EFTA01121818.pdf

1 The opportunity and obligation to strengthen national immunization programs Felicity Cutts and Robin Biellik, Consultants to BMGF, December 2010 Summary and recommendations Vaccine-preventable disease (VPD) control is one of the operationally simplest and financially most cost-effective public health interventions available. Immunization only achieves its greatest potential to avoid illness and save lives, however, when it is extended to all corners of society. New, more expensive vaccines such as pneumococcal conjugate vaccine (PCV) and rotavirus vaccine (RV) are a means to prevent diseases which commonly kill children in impoverished communities. In communities with access to high-quality curative care for pneumonia and diarrhoea, however, mortality from these conditions may already be low. It is essential for these vaccines to reach the hard-to-reach groups. The magnitude of immunization's contribution to the achievement of Millenium Development Goal 4 (MDG4: reduce by two-thirds, between 1990 and 2015, the under- five mortality rate) depends upon our ability to achieve and sustain universal vaccination coverage in all countries and districts and to overcome geographic, political, socio-economic or cultural barriers to effective provision and use of health services. Despite considerable progress since the inception of the global Expanded Program on Immunization (EPI) in 1974, routine vaccination coverage during the past 5 years (2005-09) fell or remained stagnant at inadequate levels in 21 of the lowest-income countries. Many of the countries, and areas within countries with the highest numbers of unvaccinated or incompletely-vaccinated children, including Chad, Ethiopia, India, Indonesia, Nigeria and Pakistan, also have high under-five mortality rates. In some of these countries, governments do not recognize that coverage is low and thus lack commitment to improving it. A competently-managed, well-resourced and financially sustainable routine immunization (RI) program provides the platform upon which new vaccines and vaccination technologies can be introduced, elimination/eradication initiatives implemented and other essential interventions delivered successfully. Hence, it is essential for programs to enjoy appropriately skilled and deployed human resources, an uninterrupted flow of vaccines and injection supplies, and a logistics system with appropriately maintained and utilized controlled temperature chain and transport. This infrastructure must be complemented by timely and accurate coverage and adverse events monitoring, VPD surveillance and outbreak response, social mobilization and public and professional Information, Education and Communication (IEC). To achieve and sustain programmatic success, a solid base of political commitment, effective management and reliable financing is required. Countries with this solid foundation have reduced VPD morbidity and mortality to low levels and introduced new vaccines and technologies smoothly to further protect their populations. At the World Economic Forum in Davos in January 2010, Bill and Melinda Gates announced their unprecedented commitment to realizing the potential of vaccines with a $10 billion donation for a "Decade of Vaccines". At the same time, they and others continue their commitment to the Global Polio Eradication Initiative (GPEI). As long as Wild Polio Virus (WPV) transmission has not been interrupted worldwide, all polio-free countries and areas remain at risk of re-importation. From 2003 to 2009, the World Health Organization (WHO) recorded 133 WPV importation events in 29 previously polio-free countries, leading to 60 outbreaks and a total burden of 2193 polio cases. The risk of importation with subsequent spread was highest in countries immediately bordering endemic countries, and was also higher in countries with low coverage of routine immunization. There is thus both opportunity and obligation to improve countries' capacity to reduce mortality in communities at greatest need by strengthening their routine immunization programs. EFTA01121818  2 The Bill and Melinda Gates Foundation, recognizing the importance of a strong immunization platform to the achievement of polio eradication and realization of the potential impact of new vaccines, has embarked on a strategic planning process to further define its role in improving routine immunization services. To complement the investments previously made in this area, such as the Africa Routine Immunization System Essentials (ARISE) project being conducted by John Snow, Inc., the Foundation commissioned the authors to undertake a detailed review of vaccination coverage data, review the published and grey-literature evaluations of mechanisms to overcome barriers to raising coverage and improve the quality of routine service delivery, and undertake key informant interviews with experienced national EPI managers and senior staff serving with partner agencies in headquarters, regions and countries. Our review highlighted that many successful interventions have been documented over the years. Improving program management through a comprehensive, district-based approach such as "Reaching Every District" (RED), conducting door-to-door visits (channeling) to identify and refer eligible children for vaccination at nearby vaccination sites, using flexibility in vaccine delivery through outreach vaccination strategies, use of community health workers, and the deployment of mobile teams into geographic areas with difficult access have all been successful in specific settings, although their costs and cost-effectiveness are less