...

Indonesian Village Health Institutions

by user

on
12

views

Report

Comments

Transcript

Indonesian Village Health Institutions
 IndonesianVillage
HealthInstitutions
ADiagnostic
Karrie McLaughlin TableofContents
ACKNOWLEDGEMENTS 3 LIST OF ACRONYMS AND INDONESIAN WORDS 4 1. EXECUTIVE SUMMARY 6 2. BACKGROUND AND METHODOLOGY 12 2.1. BACKGROUND AND RATIONALE 2.2. STUDY OBJECTIVES 2.3. METHODOLOGY 2.4. RESEARCH LOCATIONS AND SITE SELECTION 12 13 14 15 3. FINDINGS 17 3.1. INTRODUCTION 17 3.2. WHAT DO COMMUNITY MEMBERS KNOW ABOUT MATERNAL AND NEO‐NATAL HEALTH? WHERE DO THEY SEEK CARE?
17 3.2.1. I THINK I’M PREGNANT (OR MY WIFE/ DAUGHTER/ DAUGHTER IN LAW, ETC. IS PREGNANT). WHAT DO I DO? 18 20 3.2.2. I’M ABOUT TO GIVE BIRTH (OR BE A FATHER OR GRANDPARENT). WHAT DO I DO? WHERE DO I GO? 3.2.3. I AM THE MOTHER (FATHER/ GRANDPARENT) OF A BABY/TODDLER. WHAT DO I DO? 23 24 3.3. WHAT ARE THE ROLES OF ACTORS, INSTITUTIONS AND POLICIES FROM THE VILLAGE TO THE DISTRICT LEVEL? 24 3.3.1. WHAT ARE THE ROLES OF VILLAGE ACTORS, INSTITUTIONS AND POLICIES? 3.3.2. WHAT ARE THE ROLES OF KECAMATAN ACTORS, INSTITUTIONS AND POLICIES? 45 50 3.3.3. WHAT ARE THE ROLES OF KABUPATEN ACTORS, INSTITUTIONS AND POLICIES? 3.4. KEY ISSUES AROUND THE USE OF DATA 56 56 3.4.1. DATA COLLECTION AT THE VILLAGE LEVEL 56 3.4.2. INCENTIVES TO REPORT 3.4.3. ABILITY TO USE DATA‐ GOVERNMENT 57 58 3.4.4. ABILITY TO USE DATA‐ GENERASI 4. ANALYSIS 60 4.1. LINKAGES FOR INFORMATION SHARING ABOUT MATERNAL AND CHILD HEALTH CARE 4.2. INCENTIVES AND REALITIES AROUND ACCESSING MATERNAL AND CHILD HEALTH CARE 4.3. THE USE OF DATA 4.4. PATHWAYS TO ACCOUNTABILITY? 60 62 63 64 1 5. OPPORTUNITIES AND GENERAL RECOMMENDATIONS 65 5.1. POSYANDU SUPPORT 5.2. SHARING INFORMATION AT THE VILLAGE LEVEL (OUTSIDE OF POSYANDU) 5.3. FURTHER EXPLORE THE ROLE OF NGO ADVOCACY 5.4. POLICY DISCUSSIONS AND RESEARCH OPPORTUNITIES 65 67 68 68 6. RECOMMENDATIONS FOR PNPM GENERASI 69 6.1. IMPROVE INFORMATION COLLECTION, SHARING AND USE AT THE VILLAGE LEVEL 6.2. STRENGTHENING PNPM GENERASI LINKAGES TO AND POSITION VIS‐À‐VIS THE HEALTH DEPARTMENT 69 70 ListofBoxes
BOX 1: PNPM GENERASI HEALTH INDICATORS ............................................................................................ 12 BOX 2: VILLAGE HEALTH REGULATIONS IN SUMBA TIMOR .............................................................................. 20 BOX 3: HEALTH REALITIES IN SUMBA TIMOR ................................................................................................ 21 BOX 4: PAYING FOR DELIVERIES IN SUMBA TIMUR ........................................................................................ 28 BOX 5: DUSUN HEADS AS CADRE POSYANDU ............................................................................................... 32 BOX 6: LOCAL ELITE AS HEALTH ACTORS ..................................................................................................... 38 BOX 7: A PROACTIVE VILLAGE HEAD .......................................................................................................... 38 BOX 8: A MODEL DESA SIAGA .................................................................................................................. 44 BOX 9: HEALTH POLICY, POLINDES CONSTRUCTION AND GENERASI IN SUMEDANG ................................................ 55 ListofTables
TABLE 1: STUDY RESPONDENTS ................................................................................................................. 14 TABLE 2: RESEARCH LOCATIONS ................................................................................................................ 16 ListofFigures
FIGURE 1: VILLAGE AND KECAMATAN HEALTH NETWORKS .............................................................................. 25 2 Acknowledgements
For their support in the development of the research concept, the author would like to thank Robert Wrobel from the World Bank Generasi team, John Voss and Dewi Sudharta from the Australian Department of Foreign Affairs and Trade (DFAT) and Dewi Susanti from TNP2K. Key inputs prior to field work were provided by members of the PNPM Generasi Team, particularly Sadwanto Purnomo and Gerda Gulo, members of the Australian Community Development and Civil Society Strengthening Scheme (ACCESS) and the Australian‐ Indonesian Partnership on Maternal and Neonatal Health (AIPMNH) teams in DFAT and members of the World Bank health team. Juliana Wilson provided initial research support. Gunawan explained many aspects of the Indonesian health system and