences of all people concerned as well as to identify potential barriers to access. Interview guides were used to orientate the process of obtaining information on coverage. These interview guides were developed based on guides already used in other SQUEAC investigations but also adapted to the context and modified/upgraded by the investigation team. The following methods to gather information were used: - Focus Group Discussions - Semi-structured interviews These focus group discussions and semi-structured interviews were conducted with the following sources of information: - Health Extension Workers (HEW) and Stabilisation Centre (SC) nurses - Community Development Army (CDA) - Community leaders 5 We took advantage of these meetings in the community and health facilities to identify the local terminology used to describe acute malnutrition (Oromifa) and the key informants in the community. This preliminary research is essential to facilitate the active and adaptive case-finding methodology that is used in stages 2 and 3. 9

10 - Religious leaders - Community men - Community women - Mothers/caretakers of SAM children within the program 18 villages spread across the different kebeles in Oda Bultum were visited for the collection of qualitative information. Inaccessibility due to the rainy season unfortunately limited the selection of villages and this may be a biasing factor as these villages may also tend to have lower coverage due to this seasonal inaccessibility. The different people encountered and the various methods used allowed the investigation team to collect information about the barriers and boosters to coverage of the CMAM program. The data gathered was recorded on a daily basis using a tool called BBQ (Barriers, Boosters and Questions). This tool not only allows for the organizing of information on a day to day basis, to continue with the research of qualitative information in an interactive and directed manner, but also ensures the triangulation of information. To guarantee the exhaustiveness of the process, the research of information continued until saturation - until the same findings were obtained from different sources, using different methods. Altogether, the findings from the quantitative analysis and the conclusions from the investigation team s discussions were included in the BBQ with qualitative data collected in the field to triangulate the set of all knowledge around barriers and boosters to coverage in Oda Bultum Woreda. Identification of potentially high and low coverage areas and formulation of a hypothesis on coverage based on the evaluation of positive and negative factors. Depending on the barriers and boosters found, the hypotheses on high or low coverage areas are developed: the hypothesis about heterogeneity of coverage are based on the identification of areas of good and less good coverage. Then, small-area surveys are conducted to confirm or refute these hypotheses STAGE 2: High and low coverage areas hypothesis testing through small-area surveys The objective of the second stage of the investigation is to confirm or reject, through smallarea surveys, the assumptions on areas of low or high coverage as well as the barriers to access as identified in the previous stages of analysis. The small geographical survey method was used to test the assumption of homogeneity/heterogeneity of coverage. In this case, 12 villages were selected (6 villages in the area with potentially high or satisfactory coverage; and 6 villages in the area with low or unsatisfactory coverage) to test the hypothesis of homogeneity/heterogeneity of coverage. The villages were selected according to the criteria identified to be the most relevant, according to the information triangulated up to that point in the survey. The sample of small-area surveys was not calculated in advance; but rather was based on the number of SAM cases found. 10

11 SAM Cases were searched for using the active and adaptive case-finding method (i.e going from house to house based on key informants information to find all severely malnourished children in the village). The case definition used was: "all children aged 6-59 months with the following characteristics: MUAC <110 mm and/or presence of bilateral oedema, or who were currently in the CMAM program for the treatment of SAM". Analysis of the results was done using LQAS (Lot Quality Assurance Sampling) in order to obtain a classification of coverage compared to the threshold value set at 30%. The decision rule was calculated using the following formula: d n d 100 n: number of cases found p: standard coverage defined for the area The number of cases found and the number of cases covered was examined (see annex 3 for form to gather the data in the field) based on the following criteria: - If the number of cases covered was higher than the threshold value (d), then coverage was classified as satisfactory (coverage meets or exceeded the standard). - If the number of cases covered was lower than the threshold value (d), then coverage was classified as unsatisfactory (coverage did not meet, neither exceeded the standard). Throughout the small-area survey, a questionnaire (annex 4) was distributed to mothers or other caretakers of all non-covered SAM cases detected in order to further understand the reasons that these children had not received treatment as this allows for the identification of barriers to access. All non-covered children found during the study were referred to the appropriate health services for treatment. The information obtained through the questionnaires of the non-covered cases in the small-area survey was added to the BBQ in order to triangulate information regarding barriers to coverage in Oda Bultum Woreda. The software XMind is a powerful tool capable of displaying findings in a visual and orderly manner. It was used in Oda Bultum to develop two different Mind Maps summarizing the barriers and boosters identified during the first two stages of the investigation, as well as the different sources of information and methods utilized. Also, conceptual schemas for different barriers and boosters identified were developed by the team in order to better understand the cause and effect relationships between the various factors influencing coverage. 11