well documented. In recent years, the conduct of media- enhanced Immunization Days or Weeks at national or regional levels ("periodic intensification of routine immunization") has become common, but the contribution of such efforts to increasing coverage has been poorly documented to date. There are many tools available to help program managers improve planning and monitoring. These include methodologies for assessing missed opportunities for vaccination, templates for effective micro-planning from districts up to the national level, modules for training health workers, checklists for supportive supervision, tools for assessing vaccine storage and handling, and guidelines for conducting coverage reviews and surveillance assessments, and they have been used with positive impact in many settings. Efforts to increase political commitment, accountability and financial sustainability in developing countries have also been initiated. The risk factors associated with unvaccinated and incompletely-vaccinated children have been documented extensively. With few exceptions, the reasons for chronic under-performance among certain EPI programs are understood and the tools are available to maximize vaccination coverage and the quality of service delivery, but what has been lacking in some countries is commitment, coordination and management. When national and district managers lack a solid foundation of management training, the plethora of tools may confuse rather than aid them, and lack of career development opportunities or bureaucratic obstacles to innovation reduce the motivation to use available tools. We argue that BMGF's Decade of Vaccines should kick-start investment in management capacity and set the foundation for decades of effective health care. With an adequate project time-frame, a generation of strong managers can be built, with lasting impact on delivery of vaccines and essential public health interventions. Monitoring, learning and evaluation (MLE) of programs to strengthen managerial capacity will help convince countries to take greater domestic responsibility for EPI financing and also to utilize international funding (such as funds for Health Systems Strengthening from the GAVI Alliance) more efficiently. Synergies can be achieved through alignment of our recommended actions with the BMGF-supported Optimize Project and a recent proposal for logistics support and training from WHO to BMGF, and also with the on-going transition of GPEI field staff to monitor and analyze key process indicators on routine immunization service delivery and provide supervision and other support where appropriate. With effective coordination, these projects could make a concerted effort to reverse the chronic under-performance of critical EPI programs. Our recommendations are divided broadly into two categories: demonstration of best practices in strengthening district health services management, and investing in future generations by improving EFTA01121819  3 education as an incentive to improve health care and a means to raise the standard of human resources for health. The table shows suggested timeframes, critical indicators and approximate budgets for these, and more detail of the activities under each category is proposed below. Table: Recommended activities and outputs to support routine immunization Activity & Outputs/critical indicators Approximate timeframe Budget estimates Senior In each site, national transport and vaccine supply chain assessments $5 million management conducted, action plan developed, improved utilization and efficiency ($1.25 million advisors for documented within 3 yrs and government posts for logistics and per site) MLE in 4 transport officers created and filled by year 5 strategic In each site, management tool used at district level and improved locations for coverage of hard-to-reach communities documented 5 years Mentoring of trainees and graduates from health management training program done (>10 per site over years 2-5) At least 2 operational research studies done in each site demonstrating cost-effective interventions to increase coverage Where posts are at country level, DTP3 and measles coverage in each country reaches 90% by year 3 and is sustained to year 5 and beyond Immunization is line item in national budget and amount increases over time National personnel develop successful grant proposals to international agencies for continued funding of immunization strengthening Health Program developed (yr 1-2) $15 million ($3 management million start-up; training Students attracted and trained (>30 in year 2, increasing to >100/yr by $1.5 million per program (5 year 5) year running years) costs to host Funding attracted to continue the program institutions and $1.5 million per year tuition fees years 2-5) Development Tool developed and piloted (yr 1) $600,000 and use of Senior managers demonstrate improved effectiveness of service delivery practical after using tool (yr 2) tools to Tool made available on intemet and included in short-term and long- improve term training courses (yr 3) management Operational Studies completed and results are used to improve program planning $10 million research and monitoring, and to shape grant proposals for continued funding of studies (4 cost-effective interventions and strategies years) Include Training modules developed or adapted and translated into multiple $1 million management languages (years 1-2) in pre-service Management training incorporated in basic training in sites where senior nurse and personnel located (year 2) doctor Nursing and medical associations recommend inclusion of management