undertook a basic district budget analysis for all research locations to help guide questions. The research team was composed of Gunawan, Juliana Wilson, Nelti Anggriani, Siti Ruhanawati and Firkan Maulana. They were all active in both shaping the research and contributing their thoughts along the way, and this report is infused with their insight. The team was supported and welcomed at district, sub‐district and village levels by PNPM Generasi actors, and would like to thank them for their time and efforts. Similarly, members of the AIPMNH team in Sumba Timor and ACCESS partners in Sumba Timor and Lombok Utara provided information and facilitation that was greatly appreciated by the team. Local government officials, health staff at district offices and puskesmas, posyandyu volunteers, midwives, formal and informal village leaders and ordinary citizens all graciously endured interviews and observation, and some of them kindly housed researchers. It is the hope of the author that this report reflects their views. During writing and revision, this report benefitted from presentation feedback and peer review. Comments from participants at presentations to the PNPM Support Facility, TNP2K and DFAT were all useful. Comments from Robert Wrobel were also valuable in shaping the work. Financial support for the study came from the PNPM Support Facility and the Australian Department of Foreign Affairs and Trade. 3 ListofAcronymsandIndonesianWords
ACCESS Australian Community Development and Civil Society Strengthening Scheme Adat Traditional law AIPMNH Australian‐ Indonesian Partnership on Maternal and Neonatal Health AKI (nol) Angka mematian ibu (nol) Maternal mortality number (zero) Badan Kepegewaian Civil Service Body Bappeda Badan Perencanaan Daerah Regional Planning Body Belis Bride price Bidan Village midwife BKAD Badan Koordinasi Antar Desa Inter‐village Coordination Body BOK Bantuan Operational Kesehatan Health Operational Assistance BPMD Badan Pemberdayaan Masyrakat Desa Community Empowerment Unit Buku KIA Buku Kesehatan Ibu dan Anak Maternal and child health book Buku KMS Buku Kartu Menuju Sehat Toward Health Card Book (included in Buku KIA) Bupati Mayor Cadre posyandu Village integrated service post cadre Dasa Wisma Group of 10 ‐20 women in PKK Desa Siaga Alert Village (program to support pregnant women) DPRD Dewan Perwakilan Rakyat Daerah Regional Parliament Dukun beranak traditional midwife who helps with birthing in particular Dukun berobat traditional apothecary Dukun Traditional midwife Gerakang Sayang Ibu Mother Friendly Movement HDIs Human Development Indicators Jamkesmas Jaminan Kesehatan Masyarakat National insurance scheme for poor citizens Jampersal Jaminan Persalinan Nasional National Birth Insurance Program K1 murni first pregnancy check, when conducted in first trimester KEK Kurang Energi Kronis Chronic Low Energy (Poor maternal nutrition) Kepala desa Village head 4 Kepala dusun Hamlet heads MDGs Millennium Development Goals Mini lokakarya mini workshop (at health post) Mitra Partnership Musrenbang Bottom‐up planning process NTB Nusa Tenggara Barat West Nusa Tenggara NTT Nusa Tenggara Timor East Nusa Tenggara PIK Pagu Indikatif Kecamatan Kecamatan indicative budget limit PK Pelaksana Kegiatan Activity Implementer (Generasi village level actor) PKK Program Kesejahteraan Keluarga Family Prosperity Program (village women’s group) PMT Program makanan tambahan supplemental feeding program PNPM Generasi Sehat dan Cerdas PNPM ‘Healthy and Bright Generation’ PNPM Program Nasional Pemberdayaan Masyarakat National Citizen’s Empowerment Program PODES Potensi Desa Village Potential Survey Polindes Pos Bersalin Desa Village Delivery Post Posyandu Pos pelyanan terpadu Village integrated service post Puskesmas Health Clinic (sub‐district level) SKPD Satuan Kerja Pemerintah Daerah Local Government Work Unit SKTM Surat Keterangan Tidak Mampu “poverty letters” Tokoh masyarakat Village elders/Community leaders 5 1.ExecutiveSummary
