DFID’s Contribution to the Reduction of Child Mortality in Kenya

Contents

Executive Summary page 1

1 Introduction page 2

2 Findings page 8

Objectives page 8

Delivery page 11

Impact page 16

Learning page 22

3 Conclusions and Recommendations page 25

Annex page 28

Abbreviations page 44

The Independent Commission for Aid Impact (ICAI) is the independent body responsible for scrutinising UK aid. We focus on maximising the effectiveness of the UK aid budget for intended beneficiaries and on delivering value for money for UK taxpayers. We carry out independent reviews of aid programmes and of issues affecting the delivery of UK aid. We publish transparent, impartial and objective reports to provide evidence and clear recommendations to support UK Government decision-making and to strengthen the accountability of the aid programme. Our reports are written to be accessible to a general readership and we use a simple ‘traffic light’ system to report our judgement on each programme or topic we review.

Green: The programme performs well overall against ICAI’s criteria for effectiveness and value for money. Some improvements are needed.

Green-Amber: The programme performs relatively well overall against ICAI’s criteria for effectiveness and value for money. Improvements should be made.

Amber-Red: The programme performs relatively poorly overall against ICAI’s criteria for effectiveness and value for money. Significant improvements should be made.

Red: The programme performs poorly overall against ICAI’s criteria for effectiveness and value for money. Immediate and major changes need to be made.

Executive Summary

This review assesses DFID’s contribution to the reduction of under-five child mortality in Kenya, where donors spend more than £500 million each year on the health sector. We looked at DFID’s work in three dimensions: its influence on progress by the international community in reducing under-five mortality; direct aid programmes totalling £163 million focussed on malaria and health systems strengthening; and its funding of immunisation and bed nets, both directly and through the GAVI Alliance (GAVI) and The Global Fund to Fight AIDS, Tuberculosis and Malaria (the Global Fund).

Overall Assessment: Green-Amber

Reducing under-five child mortality is a global priority and has seen remarkable progress. Under-five mortality has fallen by 24% in Kenya since 1990 but this is less than global and regional averages and the level in Kenya remains high. DFID has played a significant leadership role globally, particularly on malaria. DFID’s work in Kenya is largely coherent; its bilateral programmes have achieved impact in a challenging context and improved immunisation and the provision of bed nets have saved lives. There are, however, significant weaknesses and regional disparities in basic health systems. DFID should focus more on neonatal mortality, sustainability, health systems strengthening and hard-to-reach areas.

Objectives Assessment: Green-Amber

Reducing under-five mortality is a priority for DFID in Kenya. DFID’s health work is based on clear international guidance and is aligned with the Government of Kenya’s priorities and other donor work. DFID has promoted effective donor co-ordination in the health sector and has supported GAVI and the Global Fund to deliver high-priority interventions. DFID has not, however, used beneficiary feedback well in project design and health systems strengthening work is not effectively co-ordinated amongst donors. DFID is just beginning to leverage the private sector more in the health sector.

Delivery Assessment: Amber-Red

There are significant weaknesses with both bilateral and multilateral modes of delivery for health services in Kenya. Beneficiaries report variable quality of services. DFID’s decision not to finance the Government of Kenya directly, due to corruption, is right but has costs. It has led to parallel delivery systems which are effective but not sustainable. It is inconsistent that DFID continues to support multilateral institutions which work through government systems. DFID has robust risk frameworks. Devolution within Kenya is a risk for delivery but offers opportunities for county strategies to benefit poor people.

Impact Assessment: Green-Amber

We found evidence of substantial impact from four of the five DFID programmes which we reviewed. DFID has played a leadership role in preventing and treating malaria, which has reduced under-five mortality in Kenya. Under-five deaths and illnesses have also been reduced through immunisation. DFID has invested heavily in basic supplies and should do more on strengthening health systems for sustainability. In the long term, this will require DFID to finance the Government of Kenya directly but it does not have clear criteria for this. DFID should have a greater focus on hard-to-reach areas and neonatal mortality. Beneficiaries have also highlighted the importance of access and quality of care issues. Monitoring and evaluation should focus more on impact.

Learning Assessment: Green-Amber

DFID has played a key role in supporting global research and learning on under-five mortality. DFID has adapted its Kenya programmes to reflect learning and to support innovation by delivery partners. New vaccines and improved bed nets have been introduced. There have been missed opportunities for learning in terms of neonatal mortality and feedback from intended beneficiaries in project design. Learning is also hampered by the poor quality of data in Kenya. Devolution in Kenya and new international targets are creating fresh learning challenges and opportunities for DFID in the future.

Recommendations

Recommendation 1: DFID centrally should specify its policy on equity more clearly and DFID in Kenya should focus systematically on the quality of – and access to – maternal, neonatal and child health services for remote and hard-to-reach populations. DFID should routinely use beneficiary feedback in its programme design.

Recommendation 2: DFID should develop a clear exit strategy for funding basic supplies in Kenya (such as bed nets) and focus instead on achieving a long-term and co-ordinated approach amongst development and financing agencies, including GAVI and the Global Fund, for health systems strengthening. DFID should develop criteria for resuming direct financing to the Government of Kenya.

Recommendation 3: DFID should engage with emerging county government structures in Kenya to mitigate the risks and to expand the opportunities of devolution for health outcomes and to help develop information systems and financial management tools that will rapidly identify and address any negative impacts.

1 Introduction

There has been substantial progress globally and some progress in Kenya in reducing under-five child mortality

Reducing under-five mortality is a global priority and has seen remarkable progress since 1990

1.1 The fourth Millennium Development Goal’s (MDG4) target is to reduce the under-five mortality rate by two thirds globally between 1990 and 2015. Although the target will be missed, MDG4 has galvanised a 47% reduction in under-five mortality worldwide since 1990, from 90 to 48 for every 1,000 live births (see Figure 1). The annual number of under-five deaths has fallen from 12.6 million to 6.6 million, meaning that 17,000 fewer children died each day in 2012 than in 1990.¹

1.2 There are large differences in under-five mortality both between and within countries. These are driven, in part, by levels of poverty, access to services, mothers’ education and geographical remoteness. The average under-five mortality rate for sub-Saharan Africa is 98 for every 1,000 live births and in 2012 the region accounted for 50% of all under-five deaths globally. Less progress has been made in reducing deaths which occur within the first month of life (neonatal period). The share of neonatal deaths amongst total deaths for under-fives globally has increased from 37% in 1990 to 44% in 2012. In sub-Saharan Africa, about a third of all deaths occur in the neonatal period. Stillbirths are also a major concern worldwide.²

Kenya has made some progress but has lagged behind regional and global averages

1.3 As the sixth largest country by population in sub-Saharan Africa, Kenya is important for the achievement of global under-five mortality targets. Kenya achieved a 24% reduction in under-five mortality between 1990 and 2008, from 98 to 74 for every 1,000 live births.³ Of these, 52 for every 1,000 live births were infants under one year old, including 31 for every 1,000 live births of babies were under one month old.⁴ Kenya’s under-five mortality rate rose in the 1990s, while it was falling amongst its neighbours and in sub-Saharan Africa as a whole. Reasons for this include malaria drug resistance, falling levels of women’s education and immunisation, reduced use of maternity services, increased malnutrition, inter-ethnic violence and the spread of HIV.⁵ These trends, however, have been reversed with a 34% reduction in under-five mortality in the decade prior to 2008.

1.4 Authoritative figures for Kenya are not available after 2008 due to a delay in the major health survey. The latest estimated under-five mortality figures (for 2012) show a further slight decline in under-five mortality to 73 for every 1,000 live births.⁶ Neonatal mortality in Kenya has remained fairly constant (see Figure 1 on page 2), as have stillbirth rates.⁷

Kenya exemplifies important issues for the reduction of child mortality

1.5 Kenya was chosen for the review because of the varying progress over time that has been achieved in reducing under-five mortality, the substantial regional disparities that remain and the large-scale donor programmes that are not dominated by budget support. This is all in the context of a country which has been heavily aided over many decades.

1.6 Childhood deaths in Kenya are most common among poor people.⁸ Children in the lowest wealth quintile are 44% more likely to die before the age of five than those in the highest quintile. The richest families are three times more likely to receive key child survival interventions than the poorest families. The worst outcomes are in rural areas and informal urban settlements. This goes to the heart of DFID’s focus on the poorest in society.

The Kenyan context is challenging for DFID

1.7 DFID has been working under difficult political and operational conditions in Kenya. Challenges to working in Kenya have included corruption, post-election violence in 2007-08 and current uncertainty about the indictment of the President and Vice President by the International Criminal Court and the Westgate shopping mall attack in September 2013. There is a timeline of events in Annex A1.

1.8 There has been substantial restructuring of the Government of Kenya. The Ministry of Health⁹ was divided into two in 2008 and reunified in 2013. A process of devolution of responsibility for health from central government and eight provinces to 47 counties is under way. This has been rushed through and has created substantial challenges for planning, budgeting and delivery of services that threaten the achievement of future health targets.

1.9 Following the 2007 election and a major fraud¹⁰ in the education sector which came to light in 2009, DFID decided not to finance the Government of Kenya directly because of concerns about corruption. This meant that health programmes, which were planned to be implemented on a sector-wide basis with budget support, had to be converted into a complex and diverse set of projects.

Our approach reviews three different dimensions of DFID’s work to reduce under-five mortality in Kenya

1.10 The approach and methodology of the review are summarised in Annex A3. DFID contributes to the achievement of MDG4 through research, policy influence, aid programmes and multilateral contributions. These are reflected in the three dimensions of our review which are as follows:

• DFID’s influence in the international system and the overall performance of the health system in Kenya;
• DFID’s bilateral programme; and
• the provision of bed nets and immunisation, including work funded by GAVI and the Global Fund. This allowed us to consider the particular role of multilateral channels – as distinct from bilateral channels – and to draw some comparisons between the two in the case of bed nets.

1.11 The key questions that the review has set out to answer, linked to these three dimensions, are:

• How coherent are DFID’s contributions to global initiatives to reduce under-five mortality and how well is DFID leveraging this expertise across the Kenyan health system as a whole?
• What has been the effectiveness and sustainable impact for intended beneficiaries of DFID Kenya’s programmes to reduce under-five mortality and to what extent are DFID’s plans for the future based on learning?
• What influence has DFID had in enhancing the impact and effectiveness of multilateral agencies, specifically GAVI and the Global Fund, in reducing under-five mortality and how well are these initiatives linked to health systems strengthening activities (by which we mean changes in policy and practice, leading to better health through improvements in access to and quality of health care) in Kenya?

