Document of The World Bank Report No: ICR00002460 IMPLEMENTATION COMPLETION AND RESULTS REPORT (TF-92181) ON A GRANT IN THE AMOUNT OF USD 6.23 MILLION TO THE REPUBLIC OF YEMEN FOR A SAFE MOTHERHOOD VOUCHER PROGRAM January 21, 2015 Health, Nutrition and Population Global Practice Middle East and North Africa CURRENCY EQUIVALENTS (Exchange Rate Effective December 8, 2014) Currency Unit: 1 USD = 214.9 YER FISCAL YEAR: January 1 – December 31 ABBREVIATIONS AND ACRONYMS AMR Annual Monitoring Report MTR Mid-Term Review BTOR Back-to-Office-Report MS Moderately Satisfactory CAS Country Assistance Strategy MU Moderately Unsatisfactory EFA External Financial Auditor NGO Non-governmental Organization FM Financial Management OBA Output-Based Aid GA Grant Agreement OM Operations Manual GPOBA Global Partnership for Output- PAD Project Appraisal Document Based Aid PDO Project Development Objectives GPOBA SAR GPOBA Semi-Annual Report POE Panel of Experts ID Identity Card S Satisfactory IFC International Finance Corporation SD Service Delivery IVE Independent Verification Expert SOUL SOUL for Development MDG Millennium Development Goal SGH Saudi German Hospital MNVP Maternal and Newborn Voucher Project SMP Safe Motherhood Voucher Program M&E Monitoring and Evaluation TTL Task Team Leader MTI Monitoring and Tracking Indicators USTH University for Science and Technology Hospital (Al Marwarid) Senior Global Practice Director: Timothy Grant Evans Acting Country Director: Poonam Gupta Practice Manager: Enis Barış Project Team Leader: Ali Ahmed Al-Mudhwahi ICR Team Leader: Ali Ahmed Al-Mudhwahi     ii  REPUBLIC OF YEMEN Safe Motherhood Voucher Program CONTENTS 1. Project Context, Development Objectives and Design ............................................................................. 1 1.1 Context at Appraisal ..................................................................................................................... 1 1.2 Original Project Development Objectives (PDO) and Key Indicators..................................... 3 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification ............................................................................................................................ 3 1.4 Main Beneficiaries ........................................................................................................................ 3 1.5 Original Project Components ...................................................................................................... 4 2. Key Factors Affecting Implementation and Outcomes ............................................................................. 5 2.1 Project Preparation, Design and Quality at Entry .................................................................... 5 2.2 Implementation ............................................................................................................................. 7 2.3 Monitoring and Evaluation Design, Implementation and Utilization .................................... 10 2.4 Safeguard and Fiduciary Compliance....................................................................................... 13 2.5 Post-completion Operation/Next Phase..................................................................................... 14 3. Assessment of Outcomes ........................................................................................................................ 15 3.1 Relevance of Objectives, Design and Implementation ............................................................. 15 3.2 Achievement of Project Development Objectives .................................................................... 18 3.3 Efficiency ..................................................................................................................................... 20 3.4 Justification of Overall Outcome Rating .................................................................................. 20 3.5 Overarching Themes, Other Outcomes and Impacts .............................................................. 20 4. Assessment of Risk to Development Outcome ....................................................................................... 21 5. Assessment of Bank and Borrower Performance.................................................................................... 21 5.1 Bank Performance ...................................................................................................................... 21 5.2 Borrower Performance ............................................................................................................... 24 6. Lessons Learned...................................................................................................................................... 26 iii  ICR DATA SHEET REPUBLIC OF YEMEN Safe Motherhood Voucher Program Report No.: ICR2460 A. Basic Information Yemen Safe Motherhood Country: Yemen, Republic of Project Name: Voucher Program Project ID: P104946 L/C/TF Number(s): TF-92181 ICR Date: 12/09/2014 ICR Type: Core ICR - SOUL for Development - Saudi Yemeni Health Care Company Lending Instrument: SIL Grantee: - Al Mawarid Company for Educational and Health Services Original Total Commitment: USD 6.23M Disbursed Amount: USD 3.56M Revised Amount: USD 3.56M Environmental Category: B Implementing Agencies: SOUL for Development Saudi Yemeni Health Care Company Al Mawarid Company for Educational and Health Services Cofinanciers and Other External Partners: GPOBA B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 01/18/2007 Effectiveness: 09/08/2008 03/22/2011 Appraisal: 06/20/2007 Restructuring(s): 05/31/2012 Approval: 06/02/2008 Mid-term Review: 09/25/2010 09/25/2010 Closing: 06/30/2012 02/28/2014 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: High iv  Bank Performance: Moderately Satisfactory Grantee Performance: Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Moderately Quality at Entry: Government: Not Applicable Unsatisfactory Implementing Quality of Supervision: Moderately Satisfactory Moderately Satisfactory Agency/Agencies: Overall Bank Overall Borrower Moderately Satisfactory Moderately Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if Indicators Rating Performance any) Potential Problem Project at No Quality at Entry (QEA): None any time (Yes/No): Problem Project at any time Quality of Supervision Yes None (Yes/No): (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Health 100 100 Theme Code (as % of total Bank financing) Other human development 25 25 Population and reproductive health 50 50 Social safety nets 25 25 E. Bank Staff Positions At ICR At Approval Vice President: Gerard A. Byam Daniela Gressani Country Director: Poonam Gupta Emmanuel Mbi Practice Manager/Manager: Enis Barış Akiko Maeda Alaa Mahmoud Hamed Project Team Leader: Ali Ahmed Al-Mudhwahi Abdel-Hamid ICR Team Leader: Ali Ahmed Al-Mudhwahi ICR Primary Author: Preeti Kudesia v  F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The objectives of the project are to: (a) provide quality maternal care to about 40,000 eligible women in targeted districts in Sana'a and (b) design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision. Revised Project Development Objectives (as approved by original approving authority) First revision of March 22, 2011: To: (a) provide quality maternal care to about 30,000 eligible women in targeted districts in Sana'a and (b) design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision. Second revision of May 31, 2012: To: (i) provide quality maternal care to (15,000) eligible women in the targeted districts in Sana’a; and (ii) design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision. (a) PDO Indicator(s) Original Formally Actual Actual Formally Target Revised Actual Value Achievement achievement Revised Baseline Values Target Achieved at Indicator (before (before Target Value (from Values Completion target target Values approval (March ion revision) revision) (May 2012) documents) 2011) Indicator 1 : Number of safe deliveries Value (Quantitative or 0.00 40,000 3,175 30,000 5,273 15,000 16,141 qualitative) Date achieved 06/30/2008 02/28/2011 03/22/2011 04/30/2012 05/31/2012 02/28/2014 Comments The revised target following two project restructurings was achieved (108%), with a (incl. % resultant weighted average of Moderately Satisfactory achievement (Annex 3 I). achievement) (b) Intermediate Outcome Indicator(s) None: As described in the ICR, this project was designed as a GPOBA project did not have a Results Framework in the designated format as required by the Bank. Therefore intermediate indicators were not defined. Performance indicators, to be measured based on targets on a quarterly basis, were defined for the grant recipients to measure service delivery and therefore reimbursements to them. These are described in Annex 3 (D&E). vi  G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 05/05/2010 Moderately Satisfactory Moderately Satisfactory 0.15 2 01/11/2011 Moderately Unsatisfactory Moderately Unsatisfactory 0.38 3 03/12/2011 Moderately Unsatisfactory Moderately Unsatisfactory 0.50 4 07/05/2011 Moderately Unsatisfactory Moderately Unsatisfactory 0.57 5 11/14/2011 Moderately Unsatisfactory Moderately Unsatisfactory 0.98 6 07/10/2012 Moderately Satisfactory Moderately Satisfactory 1.67 7 12/18/2012 Satisfactory Satisfactory 2.01 8 09/23/2013 Satisfactory Satisfactory 2.621 H. Restructuring (if any) ISR Ratings at Amount Board Restructuring Disbursed at Restructuring Approved Reason for Restructuring & Key Changes Restructuring Date(s) PDO Made DO IP in USD Change millions A revised price list for the defined safe motherhood services estimated that the allocated grant of US$5,403,960 for service provision would be able to cover the 03/22/2011 MU MU 0.57 subsidies for only about 30,000 beneficiaries; therefore the restructuring reduced the number of target beneficiaries from 40,000 to 30,000 to reflect the revised cost structure. Withdrawal of Saudi General Hospital (SGH), the second service provider, and revised estimations of the time that would be required to reach the revised end-target. The restructuring changes were: (i) amendment of the revised PDO with an amended key project indicator of 15,000 beneficiaries 05/31/2012 MU MU 1.66 based on a single service provider; (ii) extension of the project closing date from June 30, 2012 to February 28, 2014 to meet the revised target of 15,000; (iii) reallocation of Grant proceeds; and (iv) a cancellation of US$2,319, 910 from the unused grant funds because of withdrawal of SGH.                                                              1  By the end of the grace period (June 30, 2014), the total disbursement amounted to USD 3,555,224.93.    vii  I. Disbursement Profile viii  1. Project Context, Development Objectives and Design 1.1 Context at Appraisal Country and Sector Background 1. The Republic of Yemen was formed in 1990 by the unification of North and South Yemen. The country has a diverse geographical topography, ranging from high mountainous regions to coastal terrains, deserts and islands. At project appraisal, Yemen’s population of 20.9 million was predominantly rural2 (73%), with about 42% of people living in poverty3 with limited access to basic health services, and ranked 151st out of 177 countries on the 2005 United National Development Programme Human Development Index. 4 Yemen also had one of the highest maternal mortality ratios at 570/100,000 live births (2000 estimates) with only 27% of births attended by skilled birth attendants (2003) as well as a high fertility rate (average of seven children per woman). Life expectancy increased from 41.6 years in 1970 to 62 in 2006, with that for women mirroring the overall trend. While there was a significant increase in enrollment rate in basic education (3 million in 1996 to 4.1 million in 2004), female literacy rates remained low at 28.5 percent in 2002. 2. Many persisting challenges have affected the country’s development, despite achievements over the last three decades. Those that are most prominent and population-related are: (i) high maternal, infant, and child mortality; (ii) high prevalence of malnutrition particularly for children under five years; and (iii) rapid population growth. Despite these challenges, Yemen is on track towards achieving the 4th Millennium Development Goal (MDG) (reduce child mortality) by 2015 and already achieved the 5th (improve maternal health). 3. Yemen, at project preparation and today, has a fragmented healthcare system with several different public and private providers. The number of healthcare providers in Yemen (either public or private) providing quality service is very limited. The health facilities often lack adequate equipment and supplies, and this is compounded by a scarcity of physicians and other medical professionals, including the poor quality and reduced standards of medical training in the country. Amongst those Yemenis that can afford to, many travel abroad for private treatment. 4. In the poorest communities in Yemen, poor access to quality services, lack of confidence in the healthcare providers and lack of affordability of care are contributing factors to the high maternal mortality ratio and key barriers to safe maternal care and services. There are also certain cultural specificities in their health seeking behavior – pregnancy not being viewed as requiring any medical attention, the choice to seek medical care often being made by the husband or mother- in-law, the lack of female doctors, and a distrust of the providers. As a result, amongst poorer communities it is not uncommon to find women not seeking any healthcare during pregnancy and opting for delivering at home.                                                              2 And continues with the same urban-rural distribution today, but poverty has increased to 54.5% 3 1998 Household Survey estimate 4 World Bank, WHO sources   1  Rationale for Bank Assistance and Link to Country Assistance Strategy (CAS) 5. The Bank Group’s ongoing engagement in the Middle East Region and Yemen provided it with a comparative advantage to support an innovative methodology to address the challenge of high maternal mortality in the country. The Global Partnership for Output-Based Aid (GPOBA) had the ability to provide grant financing to pilot an innovative approach to address this challenge. This grant also helped to contribute towards addressing the maternal MDG which was a stated priority of the country; and this was well aligned with the World Bank’s 2006-09 CAS for Yemen: “The overall objective of the Bank Group’s program proposed in this CAS is to facilitate Yemen’s further progress towards the MDGs”. Moreover, under the third overarching goal related to improving living conditions and social indicators, the CAS stated the following: “Significant improvements have been achieved over the last years, in large part thanks to a dramatic expansion of health facilities, increased immunization coverage, and improved control of major diseases (malaria, TB). The Development Plan for Poverty Reduction aims to build on this foundation to accelerate progress towards the MDGs through a renewed focus on the modalities of service delivery”. Yemen Safe Motherhood Voucher Program (SMP) 6. The project was a performance/output-based grant financed by the International Finance Corporation (IFC), managed by GPOBA and supervised by the Bank. The IFC, GPOBA, and the Bank established one of the early partnerships at that time. Senior management was interested in monitoring this collaboration and learning lessons on how these organizations worked together. There were common objectives that brought these organizations together. IFC was interested in exploring and promoting a role for the private sector in health services provision; GPOBA was interested in expanding its subsidy program to areas beyond infrastructure, and the health sector was selected; and the Bank was looking for an innovative model for service delivery of maternal health services. These organizations agreed that this project would be a good experiment that would allow them to test and develop a model for service delivery for maternal health. Each organization, however, at that time had different and separate guidelines to manage the design and implementation of their operations.5 As part of the operating arrangements, the Bank, in addition to being responsible for project supervision, was also the Fiduciary Management Agency for the project and responsible for the recruitment and management of an Independent Verification Expert (IVE) and an External Financial Auditor (EFA). The Project Grant Agreement (GA) was signed on June 11, 2008 and became effective on September 9, 2008 after finalization of the project's Operations Manual (OM). After an extension of the original closing date of June 30, 2012, the project closed on February 28, 2014. The grant recipients were SOUL for the Development (SOUL), a national non-governmental organization (NGO), and two private service providers, the Saudi German Hospital (SGH) and the University for Science and Technology Hospital /Al Marwarid (USTH).6 The project aimed to provide a “Mother-Baby package” of essential services as defined by the World Health Organization7 to eligible poor women (based on income criteria of the household and residing in nine selected districts of Sana’a City, Yemen. The GPOBA grant subsidy was designed to be provided for a targeted number of safe child births by the two service providers.                                                              