Despite its status as a middle income country, Indonesia continues to perform poorly on a number of health indicators. As part of the broader picture of health service delivery, this study focuses on understanding the roles of village level health actors and institutions. It maps interactions between health actors and institutions at the village level and examines how they engage (or don’t engage) with supra‐village actors and institutions. In doing so, it aims identify opportunities and constraints for communities which seek to engage more effectively with local health providers and ways that communities can support the improved delivery of quality health services. It is an input to the design of PNPM Generasi Sehat dan Cerdas (PNPM Healthy and Bright Generation, ‘Generasi’), an incentivized block grant program that seeks to provide a demand‐side support to the achievement of targeted health outcomes. In line with Generasi’s emphasis, this study focuses on actors and institutions related to maternal and child health. The research took place in four provinces: West Java (pilot), Gorontolo, West Nusa Tenggara and East Nusa Tenggara. Provinces were selected to allow both for an understanding of the role of Generasi and to examine alternative approaches to improving citizen engagement in service delivery taken by the Australian Community Development and Civil Society Strengthening Scheme (ACCESS) and the Australian‐ Indonesian Partnership on Maternal and Neo‐natal Health (AIPMNH). In each province, one district was chosen. Within that district, one village in each of two sub‐districts was selected for study. The study employed qualitative research methods, gathering information through semi‐structured interviews and focus group discussions with key actors at the district, sub‐district and village levels. In each village, researchers were asked to describe actors and institutions that provided health information, supported access to maternal and child health services, or had the potential to do so. They then undertook a network analysis to determine where actors were (and were not) working together toward improved service delivery. Based on this analysis, this report focuses on three broad sets of findings. The first is a very basic picture of what community members understand about maternal and child health needs and care. It explores how that understanding contributes to family decision‐making about when to seek care and examines how other factors, such as access, costs and other family responsibilities, etc. play into those decisions. This information provides a backdrop to the core of the report – an examination of health actors and institutions at the village level. For each actor or institution, researchers considered their role in providing health services or information in the village; the amount of power that they command; the amount of trust that they enjoy; and how they link to other key actors within the village or broader health system. The final set of findings centers around the use of data, both within the health system generally, as well as within Generasi. HealthKnowledgeandAction
Two general observations can be made about health messaging. The first has to do with who can share information and how it moves through the health system. The approach is currently highly technocratic and works against a more holistic approach. For example, each health clinic (puskesmas, generally at the sub‐district level) should have a nutrition expert. However, if the position is empty, then little 6 information is provided to communities. The second has to do with how information is shared: what people need to do is generally clear, but not why. For example, while many mothers know that they should feed their children certain types of food, they have little to no understanding of the full impact of stunting or malnutrition on their children if they do not. While simple messaging can be important, the approach is consistent with an underlying attitude from the health department that community members can’t really understand why they should do certain things, but just need to be directed. At its strongest, this attitude presses beyond how information is shared into support for punitive measures against citizens who do not take desired health actions. Political attention to maternal and neo‐natal health outcomes has increased, partly driven by increased pressure to achieve the Millennium Development Goals. In most locations, the initial government response has been to encourage women to give birth in ‘adequate’ health facilities in an attempt to reduce maternal and baby deaths. There is local variation as to what is considered ‘adequate’ – in some areas it is the village health post, in others, women need to go to the sub‐district level. In addition, the State is taking strong steps to reduce the involvement of traditional midwives (dukun) in childbirth. Both with respect to going to health facilities and using dukun, the study revealed a fine line between supporting families to access care in particular ways and blaming or punishing