DFID’s multilateral contributions to health in Kenya

1.12 International development agencies are currently committing over £500 million per year¹¹ to support health programmes in Kenya. This is only slightly below the Government of Kenya’s own budget for health, which comprises 5.7% of its overall budget for 2013-14 and is on a declining trend. The scale of donor funding in the sector risks creating distortions in the Government of Kenya’s commitment to health spending. DFID is the third largest donor in health in Kenya. The United States provides more than 60% of donor funding and spends more than 85% of its money on HIV work. Current bilateral and multilateral contributions to the Kenyan health sector are summarised in Figure 2.

1.13 DFID also makes indirect contributions to the health sector in Kenya through its central funding of multilateral organisations. These amounts are lower than the size of the bilateral programme. They include:

• the Global Fund, where DFID has contributed an estimated 10% of funding for the 2011-13 period (equivalent to £7.7 million in 2013-14); and
• GAVI, where DFID has contributed a 33% share of funding for the period 2011-15 (equivalent to £7.1 million in 2013-14).

1.14 DFID has also contributed to strengthening the United Nations Children’s Fund (UNICEF) and the World Health Organisation (WHO) in Kenya by funding key staff and programmes from its bilateral funds. WHO, in turn, advises the Government of Kenya and other development agencies.

DFID’s bilateral aid programmes in Kenya

1.15 In 2013-14, the share of DFID’s health support to Kenya was planned to be £37 million, 26% of the total DFID aid programme to Kenya but, due to delays, this is now forecast to be £21 million.¹³ DFID promotes a multi-sectoral approach to addressing the causes of under-five mortality because of the many factors that affect child health. DFID does not collect data on what percentage of the programme is specifically being spent on addressing under-five mortality. Programmes on livelihoods, nutrition, HIV and reproductive health, funded by DFID centrally and through DFID in Kenya, all contribute to reducing child mortality in Kenya (illustrated in Annex A4(b)).

1.16 Although DFID acknowledges the importance of water and sanitation in reducing under-five mortality, it is not funding this work through its bilateral programmes in Kenya.¹⁴ Other development partners, including Germany, have strong bilateral programmes in this area. DFID does centrally fund the World Bank Water and Sanitation Program, which has six international policy areas, four of which include Kenya as a focus country (water and sanitation sector reform, rural sanitation, improved urban services and support to the private sector).

1.17 Humanitarian and HIV-related work were not included in this review because these have been the focus of previous ICAI reviews. Similarly, nutrition was not included, as it will be addressed in a forthcoming ICAI review. We decided to focus, instead, on five core health programmes, totalling £163 million, which address malaria and support to immunisation and ‘health systems strengthening’.¹⁵ These programmes have been running since 1999.

1.18 The programmes that we focussed on were:

• Malaria Control in Kenya: this programme was implemented through WHO¹⁶ and supported the development of the National Malaria Strategy and subsequent systems strengthening. Two technical posts were funded at WHO in Kenya including one malaria expert and one health systems expert to strengthen the Department of Malaria Control and Ministry of Health more broadly. Support was provided for improved malaria treatment, insecticide residual spraying, policy development, co-ordination, disease surveillance, research, Global Fund proposal development and budgeting capacity for the Ministry of Health;
• Social Marketing¹⁷ of Insecticide-Treated Nets: this programme was implemented through Population Services International (PSI), a US-based international not-for-profit health organisation. Bed nets were sold in public health facilities and in rural shops at heavily subsidised prices (known as ‘social marketing’). In 2005 this changed, with the distribution of free nets at public health facilities for pregnant women attending antenatal clinics and for mothers of children undergoing routine immunisation;
• Essential Health Services: this programme was funded through Liverpool Associates in Tropical Health,¹⁸ in partnership with Liverpool Voluntary Counselling and Testing, Kenya, Nuffield Centre for International Health and Development (UK) and Health Unlimited, Kenya. The programme worked at the national level and in Nyanza (one of the most disadvantaged provinces in Kenya) to support the delivery of essential health services. This included planning, co-ordination and national-level support for maternal and neonatal health and integrated care for childhood illnesses, such as pneumonia, diarrhoea and immunisations, primarily at the provincial level;
• Kenya Health Programme: this programme was implemented through a number of partners. Malaria control is funded mainly through Population Services International (PSI) and WHO with smaller inputs from The MENTOR Initiative.¹⁹ Research is delivered through non-government organisations (NGOs) and academic partners, including Family Care International. Health systems strengthening is delivered through WHO.²⁰ Elements of this broad-based programme have provided free distribution of bed nets to pregnant women and for children under five, as well as malaria prophylaxis for pregnant women and efforts to improve malaria diagnosis and treatment. It continues to support WHO’s technical assistance for the development and implementation of the National Malaria Strategy and broader support to health systems strengthening. This included development of the national health strategy, a human resource strategy and support to health planning for newly devolved counties;
• Support for Indoor Residual Spraying:²¹ this programme is funded through the US Agency for International Development (USAID) and the Kenya Medical Research Institute which works with the US Centre for Disease Control and Prevention. This programme focusses on Western Kenya, where malaria is highly endemic. Insecticide spraying of homes and institutions is planned twice a year, together with research to track effectiveness of the intervention. The spraying is yet to start because of a lack of approval for the proposed insecticide; and
• Reducing Maternal and Newborn Deaths: we looked at this new programme to assess how the design had incorporated learning from previous programmes.²² This programme will be implemented by a number of partners, including UNICEF Kenya and the Liverpool School of Tropical Medicine. The planned budget for 2013–18 is £75 million. The aim of the programme is to reduce maternal and newborn deaths in Kenya by increasing access to – and uptake of – quality maternal health care.

1.19 The timing, duration and scale of these programmes are illustrated in Figure 3.

1.20 DFID has selected a wide range of effective partners to deliver its bilateral health programmes, based on their relative strengths in Kenya and has ensured synergy in delivery between programmes.²³ PSI has a strong track record of delivering social marketing initiatives in Kenya and this programme has been complemented by strategic support to malaria control delivered through WHO. Support to strengthen the technical capacity of WHO has not only benefitted DFID-funded initiatives but has also delivered improved technical support for other health programmes and partners in Kenya.

1.21 In its choice of delivery partners, DFID has also demonstrated a strong synergy with delivery of its own centrally funded initiatives. For example, in the new indoor residual spraying programme, DFID Kenya has partnered with the Kenyan Medical Research Institute to deliver the research aspects of this programme. This institute also receives support from DFID centrally and produces research which has national, regional and global influence. In the newly approved maternal and newborn health programme, DFID Kenya builds on DFID’s success elsewhere in working with the Maternal and Newborn Health Unit at the Liverpool School of Tropical Medicine as a delivery partner.

DFID’s contributions through GAVI and the Global Fund

1.22 The review allowed us to examine GAVI and the Global Fund, two multilateral funding institutions which are funded in part by DFID. These are often described as ‘vertical funds,’ because they support results-focussed interventions in specific areas. Both organisations have a low-cost centralised model with country and regional teams based in Geneva. GAVI has no country presence and works through Government of Kenya systems. The Global Fund also works through the Government of Kenya and has local structures for oversight and financial accountability.

1.23 GAVI was founded in 2000 with the purpose of funding immunisation and reinvigorating the vaccine market. It has been very successful in both. By 2012, it had funded vaccinations for over 440 million children worldwide and estimated that 6 million deaths had been averted. It has successfully introduced new vaccines to a range of countries – including Kenya – and has driven a substantial reduction in the price of vaccines. It has an annual budget of approximately US$1 billion.²⁵

1.24 Between 2008 and 2013, GAVI provided a total of £120 million in support to Kenya for vaccines (£114 million) and health systems strengthening (£6 million). GAVI has supported the national introduction of pentavalent (2001) and pneumococcal (2010) vaccines. The Government of Kenya co-finances the vaccine introductions and delivery systems to help foster sustainability. GAVI has committed over £200 million to its Kenya programme for the period 2001-16. Of this, 94% is spent on vaccines. The UK has pioneered innovative financing mechanisms for GAVI, such as the International Finance Facility for Immunisation.²⁶

1.25 The Global Fund was founded in 2002. It is an international financing organisation operated as a public-private partnership. It is the largest global funder of programmes to combat malaria and had distributed 340 million bed nets globally by 2013.²⁷ It has an annual budget of about US$3.5 billion, of which approximately 30% is spent on malaria. The Global Fund has committed £135 million to Kenya in malaria grants since 2006, of which £128 million has been disbursed. The planned allocation for the period 2014-17 for malaria grants to Kenya is £56 million.

Our methodology

1.26 Our methodology is summarised in Annex A3. Our review included a comprehensive literature review, an expert panel and interviews with DFID and other health and development professionals in Kenya, Switzerland and the UK. The review team visited GAVI and the Global Fund headquarters in Geneva in October 2013 to interview their staff.

1.27 The review team visited Kenya in November 2013 and went to regions that had both high and low levels of child mortality. The locations visited were Nairobi (including informal settlements and suburbs), Kisumu (Western Kenya on the shores of Lake Victoria, with high malarial incidence) and Nyeri (near Mount Kenya, where respiratory illness is more common). At each location we:

• held discussions with women who were either mothers or the carer of at least one child under five;
• interviewed health providers and politicians;
• visited health centres and district and provincial hospitals; and
• held interviews in Nairobi with civil servants and representatives of multilateral organisations, NGOs and bilateral development agencies.

1.28 A beneficiary survey was conducted as part of the review. Focus group discussions were held with a total of 80 mothers or carers of children under-five in five separate locations. These covered the regions of high and low child mortality to understand intended beneficiaries’ experience of services and their role in the design and improvement of programmes. Areas selected for the survey benefitted from DFID expenditure in reducing under-five mortality and maternal health.

1.29 The review focussed mainly on programmes active during the period 2008-13. We also took a longer-term perspective to assess impact, trends in mortality and the way in which DFID had used learning from previous programmes.

1.30 Ratings for each of the DFID programmes which we reviewed and for the work of GAVI and the Global Fund are given in Annex A2. The overall ratings in the report draw from these elements but also include DFID’s wider contribution to reducing under-five mortality in Kenya; including its strategic decisions, centrally-funded programmes and wider multilateral influence. The evidence to support each of the overall ratings is given in the relevant part of the Findings section.

2 Findings

Objectives

Assessment: Green-Amber

Health is a key priority for DFID globally and in Kenya and reducing under-five child mortality is important within this priority

DFID follows global best practice and focusses on national priorities

2.1 DFID is clear in its focus on reducing under-five mortality globally as one of the MDGs. It has played a key role in the development of global best practice for reducing under-five mortality and has applied this learning in its Kenya programme. DFID has well-researched evidence and policy papers, including on malaria and child health, which inform national country programmes. DFID globally has prioritised health systems strengthening.