5 These differences did create certain challenges which are described later in the ICR. 6  The grant was eligible only to IFC clients and this criterion had to be considered while selecting the service providers. 7  Consists of antenatal care, attended childbirth, postnatal care, complicated care services and family planning services.  2  1.2 Original Project Development Objectives (PDO) and Key Indicators 7. To: (a) provide quality maternal care to about 40,000 eligible women in targeted districts in Sana'a; and (b) design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision (GA and OM). 8. The OM) further elaborates the primary objective of the project as expanding access to safe and attended childbirth to the target population to enable the GPOBA subsidy funding to result in about 40,000 safe child deliveries by two service providers over a four year period. This was the key project indicator.8 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification 9. The original PDO and the associated indicator were revised twice during the project. 10. First revision of March 22, 2011: The revised objectives of the Project: (a) provide quality maternal care to about 30,000 eligible women in targeted districts in Sana'a; and (b) design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision. The revised key indicator was about 30,000 safe child deliveries by two service providers over a four year period. 11. Second revision of May 31, 2012: The revised objectives of the Project are to: (i) provide quality maternal care to eligible women in the targeted districts in Sana’a; and (ii) design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision. The revised key indicator was 15,000 safe child deliveries by service providers over the extended project period. 12. The justification and reasons for extension are described as part of the description of Project Restructuring in Section 1.7 below. 1.4 Main Beneficiaries 13. The main project beneficiaries were to be pregnant women living in nine selected districts in Sana’a and fulfilling the following eligibility criteria: (i) maximum gestational age of 22 weeks;                                                              8  The ex-TTL informed the ICR author that since this was a GPOBA project, it did not have a detailed Results Framework complying with the Bank standard format. In the absence of which, this number (and later the revised numbers) was taken as the key indicator for the ISRs, and so also for the ICR.  3  (ii) within a reproductive age (15-49 years); (iii) with household daily income less than US$8; and (iv) with ability to pay the contributory amount of US$15.9 1.5 Original Project Components 14. The project had the following components as described in the OM: 1) Project preparation and start-up cost - to SOUL for conducting a baseline study for community assessment of the targeted districts in Sana’a, provision and installation of a biometric server and eye print/finger print equipment and office furniture and equipment. 2) Establishment of satellite clinics10– to establish 12 satellite clinics with GPOBA providing a 50% subsidy to the service providers. During the first two years of the project, each service provider expected to establish two new satellite clinics annually, compared to one each in year 3 and 4. 3) Service Delivery – to subsidize the defined safe motherhood package. Invoices for services rendered and the verification of Performance Indicators to trigger reimbursements to service providers on an output basis. 4) Community Outreach Administrative Cost - SOUL was to: (i) conduct community outreach for the identification, evaluation and documentation of eligible beneficiaries under the Project; (ii) establish, manage and make available a database of such eligible beneficiaries; and (iii) provide capacity building to support the implementation of the project and the community outreach activity. 5) Education and awareness campaigns– SOUL to conduct education and public awareness campaigns to promote quality maternal care throughout the project in targeted districts in Sana’a. 1.6 Revised Components 15. Project components remained unchanged during the life of the project, though the grant financing categories were changed at the second project restructuring (described below). 1.7 Other Significant Changes 16. The project underwent two Level 1 Restructurings, including a project extension, as described below: 17. First Restructuring, March 22, 2011: Based on an ongoing discussion and demand from the two service providers, an assessment was conducted for the Mid-Term Review (MTR) in September-October 2010, when it was determined that the unit cost per beneficiary needed to be                                                              9 This copayment by the beneficiary was a GPOBA design feature, as communicated by the ex-TTL to the ICR Author.  10  The location of satellites clinics used the following selection criteria in the priority of order: (i) population of districts; (ii) availability of other health facilities per population density; (iii) status of socio economic standards; and (iv) accessibility to the service providers.   4  increased (otherwise risking the complete withdrawal of both service providers from the project). A revised price list for the defined safe motherhood services was established for the project (Annex 3 A). With this new price list, it was estimated that the allocated grant of US$5,403,960 for service provision would be able to cover the subsidies for only about 30,000 beneficiaries of safe deliveries (referred to as ‘beneficiaries’ henceforth). This would render the original PDO unachievable, and therefore the restructuring reduced the number of target beneficiaries from 40,000 to 30,000 to reflect the revised cost structure. This change, reflected in the revised PDO was the only element of the restructuring. The Project was rated Unsatisfactory at that time. 18. Second Restructuring, May 31, 2012: Withdrawal of SGH, the second service provider, (as conveyed to the Bank) and revised estimations of the time that would be required to reach the revised end-target, made this restructuring necessary. As a result the project was restructured as follows: (i) an amendment of the revised PDO with an amended key project indicator of 15,000 beneficiaries based on a single service provider; (ii) an extension of the project closing date from June 30, 2012 to February 28, 2014 to meet the revised target of 15,000; (iii) reallocation of Grant proceeds;11 and (iv) a cancellation of US$2,319, 910 originally earmarked for SGH. The Project was rated Moderately Unsatisfactory at that time. 19. At the time of the second restructuring, the GPOBA Panel of Experts (POE) raised certain queries concerning: replication of part (b) of the PDO, sustainability of financing after project closing particularly if there were political changes, and why 20 months was required to enroll 5,000 more women. The Task Team Leader (TTL) responded by indicating that replication in rural areas by the private sector was more likely than in urban areas, social services were unlikely to be impacted irrespective of which political party was in power, and that the 5,000 women would be enrolled in the first 14 months and the remainder time would be needed to ensure that they got the safe delivery services. The POE endorsed the restructuring. 20. A letter dated May 31, 2012, which confirmed the second amendment to the GA, referred to the partial cancelation of funds and the revised PDO with no reference to the revised target of 15,000 beneficiaries and the project extension and discontinuation of services of SGH, though these are all mentioned in the Restructuring Paper and were considered formalized by all concerned. The numbers of beneficiaries were reflected in the revised OM. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 21. Project Preparation: GPOBA has the mandate to design and fund pro-poor output-based subsidies in all sectors including health. The project was prepared following an assessment by IFC of the Yemeni private health sector, and a Concept Note from the three then potential grant recipients was provided in early 2007. This was prepared as the first performance/output-based contracting program with local private health providers targeting poor communities in Yemen. It was also the first GPOBA Window 3 (subsidy-based) project in Yemen.                                                              11  To reflect a new Disbursement Category F to reflect USTH as the sole service provider and close Category E to reflect the withdrawal of SGH.  5  22. Project Design: The project design was built on the fact that affordable quality maternal services were not easily available through the public health system to poor women in Yemen resulting in high rates of maternal mortality, and was therefore seeking to explore contracting with the local private sector to provide the same. It employed the output-based-aid (OBA) approach to innovatively address this issue. Since this was being done for the first time in the country, the project interventions were restricted to selected districts in Sana’a City (urban) while targeting women from poor households (monthly income less than US$8). Services were designed to be provided by two identified service providers (through an appraisal done by IFC, and both were IFC clients), while the community outreach including health education and appropriate targeting was assigned to a reputed NGO in Sana’a. Prerequisites that each organization had to fulfill at that time resulted in certain design weaknesses. IFC could only provide this grant to its’ clients thus limiting competition with non-IFC clients, taking into consideration that GPOBA required a contribution by each beneficiary to the subsidy and did not have provisions for pre-financing providers. 23. Quality at Entry: GPOBA/IFC and the Bank fiduciary team had discussions with the three identified grant recipients. According to feedback received from SOUL during the ICR preparation, interaction with the Government was limited, with its reaction being reserved regarding external financing being garnered towards the private sector considering that the benefits would be accrued by poor women and households. A detailed OM was prepared by the Grant recipients and agreed with the Bank team by August 2008 (two months after signing of the GA) resulting in project effectiveness on September 9, 2008. However, the rules of engagement at that time between the different organizations were not clear. The Bank team followed the guidelines available at that time and adjusted its compliance to updated guidelines as they evolved during the operation. For example, except for the GPOBA Commitment Paper/Document 12 no project preparation documents (Project Appraisal Document (PAD), technical reports, and feasibility studies) were required to be prepared. Critical risks and mitigation measures were identified in the GPOBA document, e.g.: (i) the affordability and willingness to pay by customer (mitigation measure - minimum user fees and inclusion of nutritional component); (ii) inadequate control systems within the NGO (mitigation measure - external financial and organizational audit; and (iii) the potential for collusive behavior of patients and service providers (mitigation measure - service providers required to keep standardized medical records and provide them upon request for auditing. Random audits of claims’ database to be done in order to identify claiming patterns that might suggest collusive behavior). Despite this, some of the other identified risks – (i) early reluctance of population to accept services; and (ii) the possible lack of quality staff/capacity within service providers were underestimated and adversely affected implementation, particularly during the first few years. The most significant areas that seem to have been underestimated were the real understanding of/by the private providers to the project and the in-depth involvement and commitment of the government.                                                              12 It was communicated to the ICR author that for a GPOBA stand-alone project, GPOBA requires a Commitment Paper. It is up to the Bank team to decide if a PAD is needed. 