12 3.2.3 STAGE 3: Estimation of global coverage Stage 3 was not conducted in this investigation due to the very low prevalence of SAM 6 in the area assessed and thus, an estimate of the overall coverage in Oda Bultum Woreda could not be established. The prevalence of severe acute malnutrition is indirectly related to the number of villages that need to be assessed during a large area survey. This survey is necessary in order to build a likelihood curve to determine the overall level of programme coverage. In this case, the amount of villages required (n) to reach the minimum sample size of children (N) was not manageable and thus stage 3 was infeasible. N n population average between 6-59 months SAMprevalence village * * population However, for training purposes, a case study based on the actual findings of the present investigation was conducted with the team in order for them to understand the different steps to undertake during a coverage assessment and to be confident with all of the methodology involved within a SQUEAC assessment. 3.3 ORGANISATION OF THE STUDY Technical support from CMN project GOAL and MoH received technical support from the Coverage Monitoring Network (CMN) project. The CMN project is a joint initiative involving several organizations: ACF, Save the Children, International Medical Corps, Concern Worldwide, Helen Keller International and Valid International. The project aims to provide technical support and tools to CMAM programs in order to help them assess their impact and to share and capitalize on lessons learnt with regards to factors influencing their performance. As part of this assessment, the support from the CMN project involved different phases. In a first phase, technical support was provided remotely through exchanges between the team of experts from CMN, José Luis Álvarez Morán (CMN Coordinator) and Beatriz Pérez Bernabé (RECO) and GOAL staff, Hatty Barthorp (Global Nutrition Advisor), Zeine Muzeiyn (Nutrition programme coordinator) and Hailu Sitotaw (Senior Survey and Assessment Coordinator) and Lindsey Pexton 6 Prevalence of SAM (MUAC < 110mm): 0.2% (95% IC ) Report on nutrition and retrospective mortality survey conducted in Oda Bultum Woreda, West Hararghe Zone of Oromiya Region, March 2013; 0.5% according to program admissions and screening data during the period of the year when the assessment was carried out ( end of hunger gap) 12

13 (Nutrition Advisor), for the planning and preparation of the evaluation. For technical support in the field, Beatriz Pérez Bernabé was deployed to Oda Bultum to train the nutrition team in SQUEAC methodology and to carry out the coverage assessment in the area of intervention Training and investigation An investigation team composed of the national survey and assessment team, 5 members of GOAL Ethiopia CMAM program and 3 representatives of the MoH (Oda Bultum Woreda and West Hararghe Zone) were trained in SQUEAC methodology in order to be able to undertake future coverage assessments. The coordination of the present evaluation was jointly done by the CMN expert and an international Nutrition Advisor employed by GOAL. The coverage assessment took place from August 31th to September 10th 2013 (chronogram annex 2). Two days of introductory theoretical sessions concentrated on the importance of assessing coverage and the basics of SQUEAC methodology, after which the investigation began in earnest. The training process was then run concurrently with the investigation - in-classroom sessions for each key stage of the study was alternated with guided practical implementation in the field, all framed with iteractive briefing and debriefing sessions. One additional day was added to the initial planning for the collection of quantitative program data in health facilities needed for stage 1. This would ideally have been conducted in advance. Photo 1. Groupwork. (Oda Bultum Woreda, Ethiopia, September 2013) 13