2.2 In selecting health programmes, DFID Kenya considers need, health priorities and its own strengths and resources. DFID recognises under-five mortality as an important issue to address in Kenya. DFID places emphasis on addressing critical gaps in the provision of health services. Interventions with particularly low coverage by governments and donors on a global basis include family planning, antenatal care, skilled attendance at delivery, malaria prevention, children sleeping under insecticide-treated bed nets (ITN) and malaria treatment. These have a direct impact on reducing under-five mortality and are all DFID priorities in Kenya.

2.3 The theory of change for reducing under-five mortality is summarised in a range of guidance about priority interventions. The international consensus on what is required to reduce under-five mortality includes addressing:

• inequities in access and quality of healthcare;²⁸
• poor governance;²⁹ and
• weak health systems.³⁰

2.4 DFID seeks to address these factors in Kenya by working in some of the remotest regions and areas of greatest need, for example in Western Kenya where malaria is highly endemic. DFID’s actions in response to Government of Kenya fraud demonstrate its strong commitment to improving governance. DFID’s cessation of direct financing to the Government of Kenya, however, has limited its ability to work directly on governance reform or to strengthen government health systems.

2.5 These overarching approaches are complemented by a series of proven high-impact interventions that have been developed by the international community to address under-five mortality. The approach was summarised in a set of child health papers, published in the Lancet in 2003.³¹ DFID, the Global Fund and GAVI implement interventions from within this list and have made good choices in identifying priorities that reflect their strengths and comparative advantage. DFID programme designs in Kenya are generally strong and consistent with global theories of change, as evidenced by external reviews.³²³³³⁴

DFID allocates substantial resources to its health work in Kenya

2.6 Health as a whole takes the largest sectoral share in the current DFID Operational Plan for Kenya. It is not possible to track specifically, however, how much DFID is contributing towards reducing under-five mortality. Child mortality reduction is not a stated priority for DFID in Kenya but many of its programmes contribute to this goal. DFID’s current operational plan in Kenya for 2011-15 is committed to providing bed nets, maternal health and family planning services to avert 7,000 deaths from malaria and 6,000 maternal deaths over the current planning period.³⁵

2.7 Over the past five years, the share of health spending within the DFID Kenya bilateral programme has decreased. It has gone from a high of 47% in 2009 (£28 million) to 27% in 2012 (£25 million) as the overall size of the programme has increased. The share of health in DFID’s overall Kenya portfolio is illustrated in Annex 4(a). During the current operational plan for 2011-15, spending on malaria comprises 35% of DFID health spending while reproductive, maternal and newborn health is 31%.

Intended beneficiaries view public service provision as being well integrated for services such as immunisation

2.8 Intended beneficiaries we met had accessed Government of Kenya services, including local clinics and referral hospitals. Services, such as immunisation, were well integrated into core health systems and generally work well. Awareness of the role of donors in underpinning this system was low, although we did see USAID and DFID branding on bed nets.

2.9 A clear priority of women from our beneficiary surveys was to have facilities located in their community, where access is easier and they expect to be treated with more respect. One woman commented: ‘I would like that the health centres…should be provided with enough equipment…because many people go there but keep being transferred to district hospitals’.

DFID’s approach to reducing under-five mortality is reasonably coherent

DFID’s programmes are reasonably coherent and complement Government of Kenya and multilateral programmes

2.10 The Government of Kenya has a strong country child survival strategy.³⁶ This is based on an analysis of the major causes of under-five mortality in Kenya, illustrated in Figure 4. Immunisation programmes have been developed to respond to the high prevalence of pneumonia and diarrhoea. Although malaria is responsible for a small share of overall deaths, national data underestimate the importance of malaria for specific regions where DFID operates.

2.11 DFID has a reasonably coherent set of programmes of work in the health sector, which recognise the work of other donors and support Government of Kenya priorities. The programmes we reviewed are mutually reinforcing and complementary, with good links to the related areas of family planning, nutrition and livelihoods. Each sub-sector has a donor group to co-ordinate activities. DFID does not fund routine vaccination directly through its bilateral programme but it does through its core contributions to UN agencies and through GAVI, which plans and distributes vaccines.³⁷ DFID covers routine distribution of bed nets to mothers through health centres, while the Global Fund supports the mass distribution of nets. Overlap and duplication have been avoided.

2.12 There is sometimes a risk that vertical programmes distort local health priorities. We did not find evidence that the vertical nature of programmes to support vaccination and bed nets in Kenya was leading to excessive financial allocations to these high-priority interventions. Many of our interviewees argued that this was the case for HIV, however, where large-scale donor funding was compared with low Government of Kenya allocations to health issues, such as non-communicable diseases, which are not priorities for donors.

DFID has substantial positive influence in enhancing the coherence of the international system and donor co-ordination

2.13 DFID plays a significant global governance role within multilateral organisations. It is an active member of the boards of both GAVI and the Global Fund. It has promoted policy and reforms in line with global best practice. Staff in these organisations recognised the influence and professionalism of DFID’s work. A global DFID priority for GAVI in the coming planning period, for example, is to address regional disparities in vaccine coverage and equity. This is highly relevant for Kenya because of the large regional variations in immunisation coverage.

2.14 The DFID Multilateral Aid Review (MAR)³⁸ has had a substantial impact on the multilateral organisations which it assessed. Both GAVI and Global Fund staff told us that they view its recommendations and the six-monthly review very seriously and take action in response to DFID’s recommendations. We saw a DFID Kenya input for the review of WHO for the MAR, which demonstrated high-quality analysis and was a good example of co-ordination between the central DFID departments working with the multilaterals and a country office. Such inputs strengthen DFID’s case for reform in these institutions.

2.15 DFID chaired the Global Fund Board through its process of reform between 2011 and 2013. Reform included a new strategy, revised governance arrangements and a significant strengthening of senior management, including the appointment of a new Executive Director and Inspector General. The process has led to radical changes to address weaknesses in implementation and supervision of Global Fund programmes.

2.16 DFID is part of the team that represents international development agencies within the Global Fund’s Country Coordination Mechanisms in Kenya. DFID has helped to promote some reform of these processes since 2010. DFID has provided funds through WHO and UNICEF to help the Government of Kenya draft successful proposals to both GAVI and the Global Fund. The latter followed a significant period during which Kenya had not been successful in winning Global Fund grants for malaria.

2.17 The overall co-ordination mechanisms in the health sector, which involve both the Government of Kenya and development agencies, vary in their effectiveness. The high-level Health Sector Coordinating Committee for the Government of Kenya and development agencies has not met since 2011. There are seven sectoral co-ordination committees which were assessed by interviewees as being generally effective, covering healthcare financing, HIV and AIDS, malaria, child health, nutrition, reproductive health and human resources for health. DFID has played an active role in the development community in driving co-ordination between donors and is asked to chair many of the key bodies. Development Partners in Health Kenya, for which DFID took over the chair in December 2013, has an active secretariat and facilitates good co-ordination amongst donors.

There are some areas of weakness in coherence

2.18 There is a weakness in the coherence of health systems strengthening work. There is an overall Government of Kenya strategy for the health sector.³⁹ Health systems strengthening work is funded by a range of donors, including GAVI and the Global Fund – but without a clear division of labour between development partners. This is particularly a problem for these donors which do not have a local presence to allow them to co-ordinate effectively with other initiatives in the sector.

2.19 A broader weakness in the coherence of the work of international organisations is demonstrated by the proliferation of new global initiatives for reducing under-five mortality, all of which affect Kenya. One respondent admitted that ‘each international meeting seems to lead to the development of another global strategy’. Approaches, such as Every Woman Every Child, The Integrated Global Action Plan for Pneumonia and Diarrhoea and A Promise Renewed, are useful for focussing international attention on under-five mortality but can distract resources from implementation priorities and require additional country-level reporting. We also saw evidence of a lack of coherence of donor-funded training at a clinic where 13 out of 15 staff were attending training courses elsewhere when we visited.

DFID’s impact on health systems strengthening is limited by it not directly financing the Government of Kenya and by its substantial spending on commodities

2.20 DFID’s ability to apply its learning and expertise in full is limited by it not directly financing the Government of Kenya. This has reduced its ability to be fully engaged in health systems strengthening and to address governance issues. Central co-ordination of donor inputs to health systems strengthening in Kenya is not being undertaken, largely because of the lack of a sector-wide programme. This is a significant gap.

2.21 DFID programmes include substantial spending on commodities, including bed nets and medicines. ‘Commodities’ is a term that we use in this report to refer to consumable items, primarily malaria drugs and bed nets that are purchased as part of the programme. A large programme helps to give DFID credibility with the Government of Kenya and other donors. This represents spending, however, that could have been undertaken by any agency and does not maximise the potential for value to be added by DFID expertise. This money could have been spent on other, more strategic health systems strengthening work where DFID has an acknowledged global leadership role and a large portfolio of programmes worldwide. For the social marketing of bed nets programme, £50 million (over 70% of total funding) was spent on bed nets.⁴⁰ For the Kenya Health Programme, over £21 million has been spent on bed nets to date.⁴¹

DFID should strengthen its design process, including feedback from intended beneficiaries

2.22 DFID’s design and business case processes are cumbersome, leading to the slow development of new programmes. The latest programme, Reducing Maternal and Newborn Deaths, has taken over one year to reach approval stage. Project design does not include enough flexibility for DFID Kenya to respond to rapidly changing circumstances.

2.23 Intended beneficiaries are not routinely consulted by DFID on their priorities when undertaking programme design. This increases the risk that interventions will not address their concerns. Women whom we met were not aware of any communication channels to the Government of Kenya or to development agencies. Feedback mechanisms, including suggestion boxes and user committees, were not working in the places we visited. DFID does, however, collect substantial beneficiary feedback from its monitoring and review processes, which could be used more systematically to understand the needs and priorities of potential beneficiaries. More inputs from beneficiaries in project design would help DFID to identify gaps and opportunities for improved interventions.

Delivery

Assessment: Amber-Red

DFID has chosen effective delivery partners and strengthened its engagement with the private sector

2.24 When DFID decided not to finance the Government of Kenya, this inevitably meant that delivery would not be optimum in terms of DFID’s preferred option of supporting country systems directly. DFID generally applies its expertise in health systems strengthening through working closely with governments and backing this up with financial support. As DFID no longer directly finances the Government of Kenya, this has created challenges in how best to optimise support for Kenya’s health system. DFID has had to promote health systems strengthening through indirect means and to choose a range of other partners to deliver benefits to intended beneficiaries. It has selected multilateral, bilateral, NGO and not-for-profit partners who have complementary skills and have collectively delivered an effective bilateral programme in difficult circumstances. DFID Kenya has also sought to promote health systems strengthening indirectly. An example is work on the development of the national health strategy where DFID funded WHO to take on a leadership role in health systems strengthening.