6  2.2 Implementation 24. There was no delay in project effectiveness, and even prior to that, in August 2008 the Bank team visited Yemen to prepare for effectiveness and finalize the terms of reference for the competitive selection of the IVE and EFA who were appointed in June 2009. However, the project faced implementation challenges almost from the very beginning. These, including the mitigation measures undertaken, are described below: 25. Capacity/Ability/Commitment of the service providers to reach the target of 40,000 beneficiaries: Despite an appraisal by IFC of the capacity of the private health sector in Yemen (Annex 3 A), the selection of SGH and USTH as the service providers for the project. There were issues with their capacity that were identified as a risk as early as of August 2008, including that this may negatively impact the achievement of the target before the closing date of the project. In August 2009, the unsatisfactory performance of SGH was highlighted by the TTL, including the lack of business standards and practices and poor understanding of the design and implementation of the project. These continued to be highlighted as risks, as late as September 2010 (MTR). The role and performance of the grant recipients varied during the project phases. SGH was the main leader for project design and led all the technical discussions and later was the first to establish its clinic, while USTH was initially hesitant to partner in the project. Although SGH and USTH had technical and fiduciary management capacity at the hospital level, it was difficult to assess their capacity at a clinic level, given there was no prior experience. The business standards of contracting health services in Yemen were based on passive reimbursement and not on active purchasing. They did not have prior experience in managing Bank operations with its fiduciary management requirements and a learning curve was needed to reach an acceptable level of compliance to Bank guidelines. This led to delays in disbursement which further affected the cash flow of these organizations. In particular, SGH suffered more from the financial global crisis than USTH, being an international organization with more limitations from its headquarters on its cash flow. Although SGH tried to address their problems during the first half of the project by changing the CEO of the hospital and injecting some cash into the system its Board decided to freeze their participation in the project. On the other hand, USTH, which is backed by a local company and substantial charity funds, was able to meet with the set targets. 26. USTH also reported suffering from financial losses in mid-2010. According to a note provided by the Bank TTL to GPOBA, these losses resulted from the lack of experience in implementing the innovative outputs-based disbursement service delivery model at the satellite clinics since its experience was mostly related with the service delivery at the hospital level. These losses forced USTH to reconsider the degree of their participation in the project or to withdraw completely. In order to continue its participation in the project, USTH requested that a higher service delivery (SD) package be reconsidered to replace the price of US$135 per beneficiary that was originally with GPOBA as a subsidy for all cases. It was agreed that this would be reviewed at MTR. Based on the agreements subsequently reached (described below), USTH decided to continue participating fully in the Project. 27. In 2010, SOUL also suffered cash flow problems that limited its ability to pre-finance activities related to enrollment and health education/awareness as planned due to the fact that: (i) SOUL’s ability to achieve its targets was dependent on the reputation and performance of the 7  service providers as well as the clinical eligibility criteria of beneficiaries; and (ii) lack of funds that would allow SOUL to pre-finance output-based disbursements. 13 These problems were addressed by amending the OM to reflect SOUL to use: (i) the advance of the Designated Amount to fund output-based disbursements; (ii) the performance bands for performance-based indicators to apply without the need for a mandatory indicator; and (iii) expanding the eligibility criteria to include all poor pregnant women. 28. Difficulties in enrollment and participation: In April 2009, the Bank team highlighted that the initial enrollment rate was low – about 20% of the target. In a Back-to-Office-Report (BTOR) in May 2009, the TTL ascribed the low enrolment rates to two groups of factors: (i) Cultural and Marketing- refusal of husbands to enroll their wives in the project. Discussions with SOUL who was responsible for enrolment indicated that the main reason for the reluctance to enroll at the start of the project was suspicion and mistrust about the project and its motives - doubts and fears among the community about these free services that were to be offered in private clinics and hospitals. A private collaboration of this nature was being introduced for the first time in Sana’a resulting in skepticism; however, later when women who received services through the project spread the word to their community, the enrollment substantially improved. The latter was primarily because of the focused marketing strategy that was later developed and implemented by SOUL, which included a short film explaining the objectives broadcasted in a couple of Yemeni TV channels, as well as posters and brochures distributed in public places in the target areas; and (ii) Provider Performance and Reputation – especially related to SGH. It was found that the service providers were trying very early in the process to: (a) transfer the risk of the cost of complicated cases to the beneficiaries of the project or to the Bank; and (b) decrease their operating cost. While the Bank team repeatedly communicated that the cost of the SD package would be revised during the project MTR, the providers did not demonstrate confidence, and this adversely affected the identification and enrollment of beneficiaries. SGH was also facing additional problems such as irregularity of availability of a provider or unavailability of a female provider as promoted at the outpatient clinic; and some alleged instances of malpractice resulting in rumors that discouraged beneficiaries from participating in the project. In addition, SGH was responsible for preparing the training content to be used to identify and enroll beneficiaries. SGH conducted this training and provided a number of wrong messages in the training package that focused on informing the beneficiaries that they would be charged for complicated deliveries and Cesarean Sections including diagnostics and medications including those given during the antenatal care period. According to project documentation, these messages discouraged participation. 29. In mid-2009, several actions were agreed upon to increase enrollment as follows: (i) modify the eligibility criteria to enroll eligible pregnant women up to 20 weeks (first five months) of pregnancy instead of 12 weeks (first three months), and to include all households that                                                              13  GPOBA did not allow pre-financing to the other two grant recipients, and this was quoted as a constraint by the TTL. 8  have a daily income of less than US$8 as long as they can pay their contribution of US$15 per beneficiary; (ii) agree on a number of marketing techniques that would increase enrollment rates; (iii) develop criteria for selection of sites for establishment of satellite clinics within selected districts that ensure accessibility of providers to clinics; (iv) revise the current schedule for establishment of satellite clinics to avoid establishing more than one clinic in the same district within a short period of time to allow for accumulation of adequate demand; and (v) define the date of signature of a contract for rent of a clinic by the providers as the cut-off point to trigger identification and enrollment of beneficiaries in that district's clinic to ensure the timeliness of enrollment of beneficiaries with the establishment of clinics. 30. Dropouts after enrollment and issue of project criteria restricting eligibility of complicated cases – In August 2009, SOUL reported a dropout rate of almost 70%. While this rate decreased substantially over the life of the project (25%), in part because of the targeted and effective communication strategies adopted by SOUL, one serious challenge persisted. While this has not been clearly documented by the Bank team, reports and recommendations provided by SOUL clearly indicate that since the project, by design, was restricted to only cover 15% of complicated cases, this imposed a challenge for health providers to accept all complicated cases (the 15% cap was reiterated when the pricing issue was addressed). Cost of Service Delivery Package and complicated cases 31. The cost of complicated cases was a subject of extensive discussion during the finalization of the OM (before project effectiveness) and subsequently. The Bank team reiterated repeatedly to the service providers that these services were part of the package and that the risk was mitigated by capping the number of cesarean section cases to not exceed 15 percent of all cases. It had also been agreed that the cost of the package would be reviewed at project MTR. As part of the MTR assessments, a revised price list was developed based on: (i) the study of the actual costs of the two Service Providers (USTH and SGH); and (ii) the market prices of equivalent Service Providers to ensure that these prices were competitive, economical, and developed in a transparent manner. The revised SD package amended the payment mechanism to (i) pay separately for normal delivery, complicated delivery, cesarean section; (ii) add a package for neonatal care (Annex 3 A). This was agreed, and ensured the continuation of participation of the service providers immediately after the MTR. Crisis Period in Yemen in 2011 32. The popular uprising and political turmoil in February 2011 in Yemen resulted in interrupting the provision of health services including SMP operating clinics. The situation deteriorated with violence starting in March, escalating in June-July, and continued till close to end of the calendar year. The Bank officially suspended disbursements to all Yemen projects on July 28, 2011. On November 11, 2011 GPOBA wrote to the Bank’s Country Director for Yemen, requesting a special exemption from this suspension reasoning that the project was a performance- based financing not being implemented by the government, and that it was doing well despite the conflict. However this was not agreed upon, and the suspension of disbursements for all projects was lifted only in January 2012. USTH had established 6 clinics before the uprising, but after the suspension, only one clinic operated first, and later two more were added. 9  33. While SOUL and USTH tried to continue providing services during the crisis period, they described a number of challenges that they faced: (i) low enrollment and utilization rates due to the security situation; (ii) continuous disruption of electricity; (iii) high prices of fuel or its shortage; (iv) the increase in prices of medical supplies and drugs; (v) the attacks on the satellite clinics and ambulances; (vi) difficulty to communicate with the Bank team; (vii) the disruption of cash flow due to the suspension of disbursements by the Bank; and (viii) difficulties in conducting verification. 2.3 Monitoring and Evaluation Design, Implementation and Utilization Monitoring and Evaluation (M&E) Design: 34. Since this was a GPOBA project, it did not have a designed Results Framework as per the Bank’s requirements. 35. The M&E design of the project was built on the performance/output based premise of the project, and therefore, the linked subsidy reimbursement measures/conditions are also described below: 36. Performance Indicators: (i) For Service Providers: A set of seven performance indicators (which included measures of clinical practice, patient volume, client satisfaction, status of hospital equipment and waste management) (Annex 3B) was designed. The service providers would be reimbursed on a quarterly basis, proportionate to the care provided to the eligible beneficiaries and based on the performance band against the number of targets achieved (Annex 3 D), with a caveat that the targets for three indicators (% births by skilled attendants, % of women completing four antenatal visits, and % of women completing one postnatal visit) had to be met for any subsidy to be paid (up to the maximum original subsidy of US$135 which was later enhanced). This would be done by the IVE verifying the quarterly hospital records and invoices from grant recipients and providing the Bank Team with quarterly Invoice Verification Reports. The OM stated that these performance indicators and threshold targets may be modified based on revised Government indicators following the first year of operations, however there is no evidence of such discussions or review during the life of the project. (ii) For SOUL: A set of five performance indicators and performance bands was also designed to measure the performance of SOUL (Annex 3E). A portion of SOUL’s project start up and fixed costs was to be paid upfront by GPOBA. The balance of reimbursement would be based on verification of delivery of services (number of educational/promotional activities, areas of coverage, and number of new pregnant women added to the program during the quarter) by IVE. 37. Monitoring and Tracking Indicators- Additionally, in an effort to assess and improve upon the effectiveness of its Grants, GPOBA had an interest in tracking additional project indicators that were not triggers for disbursements. Therefore, the project also incorporated a series of 10  Monitoring and Tracking Indicators (MTI) to measure the overall impact that the new services had on target communities. The Bank TTL was to be responsible for reporting MTI to GPOBA semi- annually, based on information provided by the IVE and the three grant recipients. These indicators included outputs indicators for SOUL, service providers, static data and project funding sources (Annexes 3 D-G). 38. Annual Monitoring Reports (AMR) - The service providers were to provide to the Bank the AMRs specifying the social, environmental and developmental impact of service provision as described in Annex F of the OM. This required a detailed reporting including all aspects of medical and non-medical waste management (discussed further in the Safeguards Compliance Section below). 39. Project Completion Report - The grant recipients were to furnish the Bank with a Project Completion Report including details on service provision to the target community and the overall use of funds no later than six months after the Closing Date. This report was to include a summary of Monitoring and Performance Indicators throughout the project’s life. In addition, the grant recipients were also meant to prepare two additional reports (“Post Project Reports”), each of such scope and in such detail as the Bank would reasonably request, not later than one month after the expiry of one year and two years, respectively, from the Closing Date. 40. Baseline Study – SOUL was to do a project baseline to document the socio-demographic and services–related factors, and explore the perception of safe motherhood beneficiaries. 41. Identification and Verification of Beneficiaries – The project design at the concept stage had described providing vouchers for eligible women, which would then be used as a measure of verification and subsidy payment. At the Commitment stage, the voucher design was rejected as it was thought to add an unnecessary level of complexity and costs, and was replaced with eye scanning and finger printing. This task was assigned to SOUL who were meant to eye scan and finger print the eligible women at the time of enrolment, and this data was meant to be transferred to the service providers at their clinics/hospitals to ensure the right identification of beneficiary. 