training (3 training in all countries (year 3) years) Investing in Improved teacher training techniques developed $100 million education (20 Improved learning techniques for young children developed and EFTA01121820  4 years) implemented Improved learning techniques for high school children developed and implemented Child-to-child IEC programs implemented Literacy rates increased Demonstration of sustained reductions in neonatal, infant and child mortality by improving educational status of women and communities. • A. Demonstration of best practices in strengthening district health services management. I. Funding ofsenior management advisors in 4 key strategic locationsfor Monitoring Learning and Evaluation ofefforts to improve planning, management and monitoring ofvaccination at country andregional levels (and eventually ofother health interventions). They willparticipate in activities 2-5 below. There are different potential mechanisms to fund staff (e.g. via an international agency such as WHO or UNICEF, or via a non-governmental organization (NGO)), each having potential advantages and disadvantages. In EMR and SEAR, secondment of staff to WHO regional offices with a mandate to focus on specific countries is likely to be effective and appropriate. In AFR, secondment of staff to country level (Ministry of Health, WHO or UNICEF country offices) may be more appropriate to ensure maximum field-based work and MLE. An experienced NGO could be contracted to arrange these secondments and to participate in the other activities below. Potential locations where we would suggest funding for staff for MLE include: • A country in eastern or southern Africa which has coverage of 70-80% but has the potential to do more, and where improved management and increased advocacy could make a big difference. Examples include Ethiopia, Kenya, Mozambique, Uganda and Zambia. • A country in western or central Africa which has had chronically low national coverage (but may have some better-performing districts). Examples include CAR (which is included in the CASE project), Guinea, Liberia, Niger. • SEAR regional office to focus particularly on India (including disseminating lessons to be learned in the Bihar project), Indonesia and Nepal. • EMR regional office to cover the 5 larger GAVI-eligible countries which have some of the greatest challenges of geography and security, and ongoing polio transmission or threat of importation, with particular focus on Pakistan. These personnel will be critical to the implementation of the other components below, and to analyzing, documenting and disseminating lessons learned. They will improve the monitoring of inputs, outputs and impact of programs in under-performing countries. They will demonstrate how to improve the functioning of inter-agency coordinating committees. They will advocate within-country and at regional level for increasing domestic financing for immunization, learning from lessons from the BMGF-supported project conducted by the Sabin Foundation. They will also liaise with WHO, Optimize and Transaid/Riders for Health to advocate for countries to budget for posts in logistics and transport management and to develop and recognize vocational training in these disciplines. 2. Development ofa distance-basedprogram ofhealth management training that combines distance- based theoretical learning with hands-on experience in the field and is tailor-madefor low income countries, leading to a diploma (for candidates with no prior university qualifications) or degree (for doctors or other graduates). EFTA01121821  5 Such a program will develop core skills in planning, budgeting and financial management and accounting, human resource management, logistics and transport management, monitoring and evaluation, communications, grant application and advocacy. It should be predominantly field-based, but with adequate theoretical background through a mixture of distance-based learning and short face- to-face courses. It should be modular, with modules being available as stand-alone training modules for a wider audience (e.g. recent nursing and medical school graduates) that may not yet be ready for the complete masters-level training. It should have optional tracks for practical experience in different areas (e.g. based on running vaccination programs; malaria, TB or HIV control programs, Maternal, neonatal and child health programs etc) so that a wide variety of candidates and of funding sources (e.g. GAVI or Global Fund HSS; British Council; bilateral agencies) will be attracted. It should learn from the experience and approaches used by other long-term field-based training programs such as the EPIVAC management training program for Francophone countries, and the Field Epidemiology and Laboratory Training Program (FELTP) approach to capacity-building. It should also benefit from lessons learned by Sabin Institute in advocacy with developing-country parliamentarians to achieve sustainable domestic financing and accountability for public health interventions including EPI. It should be linked to ongoing mentoring and interchange of experience between graduates, e.g. by linking with Technet, Afenet and other networks. The staff placed in key countries/inter-country posts should act as mentors for the program together with other in-country persons. The development of this program can be done through partnership of schools of public health with organizations working in low income countries and with WHO/UNICEF. Competitive bidding for finance from the Foundation to develop and start-up (say, 5 years initial support including scholarships for students) the program should be sought, and applicants should demonstrate a strategy for attracting other funding to continue the program after this period. 