citizens for not taking desired actions. As pressure increases to improve maternal and neo‐natal health outcomes, there seems to be a trend toward the latter. While this may be a useful approach when facilities are easily accessible and of good quality, when they are not, it simply erodes village safety nets for pregnant women without providing an alternative. Unfortunately, there seems to be a weak understanding of (or unwillingness to recognize) obstacles to care by health officials. These often go significantly beyond laziness, stupidity or backwardness to issues of information, communications systems, transportation systems and sometimes significant up‐front costs in accessing health care outside of the village. While many women know where they should go certain points in pregnancy and with their children, other health knowledge, particularly around maternal and child nutrition, remains variable. Most women are aware that they should be checked by a midwife (bidan) when they become pregnant, try to go to a health facility to give birth, and take their children to the monthly ‘integrated service posts’ (posyandu) in the village, where they are generally weighed and immunized. However, when they do go to posyandu, there is very little discussion of how their children are doing (mapping their progress on the growth chart, for example). Nor are discussions held about ‘at‐risk’ children. Indeed, there is some disagreement on when children are considered malnourished. While knowledge of how to measure maternal undernourishment is fairly standard, it is the traditional midwives (dukun), not the state midwives (bidan), who provide the most information on what women should eat during pregnancy. Some of the information that they provide is useful, and some deeply detrimental. However, as the State is mainly focused on removing traditional midwives from any aspect of childbirth, there is no engagement with thisme on issues of nutrition. HealthActorsandInstitutions:AMap
The network analysis highlights the primacy of the posyandu, and by extension the importance of the village volunteers who support it (cadre posyandu). In all of the villages visited, it was many women’s main point of contact with the formal health system on issues of maternal health, pregnancy risks, and 7 baby and child immunizations and nutrition, among other things. While this does not mean that women only get information and care from the posyandu, or even that all women go to posyandu, it does mean the institution, and the cadre that drive the institution, have a critical role to play maternal and child health. In addition, it means that the people and institutions who support the posyandu, (namely the puskesmas and the bidan), have an important role to play. It also highlights the strength of the bidan in providing individual services and her weakness in providing institutional support or coordination at the village level. Bidan provide pre‐ and ante‐natal care, in some cases doing so with great empathy and discretion. They also conduct checks at posyandu, and are responsible for compiling and forwarding data from posyandu to the puskesmas. However, beyond rudimentary data sharing, linkages between the puskesmas, the bidan, and the cadre posyandu are often weak, and linkages with “non‐health” actors who could play important supporting roles on health issues (such as members of village government) are virtually non‐existent. There are several factors that contribute to weak linkages to and support for posyandu and cadre posyandu and disengagement of the bidan. First, the posyandu sits under the community empowerment division, as a community empowerment institution, with the health department only providing technical assistance on training to cadre. This results in a great deal of variation in support for cadre posyandu from the puskesmas. The study shows that if puskesmas are both interested and incentivized, they can provide more active support to posyandu. Unfortunately, the opposite is also true. These structural arrangements have the additional problem of creating a gap between the bidan and cadre posyandu, despite the fact that they both sit at the village level. While cadre provide the bidan with data, she has little responsibility for their training or support, which falls largely to the puskesmas. In no studied villages did the bidan have a routine way of talking to cadre posyandu outside of the posyandu itself, which tends to be quite busy. These issues contribute to a second problem, a gap between