42. The M&E design was clearly very detailed based on the requirements of the GPOBA. It was a complex design and its monitoring and validation depended solely on the IVE. M&E Implementation 43. Baseline Survey - The baseline study report completed by SOUL was available around Oct 2009. While the OM had stated that the baseline report was to be available before the satellite clinics started offering care, three clinics were established and 273 cases enrolled and received services by August 31, 2009. The baseline study used both quantitative and qualitative methodologies in four urban districts in Sana’a City. The key findings indicated that: (i) healthcare seeking behavior was far from desired mainly limited to miscarriage/abortion or stillbirth, childbirth, as well as postnatal care periods; (ii) childbirth care was preferred to be at home assisted by relatives/friends, while antenatal care was preferred at clinics assisted by skilled health providers especially females; [however, care was usually sought at public health facilities when a health problem was perceived to occur] and (iii) the use of family planning was surprisingly high at 80.5% with preference for oral contraceptive pills and intrauterine devices, reflecting increased 11  knowledge and a high demand. Any discussion on the perception of women regarding service delivery by private providers was missing however, despite being the focus of the project. 44. Identification and Verification of beneficiaries – While this is not recorded in any of the available documents, discussion with SOUL revealed that eye scanning was never implemented because of its cost and complexity. Finger printing was done by SOUL for all enrolled women, however because of technology constraints; they were never able to transfer the finger prints from their database (where it also occupied a large space) to that of the service providers. Therefore, all identification and verification of eligible beneficiaries through the life of the project was done with a physical photo-identification of the Identity Cards (ID) issued by SOUL at the time of enrolment and subsequently by the IVE in order to endorse the reimbursement to SOUL. 45. Role of the IVE - The role of the IVE was absolutely critical for ensuring compliance with the M&E framework and subsequently the subsidy reimbursement to the grant recipients of the project. The MTI reports provided by the TTL to GPOBA primarily depended on the reports provided by the grant recipients and IVE, and, thereby reinforcing the fact that the M&E compliance depended on the IVE14. The IVE was competitively selected and started providing services from July 2009 onwards. Till the MTR report (June 2010) Cardno Emerging Markets USA, Ltd. was the IVE. The frequency of their interactions was less, compared to when the IVE changed after MTR to Deloitte (who was also the EFA) and who had a Public Health specialist for technical verification based in Sana’a. In the Final IVE Report (May 2014) covering the entire project period, provided by Deloitte, there appears to be several inconsistencies in data, including misinterpretation of definitions; e.g., the number of women enrolled has been interpreted as meeting the final project target, which was also used by the last project ISR. Disbursement could not be triggered to the grant recipients without the relevant IVE reports, but the quality of the data reported by Deloitte can be further investigated. 46. SOUL and USTH provided the mandated performance reports in a timely manner, and SOUL also maintains a full database since start of the project. IVE reports are available, and while thorough, had data and definition inconsistencies, as described above. 47. A Project Completion Report from SOUL was received on October 14, 2014 and that from USTH is yet to be received though it is now close to eight months since project closing. The realism of the requirement (described above) of further reports one and two years after project is debatable, considering the significant loss of institutional memory at present, eight months after project closing (except for SOUL). M&E Utilization 48. As described above, the M&E framework for the project was primarily designed to be able to monitor and thereby provide evidence to trigger subsidy reimbursements to the grant recipients, and not as a Results Framework for the project. It was used for this purpose as well as to assess the performance of the project. SOUL did maintain a full database for the project and it expressed appreciation of the capacity building that this had facilitated. As alluded to above, the M&E design                                                              14 The TTL informed the ICR Author that the methodology of the IVE was later adopted by the Yemen Country Office for Third Party Monitoring for all projects in Yemen. 12  was complex, and some parts of it not implemented, for example the eye scan and finger printing, as a result of which only photo ID cards were used for beneficiary verification during the project. 2.4 Safeguard and Fiduciary Compliance Safeguard Compliance 49. The Project was to be implemented (specifically with reference to the establishment and operation of the clinics by the service providers) according to IFC’s Performance Standards on Social & Environmental Sustainability (Annex 3B). This compliance was to be monitored and reported by the IVE. Additionally, each service provider was to provide AMR to the World Bank, as described in Annex F of the OM. The AMR information had to be collected in a manner consistent with applicable IFC Environment and Health and Safety Guidelines (Annex 3B). The AMR included, service providers conducting and thereby reporting, on regular re-training of staff, internal-auditing and record-keeping on sanitation, infection control and medical waste management. 50. While the available IVE reports do state that medical and non-medical waste management was maintained according to standards (no details provided), this ICR preparation did not have access to any AMR, and is therefore constrained to provide a comment on the safeguard compliance. Fiduciary Compliance Financial Management and Disbursement 51. The original project design in the GPOBA Commitment Document described a Fiduciary Agent who would be responsible for financial management for the project. However according to the final design in the OM, the project did not have a Fiduciary Agent and all such functions were managed by the Bank team. Since this was a performance-based project all agreed subsidy reimbursements to the grant recipients were triggered subsequent to a confirmation of achievement of targets by the Bank TTL to GPOBA. Three types of disbursements took place during the life of the project: (i) project preparation/design and start-up costs - disbursed as subsidy to SOUL upon GPOBA grant agreement being signed; (ii) capital costs of satellite clinics - capital cost of establishing the satellite clinics split on 50:50 basis between GPOBA and the service providers. Fifty percent of this reimbursed the service providers after commissioning of the said clinics, and the remaining 50% of GPOBA share paid after the first three months of operations and subject to a satisfactory quality performance report by the IVE; and (iii) invoiced services - starting from the first quarter of operations, GPOBA reimbursed costs of service provision based on the performance indicators described above. 52. A revised disbursement letter dated January 17, 2013 replaced the original disbursement letter of June 11, 2008 confirming grant recipients as only two – SOUL and USTH, with the withdrawal of SGH. 13  53. Based on reporting in the GPOBA Semi-Annual Report (SAR) and the MTR aide memoire, SOUL and USTH Financial Management (FM) departments were well staffed and each agency had an automated accounting system, acceptable FM manual, and acceptable auditing arrangements. All Interim Financial Reports (IFRs) submitted to the Bank were reviewed by the EFA. Audit reports were satisfactory. Details of clinics’ establishment and subsidies provided are not available. It is unclear what consideration GPOBA and Bank give to the fact that while the clinics’ establishment was provided a 50% subsidy, only one of the clinics is operational at present, and this too is unable to provide subsidies to the poor targeted women. Procurement 54. Procurement under the project was done by the following: (i) the Bank – of the IVE and EFA; (ii) SOUL; and (iii) both service providers. According to the MTR aide memoire, SOUL had implemented 16 contracts (two for services and 14 for goods); USTH had completed the purchasing of equipment and furniture for the six satellite clinics; and SGH had implemented two contracts for goods. Procurement was found to be generally satisfactory, except for that done by SGH, though details of the latter were not available in the documents reviewed. The financial management and procurement specialists on the task team were not available for discussion for preparation of this ICR, therefore clarifications were sought by email, and those responses have been incorporated in the discussion above. 2.5 Post-completion Operation/Next Phase 55. In October 2012, at a regional workshop in Amman, the performance-based financing experience of the project was presented and discussed to design an appropriate RBF model for service delivery. This workshop was meant to evaluate the project’s performance based contracting service delivery model compared to the KfW Voucher Project - both models were based on contracting providers, with performance based payments for providers at the facility level. Both SOUL and USTH expressed interest in the continuation of the project and there were discussions with the public and private sectors to explore such possibilities. There were initial discussions between USTH and the Municipality of Sana’a to possibly replicate the project model in public clinics with the Solidarity Fund (described in Section 3.5) envisaged by SOUL, but so far this has not been implemented, and is uncertain given the current country context. 56. A second workshop in March 2014 that also included the government (Ministries of Health and Finance) discussed the concept of performance based financing which was adopted by the Minister of Health during the workshop. The workshop explored ways to link vouchers to performance-based financing. 57. A Maternal and Newborn Voucher Project (MNVP) was prepared by the Bank while this project was under implementation (became effective on August 28, 2014). Yemen Safe Motherhood Program, having been the first RBF intervention in the health sector, demonstrated lessons that using vouchers was likely easier than performance-based financing, and that it was important to work with a national quasi-governmental organization to be the payer for the scheme contracting public and private (profit and non-profit) health facilities. The MNVP design was also informed by the lessons learned from the ongoing KfW Reproductive Health Voucher Project, which was designed in collaboration and learning from the SMP. The MNVP PAD does make a 14  reference to align with the satellite clinics established under this project. While the MNVP pilot has started, the project has plans to subsequently include Sana’a city also, though given the time lag since SMP closed and this initiative is likely to start, it is unlikely to contribute to the sustainability of the SMP clinics established. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Relevance of the Objectives 58. Although Yemen has made great strides in reducing the maternal mortality ratio since start of the project, it remains high at 148 deaths per 100,000 live births15, which translates to some 24 women dying every week due to pregnancy and birth-related complications. At close of the project, nearly half of its estimated population of 25.5 million was living on less than US$2 per day with poverty increasing to 54.5% in 2012 and the country ranked 160th out of 186 on the 2012 HDI. Women, who were already severely disadvantaged, suffered disproportionately as a result of the 2011 crisis, with decreased access to basic and social services and increased levels of gender violence. 59. The objective of providing quality maternal care to poor women in Yemen, who otherwise cannot afford the same, continues to be highly relevant today. Given the poor quality service provision in the public sector, involvement of the private sector would no doubt substantially contribute to improved maternal and newborn outcomes in Yemen. However, what may need further demonstration is the nature of the arrangement that would make it attractive for the private sector to be involved on a sustained basis. Additionally, in a conflict/fragile country context, to what extent would the government be willing to, or could render its support to such partnerships is uncertain. Relevance of the Objectives is rated as Substantial. Relevance of Design 60. The design of the project was innovative and new for the Yemeni health sector - a performance-based approach with the private sector. It was originally designed as a pilot for the first 6-12 months in two districts of Sana’a, with a plan to learn from the experience, and if successful scale up to more districts of Sana’a. Subsequently in the OM the design was expanded to nine of the twelve districts of Sana’a, with a proposed phasing-in of the districts over four years. It is unclear why this change in design took place, and in retrospect a smaller pilot may have provided useful insights into the approach, in particular the understanding of the concept itself by the private sector, its capacity, and what would make it attractive for their sustained engagement. 61. The project was designed to be consistent with the six core concepts of OBA: (i) providing targeted subsidies to poor groups; (ii) accountability for results; (iii) encouraging innovation and efficiency; (iv) using incentives to serve the poor; (v) conducting output verification and monitoring of results; and (vi) fostering sustainability. This is further discussed below:                                                              15Yemen National Health and Demographic Survey, 2013 15  62. Targeting the poor: Income based targeting was used to ensure that subsides reached target women living in nine selected districts in Sana’a, who otherwise could not afford quality health care. SOUL, who was responsible for enrolment of eligible women, state in their final report that the targeting could have benefitted from stronger verification techniques for the poor – “with the spread of the good word about the project, middle class women who can afford the service would pretend they are poor to enroll in the program. Despite SOUL efforts in investigating their cases, logistical and financial capacity imposed a challenge for a comprehensive verification system”. According to the TTL, the GPOBA’s design (to ensure sustainability) feature of copayment by the beneficiaries may have excluded some of the poor. 63. Accountability for results: The performance-based subsidies to the grant recipients made them accountable for the agreed results, thereby shifting the performance risk to them in accordance with the OBA concept. . The IFC’s appraisal had confirmed reasonableness of the rates at project inception, yet this affected the service providers and the eventual revision of rates, also resulting two changes of the project target and delayed implementation. 