3. Harmonizing existing tools to develop field-friendly, practical tools for conducting situation analyses ("district mapping"), diagnosing the problems, planning and monitoring of essential health interventions at national and district levels. For this, BMGF could begin by hosting a convening of existing and recent grant-holders and key agencies. The primary aim of the convening is to develop a comprehensive situation assessment and planning tool, and to determine if further harmonization of monitoring tools is indicated and if so, how this should be done. A core group of experts would plan (including development of a draft comprehensive tool) and co-ordinate the convening. Participants could include the following: • Optimize (logistics and vaccine management tools and monitoring systems) • Transaid and Riders for Health: transport assessment and management tools • WHO IVB: comprehensive EPI program review tools, post-new vaccine introduction- evaluation tools, and missed opportunity surveys; experience in developing accreditation programs e.g. laboratories and National Regulatory Authorities (NRAs) • GPEI: community mapping; developing micro-plans; community involvement; tracking children; monitoring and surveillance • Centers for Disease Control, John Snow Inc. and WHO AFRO: tools for monitoring and evaluating RED implementation and using this to improve guidelines and strategies; ARISE project lessons learnt to date on drivers of strong RI programs in the African region • Emory University: Evaluating the knowledge, attitudes and practices of immunization providers in India • Vaccine Delivery Innovations Initiative — ethnographic methods to understand community perspective on barriers to immunization • PATH: approach to assessing the performance of organizational systems (current grant) and lessons learned from successful and cost-effective interventions to raise RI coverage and quality (past grants from the Children Vaccine Program). EFTA01121822  6 • Agence de Medecine Preventive: implementation of the EPIVAC training program in vaccine management, use of a training network to develop advocacy; experience to date in forming National Immunization Technical Advisory Groups and their role in improving RI • WHO and UNICEF Regional Advisors: lessons regarding utilization of tools at national and district level and how to ensure that the harmonized tool will be practicable to use. Once developed, the tool(s) will be piloted in countries where the key senior management personnel (recommendation I) are located, then further adapted and disseminated. They will also be used as part of the work experience of trainees in the program under (2). In addition, the convening will encourage translation of lessons learned into WHO policies and procedures, for example discussion with WHO IVB of the potential to develop an IVB accreditation program for transport management in a similar way to its accreditation of vaccine stores, NRAs, and laboratories. It will also offer an opportunity to forge relationships between the different agencies and potentially develop a consortium that can implement all five activities under category A. 4. Conduct operational research studies in countries where the senior management personnel are located, and where possible, involving trainees of the management training program and existing FELTP programs, to evaluate the costs and effects of using available tools and approaches to increasing vaccine coverage, including: a. Improved transport and vaccine supply chain management b. Improved situation analysis and micro-planning to reach hard-to-reach populations c. Improved monitoring of vaccine coverage at district level, including the assessment of new approaches such as serological surveys d. Door-to-door canvassing to identify and refer children eligible for vaccination and for other essential health interventions e. Methods to increase community demand for vaccination and improve accountability of health services to communities f. Different combinations of outreach, mobile teams or supplementary immunization activities, according to geographic and other characteristics. g. Evaluation of the impact of PIRIs on coverage Such studies should be co-ordinated with WHO IVB and regional offices to ensure maximum awareness and use of results for action. 5. Ensure that newly trained nurses and doctors have skills in management of immunization and other programs. Develop or adapt existing modules on modem theory and practice of EPI, including management of the vaccine supply chain and transport, for pre-service training of Medical Officers and Nurses (this component can be linked with activity (2)). Liaise with professional associations such as the proposed African Medical and Nursing School association, to ensure that this training becomes a standard part of medical and nursing schools' basic curricula. Possibility of an overall package of interventions run by a consortium of NGO-academia-WHO, identified by a request for proposals (RFP). Since all of the recommendations under (a) are best implemented by groups that combine the field experience of NGOs (local and international), the technical expertise of academia, and the public health leadership of WHO, to ensure maximum synergies between the recommendations they could all be part of an RFP for the formation of a consortium for strengthening national immunization programs. This could be the most efficient process for BMGF, instead of running