2.25 A substantial proportion of health care in Kenya is delivered outside the government system. Faith-based health suppliers are estimated to provide 40% of health services in Kenya.⁴² Many poor people remain outside health systems altogether. This increases the importance of outreach strategies. Private sector facilities, including small-scale retail outlets, are often the first point of contact for poor people in relation to the health system. DFID’s support of PSI’s work with the Tunza Family Health Network of clinics, one of which we visited, is an important initiative to strengthen the quality of care in the private sector (see Figure 5). DFID is not, as yet, however, working with and through the private sector in a comprehensive way.

There are significant weaknesses with all modes of delivery for health services in Kenya

2.26 The focus on immunisation and bed nets has allowed us to compare the DFID Kenya bilateral programmes, which work outside the Government of Kenya, with those of GAVI and the Global Fund, which are implemented through Government of Kenya systems. There are problems with both. The DFID approach involves duplication and lacks sustainability but has been generally well delivered. The multilateral agencies face higher levels of risk of fraud and suffer from the delays and inefficiencies of government systems.

2.27 There is an inconsistency in DFID’s approach of supporting multilateral organisations, such as the Global Fund, that provide direct finance to the Government of Kenya, when DFID is not willing to do this through its bilateral programme. The high risks of fraud faced by multilateral organisations affect DFID’s money spent through these agencies. DFID’s oversight is dependent on the effective functioning of the governance arrangements and policies of these institutions. DFID does not systematically seek to strengthen that oversight for countries where it has concerns that government systems are weak.

DFID and other development agencies’ decision not to finance the Government of Kenya directly is right but it has costs

2.28 Due to instances of corruption over the past five years, DFID and most other bilateral development agencies have made the right decision to work outside Government of Kenya systems. About three quarters of donor support to the health sector in Kenya does not go through the Government’s budget.⁴⁴ Establishing a parallel system, however, is almost certain to lead to an increase in aggregate costs because of the duplication of fixed costs and a loss of the economies that are realised by larger scale operations. This also leads to reduced donor influence and a lack of sustainability. It makes delivery in general – and health systems strengthening work in particular – more difficult. Parallel delivery systems have been effective but there is no clear exit strategy. There is no plan for DFID to resume direct financing to the Government of Kenya at this stage.

2.29 DFID’s options to address these challenges are limited. The Kenya programme may have been bigger in the absence of concerns over governance and accountability. In 2012-13, Kenya received a much lower allocation of DFID bilateral funding than its neighbour, Tanzania (which has a similar population and only slightly lower income levels). Kenya, nonetheless, has received an increasing amount of UK aid over the past two years and it has risen from a ranking, in 2012-13, of 15 to one of 13 on the list of DFID’s top 20 recipients.⁴⁵

2.30 DFID’s impact and influence depend critically on the quality and quantity of its in-country health professionals. The existing team, which includes two health professionals, is widely respected but heavily stretched. There are additional high-priority tasks in donor co-ordination, engaging with county-level governments and beneficiary interaction that could be undertaken if capacity were increased. Another donor strongly contended in discussions with the review team that DFID should increase this capacity. Additional capacity will also be needed if DFID undertakes more work on health systems strengthening.

Beneficiary and health providers’ feedback on health service delivery in their areas was mixed

2.31 Beneficiaries generally had a good understanding of services available to them and the benefits of using these services. Specifically, they:

• had good knowledge of antenatal care, immunisations and the benefits of sleeping under a bed net;
• confirmed receiving free bed nets at their antenatal clinics and free nets for their baby at immunisation sessions, with some also reporting free malaria prophylaxis;
• reported a number of difficulties in accessing health services, including transport, high user fees and lack of staff, especially during the night; and
• reported a lack of availability of some drugs, as well as being asked to purchase drugs outside health facilities and at high prices, when these drugs should have been freely available.

2.32 Health providers reported good availability of bed nets in malarial areas but a lack of some essential drugs. There were also reports of a lack of adequate staff, equipment and medical supplies, as well as a lack of appropriate training for some staff. These issues are monitored through routine Government of Kenya reporting systems.

GAVI’s model has advantages over the Global Fund in Kenya but both have weaknesses

2.33 Both GAVI and the Global Fund respond to country-led strategies and deliver through Government of Kenya systems. GAVI primarily supplies physical vaccines, whereas the Global Fund gives grants and thus requires more intensive supervision. GAVI has created an effective distribution system for vaccines, which ensures that they are kept refrigerated as required. We saw little evidence of vaccines or bed nets being out of stock. This compares with other medicines, especially high-value ones, which were often not available at the facilities that we visited.

2.34 The Global Fund’s allocations were based in the review period on a competitive bidding process, which meant that outcomes were uncertain and planning was difficult. Kenya was unsuccessful in its bids for malaria funding in the bidding rounds between 2006 and 2010. Money was primarily spent during the period 2008 to 2013 from the 2006 grants. This will be addressed in future by the Global Fund’s ‘New Funding Model,’ which provides greater predictability of funding for recipient countries, based on disease burden and income levels.

2.35 The Global Fund’s own monitoring system judged the performance of the main grant for malaria during this period as unsatisfactory. Delivery of the Global Fund’s support has been costly and inefficient during the review period, although people whom we interviewed said that the quality of the portfolio is now improving. The DFID 2011 MAR noted that the Global Fund requires countries to devote substantial resources to management, co-ordination and impact assessment for its grants and that its own requirements often take precedence over national priorities.⁴⁶

2.36 An independent donor review concluded that the delays for implementation of Global Fund grants in Kenya typically ranged between 6 and 24 months. Financial management issues included audit delays and accounting anomalies, as well as conflicting policies.⁴⁷ The Global Fund Board does not generally address country-specific issues at this level. DFID follows up by highlighting issues to the Country Team but this process is not systematic and depends on DFID having capacity and a role in Global Fund governance at the country level.

2.37 GAVI has had problems with the introduction of the rotavirus vaccine. This was identified as a priority on the basis of global research and the needs of Kenya due to the high share of diarrhoea in under-five deaths. The vaccine has only been available on a global basis since 2011. The launch was scheduled for 2013 and the Government of Kenya had prepared for this. Due to the choice of vaccine, however, suppliers were unable to meet the required level of delivery and the launch has been delayed until 2014.

We found good examples of value for money

2.38 Bed nets and immunisation are amongst the most cost-effective of any health interventions worldwide. Both GAVI and the Global Fund have achieved reductions in purchase prices of these commodities through improved international procurement. The Global Fund is taking a lead to help development agencies to co-ordinate their procurement of bed nets in order to achieve lower prices.

2.39 It is important, however, that global procurement of bed nets takes into account the impact of pooled purchasing on the structure of supply, both globally and in terms of local producers. Local private sector manufacturers and distributors of nets in Kenya have suffered from the widespread availability of free bed nets, which is bad for the long-term sustainability of the sector.

2.40 Savings made by GAVI in vaccines, such as the 67% reduction in the price of the rotavirus vaccine achieved in 2012, are passed on to the Government of Kenya. These savings, combined with GAVI’s progressive graduation policies for countries, increase the potential for long-term sustainability. Figure 6 illustrates the 36% unit cost saving for fully immunising a child with pentavalent, rotavirus and pneumococcal vaccines, achieved globally by GAVI between 2010 and 2012.

2.41 With support from DFID and USAID, PSI has established systems for international procurement, warehousing and the use of an international delivery company for distribution of bed nets within Kenya. These run alongside the Government of Kenya distribution system, which has also been strengthened with donor support over the past five years.

2.42 A comparative cost analysis, undertaken by the review team, between DFID-funded bed nets distributed through PSI and government-distributed bed nets, has highlighted the cost reductions achieved in 2012-13. The results also suggest that the unit costs of PSI nets for purchase and distribution were 16% below those of the Government’s system in 2011-12 (see Figure 7).

2.43 DFID also works to achieve value for money in its own programmes. It provides training for its staff and partners. Unit costs and overheads in projects are routinely reviewed. The PSI overhead in its DFID contract was reduced from 12% to 8%, partly based on our Zimbabwe Health Sector report.⁴⁸

Corruption is endemic but DFID’s financial management and systems are strong

2.44 Kenya has high levels of corruption. It is ranked 136 out of 175 countries in the 2013 Transparency International Corruption Perceptions Index. DFID has a comprehensive and proactive approach towards risk identification, prevention and mitigation at several levels. It communicates regularly with implementing partners on these issues and has a portal which allows public access for disclosure of fraud. Fourteen cases across all sectors of the DFID Kenya portfolio were under review during our visit. We judge that DFID Kenya has done well to encourage disclosure while maintaining zero tolerance for fraud.

2.45 Several minor frauds had been detected and reported by PSI in DFID programmes. These were well handled and resulted in repayment of monies lost. The Global Fund has experienced more systemic weaknesses, outlined in Figure 8 on page 16, which put DFID money at risk.

Devolution creates substantial new risks and opportunities for delivery

2.46 The 2010 Kenyan Constitution outlined plans to abolish the provincial layer of government and to devolve various functions to 47 counties. This is particularly important for the health sector, which is one of the first and most devolved of the government functions. Many interviewees argued that this transition is putting at risk the progress made to date in reducing child mortality and the likely future trajectory. It also, however, creates new opportunities to focus on the priorities of beneficiaries at the county level.

2.47 A three-year transition period was envisaged for devolution but most of the process was subsequently shortened to five months, to be finalised by July 2013. This was too rapid to be fully effective. At the time of our visit in November 2013, a six-month extension had been granted for a number of functions including human resources management and health worker salaries which were due to be devolved by 1 January 2014. This deadline has been met but counties, central government and health worker unions have differing positions on the number of health staff transferred and the associated salary costs.

2.48 It is estimated that the Government of Kenya will devolve 65% of the health budget in 2013-14 and that this will comprise around 40% of the budgets of the new counties.⁵⁰ Our discussions with development agencies and health experts confirm that the risk of governments spending this money on non-health budgets is high. There is also a political incentive to fund visible items, such as hospitals, ambulances and bed net distributions, even when these might not be the highest priority to achieve improved health outcomes. This poses risks to spending on essential services that promote child survival.

2.49 DFID has responded to the risks presented by devolution by making important investments in the work of WHO to help planning for the transition. DFID supported WHO to undertake vital work to map the preparedness of the counties for devolution and to help them to develop planning tools. This work is having a system-wide positive effect. We saw examples of new planning processes in action in two county contexts.

2.50 Devolution will be particularly difficult for GAVI and the Global Fund, which rely on government systems and do not have the capacity or presence to engage with county-level governments. GAVI commented to us that they expect procurement of vaccines to be unchanged but they are concerned about the impact of change on the local systems for delivering immunisation.

2.51 The upside of devolution is that needs vary markedly in different parts of the country and devolution provides opportunities to have much more customised health strategies for each county. We met a new county Health Minister who was clearly keen to prioritise local needs. Devolution also provides an opportunity for DFID to engage with these new administrations to prioritise the neediest areas and to promote effective approaches to financial management and health systems strengthening in the new political context.