14 4. RESULTS According to the methodology explained above we present here the main results emerging from our investigation: 4.1 STAGE 1: IDENTIFICATION OF AREAS OF LOW AND HIGH COVERAGE AND BARRIERS TO ACCESS The objective of this stage was to identify areas of high and low coverage and to have an initial understanding regarding the reasons for poor access to treatment, using the program s existing quantitative data, together with qualitative information collected from the various stakeholders. Although first admissions were registered in August 2008, the analysis of the quantitative information was carried out based on data corresponding to the last year of program activities (July 2012 July 2013). Routine program data was easily available and mainly extracted from monthly reports. Individual monitoring data was collected by the team at the health facilities during one working day at the beginning of the investigation. Inaccessibility due to the rainy season and the absence of HEWs at certain health facilities as a result of a concurrent training limited the collection of the full set of data. Information was collected from registration books and outpatient record cards from 23 out of the total of 37 in the woreda Analysis of quantitative programme data routine data: monthly statistical reports A. Admissions: trends over time and capacity to meet needs A seasonal calendar for the various seasonal events (child morbidity, climatic and agricultural activities) was developed by the team and compared to the curve of admissions during the period July July 2013 (227 children) to assess the capacity of the program to meet needs (Figure 1). According to the results of previous nutrition surveys conducted in the area, malnutrition rates tend to increase between May and September (hunger gap period). The expected rise in cases was clearly reflected in admissions trends: admissions experienced a gradual increase starting in April, and peaking in July. When compared to the Malaria calendar, the increase also matched the higher prevalence rates during the rainy season (from July to September). This suggests that the CMAM program is responding to the seasonal increase in SAM cases each year. 14

15 Number of cases Admissions over time (OTP & SC) SEASONAL CALENDAR Jul-12 Aug-12 Sep-12 Oct-12 Nov-12 Dec-12 Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13 Jul-13 SEASONS CHILD MORBIDITY AGRICULTURAL ACTIVITIES Rains Malaria Land preparation, sowing & weeding OTHER ACTIVITIES HUNGER GAP Harvest Pety trade LOCAL EVENTS Ramadan Figure 1. SAM admissions in OTP and SC and seasonal calendar (Oda Bultum Woreda, Ethiopia, September 2013) B. Admissions in OTP/SC and admission criteria From the total of 227 admissions, 88% correspond to SAM children admitted in OTP and 12% to SAM children presenting complications and being treated at the SC. The percentage of cases treated at SC level is unfavourably high. In programs with strong community outreach and early case finding we would expect to see a lower rate of around 5%. The analysis of the admission criteria for all the severely acute malnourished children in the program shows that an extremely high proportion, 60%, have been admitted under oedema 15

16 criteria (Kwashiorkor cases) and 40% corresponds to Marasmus cases identified only by MUAC (<110 mm) Although WHZ is considered as an independent criterion of admission according to the national protocol (only at HC level, not in HP) no child was found to be admitted under WHZ criterion. C. Performance indicators The performance indicators for OTP were calculated both based on the data from the monthly statistical reports and from the data coming from the outpatient record cards. Although results were the same and very satisfactory compared to SPHERE standard values, it remains uncertain how often hidden deaths may have occurred as defaulter tracing records were not available. GOAL Programme SPHERE Standards Cured rate 98.3% >75% Defaulter 0.4% <15% Death 0.0% <10% The proportion of children who did not respond to the treatment was only 0.1% and 1.1% were transferred to SC. E. Source of referral Roughly equal numbers of cases were admitted into the program either as self-referrals (32%) or as referrals by CDAs (33%). Unfortunately, 22% of admissions did not have the referral mechanism registered on their outpatient record card. Referrals by HEWs were noted for 12% cases, with only 1% from the mass campaigns. 0% 1% Campaign 32% 33% CDA Neighbour Health extension worker Information not available Self-refered 22% 12% 0% Village leader Figure 2. Source of referral of SAM admissions (Oda Bultum Woreda, Ethiopia, September 2013) 16