multiple small projects. EFTA01121823  7 B. Investing in future generations — improvement of education as an incentivefor strong preventive health programs and a means to transform their effectiveness Low educational status of communities and especially of women is a major stumbling block for effective health care. Firstly, studies have consistently shown lower uptake of proven effective interventions including vaccination among families where the mother (and to a lesser but still significant extent, the father) is illiterate. Second, school-based, or "child-to-child" programs offer the potential for delivering a range of interventions to school-children and their siblings. Even in remote areas, there is often a school available in areas where there is no health facility and schools offer contact points for delivering health promotion. Third, new vaccines such as HPV and in future, HIV vaccine will need to be delivered to school-age children and improving the quality of schooling will encourage higher school attendance and facilitate school-based health delivery. Fourth, districts where educational achievement is low have few candidates for professional or vocational training schemes and do not attract good quality health care workers, as trained professionals prefer to live in areas where their own children will have access to education. They are thus stuck in a vicious cycle and do not benefit from advances in other parts of the country. Fifth, the quality of education is poor in many low income countries so that, while having a school-leavers certificate offers access to further training, it does not equate with having the ability to be innovative, self-learning, or problem-solving. Efforts to assist countries to attain MDG 2 (universal primary school education) mostly focus on increasing access to school. The quality of education, however, also needs great improvement. Modern technology and communications options are expanding rapidly, offering an opportunity to transform the learning experience even in difficult environments. BMGF has invested over $4 billion in schools and scholarships in the USA, from early learning (preschool) to college preparation, and shown that effective teaching is the most important school-based factor in student achievement. The experience from the USA should be translated into improving learning in poor countries. We propose that BMGF conduct a demonstration project in at least one country, with simultaneous investment in preventive health care (driven by investment in strengthening RI and bringing along other preventive interventions) and in school education (initially primary school then extending to secondary education), with a predicted lifespan of at least one generation. This length of investment may seem long but to put it in perspective, it is the length of time that Rotary and others have been supporting polio eradication. A program to develop innovative teaching techniques for rural areas should be supported, with training of current teachers and use of distance-based techniques, and adapted over time as technology advances. Health promotion messages would be an important part of the curriculum. Tools, techniques and lessons learned from the project would be disseminated over its course. Beginning investment today to transform the learning experience of school-entry children means that within ten years, a generation of literate school-children would enter their reproductive years able to access and understand information, and by then strengthening secondary education, some of these children will become the next generation of teachers. Within 20 years, the impact of high literacy and education on sustaining immunization coverage in the next generation can be measured. Additional aspects could be addressed and evaluated, such as giving incentives to mothers when their children complete the immunization series, for example paying the costs of school attendance through primary school. Since the project would mainly be investing in improving methods and quality, it should be self-sustaining as other sources of funds would be used for the "bricks and mortar" component and staff would be paid by government (but incentivised by access to improved tools and technology and mentoring by project staff). EFTA01121824  8 I he opportunity and obligation to strengthen national immunization programs Felicity Cutts and Robin Biellik, Consultants to BMGF, December 2010 Background information List of contents I. Introduction 2. Coverage among GAVI-eligible countries in 2009 and recent coverage trends 3. Current approaches to the delivery of vaccination 4. Who are the under-served and hard-to-reach? 5. Summary of literature reviews on the effectiveness of interventions to raise coverage 6. Logistics and management 7. Monitoring and operational research 8. Conclusions 9. References 10. List of persons interviewed II. Tables Table 1. Recent coverage trends in GAVI-eligible countries, and resources for health la. High (>80%) coverage >4 yrs lb. Medium (60-80%) coverage in 2005 and/or 2009 lc. Increasing coverage Id. Low (<60%) coverage Table 2. Main countries with internally displaced populations and/or people in refugee-like situations due to conflict, 2007-8 Table 3. Indicators to monitor immunization program performance Table 3. Advantages and disadvantages of methods to measure vaccination coverage 12. Figures Figure 1. Global immunization 1980-2009, DTP3 coverage Figure 2 (a-i): Coverage trends in countries in the RED evaluation Figure 3: Coverage trends in Bangladesh Figure 4: Coverage trends in Sudan EFTA01121825  9 I. Introduction