Impact

Assessment: Green-Amber

Under-five mortality has been reduced in Kenya

2.52 Kenya has achieved a turnaround in its reduction of under-five mortality. It saw an increase in under-five mortality during the 1990s, albeit from a relatively low base, while other African countries saw a reduction. This was then reversed and figures fell by over a third in the decade to 2008. The under-five mortality rate in Kenya is currently a little lower than the Eastern and Southern African average of 77 deaths per 1,000 live births.⁵²

2.53 Since 2000, Kenya’s infant mortality rate (under one year old) has shown the fastest rate of decline among the 20 countries in sub-Saharan Africa for which recent data is available.⁵³ Neonatal deaths (under 28 days old) in Kenya, however, have remained largely unchanged over the same period. These deaths now account for 39% of all under-five deaths.⁵⁴

2.54 The overall reductions in under-five mortality also mask significant regional variations (see Figure 9). In Central province, the numbers are as low as 51 per 1,000 live births but, in Nyanza province, they are 149 per 1,000 live births. There are also variations within regions; for example, individual informal settlements in Nairobi have an exceptionally high rate of under-five mortality, at 156 per 1,000 live births.⁵⁵

2.55 We gathered data on impact from a range of sources but there is no authoritative country-wide data after 2008. This is a problem for DFID and for the Government of Kenya. The major Kenya Demographic Health Survey has been delayed from 2013 to 2014 due to the election and planning and funding issues. When completed, it will provide data on Kenya’s achievements in the run-up to the 2015 target date for the MDGs. DFID is supporting this process.

Malaria prevention and treatment has been key to the impact that has been achieved

2.56 Bed nets and malaria treatment have been key elements of this progress in Kenya. Malaria has been reduced globally, with strong leadership from DFID over a number of years. This includes DFID’s commitment to halve malaria in 10 countries, including Kenya, by 2015. DFID’s role has been documented in a recent report by the National Audit Office (NAO) on Malaria.⁵⁶ The World Malaria Report 2013 notes that, between 2000 and 2012, malaria deaths among under-fives, worldwide, fell by 51%.

2.57 The latest malaria survey in Kenya shows sharp declines in the under-five prevalence of malaria from 80% in 1996 to 20% in 2007 and 12% in 2010.⁵⁷ Malaria was still the leading cause of outpatient visits for under-fives in 2011, accounting for 24% of visits.⁵⁸ The impact of specific malaria programmes is illustrated by areas on the coast, such as at Kilifi, where child mortality fell from 115 per 1,000 live births in 1990 to 60 in 2009. Malaria fell as a share of under-five deaths in hospital from 16% to 5% (see Figure 10 on page 18). Latest research has demonstrated that decreases in malaria transmission have been substantially driven by increases in bed nets and malaria treatment.⁵⁹

2.58 Declining child mortality between 2003 and 2008⁶⁰ is linked to hugely increased bed net coverage. The World Bank found that ‘increased ownership of insecticide-treated nets in endemic malarial zones explains 39% of the decline in post neonatal mortality and 58% of the decline in infant mortality’.⁶¹

Access to services has improved but further improvements are still required

2.59 Access to basic services in Kenya has improved over the past five years. Kenya has begun to roll out the Integrated Management of Childhood Illness approach,⁶³ which we saw in operation and promises to be a more effective response to the illnesses faced by young children because it is both comprehensive and systematic. There are large regional variations in delivery of services, however, as shown in Figure 11.

Quality of care is a significant problem

2.60 Heath professionals we interviewed contend that free healthcare for mothers and children has increased demand and access but has led to a declining quality of care in an already stretched system. The consistent messages from our beneficiary surveys included positive feedback for routine care at local health centres but there were reports of poor quality of care at hospitals. Resource gaps in local facilities include staff, equipment and drugs.

2.61 Beneficiaries complained of petty corruption and drugs not being given out, despite being available. They wanted to be treated with respect by health professionals but were often subjected to physical and emotional abuse. Many women with whom we spoke reported knowing someone who had died in labour or who had lost her baby during pregnancy or labour. Extreme examples cited to us included children dying in queues at the hospital. One woman described how her cousin died after delivering late in the evening when there were few staff. Staff had failed to deal with this woman properly after delivery: ‘No one noticed the tear. My cousin died because of the bleeding’.

Performance of the DFID direct aid programme to Kenya is strong

2.62 Overall, DFID Kenya’s existing programmes are delivering against agreed results. Four out of five projects that we reviewed are performing well, with outputs meeting or exceeding expected DFID targets. Our project scorings are shown in Annex A2. The poorest performing element of the portfolio is the Indoor Residual Spraying programme where spraying has been delayed. Key results are shown in Figure 12 on page 19.

2.63 DFID’s support to WHO and PSI was pivotal in scaling up ownership and the use of bed nets by at-risk communities. Research has led to more effectively targeted distribution of nets at antenatal and immunisation clinics. DFID’s support to the national malaria strategy, through WHO, provided one framework used by all partners and ensured a co-ordinated approach to tackling malaria in Kenya. Current research under the Kenya Health Programme has led to a shift in communication and marketing techniques by PSI to focus on encouraging people to use their nets consistently.

2.64 The programme has also demonstrated the ability to move into new areas. The savings of £1.2 million made in the procurement of bed nets have been used to fund rapid diagnostic testing (RDT) for malaria. RDT provides accurate diagnosis and treatment of fever management. RDT reduces over-prescription of anti-malarials, estimated at 70% in adults and 50% in children, with potential cost saving implications of millions of pounds every year.⁶⁷

2.65 DFID funding for national data collection systems, such as the Malaria Indicator Surveys (MIS) and Demographic Health Survey, has provided more accurate data for planning. DFID has indirectly promoted heath systems strengthening. The Government of Kenya said that DFID played a ‘crucial’⁶⁸ role in the implementation of initiatives to gather beneficiary feedback on the quality of their healthcare. Through its support to WHO, DFID has helped to integrate support for malaria control with a broader strengthening of health services. DFID has also demonstrated success in health systems strengthening through its work in Nyanza (see Figure 13).

There have been setbacks to DFID’s bilateral work

2.66 Broader health sector reform has proved challenging. DFID has worked to influence health worker skills, management competencies and referral processes. Good progress has been made in developing policy and strategy, with less progress on the longer-term issues of putting strategy and policy into practice, behaviour change and improving skills and attitudes amongst some health professionals. Changes of this nature will require a well-funded and integrated programme implemented over an extended period. This in turn will require donors to work together in close co-operation with the Government of Kenya.

2.67 The DFID residual spraying programme has not started due to the lack of approval by Kenya’s Pest Control Products Board for the insecticide to be used. This is not due to any failure by DFID but has absorbed a huge amount of staff time and effort. The delays in starting the project mean that two rounds of crucial spraying have been missed, which will lead to increased sickness and deaths from malaria.

Monitoring and evaluation need to focus more on impact

2.68 DFID’s internal systems for monitoring and evaluation of individual programmes are good. DFID regularly tracks results with independent annual reviews and progress reports, which are completed regularly and on time. Other than in Project Completion Reports, however, DFID’s monitoring of results focusses on activities and on outputs rather than on outcomes for intended beneficiaries. Assessment also focusses primarily on individual projects, such as those selected for our review. There is less attention on the overall impact of the full set of health programmes in terms of outcomes for intended beneficiaries and the performance of the health system.

GAVI and the Global Fund have had impact through commodity delivery

2.69 The DFID MAR⁷⁰ has assessed that GAVI is a highly effective organisation. GAVI has had positive and demonstrated impact in Kenya through the delivery of immunisation. GAVI has a clear progression and exit strategy for countries receiving its support. GAVI obtains reduced prices for vaccines for the Government which, combined with its progressive co-financing and graduation policies, improves the prospect for sustainability.

2.70 Between 2011 and 2013, GAVI funded more than 14 million doses of both pneumococcal and pentavalent vaccines in Kenya. The pneumococcal vaccine introduced in 2010 has demonstrated strong results. Kilifi has achieved a reduction in pneumococcal disease in under-fives of two thirds since 2011.⁷¹ The measles vaccination has played a significant role in mortality reduction. Data published in a peer-reviewed journal (Vaccine) suggests that, with GAVI support, Kenya will be able to deliver a range of life-saving vaccines that will avert over 900,000 deaths in the period 2011-20.⁷²

2.71 Overall, Kenya’s coverage for the full package of basic vaccines has risen from a low of 57% in 2003 to 85% in 2011 (see Figure 14). The weakness remains the hard-to-reach areas, although this data demonstrates that substantial progress has also been made in these areas over the past decade.

2.72 The Global Fund has distributed over 11 million bed nets in Kenya over the last five years. Bed nets, funded by DFID, the Global Fund and the United States, have transformed the statistics for people sleeping under nets, as shown in Annex A4(c). More specifically, the use of bed nets by children under five and pregnant women has increased from 5% in 2003 to over 40% in 2010.⁷⁴⁷⁵ A challenge remains to assure appropriate use of bed nets from dusk to dawn in both the dry and rainy seasons, as outlined in the recent NAO report on malaria.⁷⁶

DFID, GAVI and the Global Fund are funding health systems strengthening (HSS) but there are weaknesses in management and effectiveness

2.73 Health systems strengthening is a core area of expertise for DFID on which it has worked in many countries and which it has promoted in global policy discussions and in Kenya. Its role in this area in Kenya has been limited by not directly financing the Government of Kenya’s systems and by the complexities introduced by devolution and health ministry restructuring.

2.74 There is limited co-ordination of donor work on health systems strengthening. GAVI and the Global Fund have both implemented health systems strengthening programmes through the Government of Kenya, which have had some success, particularly around their core areas of expertise – but it is very difficult for them to have effective programmes with remote management.

2.75 GAVI’s health systems strengthening work has helped to establish an effective system for vaccine distribution, which we saw in operation. This is not connected, however, to wider health sector reform, as illustrated by the conclusions of an evaluation of GAVI’s health systems strengthening work, summarised in Figure 15.

Sustainability is vital and gains are reversible

2.76 Sustaining the gains of child mortality reduction is essential. The core of sustainability lies in strengthening basic government health systems. The sustainability of GAVI’s and the Global Fund’s impact also relies upon continued large-scale funding for basic commodities. Each new generation requires immunisation. Malaria gains can easily be reversed, with catastrophic results, if spraying, treatment or bed net usage are reduced. Bed nets need to be replaced about every three years.

2.77 Sustainability requires a vision for how the current subsidised system will evolve and how the Government of Kenya will take over responsibility for the funding of essential services. Parallel systems, developed by the bilateral and multilateral development agencies, reduce the incentives for self-sufficiency.