17 4.1.2 Analysis of quantitative program data individual monitoring records: register books and outpatient record cards F. MUAC at admission In Ethiopia, the cut-off point for admission of SAM children is MUAC <110 mm. The analysis of MUAC at admission of those children admitted based on MUAC <110 mm (40% of the total) shows that the majority of cases were admitted with a MUAC quite far from the admission criteria, with a low median value of MUAC of 105 mm (figure 6). This distribution suggests poor performance in the ability to identify marasmus cases in a timely manner. The number of critical cases (MUAC <= 90mm), 16, is of particular concern in this regard. Most cases are reached and admitted to the program late in the process of the disease which has a negative impact on the chance of recovery and length of stay. Figure 3. Distribution of MUAC at admission for SAM cases with MUAC <110 mm (Oda Bultum Woreda, Ethiopia, September 2013) F. Distance from village of origin to OTP The distance (time to travel) between the village of origin of SAM children admitted in OTP was analysed for 77.92% of total admissions in OTP. The villages were grouped into six categories according to the time to travel (in minutes) to the HP/HCI. Figure 4 shows, from left to right the closest to farthest: from 0 to 15 minutes; 16-30; 31-45; 46-60; 61-90, and greater than 2 hours. 17

18 Number of MAS cases The analysis shows that indeed distance may influence the number of admissions: the number of admissions dramatically decreases when the time to travel to the OTP is over 1 hour. However, this could be due to a low number of villages located at such a distance as the geographical spread of health facilities in Oda Bultum is generally good. Distance as a factor therefore required further investigation > 2 hours Time to travel (mins.) from village to OTP Figure 4. Distribution of admissions according to distance (time to travel) to OTP (Oda Bultum Woreda, Ethiopia, September 2013) Analysis of qualitative data The qualitative data was collected in 18 villages spread throughout the intervention area, except for those non-accessible due to the rainy season. The methods and sources of information used were those described in the methodology section (chapter 3) and findings were triangulated using the BBQ on a daily basis. Table 1 shows the list of the main barriers to coverage identified through the completion of qualitative work in the field and the subsequent triangulation and analysis of information. BARRIERS Lack of specific information about the program Although there is a good awareness about the existence of the program, there is a lot of specific information about it that often prevents the community from seeking treatment such as not knowing when the OTP days are, that the service is free or that Marasmus can be also treated within the program. 18

19 Lack of screening in the community Wrong admission criteria Distance Service not available Lack of staff at HP/HC level Lack of training of CDA /program staff There are insufficient CDAs and in some cases, where they are identified, they are not active. Crucially, CDAs are not allowed to use MUAC tapes to screen children and therefore identification of SAM cases is done using visual inspection only. Proper screening in the community is usually reduced to mass campaigns and mainly done by HEWs. Due to their various other responsibilities and the distance to some communities, screening is not conducted routinely outside of the mass campaigns. From the various interviews in the community it was confirmed that most of the time CDAs only identify and refer to the OTP the oedematous and the most severely wasted children (those that can be easily identified without measuring). This is linked to the limitations CDAs have in the execution of their responsibilities as noted above. Distance to the health facility was reported by most of the members of the community as well as the HEWs as an obstacle to access and finalizing treatment. Either due to HEW meetings or mass campaigns taking place in the community, at times the OTP remains nonfunctional/closed even during OTP days. It was also found that some HPs are temporarily closed because of the poor condition of the building and others are still under construction. An insufficient number of HEWs and nurses in some health facilities (sometimes only 1 HEW per HP) leads to work overload and non-appropriation of the CMAM program; minimizing time spent with caregivers and the level of information provided during weekly visits. There is a complete lack of formal CMAM training of the CDAs since the creation of this cadre in January CDAs that are active for CMAM perform this role based on their pre-existing knowledge of malnutrition and of the program. HEWs also mentioned the need for refreshment training on CMAM for themselves. Weak referral system CDAs do not use MUAC tapes for screening children nor referral slips to follow those children referred to the OTP/SC. Together with the lack of feedback from HEWs to CDAs, this contributes to a weak level of case monitoring. Table 1. Barriers to coverage emerged from the qualitative research (Oda Bultum Woreda, Ethiopia, September 2013) 19