Immunization is one of the most cost-effective interventions available to improve health. In GAVI- eligible countries, traditional vaccines against tuberculosis, diphtheria, tetanus, pertussis, polio and measles have the potential to save 4.5 million lives worldwide each year (WHO, 2010a) at an average cost of US$24 per fully vaccinated child (Wolfson et al 2008). This astounding potential will only be realized, however, if vaccines reach the communities at greatest risk of dying from vaccine- preventable infections. Infant and child mortality rates are highest in sub-Saharan Africa, where only two countries (Eritrea and Malawi) are on-track for Millenium Development Goal (MDG)4,' and 11 countries have made no progress in reducing child mortality since 1990 (Bhutta et al 2010). They are also high in impoverished communities of Asia and the Middle-East, where despite impressive overall progress, child mortality remains unacceptably high among residents of hard-to-reach and under-served communities. Studies have shown that certain groups are missed by all health interventions, so that while the majority of a population may benefit from powerful health technologies, those in greatest need receive none (Victora et al 2005). Although global immunization coverage is high (82% in 2009 according to WHO-UNICEF estimates of national immunization coverage (WUENIC) — Fig I), coverage has stagnated at ≤80% levels or fallen in 21 GAVI-eligible countries between 2005-9, often in the very areas with highest child mortality rates. Assumptions about the benefit of the introduction of new and underutilized vaccines (e.g. Sinha et al 2007) are over-optimistic unless national immunization programs are empowered to reach high-mortality communities. At the World Economic Forum in Davos in January 2010, Bill and Melinda Gates announced their unprecedented commitment to realizing the potential of vaccines with a $10 billion donation for a "Decade of Vaccines". At the same time, they and others continue their commitment to the achievement of polio eradication. The investment, and the skills being developed through the Global Polio Eradication Initiative (GPEI), offer renewed opportunities to improve countries' capacity to reduce mortality in communities at greatest need. To date, these potential opportunities have not been seized. New, more expensive vaccines such as pneumococcal conjugate vaccine (PCV) and rotavirus vaccine (RV) are a means to prevent diseases which commonly kill children in impoverished communities (Cutts et al 2005, Parashar et al 2009). In communities with access to high-quality curative care for pneumonia and diarrhoea, however, mortality from these conditions may already be low (Klugman et al 2003). It is essential for these vaccines to reach the hard-to-reach groups. GAVI Alliance supports low income countries to introduce more expensive vaccines under the expectation that countries will take responsibility for purchasing these vaccines when GAVI support ends in 2015. There is thus an obligation to ensure that countries accelerate efforts to reach the most under-served communities in a systematic and sustainable fashion. The GPEI has recognized the importance of strong routine immunization programs. In settings where the national OPV3 coverage rate is >80%, indigenous polioviruses are more rapidly interrupted, there is a statistically lower risk of having a polio outbreak following a wild poliovirus importation, and there appears to be a lower risk of both the emergence and spread of circulating vaccine-derived On track defined as mortality rate in children younger than 5 years of less than 40 deaths per 1000 live births, or less than 39 deaths per 1000 live births plus average annual rate of reduction (AARR) higher than 3-9%; insufficient defined as under-5 mortality rate greater than 29 deaths per 1000 live births plus AARR between 0.9% and 4.0%; no progress defined as under-5 mortality rate greater than 29 deaths per 1000 live births plus AARR lower than 1.0%. Ethiopia, Mozambique, Niger come close to an AARR of 3.9% but still have high mortality (104, 130, 167 per 1000 live births, respectively) EFTA01121826  10 polioviruses (cVDPVs) (WHO 2010b). In addition, it is anticipated that during the final stages of global eradication, some countries will switch to using inactivated polio vaccine (IPV) alone or in combination with OPV, to obviate some of the problems associated with OPV including low vaccine effectiveness in certain settings and the risk of vaccine-associated polio. IPV will likely be delivered through RI services, possibly as a combined hexavalent or heptavalent vaccine. Since IPV alone does not provide indirect protection to contacts, very high coverage is needed to sustain population immunity to polio. For that, strong national immunization programs must be built and sustained. A national immunization program needs to develop appropriate policies, select appropriate strategies to implement those policies, monitor and evaluate implementation, and modify policies and strategies according to the findings. At global and regional levels, inter-agency coordination committees (ICCs) and national immunization technical advisory groups (NITAGs) assist governments to formulate policies, in part with support from BMGF. Having appropriate policy is a good first step but action is also needed to ensure that those policies are disseminated, accepted, and followed. Recognizing the importance of a strong immunization platform, the BMGF has embarked on a strategic planning process to further define the role of the foundation in improving routine