2.78 In the long term, DFID needs to increase its engagement with Government of Kenya systems if programmes and impact are to be sustainable. Current concerns about corruption make this inappropriate but we did not see any evidence of planning or criteria for resuming direct financing to the Government of Kenya.

Learning

Assessment: Green-Amber

DFID has demonstrated and promoted learning

DFID research has been significant

2.79 DFID has a strong global research programme on health, combined with initiatives from the centre of DFID to turn theory into practice. These include a planned emphasis on implementing change in the field, research hubs and ‘evidence-into-action’ teams. DFID research has helped in implementing new vaccines and drugs for diarrhoea, developments in oral rehydration treatments and HIV drugs for children. DFID has had a major focus on health systems and leads the international community in this. There is also a strong pull from DFID’s professional health advisers to access knowledge and best practices for country programmes.

DFID has adapted its programmes in Kenya to reflect learning

2.80 There is substantial evidence of learning in DFID programmes in Kenya. DFID has encouraged and leveraged global best practices and there has been a strong strand of applied research in its work relating to Kenyan health. The introduction of long-lasting bed nets was driven by research. The move from social marketing to targeted free distribution of bed nets was based on results showing low levels of take up. The roll-out of RDT and indoor residual spraying pilots have provided further opportunities for learning.

2.81 DFID has promoted learning and sharing between the new counties with its funding of the WHO County Health Readiness process and the templates for County Health Strategies. The new programme on Maternal and Newborn health shows evidence of learning (see Figure 16). Given the extended period over which it has been clear, however, that neonatal deaths are a rising share of the under-five mortality rate, DFID did not act as early as it could have done to address this.

DFID has supported innovation by delivery partners

2.82 There is evidence of innovation in the DFID-funded PSI programme on more effective delivery methods for bed nets and new methods to encourage their use. The development of the Tunza clinics and the further expansion of this effort through a centrally-funded DFID programme, AHME, is further evidence of innovation by taking best practice on health into a large-scale, public-private partnership.

There are weaknesses in learning

2.83 There have been missed opportunities by DFID for internal learning. These include delays in responding to evidence that not all poorer women were buying socially marketed bed nets even at subsidised prices and moving more quickly from this approach to a free, targeted distribution of nets. DFID has also been slow to respond to the growing importance of neonatal mortality.

2.84 DFID does not have a central and simple system to show all of its projects that are spending money in a given country. DFID Kenya staff were not aware of all the centrally funded DFID projects working in Kenya. These include an important project to reduce maternal and neonatal mortality in Mandera County, funded by the Civil Society Department, which will shed important light on good practice in this work and the issues related to working at the county level in a remote area. The DFID Kenya team became aware of this in preparing documentation for our review and flagged it to our review team.

There are gaps in learning from beneficiaries

2.85 We saw limited evidence that beneficiary views and priorities were incorporated into learning that led to change in programmes. Through its work with WHO, DFID has played a critical role in ensuring client satisfaction surveys take place and beneficiary views are now part of the annual review process. These provide important data sources, which could be incorporated more by DFID and other development agencies into design to ensure that programmes have a real impact on the people they are designed to serve.

2.86 Examples of where beneficiary feedback is essential to programme design include:

• understanding how people use bed nets;
• understanding why mothers sometimes prefer commercial services over free public services;
• what charges are actually levied at public facilities; and
• the impact of location, waiting time and staff attitudes for access to services.

The quality of data in Kenya is poor

2.87 The quality and timeliness of data, which DFID and others rely on for planning in Kenya, is poor. The last Kenya Demographic Health Survey (DHS) was completed in 2008. DFID is rightly committed to supporting existing data collection systems rather than developing parallel studies. It is supporting the latest round of the DHS but this has been delayed from 2013 to 2014. Routine health data is collected by the Government of Kenya but quality is poor and private services do not contribute to regular Government surveys.

GAVI and the Global Fund both demonstrate learning but change has been slow

2.88 GAVI and the Global Fund both demonstrate learning from external research and from their own internal evaluations. The roll-out and prioritisation of their programmes are driven by research. GAVI, in particular, combines global testing and verification of new vaccines with country-level needs analysis. The Global Fund has demonstrated learning in procurement issues, leading to significant new initiatives for pooled procurement in which DFID is an active participant.

2.89 Much of the Global Fund’s learning is based on previous grant performance. Its financial mis-management of grants was part of a broader poor performance in delivery which came to a head in 2011 and triggered a process of reform. DFID worked effectively to lead this process at the Global Fund Board and reforms were implemented rapidly at the head office level. The Global Fund works at arm’s length and gives a large measure of autonomy to countries. Reforms were therefore slow to have an impact at the country level and were too late to demonstrate results in Kenya in the review period. Significant improvements have now been made, including the strengthening of the Global Fund’s Kenya country team and the introduction of the new funding model, which will increase the role of the Global Fund in planning and monitoring country programmes in future.

DFID programme design is strongly focussed on its own performance targets rather than the priorities of intended beneficiaries

2.90 Since the Bilateral Aid Review in 2010, DFID Kenya has focussed closely on its own agreed Results Framework. This incorporates the MDGs at the highest level and specifically monitors the under-five mortality rate. The chosen parameters at the country level, however, which guide programme design, focus on a contribution of measurable inputs and outputs to a global target. These may or may not reflect priorities for individuals at the country or county level in Kenya and do not involve country-specific beneficiary consultation.

2.91 The focus on quantitative outputs risks losing connection with the most important outcomes for intended beneficiaries. It promotes a focus on the provision of bed nets and other commodities rather than working on core health systems. A good example of this is the need for a greater focus on quality of care, as measured both in terms of outcomes and of patient satisfaction. We heard numerous complaints from beneficiaries about the quality of care in Kenyan hospitals. One women commented, ‘I took my daughter to a hospital…the doctors were on strike….The way they handle patients is not encouraging’.

2.92 The relevant indicators for under-five mortality, which DFID Kenya uses in designing programmes, are the numbers of bed nets distributed, skilled birth attendants and women using family planning.⁸⁰ These are important indicators but do not specifically track reductions in under-five mortality. New tools that are being implemented from January 2014⁸¹ will help track these figures more readily.

2.93 For future programmes, DFID staff in Kenya have demonstrated a commitment to equity in programme design, meaning that they are prioritising hard-to-reach communities, even though this involves higher unit costs. They have chosen to work in Turkana County, for example, which is more expensive than less remote locations. The team was not able to show us any central DFID guidance for these decisions but highlighted equity elements within a quality assurance checklist used for programme design.⁸² If equity is to be central to the programme, then it is important that such decisions are made on a systematic basis.

There are new learning issues for the future

2.94 Emerging issues in the health sector in Kenya include devolution, neonatal health, quality of service provision, stunting and urban health. Follow up will be needed on the links between maternal and child mortality being explored in the research undertaken by Family Care International and funded by DFID.

2.95 The objectives and indicators that DFID uses will need to be broadened to reflect the new international targets agreed for the period after 2015 when the MDGs will have expired. With an emphasis on ‘no one left behind’, it is likely that monitoring the health status of hard-to-reach population and a strong people-centred focus within evaluation will be central to these approaches.

2.96 Devolution will also require changes to DFID’s monitoring and evaluation systems. In the short term it is vital that all parties are able to monitor the impact of devolution. This requires urgent action to put in place systems which provide timely data on what is happening to health services and outcomes at the county level.

3 Conclusions and Recommendations

Conclusions

3.1 We conclude that DFID has been effective in reducing under-five mortality in Kenya through its wider influence in the international system and through its own bilateral work. It has implemented proven interventions, identified by global research and incorporating cross-country learning, particularly for malaria reduction. It has played an important role in the governance of the relevant multilateral institutions. Our reviews of bed nets and immunisation programmes both demonstrate that many lives have been saved as a result.

3.2 Kenya, nonetheless, has lagged behind many other countries in the rate of reduction of under-five mortality. Levels of child mortality in remote rural areas and informal urban settlements remain very high. Immunisation has increased but has not been delivered to all counties. Quality of care is poor and core health systems are weak. There is a need to invest in strengthening health systems to deliver further reductions in under-five mortality. This is made more urgent by the risks posed by the devolution process.

3.3 Our conclusions on the three questions that this review sets out to answer are as follows:

How coherent are DFID’s contributions to global initiatives for reducing under-five mortality and how well is DFID leveraging this expertise across the Kenyan health system as a whole?

3.4 DFID has demonstrated a coherent approach to reducing under-five mortality that spans research, policy work, multilateral governance and direct aid programmes in Kenya. UK research has been highly significant. DFID has exercised a global leadership role in malaria reduction and health systems strengthening. DFID’s programmes are complementary to GAVI and the Global Fund, focus on national priorities and complement other donors’ support.

3.5 DFID’s decision not to finance government directly limits its influence in Kenya. We support the choices that have been made not to finance the Government of Kenya but this has limited DFID’s ability to use its global knowledge across the Kenyan health system. DFID has found innovative ways to fund and strengthen other organisations, notably WHO and UNICEF, to engage in important work in supporting the Ministry of Health and planning for devolution. It is inconsistent, however, for DFID to be funding other organisations which finance government systems when it has chosen not to do so directly. DFID should either be pressing for multilaterals to withdraw in a similar way or ensuring that there are strengthened processes and systems in place to handle the higher level of risk.

What has been the effectiveness and sustainable impact for intended beneficiaries of DFID Kenya’s programmes to reduce under-five mortality and to what extent are DFID’s plans for the future based on learning?

3.6 DFID Kenya’s programmes have generally performed well against targets. DFID has pioneered and implemented proven interventions. Lives have been saved and improved, particularly through the reduction in illness and deaths due to malaria. DFID’s choice of not working through the Government of Kenya has added to costs and reduced sustainability. The start of the Indoor Residual Spraying programme has been delayed but this was not due to shortcomings from DFID, which has handled a difficult situation well.

3.7 New programmes incorporate learning, particularly on neonatal mortality, although we conclude that the need for intensified action on this could have been identified earlier. DFID has been relatively slow to develop its new programmes, based on limited staff capacity and a cumbersome approval process.

3.8 Sustainability is vital. Sustaining the gains of child mortality reduction is essential. This requires continued DFID funding in the short term and a clear plan for engaging with – and transferring responsibility to – the Government of Kenya in the medium term. This includes an exit plan for parallel distribution systems and criteria for resuming direct finance to the Government of Kenya. The core of sustainability lies in strengthening basic health systems. This is an area of DFID expertise and should be an increasing focus of its work.

3.9 Devolution poses significant threats to the Kenyan health system and health outcomes. Many of the benefits of under-five mortality reduction are reversible, particularly for malaria. Devolution threatens health budgets and outcomes but also provides new opportunities. It is vital that DFID engages in this process to strengthen financial and health systems and to ensure that accurate and timely data is being generated about what is happening on the ground.