20 Other factors associated to the non-attendance or defaulting were mentioned during the qualitative research but were not so significant/relevant in the context of Oda Bultum. Those directly related to the community were: inaccessibility to the health facility during the rainy season; the caregiver being busy due to household activities and/or other family responsibilities; seasonal population movement due to cattle herding or agricultural requirements and lack of money. Sharing RUTF was reported only once by the community and was not considered a major problem by service providers. Regarding barriers related to service delivery, rejection of healthy children, long waiting times and poor conditions in health facilities were reported. Lack of absentee and defaulter tracing (although very few) was recognized as a negative factor to coverage, but more significantly, poor follow up of discharged cases leads to many cases of relapse from MAM to SAM. This was also linked to the interruption of TSFP services and stock breakouts of CSB. Finally, insufficient supervision by the woreda was reported from the staff working in some health facilities. On the positive side, the program seems to be well known and the entire community, including the mothers of the children enrolled in the programme, have a very positive perception/opinion of it as well as the efficacy of the treatment itself. In fact, mothers of children discharged often act as informal CDAs, referring possible cases by encouraging mothers to go to the HC/HP which reflects a very positive peer-to-peer influence. Family support and the involvement of key community figures such as village and religious leaders are also constructive mechanisms that contribute to coverage: leaders often use community gatherings to provide sensitization messages regarding both malnutrition and the program itself, when possible they refer cases and generally maintain regular communication with HEWs. Awareness regarding the causes and signs of malnutrition in the community was confirmed through the data collection and the OTP was repeatedly stated as the first option considered for treatment as the child s health was viewed as a priority. But health-seeking behaviour may not be considered optimal many physical signs of malnutrition were seen in some children and amulets (such as bracelets with leaves inside) from traditional medicine were found during household visits. The interface at the health facilities was highly valued by the beneficiaries, with HEWs often praised for being friendly and welcoming. Routine screening at the HP is done for all children that arrive for consultation and campaigns are used to sensitize communities regarding malnutrition. The program profits from these situations to improve coverage. Finally, coordination and support from GOAL was much appreciated by HEWs High and low coverage zones Given the different positive and negative aspects influencing access to treatment and the burden carried by the HEWs in this program, distance from home villages to the OTPs appeared to be a factor influencing coverage. 20

21 It was thus decided to test the following hypothesis regarding the potential areas of high and low coverage: - Coverage is probably satisfactory in areas where distance from the village to the OTP is low (less than 1 hour). - Coverage is probably unsatisfactory in areas where distance from the village to the OTP is high (more than 1 hour). 4.2 STAGE 2: VERIFICATION OF HIGH AND LOW COVERAGE AREAS HYPOTHESIS SMALL AREA SURVEY To test the hypotheses of high and low coverage areas, 12 villages (six in the area of potentially satisfactory coverage and six in the area of potentially unsatisfactory coverage) from different kebeles were selected on the basis of the criteria identified: the distance to the OTP. Satisfactory coverage area Ahmed Mohamed Burka (Sefera) Kurikura Weba (Harereti) Elelie (Dida Dalo) Village (Kebele) Cheruye (Besoso) Chira (Dida Dalo) Dida Oda (Gebida) Distance + Unsatisfactory coverage area Dadhi (Guba Gutu) Weketa (Oda Roba) Husen (Guba Gutu) Keradi (Bate) Guda Burka (Ido Beriso) Berkele (Suri) - Table 2. Villages in potentially satisfactory and unsatisfactory coverage areas according to the selected criteria (Oda Bultum Woreda, Ethiopia, September 2013) Results from the active and adaptive case-finding are presented in table 3 and the analysis of the results in table 4: Satisfactory coverage area Unsatisfactory coverage area Total number of SAM cases found 7 Covered SAM cases 2 Non-covered SAM cases 5 Recovering cases 4 Total number of SAM cases found 7 Covered SAM cases 2 Non-covered SAM cases 5 Recovering cases 2 21