64. Encouraging innovation and efficiency: As alluded to earlier, this project was an innovative approach for Yemen and the health sector in the country. The original GPOBA subsidy per capita of US$135 was designed to ensure “efficiency”/value for money for delivering the safe motherhood SD package, and this was likely the reason why it took GPOBA and the Bank till MTR to enhance the per capita subsidy, despite the private providers perceiving the subsidy to be unviable from start of the project. While as an OBA concept, efficiency is achieved through competition or benchmarking leading to value for money, the two private service providers in this project were not selected through competition, instead through an appraisal of the private health sector in Sana’a carried out by IFC, possibly because it being a novel approach in the country’s health sector. That appraisal reviewed five hospitals in Sana’a and concluded that the quality (services, staff, facilities, equipment) and capacity of the two providers’ private hospitals far exceeded that of the other local Yemeni private providers (Annex 3 A). 65. Use of incentives to serve the poor: Through the project, poor women were able to avail services which were of better quality than the public sector by contributing US$15.16 Discussions with SOUL gave some insight into the maternal care seeking behavior in Yemen. It was not that women were completely unaware of the benefits of delivery at a health facility, and in the 1970s women would avail of these services at public hospitals. Over time, there was a marked decline in the quality of services provided at these hospitals, including their inappropriate behavior with the poor. This resulted in women preferring to deliver at home and going to a public hospital when complications forced them to (as shown in the baseline study for this project also). During the later part of the project, based on the concerted community outreach that SOUL and positive the experience of women treated by USTH, the numbers of women wanting to enroll and avail services increased substantially. Discussions with USTH and SOUL revel that the targeted poor communities are once again bereft of this benefit since project closing. This is an important indication of the significance of ensuring sustainability of interventions.                                                              16  According to the appraisal, this was reasonable, however the TTL reported that this could have resulted in some exclusions. 16  66. Output verification and monitoring of results. The project M&E framework was designed for this purpose and the IVE performed this function. 67. Sustainability: The project implemented a performance-based program with selected private service providers. However the design failed to ensure sustainability of the project interventions. Due to the many challenges that it faced and the resultant changes that ensued, the project closed with three satellite clinics (instead of 12 as envisaged); and even these, stopped providing services to the eligible group of women as no grant subsidies were available in the absence of a financial sustainability model17 In fact, SOUL stopped their enrolment process several months before project closing so as to ensure that at least those enrolled would be assured delivery services before completion of the project. According the final report provided by SOUL (October 2014), the profitability concern of the private sector threatened sustainability of the program. As per the original design, the satellite clinics were to be operated as profit centers. The providers agreed to deposit profits (if any) in excess of 5% margin into an escrow account/fund to be used to fund any legitimate unexpected cost-overruns in the short-run e.g. larger than estimated percentage of women requiring more expensive complicated and tertiary care. In the long run, the main purpose of these funds would be to serve as a ‘sustainability fund’ that could attract co- financing. While the program was designed in a way that the satellite clinics would provide services at cost, these clinics could potentially generate additional revenues for paying patients outside the scheme and not eligible to benefit from GOPBA subsidy. It was anticipated that successful implementation of the first phase would attract local private large corporate companies e.g. oil and industrial groupings to participate in the program. Specifically, it was meant to provide a good demonstration of how private businesses in Yemen could make investments that were socially responsible and contributed to the health of poorer communities. It was envisaged that based on the pilot, the Government of Yemen may adopt the model as well as attract other bilateral/multilateral donors and healthcare and finance providers to ensure sustainability, and scale up to other surrounding governorates in Yemen. While some elements of this future program design relate to successful projects from other regions in a non- fragile context, the sustainability fund concept is the main element that SOUL plans to implement in the future. 68. The design of the project was innovative and considerable work went into developing the performance-based monitoring mechanism. In hindsight, this was possibly an ambitious assumption that the private sector would participate fully, when financial sustainability was not built into the design. As a model to demonstrate private participation to potentially influence Yemeni public policy, this project has shown the means, but a substantial buy-in from the public sector will likely take more time, given the past and current conflict and political situation in Yemen. The rating for Relevance of Design is rated as Substantial. Relevance of Implementation 69. Implementation of the project, particularly in the early stages, was good and demonstrated proactivity in identifying challenges and bottlenecks, e.g., poor enrollment and the unsatisfactory performance of SGH. SOUL was assigned to better understand the issues for poor enrollment, and subsequently implemented an aggressive and concerted marketing and communication strategy                                                              17  One of the three clinics by USTH is currently being operated by their University Hospital. However no documentation is available that the poor women are getting the same benefits.  17  which clearly reaped positive benefits. While the poor performance and unwillingness of SGH to be a genuine partner in the program was evident during the first year of the project itself (though as alluded to earlier, it had taken the lead during project preparation), the Bank team gave them some more time to improve performance because of a change in their CEO and a renewed expressed commitment. However, due to the impact of the global crisis (described earlier) they exited from the project, closing their only clinic in May 2011, and more formally in May 2012. The dissatisfaction with the price of the SD package was evident from the start, leading even the well-performing USTH to withdraw briefly from the project, before the revised pricing was agreed to. As the service delivery model continued to develop and to be well established, discussions were then initiated as to how to sustain the program beyond the grant financed project. A number of sources of funding were identified, and measures on how to establish this potential/future fund were undergoing. Later, the project faced two major challenges. The first was the turmoil that hit the region and deeply affected Yemen. The second was the suspension of disbursement to the service providers. Both of them strongly affected the model that had just started to stabilize. The security factor was in particular significant because it led to closing of many health facilities, and the suspension of disbursement deprived grant recipients from the necessary cash to advance their operation. Despite that, USTH started to reopen these closed facilities gradually and reaching three clinics during the extended project implementation period. 70. The Bank team attempted to rectify several of the design weaknesses, the political turmoil and ensuing interruption of Bank oversight was a constraint in establishing a more robust dialogue with the government during the project period to ensure sustainability of project interventions. The rating for Relevance of Implementation is rated Substantial. 71. The overall rating for Relevance of Objectives, Design and Implementation is Substantial. 3.2 Achievement of PDOs 72. The achievement of the PDOs is discussed with reference to the three amendments that were made during the life of the project, and the achievement during those periods, before arriving at an overall assessment of the achievement. It is to be noted, as described earlier, that since the project did not have a defined Results Framework and PDO indicators, the most appropriate indicator of number of safe deliveries by the service providers was taken as the key indicator (as described in the OM). While the percentage of births by skilled attendants was a performance indicator for the service providers, through the project life this was not taken as the key indicator by the project, and therefore by this ICR. The second part of the PDO, “design and implement a model of maternal care which demonstrates how Yemeni public policy on maternal and child care can be effectively integrated with private health provision”, was not amended during the life of the project, and no specific indicator or measure was defined. (i) Phase 1: September 9, 2008 - March 22, 2011 (2 years 6 months): number of eligible women who had safe child deliveries -5,049 18 (12.62% of the target of 40,000); and disbursement -US$0.57 out of 6.23 million (9%).                                                              18 % births by skilled attendants – 3,175 (7.9%) 18  (ii) Phase 2: March 22, 2011 – May 31, 2012 (1 year 2 months): cumulative number of eligible women who had safe child deliveries – 8,243 19 (27.48% of the target of the target of 30,000); and cumulative disbursement - US$0.98 out of 6.23 million (16%). (iii) Phase 3: May 31, 2012 – February 28, 2014 (I year 9 months): cumulative number of eligible women who had safe child deliveries - 16,14120 (15,612 by USTH+529 by SGH) (108% of the target of 15,000); and disbursement – US$3.56 out of 3.92 million (91%). 73. The project had a detailed M&E framework with performance indicators for both SOUL and USTH, and these indicators were reported and verified by IVE against quarterly and semi- annual targets of percentage achievement (Annex 3 D-E). 74. At project MTR, a total of 3,989 women were enrolled of which 1,189 women (1026 by USTH and 163 by SGH) had safe child deliveries (2.98%) with a disbursement of US$0.29 million (4.8%). Extensive discussions on the pricing of the SD package, its impact on the need to reduce the target of beneficiaries, and the unsatisfactory performance of SGH took place during the MTR, including the potential need for a project extension (a decision on extension was deferred to July 2011). Following these conclusions, the first restructuring took place five months after the MTR and addressed the SD package pricing, and therefore the need to reduce the end-project target of beneficiaries. The second restructuring done 3 years and 8 months of the (till then) four year project, when disbursement was 16% and cumulative achievement of the revised target was 27.48%, formalized the exclusion of SGH as a service provider resulting in a further reduction of end-project target, an extension of the project, and partial cancelation of the Grant. It can be argued the project could have benefitted from an earlier comprehensive restructuring of the project soon after the MTR that would have realistically taken into account all the challenges that the project was encountering, however given the socio-political context prevailing in the country, and some of the constraints of working with three grant recipients,21 the Bank team made appropriate and feasible decisions. 75. In the absence of a measurable indicator defined for part (b) of the PDO, this ICR attempts on assessing the extent to which the project enabled designing and implementing a model of maternal care which demonstrated a potential engagement with the private sector and a subsequent dialogue with the government. At MTR, the Bank team requested the original IVE, Cardno, to provide an evaluation report on all aspects of the project including an analysis of the service delivery model; and also contracted Options, a consulting firm, to provide advice on the most suitable service delivery model for maternal health services based on experiences in Yemen. In addition to the review report, Options helped organize the Amman workshop (referred to earlier). Further discussions and sharing of the KfW experience resulted in the Bank’s MNVP project, thereby initiating garnering of the government’s buy-in into a focus on results and contracting with service providers for service delivery. Therefore, this project, as an innovative model and through a difficult implementation in a fragile conflict country context, provided several lessons for the future engagement in Yemen towards addressing maternal mortality, though it may take more time                                                              19 % births by skilled attendants – 5,273 (17.5%) 20 % births by skilled attendants – 12,282 (81.9%)  21  There were three grant recipients bound by one legal agreement which imposed inflexibility to reach a consensus during design and implementation, and the requirement of getting written formal requests for restructuring and cancellation, thus the time lag between reaching agreements on the ground and reflecting them in aide memoires and the actual finalization of restructuring. 19  to build sustainable models of engagement with the private sector. There was a missed opportunity of amending part (b) of the PDO during either or both of the project restructurings, to make it more measurable and possibly realistic. 76. Taking into account the weighted average of the PDO during the three phases of the project (Moderately Satisfactory) and the assessment of part (b) of the PDO, the overall achievement of the PDO is rated as Substantial. 3.3 Efficiency 77. At the Commitment stage, a preliminary analysis was conducted based on secondary research and the Program’s cost assumptions; and it was deemed that a quantitative economic and financial analysis of the Program would not provide useful insights into the program, primarily as there were minimal revenue streams (minimal user fees of US$15 per patient) as compared to the per patient investment of US$135 made by GPOBA. Further, in the absence of detailed pricing and household spending data, quantifying the expected economic benefits and conducting a robust economic analysis was deemed not feasible and inappropriate. An IRR was calculated for the satellite clinics and estimated to be 11%. Details of the costing and pricing are provided in Annex 3 A. 78. Only a single maternal mortality was recorded amongst the deliveries reported under the project, which indicates substantial gains for the limited group of project beneficiaries, given the high maternal mortality ratio in Yemen. The assisted delivery rate reached 80%, ante and post natal care 70%, and cesarean sections ranged between 5-15% for the enrolled women. Clearly a group of poor women in their reproductive age benefitted from the project, however a more detailed analysis was not possible as the population of beneficiaries was difficult to define (particularly because the number of complicated pregnancies that could enroll with the project was capped at 15%). In addition, due to the overall limited time and significantly delayed access to data, and inability for a visit to or relevant discussions with stakeholders in Yemen the analysis was limited. Efficiency is therefore rated as Modest. 