What influence has DFID had in enhancing the impact and effectiveness of multilateral agencies, specifically GAVI and the Global Fund, in reducing under-five mortality and how well are these initiatives linked with health systems strengthening activities?

3.10 DFID has been extremely influential in the governance of multilateral agencies and, in particular, GAVI and the Global Fund. It is a major funder, particularly of GAVI. It has strong representation and is engaged on the Boards of both multilateral agencies. The DFID MAR process is effective in promoting policy change. DFID led the Global Fund’s Board through the process of reform. This was driven by weaknesses in the Global Fund’s management and systems and has been effective in introducing a new model of working. Interviewees said to us that this is beginning to improve performance but it is too soon to have evaluation evidence to demonstrate improved results. It will be important that the new Global Fund Country Team strengthens supervision so that issues can be highlighted in real time, rather than waiting for an evaluation.

3.11 GAVI and the Global Fund both have evidence of impact. Immunisation and bed nets have saved lives. GAVI estimates that immunisation is saving over 90,000 lives per year in Kenya. The Global Fund’s distribution of bed nets has contributed to significant falls in under-five mortality in some areas where malaria is endemic. The best results have been achieved in coastal regions of Kenya but malaria prevalence in Nyanza has proved harder to reduce, even where nets have been distributed. This is for reasons that are not fully understood.

3.12 No longer directly financing the Government of Kenya has made it harder for DFID to exercise necessary leadership among donors on health systems strengthening. Health systems strengthening is the key to addressing, on a sustainable basis, the challenge of reducing child mortality in Kenya and it is an area where DFID is strong. This – rather than bed nets – should be an increasing focus of the DFID programme, including development of a clearer division of labour between development agencies.

3.13 Devolution could provide opportunities for DFID to engage with county-level government. DFID already has strong links to some county governments and a clear focus on hard-to-reach counties. By co-ordinating its expertise on health and governance, DFID could provide much-needed support to counties which are committed to improving outcomes. This, in turn, could provide models for approaches to be adopted across the country.

Recommendations

Recommendation 1: DFID centrally should specify its policy on equity more clearly and DFID in Kenya should focus systematically on the quality of – and access to – maternal, neonatal and child health services for remote and hard-to-reach populations. DFID should routinely use beneficiary feedback in its programme design.

3.14 The DFID Kenya team has a commitment to equity and working in challenging environments. We were not shown any DFID guidance on equity which would help to guide decisions. DFID should develop guidance in this area, particularly since the post-2015 environment is likely to stress the importance of ‘no one left behind’.

3.15 We recommend that DFID should be using its limited programme funds to target hard-to-reach areas and to model improved services which can be rolled out through the system. This is particularly important on issues of quality where not only are the required changes linked to investment but also to behaviour change, such as the correct use of bed nets, exclusive breast feeding and staff attitude to patients.

3.16 If DFID is to address these areas, then it is vital that it seeks and uses feedback from beneficiaries on the priorities and issues that they face. DFID, the Government of Kenya and other agencies already collect substantial material through surveys, reviews, specific consultations and field visits but DFID should use it more systematically in programme design.

Recommendation 2: DFID should develop a clear exit strategy for funding basic supplies in Kenya (such as bed nets) and focus instead on achieving a long-term and co-ordinated approach amongst development and financing agencies, including GAVI and the Global Fund, for health systems strengthening. DFID should develop criteria for resuming direct financing to the Government of Kenya.

3.17 The core challenge facing Kenya in reducing under-five mortality is to improve its basic health system in all parts of the country, especially in remote rural areas and urban slums. This is an area where DFID has particular strength. DFID has done some work on health systems strengthening but that has been limited by its decision not to finance the Government of Kenya directly. GAVI and the Global Fund are both funding health systems strengthening. Their programmes have good objectives but are difficult to manage and co-ordinate from a distance.

3.18 With the very low levels of Government of Kenya expenditure in health, we recommend that DFID and other development agencies encourage the Government of Kenya to increase funding to the sector, including gradually taking over the funding of bed nets and malaria drugs. This would allow DFID to focus its money on health systems strengthening, where it can make the biggest difference to poor people and where it has particularly strong expertise. If DFID is more able to facilitate a more integrated approach to this work across development agencies, it would provide a framework in which all contributions, including those from GAVI and the Global Fund, could be productive. We would expect these changes to be phased in over time to allow an orderly transition and ensure that development gains are not put into jeopardy.

Recommendation 3: DFID should engage with emerging county government structures in Kenya to mitigate the risks and to expand the opportunities of devolution for health outcomes and to help develop information systems and financial management tools that will rapidly identify and address any negative impacts.

3.19 Devolution is a radical change in governance structures that is being implemented at a very rapid pace. The transfer of budgets to counties creates risks for health programmes, given that 65% of health budgets are being devolved and perhaps will comprise 40% of total county budgets. Although outcomes are unclear, the risks to health expenditure and outcomes are considerable. There is also an upside that devolution may allow for more focussed health planning, centred on the very different needs of the individual counties.

3.20 DFID has already taken a lead in this area by funding assessments of county readiness and the development of planning tools through WHO. Devolution provides an opportunity for DFID to engage with county-level structures, systems and processes by working with progressive counties.

3.21 There is an urgent need to develop financial management skills in parallel with health expertise. DFID is well placed to develop an integrated approach and this would provide important access to government systems. The most urgent immediate priority, however, is to recognise the risks of devolution and to ensure that high-quality and timely data is collected to identify and address any negative impacts, as soon as possible.

Annex

This Annex provides more detailed background information to the review. This includes:

• a timeline of events in Kenya (Annex A1);
• a summary of the traffic light scores of the five programmes, GAVI and the Global Fund (Annex A2);
• an overview of the methodology applied to this review (Annex A3);
• the health sector share of DFID bilateral spending in Kenya, 2008-12 (Annex A4a);
• a mapping of the DFID bilateral portfolio in Kenya (Annex A4b);
• progress in the percentage of people sleeping under an ITN, 2003-11 (Annex A4c);
• additional programme-level recommendations (Annex A5);
• a bibliography (Annex A6); and
• a list of consultations (Annex A7).

Abbreviations

AHME – African Health Markets for Equity

DFID – Department for International Development

DHS – Demographic Health Survey

GAVI – Global Alliance for Vaccines and Immunisation

GIZ – Deutsche Gesellschaft für Internationale Zusammenarbeit GmbH

HIV – Human Immunodeficiency Virus

HSS – Health Systems Strengthening

ICAI – Independent Commission for Aid Impact

ITN – Insecticide-treated bed nets

KDHS – Kenya Demographic Health Survey

KEMSA – Kenya Medical Supplies Agency

MAR – Multilateral Aid Review (DFID)

MDG – Millennium Development Goal

MIS – Malaria Indicator Surveys

NAO – National Audit Office

NGO – Non-governmental organisation

NHS – National Health Service

PSI – Populations Services International

RDT – Rapid diagnostic testing (for malaria)