22 Table 3. Results from active and adaptive case-finding small-area survey (Oda Bultum Woreda, Ethiopia, September 2013) Satisfactory coverage area Unsatisfactory coverage area Calculation of decision rule/results Target coverage 30% n 7 Decision rule (d) = n * (30/100) d = 7 * 0.30 d = 2.1 d = 2 Covered SAM cases 2 Target coverage 30% Deductions Number of covered cases (2) = decision rule (2) Point coverage 30% Satisfactory coverage hypothesis NON CONFIRMED Number of covered cases (2) = decision rule (2) Point coverage 30% Unsatisfactory coverage d = 2 hypothesis NON CONFIRMED Covered SAM cases 2 Table 4. Analysis of survey results of the small-area survey Classification of coverage (Oda Bultum Woreda, Ethiopia, September 2013) n 7 Decision rule (d) = n * (30/100) d = 7 * 0.30 d = 2.1 The hypothesis of heterogeneity was therefore not confirmed suggesting that distance does not influence in the spatial distribution of coverage. In fact, contrary to the initial hypothesis, none of the mothers of those SAM cases found not to be covered mentioned distance as the reason for their child not being in the program. The reasons that emerged from the analysis actually related mostly to problems with a specific health facility and to the performance/communication misunderstandings with the HEW assigned. Such communication misunderstandings included: - Rejection at the OTP site: three cases - two of them because the child had no oedema - The child was previously identified as healthy by a HEW: two cases - The mother waited for the HEW to come to the village for her child to be admitted: one case - The mother thought the child could not be in the programme after being discharged as cured: one case - OTP closed: one case - Negligence of the HEW: one case Only one mother was not aware that her child was malnourished. 22

23 Negligence The mother thinks the child cannot be in the programme after being discharged as cured Waiting for the HEW to come to the village to be admitted OTP closed Lack of Awareness about Malnutrition The child was previously told to be healthy by HEW Rejection Figure 5. Reasons of the non-covered cases found in the small-area survey (Oda Bultum Woreda, Ethiopia, September 2013) Results from this stage suggest that coverage is quite homogeneous throughout the area of intervention and that motives for defaulting and/or non-attendance are frequently linked to service delivery. The information obtained throughout this stage was added to the previous findings and conclusions from quantitative and qualitative data. The table 6 below show the barriers and boosters identified along the first two stages of the investigation with those considered as the main negative factors to coverage in bold. The MindMaps of annexes 5 and 6 show respectively the sources of information and methods used to identify each of them. 23

24 Positive factors VALUE Negative factors Community Awareness regarding malnutrition Awareness regarding the existence of the program Good health-seeking behaviour Positive perception of the program Involvement of key community figures Family support Positive peer-to-peer influence Insufficient specific information about the program Lack of screening in the community Distance Seasonal barriers (rains) Caregiver busy Population movement Lack of money Sharing PPN Service delivery Lack of defaulters Interface at health facility Sensitization at HP Screening/sensitization done out of the program Wrong admission criteria Weak referral system Lack of training of program staff/volunteers Rejection Lack of staff at HP/HC level Service not available Long waiting times Absents and defaulters tracing Conditions in health facility Relapse Coordination/collaboration Coordination and support from GOAL Table 6. Barriers and boosters to coverage in Oda Bultum Woreda (Oda Bultum Woreda, Ethiopia, September 2013) 24

25 Figure 6 below shows one of the conceptual schemas developed by the team: how the involvement of key community figures relates to other positive factors of the program and ultimately has an impact on coverage. Figure 6. Concept map (boosters) (Oda Bultum Woreda, Ethiopia, September 2013) The involvement of key community figures in efforts to sensitize on malnutrition is highly beneficial in terms of raising levels of awareness and appreciation of the significance of malnutrition. Their engagement also has a bearing on community acceptance of the program and these two factors help bolster treatment seeking behaviour. When community acceptance of the programme is translated into support from family members completion of the treatment (lack of defaulters) can be more easily achieved. 25