3.4 Justification of Overall Outcome Rating 79. Given the fact that the project was indeed a bold endeavor to test a new model of engagement in the reproductive health sector in Yemen, combining the Substantial rating for the relevance of objectives, design and implementation, and achievement of the PDOs and Modest rating of efficiency, the Overall Outcome rating is Moderately Satisfactory. 3.5 Overarching Themes, Other Outcomes and Impacts 80. Discussion with, and reports from SOUL, indicate that their participation in the project has contributed to their capacity enhancement, and that they now feel better equipped and more confidant to support similar initiatives, particularly working with external/international organizations. Influenced by their experience, they have submitted a proposal to the Islamic Bank for Development to Finance a ‘Solidarity Fund’ that would potentially support poor women to avail the much needed maternal and child care services. This Fund was potentially meant to be linked to the project beneficiaries and service providers, but has not happened so far. 20  81. It is debatable the extent to which the project contributed to the capacity enhancement of the two private service providers – one opting out during the project, and the other, USTH performed well because of a champion (he had even initiated some discussions with the Municipality /Government to operationalize some public clinics based on this model). However, unfortunately this gentleman passed away recently, and USTH no longer seems to have a champion. However, the project did build USTH’s capacity in working at the clinic level and introducing the concept of performance-based financing, provided that institutional memory is retained. The conflict of 2011 and current political uncertainty in Yemen have no doubt had a negative on timely implementation. 4. Assessment of Risk to Development Outcome 82. As discussed above, Part (a) of the project was met with the substantially revised and reduced target during the extended period of the project. However, according to available information, no part of the intervention is being continued after project closing (no enrollment of poor women or provision of subsidized maternal care to them by the private service provider) because the grant subsidy is no longer available and the financial sustainability model envisioned was not realized. Because of the prevailing uncertain country context, sustained Government commitment is yet to fully materialize, as it may be less of a priority for the Government, given the situation. The follow-on Bank project which supports maternal and newborn care is a performance-based voucher program working with public hospitals in the rural areas and urban slums of selected parts of Yemen, so far is not linked to project intervention in Sana’a. Based on this assessment, the Risk to the Development Outcome is rated High. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance Bank Performance in Ensuring Quality at Entry: 83. The project was prepared based on a GPOBA Concept Note. For the period following the Concept Note and till the GPOBA Commitment Document and GA (signed on June 8, 2008), no other documents were prepared. Subsequently a detailed OM was prepared (providing all necessary details, but differing from the Commitment Document at several places) enabling Project Effectiveness on September 9, 2008. Feedback from SOUL indicates that while the project was being prepared (and till MTR) despite many interactions with the Bank team, there was a near complete lack of understanding of the project concept itself on part of the two selected private service providers (IFC clients). Therefore, the service providers did not comprehend fully the nature of the innovative model and the capacity requirements of implementing the project, particularly setting up and operating the satellite clinics. And most importantly, the service providers were dissatisfied with the pricing of the SD package, perceiving it as being financially unviable for them. This was dealt with by the Bank and IFC team by capping the complicated cases/cesarean sections at 15% of all cases received, thereby forcing SOUL not to/be able to enroll the high risk pregnancy cases, thereby defeating the very objective of the project of providing access to quality services to all eligible women As per discussions with the TTL, GPOBA guidelines placed certain constraints during implementation, e.g., being unable to provide pre- financing to the service providers and the co-payment by the beneficiaries. This raises the concern 21  of whether GPOBA guidelines were flexible enough to the respond to the evolving project context and increasing fragility to meet project objectives by all stakeholders. The setting of the original target of 40,000 beneficiaries may have been ambitious given the overall context. Given the above, the project could have been better prepared to ensure the full commitment and engagement of all parties, including setting of a realistic end-project target 22 , thereby mitigating many of the implementation challenges and project restructurings that ensued. Therefore, on balance, given an innovative design, but lack of clarity with the main stakeholders, the rating of Bank Performance in ensuring Quality at Entry is rated as Moderately Unsatisfactory. Quality of Bank Supervision: 84. As alluded to earlier in the ICR, the Bank team identified several of the key implementation challenges early during implementation, and was proactive in supporting some remedial measures, such as working with SOUL to improve enrollment rates. The revision of the SD package and rates, and thereby the first revision to the number of target beneficiaries was done soon after MTR. However, it took till almost (about three months before) till the original project closing, to bring the other necessary changes into effect, which resulted in lost time and limiting the model for demonstration to a single service provider, finally having to extend the project with a partial cancelation of the Grant to be able to reach the revised the end-project target of beneficiaries served. Discussions with the TTL revealed that these delays were a result of the inability of convincing an IFC client to withdraw from the project and legal issues related to that, as described earlier. 85. The early confusion regarding IFC, GPOBA vs Bank requirements to be followed resulted in only GPOBA Semi-Annual Reports (SAR) from the TTL to GPOBA for missions in August 2008, December 2008 and May 2009. Following this, and a clarification from OPCS, the first aide memoire was provided for a mission in March 2010.23 Following that four more missions are recorded by their respective aide memoires, the last one being in November-December 2013 (no final project closing mission is documented through an aide memoire). GPOBA SARs were also produced resulting in a duplication of effort by the task team. The ICR was delayed beyond six months of closing due to inability to visit the country.24 86. As per the September 2010 GPOBA SAR, a Quality Assessment of Lending Portfolio was conducted for the project in May 2010 and its recommendations included a two year extension of the project. On September 16, 2010, GPOBA provided a no objection to the Bank to make amendments to the payment mechanism (described earlier in the ICR) and a two year project extension. A MTR for the project started in September 2010 and was completed the following month after an agreement reached on the pricing of the SD package. The implementation challenges that the project was facing were highlighted at MTR, and the need to restructure the project (including a request from the Grant recipients) was referred to in the January 2011 mission.                                                              22  August 2008 GPOBA SAR, before project effectiveness, states the ability of the grant recipients to reach the target of 40,000 as one of the risks.   23  Discussion with the TTL indicated that he would combine missions for different tasks in Yemen – preparation of the Health and Population Project and a Schistosomiasis Project, however unfortunately those documents had little reference to this project. 24  This ICR Author was requested to take the task in mid-July 2014 and no ICR mission was possible because of the security situation in the country. 22  This mission rated the project as ‘Unsatisfactory’25and indicated that a project extension of two years would likely be required to reach the revised proposed target of 30,000 women. This proposed extension and upgrading of project rating to ‘Satisfactory’ was made conditional (by the Bank and GPOBA) to the formal revision of the PDO target and an enrollment of 10,000 beneficiaries by June 30, 2011. The next reported mission (reverse mission in Amman, Jordan) after that was in October 201226 after both the first and second restructurings were completed and rated the project as ‘Satisfactory’. The government also participated in this mission and the project experience and implementation model was shared with them. No other discussions with the Government are recorded through the supervision period. Management letters for two of the five aide memoires were available for review for the ICR. ISRs were also prepared following the OPCS clarification, and a total of eight ISRs (Seq. 2-9) from December 2010 to June 2014 were available for review - with PDO/IP ratings being Moderately Unsatisfactory (MU)/MU till Nov. 2011, with an upgrading to Moderately Satisfactory (MS)/MS in July 2012 and further to Satisfactory (S)/S in December 2012 which was maintained till project closing. The ISRs however provided no discussion or justification of the ratings and their change, and no issues were brought to Management’s attention. Comments by Management were provided in the last ISR. While the main dependence was on IVE reports, there is no documented evidence of the Bank team following up on issues such as only the finger-printing being used as the verification means (and in reality only IDs), or safeguards compliance (for which documentation in the required format was not available). 87. Performance of the Independent Verification and Evaluation Agency- the IVE performed its designated tasks through the project (though there was a change in the IVE as described earlier). Given the critical importance that the IVE monitoring and reporting played in the project design, and that there are some data discrepancies and interpretation of key definitions, including in its final report, IVE’s Performance is rated as Moderately Unsatisfactory. 88. Overall, the project design and implementation challenges were identified early on and substantial hand-holding was provided by the Bank team till MTR. This resulted in two project restructurings which enabled part (a) of the PDO to be achieved. There was likely a missed opportunity in addressing the ambiguity of measurement of part (b) of the PDO, however the Bank team did work towards learning from the demonstration model. Subsequently the conflict period was difficult for supervision and monitoring, and these events likely adversely affected a more meaningful engagement with the government. In conclusion, the Quality of Bank Supervision is rated as Moderately Satisfactory. Justification of Rating for Overall Bank Performance: 89. Proactive identification and addressing of the challenges early during project supervision helped achieve part of PDO following the restructurings. The highly satisfactory performance of SOUL, described below, was no doubt facilitated and supported by the Bank team. Lessons were learned and appropriately incorporated into the MNVP project including the decision to use vouchers instead of performance-based financing. Therefore, Overall Bank Performance is rated Moderately Satisfactory.                                                              25  Though the February 2011 ISR rates are MU/MU  26  The hiatus due to the conflict in 2011, but does not explain the gap of almost one year and nine months.  23  5.2 Borrower Performance 90. The Government did not substantially participate in the project beyond project approval. Performance of the Grant Recipients (Implementing Agencies): 91. To begin with the project was implemented by three grant recipients, however with the withdrawal of one of them, the project closed with two. Their individual performance is described below. 92. SOUL - SOUL was responsible for three main tasks - community outreach to promote and execute marketing for the project; enrollment of eligible women; and education/awareness campaigns on maternal and child health. At the start, the project baseline was also done by SOUL. Throughout the project, SOUL was a willing and enthusiastic partner, understood the project, and met its performance targets on a regular basis. SOUL represented the grant recipients as the Project Coordinator throughout the life of the project (though originally envisaged to be on a six-monthly rotational basis). It, likely went even beyond its terms of reference to help resolve challenges, e.g., when it played the role of an effective mediator between the Bank team and USTH when the latter was proposing to withdraw from the project in 2010. The tailor-made communication campaign developed by SOUL and implemented in response to the doubts and suspicions that the community had about the project paid dividends in terms of marked improvements in enrollments. Even later in the project during its extended phase, SOUL launched a three-track health education campaign targeting – the field (schools, mosques, other community places); medical staff to increase awareness of maltreatment of patients; and media.27 By the end of field promotion for enrollment in September 2013, according to their database (which is maintained till date) they had approached and registered 34,271 women. However the total number of ID issued was 20,499 - around 25% of women dropped out for personal reasons or were declined enrolment as they had complicated pregnancies. SOUL’s final report (October 2014) documents their feedback that this project had a unique and untraditional design (and first for Yemen) that created a partnership between NGOs and the private sector to address the maternal mortality challenge; and that the results-based design contributed to the achievements. Inspired by the project experience, SOUL has initiated the process of establishing a Fund to enable poor women to get critical maternal services. SOUL’s Performance is rated as Highly Satisfactory.28 93. USTH – In terms of performance as measured by timely setting up of the satellite 29 clinics and service delivery indicators, USTH performed well through the project. However, because of financial losses that it was facing, USTH expressed its desire to withdraw from the project around September 2010, and its 5 clinics stopped providing services. According to the                                                              27  In the last two weeks of December 2013, 8 Radio short dialogues were prepared and broadcasted on Sana’a Radio. The dialogues were repeated three times a day to total up to 18 minutes a day. Female audience is estimated to have reached around 100,000 only in Sana’a, besides more audience in different parts of the country where Sana’a Radio reaches. SOUL website and blog. Facebook SMP page got 1700 likes.  