immunization services. To assist in this effort, BMGF asked the authors to help analyze options for foundation investment in this area. This work will complement the investments previously made in this area, such as the Africa Routine Immunization System Essentials (ARISE) project which is examining factors contributing to strong immunization programs in the African region and is being conducted by John Snow, Inc. We reviewed data on trends in immunization coverage since 2000 among GAVI-eligible countries as reported by governments to the World Health Organization (WHO) and the WUENIC estimates of coverage. We gathered information on current strategies and approaches to organising routine immunization and risk factors for low vaccine coverage from the published and grey literature, regional and country plans of action and progress reports, and presentations by national EPI managers and regional and country staff of WHO, UNICEF, PATH and the U.S. Centers for Disease Control (CDC) to international and regional WHO immunization meetings. Interviews were conducted with immunization staff at WHO headquarters, with WHO regional staff and country EPI managers in SEAR and EMR, and with CDC. Structured questionnaires were used in these interviews, to determine past experience and current policies and strategies being used to strengthen routine immunization in low income countries. This was complemented by discussions on specific topics, e.g. transport for health, vaccine supply chain management, with experts in these areas. To avoid duplication of effort with the ARISE project, only two EPI managers in the African region were interviewed (Ghana and Kenya), and results of stakeholder interviews conducted by the ARISE project were reviewed rather than re-interviewing the same experts. 2. Coverage among GAVI-eligible countries in 2009 and recent coverage trends Table 1 shows trends for DTP3 coverage among GAVI-eligible countries since 2000, according to WUENIC estimates for 2009 (data downloaded 5 August 2010). DTP3 coverage (often taken as an indicator of utilization of vaccination services) is compared to DTP1 coverage (an indicator of access to health services) and the dropout between the two vaccines, expressed as a percentage of DTP1 coverage, is shown. Countries are arbitrarily classified into four groups: those with high (>80%) and relatively stable coverage since 2005; those with medium coverage (60-80%) in 2005 and/or 2009; those with a marked increase in coverage between 2005 and 2009 (some of which began the increase prior to 2005); and those with ongoing low coverage. Less than one-third of GAVI-eligible countries have had stable coverage levels for at least 4 years that are close to or above the GIVS target of 90% for DTP3 and MCV1 (MCV1 data not tabulated). In AMR, EUR and WPR, most countries have high coverage, exceptions being Haiti (low) in AMR; EFTA01121827  11 Azerbaijan (medium) in EUR, and Lao PDR and Papua New Guinea (low) and the small islands of Kiribati and the Solomons (medium) in WPR. The birth cohorts in countries in AMR, EUR and WPR that have not yet sustained high coverage are very small. There are several countries in AFR and EMR with low or medium coverage, however, and of these Nigeria in AFR and Sudan in EMR have the highest birth cohorts; Chad and Somalia while having much smaller total populations also have large numbers (>200,000 each) of unvaccinated children. In SEAR, while most countries have high coverage and only Timor-Leste has low coverage, among those with medium coverage are India and Indonesia with very large birth cohorts, and therefore the highest numbers of unvaccinated and under- vaccinated children are in SEAR. Countries with higher DTP3 coverage tend to spend more on health, as shown by total health expenditure (THE) per capita, though with some exceptions. For example in AFR, Burundi has very low THE but high coverage, whereas Equatorial Guinea has very high THE but low coverage. THE in other regions is generally much higher than in AFR although it is surprisingly low in Myanmar in SEAR and in Pakistan (EMR) which have high coverage. Several countries (Afghanistan, Angola, Congo, Democratic Republic of Congo (DRC), Ethiopia, Madagascar, Myanmar, Nepal, Niger and the Sudan) have greatly increased coverage in the last 4-8 years despite low overall THE, including countries classified by GAVI as fragile and containing substantial populations that are internally displaced due to conflict (Table 2). Some, but not all of these countries have received substantial official development assistance for child health, including GAVI Immunization Services Support (ISS) funds. Most of the 9 countries with low coverage for DTP3 and low or medium coverage for DTP1 over the last 4 years or more are classified by GAVI as "fragile" or "poorest" states. The exceptions are Papua New Guinea and Nigeria which are lower middle income countries. In most countries with medium (60-80%) DTP3 coverage in 2005 and/or 2009, DTP1 coverage is high and the main problem is high dropout between these vaccines. This is also the case in some countries with low or increasing coverage. This suggests that health service delivery factors may play a large role in the failure of children to complete the vaccination series. Other major health service delivery factors recognized as important include vaccine stock-outs (see section 6), to which for example a decline in coverage in Kenya in 2008 was attributed (http://www.who.int/immunization_monitoring/data/ken.pdf, accessed 6 August 2010). Some countries continue to have large differences between government