26 5. DISCUSSION Although, due to the low prevalence of SAM in the area of intervention, an overall estimation of coverage was impossible, findings from the SQUEAC assessment suggest a low coverage for GOAL s program in Oda Bultum Woreda. The coverage assessment did however reveal a number of addressable barriers, as well as important boosters to program access coverage. The program is well known and much appreciated by the community. However, they often lack specific information on how the program runs or there are sometimes misunderstandings surrounding the messages disseminated by HEWs, which prevent caregivers from seeking treatment in an active manner. This leaves many malnourished children uncovered and untreated by the programme. The current status of CDAs and their limited involvement in the CMAM program has been determined to be directly linked to unsatisfactory program coverage: not only the limited quantity of CDAs in some areas but most importantly the low level of activities they perform. Their inability to screen children using MUAC tapes means that the less severe marasmic children are not identified and referred for treatment. On the other hand, the referral of children is done in a completely informal way, which makes follow up and monitoring of these children difficult, if not impossible. CDAs require urgent training and a revision of their tasks needs to be conducted in order to contribute to increased coverage, improved treatment and prevention of malnutrition. Also, refreshment training for HEWs is required with a focus on admission and discharge criteria - from different stages of the investigation it has emerged that both CDAs and HEWs are predominantly identifying oedematous children, with marasmic cases remaining unidentified or even rejected at the health post. Currently HEWs have a double level of responsibility, at HP and also screening in the community. This high workload prevents them from providing the necessary attention to the caregivers during their weekly visits - and sometimes even closing the HP during OTP days - and also from routinely screening in the community (which is usually reduced to mass campaigns) as desired. Community level screening is usually reserve for mass campaigns. CDAs and HEWs need to work closer and re-define their tasks in order to ensure quality service delivery and ultimately, to improve coverage. Distance has proved not to be a major barrier to coverage. On the contrary, caregivers have shown a very positive attitude and awareness with regards to malnutrition that encourages them to overcome this obstacle in those cases where the OTP is further from the village. Other key boosters to coverage that have become apparent during the investigation include awareness of malnutrition, involvement of key community leaders and passive case finding at HP/HC level. 26

27 6. RECOMMENDATIONS In line with what the study has found, recommendations and activities to improve coverage are: Insufficient specific information about the program Barrier Key recommendation Actions Clarify understanding of the program functioning in the community Increase sensitization about the program Lack of screening in the community Wrong admission criteria Lack of training of program staff/volunteers Weak referral system Strengthen existing outreach activities Promote adequate application of admission and discharge criteria Increase capacities of CDA and HEW Strengthen the referral system Revise CDA activities/strategies Increase the number of CDAs Train CDA to screen with MUAC to improve case detection Refreshment training for HEWs on admission and discharge criteria Develop a mechanism to monitor cases referred at community level, from OTP to SC (and vice versa) and for discharged children 27

28 ANNEX 1: EVALUATION TEAM Investigation team BEATRIZ PÉREZ BERNABÉ, CMN s Regional Coverage Advisor HKI LINDSEY PEXTON, GOAL Nutrition Advisor HAILU SITOTAW, GOAL Ethiopia - Senior Survey and Assessment Coordinator SEIFU SISAY, GOAL Ethiopia - Senior Survey & Assessment Officer AFERA ASMEROM, GOAL Ethiopia - Senior Survey & Assessment Officer SHIFERAW TADESSE, GOAL Ethiopia - Survey & Assessment Info. Officer ABRAHAM LELANGO, GOAL Ethiopia - Survey & Assessment Info. Officer ASSEN SEID, GOAL Ethiopia - Survey & Assessment Info. Officer TIGABU HAILU, GOAL Ethiopia - Survey & Assessment Info. Officer TEMIR MOHAMMOD, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer MEAZA MITIKU, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer TADELECH GEBITA, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer BIRTUKAN AYALEW, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer YEBERGUAL MEKONEN, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer ZENEBU GEBRESILASE, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer CHUCHU TADESSE, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer TURENESH LEGESSE, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer DAYAN TAYE, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer MASERESHA BOGALE, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer SEID MOHAMMED, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer YIMER MOHAMMED, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer SHEWANGIZAW TESHOME, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer BESFAT ABERA, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer ABIY ALEMU, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer DEMIS TEKLU, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer YOHANNES GIRMA, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer ALEMAYEHU GEZAHEGN, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer HENOK LEGESSE, GOAL Ethiopia - Ass. Survey & Ass. Info. Officer MEKDES GEBREYSUS, GOAL Ethiopia - Data Entry Officer ZINASH BOCHA, GOAL Ethiopia - Senior CMAM Programme Officer ROZA DAGNE, GOAL Ethiopia - Senior CMAM Programme Officer TENADAM AMEDIN, GOAL Ethiopia - CMAM Nurse TIGIST BIRATU, GOAL Ethiopia - CMAM Nurse ALIYE ABDUREHIMAN, MoH-West Hararghe Zone MOHAMMED SEID, MoH-West Hararghe Zone TEFERA GIRMA, MoH-Oda Bultum Woreda - CMAM Focal Person AMIRA MOHAMMED, MoH-Oda Bultum Woreda, CMAM Focal Person 28