28 While the ICR author was informed that, SOUL, during the same period of project implementation was rated Unsatisfactory for two other projects in Yemen (the Healthy Mother Project, and the Youth Project), this assessment is based only on what SOUL was meant to do in this project and how it performed. 29 Two satellite clinics established by August 2009, 5 by MTR and one after that, all 6 closed in 2011, 2 reopened in April 2012, 1 more reopened in October 2012, therefore 3 clinics then operated till closing.  24  Bank team, these losses resulted from the lack of experience in implementing the innovative outputs-based disbursement service delivery model at the satellite clinics, as USTH’s was mostly related with service delivery at the hospital level. USTH also successfully implemented certain cost containment measures at their clinics after the MTR. Since post-MTR agreements, USTH was fully engaged in the project, and by closing of the project had provided safe delivery services to 15,612 women of which 1,172 were complicated. A total of 10,316 (66%) of these women availed of four antenatal visits, and 10,407 (66.7%) had one postnatal visit. USTH’s Performance is rated as Satisfactory. 94. SGH – The unsatisfactory performance, and doubts of its commitment to the project was highlighted by the Bank team from August 2009 onwards30, despite SGH having been an IFC client. SGH seemed to be the most dissatisfied with the pricing of the SD package. At MTR, SGH was operating one clinic and had provided delivery services to 163 women. At the close of MTR, SGH expressed a renewed interest to participate, and the Bank team agreed on time-bound actions which included opening of more clinics; however in January 2011 it conveyed to the Bank that it would not be opening any new clinics. The Bank even agreed that SGH would continue operating only one clinic and that their end-project target would be reduced to 2,000. While the formal withdrawal of SGH from the project was in May 2012 (second restructuring),31 it had stopped providing services since May 2011 when its sole clinic was closed. However SGH delayed the formal agreement because of their reluctance to formally withdraw from the project. Till then, SGH had provided safe delivery services to 529 women of which 56(10.6%) were complicated. A total of 207 (39%) of these women availed of four antenatal visits, and 204 (39%) had one postnatal visit. While SGH’s withdrawal was attributed largely to the financial crisis, based on the assessment, SGH’s Performance is rated as Highly Unsatisfactory. Justification of Rating for Overall Borrower Performance: 95. Despite the highly unsatisfactory performance of SGH, given the fact that the project was able to reach its revised end- target of reaching eligible poor women with safe maternal care services despite many implementation challenges, significantly because of the commitment of SOUL and USTH, the Overall Borrower Performance is rated as Moderately Satisfactory.                                                                30  In 2009 when SHG had not yet established its first satellite clinic and it was providing services to eligible beneficiaries through an outpatient clinic in its hospital, it was working with less than minimum staff which increased its reputational risk thereby negatively affecting enrollment rates. SGH was allowed a transitional period up to May 31, 2009 to continue to operate through this outpatient clinic to provide services only for the already enrolled cases, and it was agreed that enrolled new cases would be referred to the new clinic, once established and fully operational.   31  The TTL informed the ICR that SGH took a long time to sign the legal papers that marked their withdrawal from the project. 25  6. Lessons Learned   Ensuring a Standard of Practice for the World Bank Group: 96. This project has an important lesson demonstrating the importance of a clear understanding and agreement of guidelines, requirements and its implications for project preparation and implementation by different members of the Group. A better clarity and realism on the applicability of GPOBA policies and procedures to Bank projects is critical. Some of the guidelines, e.g., inability to pre-finance service providers, and the implications of working with an IFC client later caused considerable implementation delays, including that GPOBA’s principle of copayment by the beneficiaries may have excluded some of the poor. Establish a more thorough understanding of the country context, capacity and requirements of all project participants during project identification and continue to promote this understanding with sustained collaboration throughout implementation: 97. While the project design was no doubt innovative for the Yemen health sector and benefitted from this pilot, with hindsight it could have benefited from a much more robust assessment of readiness of all parties, and, given the innovative project approach and fragile context, from more intensive collaboration amongst all parties during preparation and implementation. Although the project clearly addressed needs identified in the CAS, Government was not involved throughout the project to the level envisioned at inception, yet it was assumed that they would be positively influenced by the experience. Non-governmental participants, specifically the private sector and NGO service providers, were consulted during project design and inception; however, it is unclear that the nature and extent of discussions were sufficient for full understanding of the requirements involved in such an innovative program and within the fragility context. This lack of clarity was reported to have diminished the original enthusiasm and interest of the providers. Introducing innovations through smaller pilots with regular reviews, particularly in a conflict- fragile country: 98. When introducing innovations, which are no doubt needed to influence desired development outcomes, smaller pilots with regular reviews may be more effective. The project may have benefitted had it retained the original project design of a pilot of six months to one year followed by a critical review, and then gradually expanded with additional service providers and more government participation based on lessons from the pilot and any adjustments needed to account for the increasingly fragile context. This is particularly important in this instance where the project attempted a number of simultaneous innovations, including influencing change in health seeking behaviors that have deep cultural roots, introducing an output based approach to service delivery and inviting collaboration between the government and the non-governmental sector, both for-profit and not-for-profit, with an intent to influence public policy, all against the context of escalating fragility. Clear targeting of the poor and opportunities for cross subsidization: 99. The PDO included the distinct goal focused on providing quality maternal care to a targeted group of poor women. The project used both geographic targeting and a variety of 26  marketing methods to reach this goal. Poverty and fragility levels continued to increase during the project period, yet it is not fully clear the extent to which this affected the project, as this requires deeper examination beyond the scope of this report. 100. The project design envisioned that satellite clinics would eventually operate as profit centers in that future phases would see participation by the non-poor (specifically through company and industry health programs) at differentiated pricing levels. This non-poor participation would generate additional revenues earmarked for a sustainability fund, a mechanism used successfully in other regions. For the complex contextual reasons presented throughout, this vision was not realized; however, SOUL plans to adopt this sustainability fund approach. 101. There was evidence of sustained demand for these services, as reported by SOUL, both from the poor and, more surprisingly, from non-poor individuals (not through companies), who tried to enroll in the last two years of the project. There appears to have been no project mechanism for capitalizing on the desired participation of non-poor individuals, which may have provided an alternative source of some cross subsidization and project strengthening. Complexity vs Need for a Robust M&E System: 102. There is no doubt that all projects require M&E; that robust M&E is a core concept for successful performance based projects; and, that independent verification is required because the payment of grant subsidies is based on achievement of individual outputs. There is an important distinction to be made between a robust system and a complex system. The set of performance indicators, monitoring and tracking indicators and performance bands were all reported; however, heavy reporting requirements (all of which translate to increased costs for all parties), data inconsistencies and misinterpretation raise the issue whether a simpler system would have been more useful, provided for more straightforward evaluation and ensured better oversight. Additionally, the absolute dependence on the IVE reports may merit an adjustment for future similar projects. A more structured and concerted engagement with Government, the Ministry of Health (required) and also the Ministry of Finance (desirable), is required for any project that wishes to effectively influence public policy, especially one that envisions growth using government and donor mobilized resources, as in this instance. 103. Involvement could be facilitated by exposure to similar experience and technical dialogue with counterparts from other countries, as well as close monitoring of the evolving country context, and within that context, the changing priorities for government. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners a) Borrower/Implementing Agencies: A completion report by SOUL is available and can be accessed separately. b) Co-financiers: Not applicable. c) Other Partners and stakeholders: Not applicable. 27  REPUBLIC OF YEMEN Safe Motherhood Voucher Program Annex 1. Project Costs and Financing a) Project Cost by Component (in USD Million equivalent) Prior to the restructuring: Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Project Preparation and Start-up 0.049440 0.047159 95.39% Establishment of Satellite Clinics 0.466200 0.16968324 36.40% Service Delivery 5.403960 2.54810569 47.15% Community Outreach Administrative 0.356600 0.55276798 155.01% Education and Awareness Campaigns 0.189000 0.23750902 125.67% Beneficiaries contributions 0.612000 0.24531032 40.08% Total Financing Required 7.077200 3.800535 53.7% After restructuring: Appraisal Actual/Latest Percentage of Components Estimate (USD Estimate (USD Appraisal millions) millions) Project Preparation and Start-up 0.049440 0.047159 95.39% Establishment of Satellite Clinics 0.169690 0.16968324 100.00% Service Delivery 2.815480 2.54810569 90.50% Community Outreach Administrative 0.557740 0.55276798 99.11% Education and Awareness Campaigns 0.319840 0.23750902 74.26% Beneficiaries contributions 0.612000 0.24531033 40.08% Total Financing Required 4.52419 3.800535 84%                                                              32 As per the FTR 33  As per the FTR  28  b) Financing   Actual/Latest Source of Appraisal Estimate Percentage of Type of Co-financing Estimate Funds (USD millions) Appraisal (USD millions) WB Grant 6.232100 3.56 62.9% (GPOBA) 41.4% Beneficiaries Copayment 0.612000 0.245310 40% Service Contribution to 0.233100 0.169690 72.8% Providers Construction of SCs Total 7.07 4.34 61.3 %     29  REPUBLIC OF YEMEN Safe Motherhood Voucher Program Annex 2. Final Disbursement Table     Category Original Amount of Actual Percentage Grant the Grant Disbursement of Allocation Allocated Expenditures after to be (US$) restructuring Financed after (US$) restructuring (A) Project Preparation and Start- 49440 49440 47,159 95.39% up - SOUL (B) Community Outreach 356600 557740 552,767.98 99.11% Administrative Cost-SOUL (C) Education and Awareness 189000 319840 237,509.02 74.26% Campaigns-SOUL (D) Establishment of Satellite Clinics D-1 USTH-AL MAWARID 116550 150260 150,258.24 100.00% D-2 SYHC 116550 19430 19,425 99.97% (E) Service Delivery Subsidies 5403960 2815480 2,548,105.69 90.50% Total 6232100 3,912,190 3,555,224.93 90.88% 30  REPUBLIC OF YEMEN Safe Motherhood Voucher Program Annex 3. Outputs by Components A. Pricing of Safe Motherhood Service Delivery Package Appraisal Survey by IFC - Summary of costs from five Yemeni hospitals (US$) Saudi German Mother Azal Hospital German UST Hospital Yemen Hospital Hospital Hospital Antenatal Care 54 57 64 64 34 Delivery 155 180 149 180 68 Family Planning 18 29 20 20 12 Basic Package 227 266 233 264 114 C-Section 300 750 390 450 298 1. According to the Concept Note, a preliminary local costing market research was conducted to determine an indicative average cost per patient of the defined package of services that could be used as a basis to launch the pilot. The average cost amounted to US$160 ($120 for antenatal care + delivery + family planning, $31 built-in subsidy to allow for a maximum of 20% caesarian sections, $5 built-in subsidy for other diseases during pregnancy, and $4 for the Voucher Management Unit fees (original design was voucher-based). The Panel of Experts at the Concept Stage endorsed a package price of US$150 that was negotiated and agreed with the service providers. While the package price was discounted to current quoted prices, it is deemed fair given that the bulk of services would be delivered at the community and not hospital level. A minimal user fee of US$15 would be contributed by each targeted women, which would be paid in month 8 or 9 of the pregnancy, prior to delivery. The US$15 was determined based on the current practice of Yemeni families giving US$10-US$20 on the day of delivery to whoever helped the pregnant woman deliver. The costs of treatment and medication in the package price including an assumed estimated 15% of patients requiring caesarian sections/emergency care. This price was also compared to local and international peer pricing. While the package price agreed by the two 31  providers was within the range of pricing of other private Yemeni providers, the quality (services, staff, facilities, equipment) and capacity of the two providers’ current private hospitals far exceeded that of the other local Yemeni private providers. Internationally, the package pricing was also comparable e.g. costs of similar quality services by similar quality providers in India range from US$350-1500. 