UN – United Nations

UNICEF – United Nations Children’s Fund

US – United States

USAID – United States Agency for International Development

WHO – World Health Organisation

Footnotes

1. Millennium Development Goals Report, Addendum, MDG 4, Reduce Child Mortality, United Nations, 2013, http://mdgs.un.org/unsd/mdg/Resources/Static/Products/Progress2013/Goal4_Addendum.pdf.
2. In 2009 there were over 2.6 million stillbirths globally with more than 8,200 deaths a day. They are not mentioned in the MDGs. Source: World Health Organisation, http://www.who.int/maternal_child_adolescent/epidemiology/stillbirth/en/.
3. T. Negussi, O. David and Z. Matthews, Teenage pregnancy experiences in rural Kenya. International Journal of Adolescent Medicine and Health. 2003, 15(4):331-340. Although the term ‘women’ is used within this report, this may include what could more accurately be described as adolescents or girls. One research study of 1,247 (12-19-year olds) in rural Kenya showed that 572 (45.9%) had had sexual intercourse and, of these, 245 (42.8%) had been pregnant at least once.
4. Child Info: Monitoring the Situation of Children and Women, United Nations Children’s Fund (UNICEF), http://www.childinfo.org/mortality_ufmrcountrydata.php and Kenya National Bureau of Statistics, Kenya Demographic Health Survey, 2008-09.
5. L. D. E. Ikamari, An Upsurge in early childhood mortality in Kenya: A search for explanations, African Journal of Health Sciences, 2004, Vol. 11, www.ajol.info/index.php/ajhs/article/viewFile/30774/23107; S. O. Rutstein, Factors associated with trends in infant and child mortality in developing countries during the 1990s, WHO, Bulletin of the World Health Organisation, 2000, 78, http://www.who.int/bulletin/archives/78(10)1256.pdf.
6. Child Info: Monitoring the Situation of Children and Women, UNICEF, 2013, http://www.childinfo.org/mortality_ufmrcountrydata.php.
7. S. Cousens et al., National, regional and worldwide estimates of stillbirth rates in 2009 with trends since 1995: a systematic analysis, The Lancet, 2011.
8. Kenyan Parliamentarians meet in Nairobi to address women’s and children’s health, The Partnership for Maternal, Newborn and Child Health, December 2011, http://www.who.int/pmnch/media/news/2011/20111205_kenyan_parliamentarians/en/index.html.
9. The Ministry of Health is officially called the State Department of Health in Kenya although it is more commonly referred to as the Ministry of Health.
10. Information from interviews with DFID Kenya in November 2013.
11. Development Partners in Health Kenya estimated data for 2013-14, covering most but not all of the significant donors. We have translated from US dollars into pounds sterling throughout the report using the applicable average annual exchange rate for 2013, US$1 = £0.640 (http://www.oanda.com/currency/average).
12. Figures collated by Development Partners in Health Kenya, 2013.
13. Operational Plan 2011-2015, June 2012, DFID Kenya and update from DFID Kenya staff, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/209268/Kenya.pdf.
15. Health systems strengthening includes many issues which are common to strengthening the effectiveness and efficiency of any organisation (e.g. governance, effective costing and planning, capacity building and HR planning and management).
16. Except, initially, when funding was through a malaria consortium called the Malaria Centre (based in Uganda but part of the London School of Hygiene and Tropical Medicine), as per http://malaria.lshtm.ac.uk/.
17. Social marketing integrates marketing concepts with other approaches to influence behaviours for the benefit of individuals and communities; in this case, encouraging the use of bed nets to prevent malaria.
18. Liverpool Associates in Tropical Health is the dedicated technical assistance company of the Liverpool School of Tropical Medicine, UK.
19. The MENTOR Initiative is an international organisation designed to strengthen the capacity of emergency-focussed agencies and national partners to implement more effective and co-ordinated action to reduce malaria-related morbidity and mortality.
20. Initially delivered through Danish and German development programmes and now delivered by WHO.
21. Indoor residual spraying is the process of spraying the inside walls and ceilings of buildings with an insecticide to kill mosquitoes that spread malaria.
22. The programme is yet to start but DFID Kenya received notification of Ministerial approval in December 2013.
23. Delivery partners are discussed here and in Annex A2.
24. For the Kenya Health and Essential Health Services programmes, which contain interventions which we did not include in our review (e.g. family planning, HIV and AIDS), budget figures are based on estimates of the spend apportioned to malaria control, immunisation and health systems strengthening.
25. Annual Progress Report 2012, GAVI Alliance, 2012, http://www.gavialliance.org/results/gavi-progress-reports/.
26. For more details, see the International Finance Facility for Immunisation website: http://www.iffim.org/about/.
27. Breakthrough on procurement to save $140 million, The Global Fund (Press Release), November 2013, http://www.theglobalfund.org/en/mediacenter/newsreleases/2013-11-05_Breakthrough_on_Procurement_to_Save_USD_140_Million/.
28. Every Woman, Every Child: From commitments to action, The First Report of the independent Expert Review Group on Information and Accountability for Women’s and Children’s Health, WHO, 2012, http://www.who.int/woman_child_accountability/ierg/reports/2012/IERG_report_low_resolution.pdf.
29. Global Burden of Disease, Injuries and Risk Factors Study, Institute for Health Metrics and Evaluation, 2010, http://www.healthmetricsandevaluation.org/sites/default/files/country-profiles/GBD%20Country%20Report%20-%20Kenya.pdf.
30. Every Woman, Every Child: From commitments to action, The First Report of the independent Expert Review Group on Information and Accountability for Women’s and Children’s Health, WHO, 2012, http://www.who.int/woman_child_accountability/ierg/reports/2012/IERG_report_low_resolution.pdf.
31. Child Survival series, The Lancet, 2003, http://www.thelancet.com/series/child-survival.
32. Liz Ollier and Cindy Stanton, End of Project Evaluation of DFID Support to the Delivery of Essential Health Services, DFID Human Development Resource Centre, March 2011, http://www.heart-resources.org/wp-content/uploads/2012/10/285649_KE-Project-Completion-Report-for-Essential-Health-Services-_EHS_-Project_Report3.pdf.
33. Bruce Mackay, Social Marketing of Insecticide Treated Nets in Kenya: Project Completion Review, August 2010, http://hdrc.dfid.gov.uk/wp-content/uploads/2012/05/273386_DFID-Kenya-ITNs-PCR-Final-Narrative-Report.pdf.
34. Stephanie Simmonds et al, Annual Review of the Support to the Kenya Health Programme, January 2012.
35. Operational Plan 2011-2015, June 2012, DFID Kenya, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/209268/Kenya.pdf.
36. Child Survival and Development Strategy 2008-2015, Government of Kenya Ministry of Public Health and Sanitation, 2008.
37. DFID Results Framework: Managing and reporting DFID Results, DFID, 2013, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/175715/DFID-external-results.pdf.
38. Multilateral Aid Review Update: Driving reform to achieve multilateral effectiveness, DFID, December 2013, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/264615/MAR-review-dec13.pdf.
39. Kenya Health Sector Strategic Plan III (KHSSP III) 2012-2018, Government of Kenya Ministry of Medical Services and Ministry of Public Health & Sanitation, 2012.
40. Bruce Mackay, Social Marketing of Insecticide Treated Nets in Kenya: Project Completion Review, August 2010.
41. Figures provided by PSI, December 2013.
42. A Key Piece of the Puzzle: Faith-Based Health Services in Sub-Saharan Africa, United States Agency for International Development (USAID), http://www.capacityproject.org/images/stories/Voices/voices_14.pdf.
43. The African Health Markets for Equity (AHME) partnership between DFID and the Bill & Melinda Gates Foundation increases private quality care for the poor. DFID is providing £22 million between 2012 and 2017. The programme operates in Ghana, Kenya and Nigeria, http://www.hanshep.org/member-area/programmes/african-health-markets-for-equity-ahme/ahme-exec-summary-2012.pdf.
44. Figures collated by Development Partners in Health Kenya, 2013.
45. Transparency data, Table 6: Top 20 Recipients DFID Bilateral Trade (excluding Humanitarian Assistance 2010-11 and 2012-13), DFID, 2013, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/260993/Table6.csv/preview.
46. Multilateral Aid Review: Ensuring maximum value for money for UK aid through multilateral organisations, DFID, 2011, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/67583/multilateral_aid_review.pdf.
47. Risk and Impact Analysis for Kenya of Changes in Global Fund Financing Modalities: Phase 1 Report, USAID, Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH, (Draft), 2012.
48. DFID’s Support to the Health Sector in Zimbabwe, ICAI, November 2011, http://icai.independent.gov.uk/wp-content/uploads/2013/12/DFIDs-Support-to-the-Health-Sector-in-Zimbabwe.pdf.
49. Distributed by PSI (for DFID) and by Kenya Medical Supplies Agency (KEMSA) (for the Global Fund/Government of Kenya).
50. Kenya’s 2013/14 Health budget cuts threaten to cripple sector, Standard Digital, June 2013, http://www.standardmedia.co.ke/?articleID=2000086000&story_title=kenya-s-2013-14-health-budget-cuts-threaten-to-cripple-sector&pageNo=2.
51. Global Fund Audit and Investigations of grants to the Republic of Kenya, The Global Fund, 2012, http://www.theglobalfund.org/en/oig/reports/.
52. Trends in under-five mortality rates, 1960–2012. UNICEF Child Info, http://www.childinfo.org/mortality_ufmrcountrydata.php.
53. What has driven the decline of infant mortality in Kenya? The World Bank, Policy Research Working Paper 6057, 2012, http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2012/05/03/000158349_20120503152728/Rendered/PDF/WPS6057.pdf.
54. Countdown to 2015 Maternal Newborn and Child Survival, WHO, 2013 Update, http://countdown2015mnch.org/documents/2013Report/Countdown_2013-Update_withprofiles.pdf.
55. N. Taffa and G. Chepngeno, Determinants of health care seeking for childhood illnesses in Nairobi slums, Tropical Medicine and International Health, 2005, Volume 10.
56. Malaria, Report by the Comptroller and the Auditor General, DFID, July 2013, http://www.nao.org.uk/wp-content/uploads/2013/07/10181-001-Malaria-Book.pdf.
57. Kenya Malaria Indicator Survey 2010, Kenya National Bureau of Statistics, 2011, http://measuredhs.com/pubs/pdf/MIS7/MIS7.pdf.
58. Health Facts and Figures, Kenya Ministry of Medical Services, 2012.
59. B.C. Kalayjian et al., Marked Decline in Malaria Prevalence among Pregnant Women and Their Offspring from 1996 to 2010 on the South Kenyan Coast, American Journal of Tropical Medicine and Hygiene, 2013. The study found a ‘profound’ decrease in malaria transmission in coastal Kenya and this was ‘concordant with increased bed net and malaria’ prophylaxis.
60. Kenya Demographic Health Surveys, 2003-04 and 2008-09.
61. The World Bank, What has driven the decline of infant mortality in Kenya? Policy Research Working Paper 6057, 2012, http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2012/05/03/000158349_20120503152728/Rendered/PDF/WPS6057.pdf.
62. Kevin Marsh, Malaria and the Epidemiological Transition in Africa, Slide presentation to Kenya Medical Research Institute, 2009.
63. Integrated Management of Childhood Illnesses is an integrated approach to child health that focusses on the wellbeing of the whole child, see http://www.who.int/maternal_child_adolescent/topics/child/imci/en/.
64. Kenya County Fact Sheets, 2nd Edition, Commission on Revenue Allocation Kenya, 2013, http://www.crakenya.org/wp-content/uploads/2013/07/CRA-County-Fact-sheets-2nd-Edition-June-2013.pdf and Kenya Demographic Health Survey 2008-09.
65. R.W. Snow and K. Marsh, Malaria in Africa: progress and prospects in the decade since the Abuja Declaration, 2010.
66. W.P. O’Meara et al., Effect of a fall in malaria transmission on morbidity and mortality in Kilifi, Kenya, The Lancet, 2008, Vol. 372, pages 1555 – 1562, http://download.thelancet.com/pdfs/journals/lancet/PIIS0140673608616554.pdf?id=eaallsXrMEqGd-mqraDsu.
67. Kenya: Support to the National Malaria Strategy, DFID Resource Centre, 2009.
68. Client Satisfaction Survey Report 2008, Ministry of Medical Services, 2009. DFID, through WHO, has supported further Client and Employee Satisfaction Surveys in 2010, 2011 and 2012.
69. Evidence gathered from interviews during our review and from an Independent evaluation conducted in 2011: L. Ollier and C. Stanton C., End of Project Evaluation of DFID Support to the Delivery of Essential Health Services (EHS), DFID, 2011.
70. Multilateral Aid Review Update: Driving reform to achieve multilateral effectiveness, DFID, December 2013, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/264615/MAR-review-dec13.pdf.
71. Delivering Together: Kenya Big Data, GAVI, 2013, http://www.gavialliance.org/delivering/kenya/.
72. L.A. Lee et al., The estimated mortality impact of vaccinations forecast to be administered during 2011–2020 in 73 countries supported by the GAVI Alliance, Vaccine, April 2013, http://www.ncbi.nlm.nih.gov/pubmed/23598494.
73. Facts and Figures 2012, Ministry of Medical Services, October 2012.
74. Kenya Demographic and Health Survey: Key Findings, Kenya Central Bureau of Statistics, 2003, http://www.measuredhs.com/pubs/pdf/SR104/SR104KE03.pdf.
75. 2010 Kenya Malaria Indicator Survey, Kenya National Bureau of Statistics, 2010, http://measuredhs.com/pubs/pdf/MIS7/MIS7.pdf.
76. Malaria report, NAO (Report by the Comptroller and Auditor General), DFID, HC 534, July 2013, http://www.nao.org.uk/wp-content/uploads/2013/07/10181-001-Malaria-Book.pdf.
77. A. Alebachew and C. Ortendahl, GAVI Health System Strengthening Support Evaluation, Kenya Desk Study, HLSP, 2009.
78. Maternal and Neonatal Health Human Resource Capacity Building – ‘Making it Happen’. Works across 12 countries in sub-Saharan Africa and South Asia.
79. Kenya’s eight provinces, which were sub-divided into districts, are being replaced by 47 counties.
80. These are in line with DFID global results framework health indicators, see https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/175715/DFID-external-results.pdf.
81. Through the Lives Saved Tool (LiST), see http://www.jhsph.edu/departments/international-health/centers-and-institutes/institute-for-international-programs/list/.
82. The quality assurance checklist asks how the intervention will deliver improvements in the lives of the poorest and most vulnerable, including women and girls and more broadly asks programmes to ensure that there is due regard to dimensions of equality and diversity.