reports of coverage and WUENIC estimates, and as shown in the table, this appears particularly so in countries with low or medium coverage (e.g., 22-52 percentage points difference in the two sources in the 5 African countries with low coverage). This is potentially a major problem because if governments are unwilling to recognize that coverage is low, they are unlikely to design or fund interventions to increase coverage. Data for India are not shown because the government official coverage data for 2009 are not yet available but in 2008, India reported DTP3 coverage 18 percentage points higher than the WUENIC estimates. Most of the countries which have markedly increased coverage in recent years show good concordance between government reports and WUENIC best estimates (Table 1), though with important caveats. The WUENIC process can only lead to improved estimates if alternative sources of data to administrative reports are available for comparison, and/or if there have been audits of the quality of the administrative data. In countries affected by conflict, often no national population- based surveys are conducted and hence the WUENIC process has little way to verify data. For example, in Afghanistan, the detailed country report (http://www.who.int/immunization monitorinWdata/afg.pdf, accessed 6 Aug 2010) shows that no survey data have been used since 1999, as the one survey conducted in 2006 excluded children without vaccination cards and hence data were not considered valid. Therefore, the WUENIC EFTA01121828  12 estimates are based on the coverage reported by the government. Similarly, among those shown in the table as increasing coverage, WUENIC have recommended that nationally representative, high-quality surveys be conducted in Angola and Sudan due to uncertainties about the reported data and lack of recent surveys in those countries. 3. Current approaches to the delivery of vaccination Vaccination may be delivered at a variety of sites, using strategies that range from fixed sites to outreach, to mobile teams, to campaigns (Box 1). Box 1: contact points and delivery strategies for vaccination Fixed sites: health facilities (HF), e.g. hospitals, health centers, health posts: usually vaccinate at least one day a week; frequency depends on catchment population size Outreach sites: visited regularly (weekly, monthly, or less) by health staff, usually from the nearest fixed site who return the same day. Vaccination may be done at a health post, school, community building, or community representatives' house. Mobile teams: teams travel and stay at least one night in remote locations. May administer vaccines at outreach-type sites and/or do door-to-door vaccination. Campaigns: mass immunization activities e.g. polio national immunization days (NIDs); supplementary immunization activities (SIAs) for measles, tetanus toxoid, yellow fever, etc. Vaccination done during a short period of time via multiple teams that vaccinate at community and/or household levels. Often with expanded age ranges for eligibility. To strengthen routine immunization services, WHO recommends that countries use the Reaching Every District (RED) approach2, a bottom-up approach to district-based planning of immunization and other health services based on data, which was launched in the African region in 2002 (http://www.who.int/immunization delivery/systems policy/red/en/index.htmp). Box 2: The RED approach to organizing vaccination and other health services Initially developed for use in districts with low immunization coverage and subsequently promoted for nationwide use. Intended as a broad-based approach to planning and delivering routine immunization, with 5 main components: • Re-establishment of outreach services (now "reaching target populations" in AFRO) • Supportive supervision • Enhancing community links with service delivery • Monitoring and use of data for action • Improved planning and management of resources, with community involvement 2Microplanning for immunization service delivery using the Reaching Every District (RED) strategy. WHO/IVB/09.11 World Health Organization, Geneva 2009 EFTA01121829  13 In 2007, WHO, UNICEF, CDC and IMMbasics reviewed progress toward improving immunization services and coverage in 9 countries (Benin, Cameroon, Democratic Republic of the Congo (DRC), Ethiopia, Ghana, Madagascar, Sierra Leone, Togo and Uganda) which had implemented RED for at least 24 months (Box 3). The evaluation showed substantial progress in implementing RED, but an uneven emphasis on its different components and in general, an over-emphasis on the extension of outreach services without sufficient attention to strengthening the planning, management and monitoring of services. In most of the 9 countries, coverage had increased, but cause and effect could not be attributed (Figs 2a-2i). RED had been scaled up using multiple funding sources, including GAVI discretionary funding in 8/9 countries evaluated, and there were concerns about sustainability if Immunization Services Support (ISS) is discontinued. Following the evaluation, the revised AFRO guide (WHO 2007) incorporates "best practices" and "lessons learned", and gives more emphasis to planning, management and the use of monitoring data, and to reaching all target populations using a mix of sites, not just outreach. In Sudan, RED implementation is considered a success story, with capacity building for improved microplanning, monitoring and evaluation, and investment in the cold chain and transport. Within northern Sudan, the degree of coverage increased correlated with the score for the level of implementation of RED