2. Original: Service Delivery Package price - up to a maximum of US$150.00 – with GPOBA subsidy up to maximum of US$135.00 and Beneficiary Contribution of US$15.00. Revised Pricing of Safe Motherhood Service Delivery Package agreed at MTR34 Revised Price of Service Beneficiary GPOBA Subsidy Unit Cost (US$) Package Contribution (US$) (US$) Category A: Normal 15.00 120.00 135.00 Delivery Category B: Special 15.00 200.00 215.00 Pregnancies and Deliveries Category C: Cesarean 15.00 450.00 465.00 Section and Emergency Operations Optional: Neonatal Care in 125.00 incubator for unlimited number of days B. Environmental and Social Performance Requirements as per Grant Agreement 3. Throughout the duration of the project: 4. The Service Providers shall: (a) design, construct, operate, maintain and monitor the Project in compliance with the Government of Yemen and local requirements as well as the applicable IFC Performance Standards on Social & Environmental Sustainability), and the applicable Environmental, Health, and Safety General Guidelines (General Environmental, Health and Safety, and Health Care Facilities), and (b) provide AMR to the World Bank, as specified in section 6.3.6 and Annex F of the OM. The AMR information has to be collected in a manner consistent with applicable IFC Environment and Health and Safety Guidelines (which include Environmental and Health and Safety – General Guidelines (April 2007) and Environmental and Health and Safety Guidelines for Health Care Facilities (April 2007) and the requirements of the Country.                                                              34  The revised price list was developed based on: (i) study of the actual costs of the two Service Providers (USTH and SGH) and (ii) the market prices of equivalent Service Providers to ensure that these prices were competitive, economical, and developed in a transparent manner. The price list also provided incentives for efficiency to contain costs and ensure the continuation of Service Providers in the delivery of the package of services under the Project.  32  5. For each clinic, the Service Providers will ensure the following: a) each clinic has, prior to opening, prepared written procedures dealing with sanitation, infection control and medical waste handling, and has trained the staff in application of these procedures; and b) relevant aspects of the Environmental and Social Management Systems for the Service Providers (as detailed in the OM) will be applied in all aspects of the day-to-day operations of the clinics, including in particular: i. regular re-training of staff on sanitation, infection control and medical waste management; ii. regular internal auditing of environmental and social performance; and iii. Appropriate record-keeping and reporting on sanitation, infection control and medical waste management. C. Details of Targeted Districts and Clinics Established and their Operationalization No. District No. of Satellite Service Providers Clinics Planned Actual Planned Actual 1.   1 1 USTH Bani-Al Hareth SGH 2.   2 0 - Al-Thurah SGH 3.   2 1 SGH SGH Shua'ob 4.   1 0 SGH - Sana'a Old City 5.   2 1 USTH Ma'een USTH 6.   1 1 USTH Al-Safia USTH 7.   1 1 USTH USTH Azal 8.   0 1 - Al-Tahrir USTH 9.   0 1 - Al Wehda USTH 10.   1 0 USTH - Al-Sabeen   12 7 6 SGH – 6 USTH 1 SGH – 6 Total USTH 33  Operationalization of Satellite Clinics over the 3 Project Periods Service Period 135 (Sep 2008 – Mar Period 236 (Apr 2011 – May Period 3 (Jun 2012 – Feb Provider 2011) 2012) 2014) Planned Actual % Planned Actual % Planned Actual % USTH 6 6 100 6 2 33.3 6 3 50 SGH 6 1 16.6 Total 12 7 58.3 6 2 3.33 6 3 50                                                              35 Represents the period prior to the first amendment to the GA. 36 Represents the period prior after the first amendment till the second amendment to the GA.  34  D. Performance Indicators and Bands for Service Providers USTH output table Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Q19 Q20 Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Cumulative Total Indicator Sep 08-Jun 09 Oct-Dec 09 Oct-Dec 10 Oct-Dec 11 Oct-Dec 12 Oct-Dec 13 Jan-Mar 10 Jan-Mar 11 Jan-Mar 12 Jan-Mar 13 Apr-Jun 10 Apr-Jun 11 Apr-Jun 12 Apr-Jun 13 Jan-Feb 14 Jul-Sep 09 Jul-Sep 10 Jul-Sep 11 Jul-Sep 12 Jul-Sep 13 Unit Performance based indicators Actual 0 28 99 171 482 999 664 523 489 805 591 212 306 587 651 782 1146 1677 1608 252 12,072 % of births assisted by skilled Target 0 46 147 215 618 1343 1259 787 649 1125 797 281 342 705 806 922 1344 1927 1857 337 15,507 attendants (Target of >= 80%) % 0% 61% 67% 80% 78% 74% 53% 66% 75% 72% 74% 75% 89% 83% 81% 85% 85% 87% 87% 75% 78% % of women with potential or Actual 0 3 13 19 55 80 28 29 26 60 74 45 44 71 56 69 89 190 173 48 1,172 acute obstetric complication Target 0 46 147 215 618 1343 1259 787 649 1125 797 281 342 705 806 922 1344 1927 1857 337 15,507 referred to the hospital (Target % 0% 7% 9% 9% 9% 6% 2% 4% 4% 5% 9% 16% 13% 10% 7% 7% 7% 10% 9% 14% 8% of 5-15%) Actual 0 44 121 205 387 554 631 459 455 702 464 66 285 590 488 851 1141 1431 1105 337 10,316 % of women that complete Target 0 46 147 215 618 1343 1259 787 649 1125 797 281 342 705 806 922 1344 1927 1857 337 15,507 basic antenatal care visits (4 visits) (Target of >= 75%) % 0% 96% 82% 95% 63% 41% 50% 58% 70% 62% 58% 23% 83% 84% 61% 92% 85% 74% 60% 100% 67% Actual 0 36 107 213 475 816 602 580 414 533 376 157 239 447 451 716 1154 1545 1546 0 10,407 % of women that complete Target 0 46 147 215 618 1343 1259 787 649 1125 797 281 342 705 806 922 1344 1927 1857 337 15,507 basic postnatal care visits (1 visit) (Target of >= 75%) % 0% 78% 73% 99% 77% 61% 48% 74% 64% 47% 47% 56% 70% 63% 56% 78% 86% 80% 83% 0% 67% 35  USTH output table as per the restructuring periods Period 137 (Sep 2008 – Mar 2011) Period 238 (Apr 2011 – May 2012) Period 3 (Jun 2012 – Feb 2014) Indicator Target Achievement % Target Achievement % Target Achievement % 4415 2966 67.18 3899 2990 76.69 7898 6703 84.87 % of births assisted by skilled attendants (Target of >= 80%) 4415 227 5.14 3981 278 6.98 7898 696 8.81 % of women with potential or acute obstetric complication referred to the hospital (Target of 5-15%) % of women that complete basic 4415 2401 54.38 3194 1972 61.74 7898 5943 75.25 antenatal care visits (4 visits) (Target of >= 75%) % of women that complete basic 4415 2829 64.08 3194 1719 53.82 7898 5859 74.18 postnatal care visits (1 visit) (Target of >= 75%)                                                              37  Represents the period prior to the first amendment to the GA. 38 Represents the period prior after the first amendment till the second amendment to the GA.  36  SGH Output table Q2 July - Sept 2009 Q5 Apr-June 2010 Q9 Apr-June 2011 Q6 July-Sept 2010 Q4 Jan-Mar 2010 Q8 Jan-Mar 2011 Q3 Oct - Dec 2009 Q1 Sep 2008-June Q7 Oct-Dec 2010 Cumulative Total 2009 Indicator Unit Actual 0 9 24 25 27 5 81 38 1 210 Target 0 31 49 39 44 56 126 114 69 528 % of births assisted by skilled attendants ((Target of >= 80%) % 0% 29% 49% 64% 61% 9% 64% 33% 1% 39.77% Actual 0 2 5 6 16 2 15 10 0 56 % of women with potential or acute obstetric complication referred to the hospital Target 0 31 49 39 44 56 126 114 69 528 (Target of 515%) % 0% 6% 10% 15% 36% 4% 12% 9% 0% 10.61% Actual 0 13 11 21 17 36 36 54 19 207 % of women that complete basic antenatal care visits (4 visits) (Target of >= Target 0 31 49 39 44 56 126 114 69 528 75%) % 0% 42% 22% 54% 39% 64% 29% 47% 28% 39.20% Actual 0 6 22 32 31 6 66 40 1 204 % of women that complete basic postnatal care visits (1 visit) (Target of >= 75%) Target 0 31 49 39 44 56 126 114 69 528 % 0% 19% 45% 82% 70% 11% 52% 35% 1% 38.64% 37  SGH output table as per the restructuring periods Period 139 (Sep 2008 – Mar Period 240 (Apr 2011 – May Period 3 (Jun 2012 – Feb 2011) 2012) 2014) Indicator Target Achievement % Target Achievement % Target Achievement % 528 210 39.77 % of births assisted by skilled attendants (Target of >= 80%) % of women with potential or acute obstetric complication referred 528 56 10.61 to the hospital (Target of 5-15%) % of women that complete basic antenatal care visits (4 visits) 528 207 39.20 (Target of >= 75%) % of women that complete basic postnatal care visits (1 visit) 528 204 38.64 (Target of >= 75%) NB: the targets were calculated for the respective periods from the different updates and not the cumulative                                                              39 Represents the period prior to the first amendment to the GA. 40 Represents the period prior after the first amendment till the second amendment to the GA.  38  E. Performance Indicators and Bands for SOUL Table 1: Output Table (SOUL) SOUL Unit Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Q19 Q20 Sep 08-Jun 09 Jul 09-Sep 09 Project Total Jan-Mar 10 Jan-Mar 11 Jan-Mar 12 Jan-Mar 13 Apr-Jun 10 Apr-Jun 11 Apr-Jun 12 Apr-Jun 13 Oct-Dec 09 Oct-Dec 10 Oct-Dec 11 Oct-Dec 12 Oct-Dec 13 Jan-Feb 14 Jul-Sep 10 Jul-Sep 11 Jul-Sep 12 Jul-Sep 13 Indicator Intermediate Indicators Number of women enrolled in the project 406 245 575 968 1795 1231 907 1157 896 346 0 854 936 1051 1449 1979 1833 240 10 0 16,878 Performance based indicators Target 720 900 1080 1200 1736 2128 1500 2100 1504 NR 489 NR 840 NR 708 NR 768 NR 704 704 19,053 % of total women targeted per six months which are enrolled/registered in the project (Target of 75%) Actual 406 245 575 968 1795 1231 907 1157 896 NR 346 NR 1790 NR 1449 NR 1833 NR 10 10 16,888 % 56% 27% 53% 81% 103% 58% 60% 55% 60% N/A 71% N/A 213% N/A 205% N/A 239% N/A 1% 1% 89% % of total women as per the annexed # Reached 43 137 139 380 1080 NR NR NR NR NR 785 NR 1658 NR 1976 NR 1976 NR 1404 1404 23,429 table for the health education campaigns reached in sessions #Targeted 406 245 575 968 1795 NR NR NR NR NR 346 NR 1790 NR 1449 NR 1833 NR 10 10 16,888 (Target of 70%) 14040 14040 % 11% 56% 24% 39% 60% N/A N/A N/A N/A N/A 227% N/A 93% N/A 136% N/A 108% N/A % % 139% NR = Not required to report 39  SOUL output table as per the restructuring periods Period 141 (Sep 2008 – Mar 2011) Period 242 (Apr 2011 – May Period 3 (Jun 2012 – Feb 2014) 2012) Targ Indicator Target Achievement % Target Achievement % Achievement % et Number of women enrolled in the 7647 2388 31.22 1482 3032 204.59 5967 6562 109.97 project (until Dec 2011) % of total women targeted per six 11364 7284 64 2833 3032 107 4856 6572 135.3 months which are enrolled/registered in the project (Target of 75%) NB: the targets were calculated for the respective periods from the different updates and not the cumulative                                                              41  Represents the period prior to the first amendment to the GA. 42 Represents the period prior after the first amendment till the second amendment to the GA. 40  F. Payment Recommended for SOUL and Service Providers Based on Performance Targets Met   Year 2009 2010 2011 2012 2013 Entity 1st 2nd 1st 2nd 1st 2nd 1st 2nd 1st 2nd SOUL missing 1/5 missing missing missing missing missing 5/5 5/5 3/5 USTH missing 4/5 missing missing missing missing missing 5/7 7/7 6/7 Year 2009 2010 2011 2012 2013 WB Rating SOUL Moderately Moderately Moderately Moderately Satisfactory Satisfactory Satisfactory Unsatisfactory Satisfactory USTH Moderately Moderately Moderately Moderately Satisfactory Satisfactory Satisfactory Unsatisfactory Satisfactory   G. Project Funding Sources (per period)    Table 4: Project Funding Sources   Unit Unit cost Total Planned USD 6,232,100 GPOBA subsidy Actual 3,560,000 Planned USD 612,000 User contribution Actual 245,310 Planned USD 233,100 Private financing Actual 169,690 Planned USD 7,077,200 Total Actual 3,975,000           41    H. Education and Awareness Component – Planned and Actual  % of women in the target areas reached by health education Period Targeted Reached campaigns 2008+2009(a) 1,226 319 26% 2010 4,901 3,851 79% 2011 2,399 2,943 123% 2012 4,290 5,310 124% 2013+2014(b) 4,072 11,006 270% Total 16,888 23,429 139% I. Assessment of PDO Achievement   Original PDO Revised PDO Revised PDO (2.6 years from (1.2 years after (1.9 years after Overall start) original) first revision) 1 Rating U MU S 2 Rating value 2 3 5 0.41with 2.58 with Total disbursed 3 0.57/6.23 cumulative cumulative 3.56 US$million 0.98/6.23 3.56/3.92 Weight % (total disbursed/final 0.41/3.56%=11% 2.58/3.56%=73% 4 0.57/3.56%=16% 100% disbursed amount of US$million) Weigh value 3x0.11=0.33 5 2x0.16=0.32 5x.73=3.6 4.25 (2 X 4) 6 Final rating MS A. Note: HU (1); U (2); MU (3); MS (4); S (5); HS (6) 42  REPUBLIC OF YEMEN Safe Motherhood Voucher Program Annex 4. Bank Lending and Implementation and Support/Supervision Process (a) Task Team Members Responsibility/ Names Title Unit Specialty Lending/Grant Preparation Carmen Nonay TTL GPOBA Olaf Smulders STC GPOBA Samantha Naidoo IFC Investment Officer IFC Investment Officer Salah-Eddine Kandri Senior Investment Officer IFC Sr Investment Officer Mikael Sehul Mengesha Procurement Specialist WB Procurement Spec. Jamal Abdulla Abdulaziz Procurement Specialist WB Procurement Spec. Josephine Masanque FM Specialist WB FM Specialist Ai Chin Wee Senior Operations Officer WB Sr Operations Officer Meskerem Brhane Senior Social Development WB Sr. Social Development Specialist Specialist Ghada Youness Legal Counsel WB Legal Counsel Lars Johannes Peer Reviewer GPOBA Peer Reviewer Supervision/ICR Financial Management Moad M. Alrubaidi Financial Management Specialist MNAFM Specialist Samira Al Harithi Procurement Specialist MNAPR Procurement Specialist (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including travel No. of staff weeks and consultant costs) Lending 10.32 72,219.56 Supervision/ICR 55.98 818,300.88 Total: 66.30 890,520.44 43  REPUBLIC OF YEMEN Safe Motherhood Voucher Program Annex 5. Documents Made Available for the ICR 1. Operations Manual (August 2008). 2. Draft GPOBA Concept Note, January 2007. 3. GPOBA Commitment Document, June 2007. 4. GPOBA Grant Agreement June 11, 2008. 5. Grant Amendment Letters i. March 22, 2011, and ii. May 31 2012 with Restructuring Paper. iii. Revised Disbursement Letter Jan 17, 2013. 6. Aides Memoires and Management Letters (ML) i. AM dated March 30-31, 2010 with ML dated April 15, 2010; ii. MTR AM Sep 25-29 & Oct 16-27, 2010, no ML; iii. Post MTR Mission AM Jan 15-16, 2011, no ML; iv. AM dated October 14-16, 2012 with ML dated Nov. 10, 2012; v. AM (without Annexes) dated Nov 28-29 and Dec. 17-18, 2013, no ML. 7. BTORs – May 2-6 2009. 8. Semi-Annual Status Report GPOBA – Aug 2008, Feb 2009, Aug 2009, Feb 2010, Sep 2010, April 2011, Oct 2011, and Aug 2012 9. ISRs – (i) Seq 2 - Dec. 29, 2010; (ii) Seq. 3 – Feb. 20, 2011 (iii) Seq. 4 – Jun. 14, (iv) Seq. 5 - no date specified (possibly Nov. 14, 2011 as inferred from next ISR); (v) Seq. 6 - Jul. 10, 2012; (vi) Seq. 7 – Dec. 18, 2012; (vii) Seq. 8 - Sep. 23, 2013; (viii) Seq. 9 – Jun. 3, 2014. 10. IVE Reports (i) SOUL Quarterly Invoice Verification Reports - Q14, Q16, Q 16-17, Q 18&19; and one draft management letter for period Jan1 - June 30, 2012. (ii) USTH- Quarterly Invoice Verification Reports - Q 14, Q 16, Q 16-17, Q 18, Q 18&19, Q20. (iii) Final Term Review for SOUL, USTH and SGH - Sep 2008 - Feb 2014 11. Project Completion Report from SOUL October 2014. 12. Others: PAD Maternal and Newborn Voucher Project, December 2013 44  REPUBLIC OF YEMEN Safe Motherhood Voucher Program COUNTRY MAP 45