the normative conception of posyandu (integrated service delivery, with many ‘tables’ of information, counseling etc) and what it actually does (weigh children, maybe provide group counseling). There are no attempts to support village analysis of problems, through, for example, routine analysis of village data. This lack of support stems from a variety of sources, including approaches to training which emphasize the formal over activities ‘on the job’ and in the village and a technocratic understanding of health in which some information isn’t shared with cadre because it is considered too complicated. Some of these problems are being addressed by supplemental training by programs like Generasi, but without ongoing engagement of cadre, it is difficult to sustain. In addition, even if cadre are more effectively empowered, the lack of resources must be recognized as a constraint: village‐level actors may face frustration if they identify problems that they cannot solve or, in the worst case, are penalized for even discussing. Beyond the roles of cadre posyandu and bidan, the network analysis shows that village heads or hamlet (dusun) heads have the power within the village to support health issues, but do not usually feel that they have either the authority or the expertise with health actors to advocate for better information sharing. 8 Despite the challenges discussed above, it must be noted that many of the problems do not stem from a lack of village capacity, but rather from a lack of support and coordination to develop what already exists. While linkages are currently weak, the bidan, the cadre posyandu (including those among them who are particularly motivated and knowledgeable), and the village government are all sitting at the village level. It would seem that support for capacity building, routine interaction and problem solving at the village level could help citizens access both health information and services, even in the absence of supply side reform. Some of this work has already begun: both Generasi and ACCESS are supporting additional training for cadre posyandu. District governments are increasingly recognizing the value of village head engagement on key issues. However, supporting the capacity that does exist at the village level is currently constrained by the fact that a large disconnect remains between community perceptions of their health needs and the approaches taken by the puskesmas. The ability for citizens to talk to (and be heard by), health service providers is fundamental to any future discussion about accountability and improved service delivery. The current gap may stem from a number of things: training, concepts of modernity, feelings of superiority, etc, but the result is that the health department is often reluctant to value the opinion of anyone without a health background, and does not fully explore community obstacles to access. The lack of recognition of the obstacles and costs communities face in accessing health services has a huge impact on potential pathways to accountability by denying community experience in planning, budgeting and policy‐making. As explored in the recommendations, there are opportunities for programs like Generasi to support increased sharing between the health department and community members. TheUseofData
The study identified a number of issues around the use of data, both generally and within Generasi. Data collection at the village level is a case study in how a lack of trust compounds and perpetuates poor data quality. Nearly every health program at the village level collects its own data, often requiring a door‐to‐
door survey, because they don’t trust other sources of data. These multiple data requests may actually decrease collection. At the same time, there are some very valid reasons that data should be treated carefully, including the fact that increasing sanctions for poor results may push some individuals to falsify data. As noted above, while political attention to health data can help drive results, it can also create perverse incentives for data falsification. Data is increasingly being used to monitor progress on certain indicators (notably maternal mortality). This turn toward evidenced‐based policy making is promising, and it opens space to start thinking about how data can be marshalled more effectively, be it by the health system, programs like Generasi or NGOs. For example, while stunting is increasingly recognized as a problem, no data is collected on the at‐risk category of children in the ‘yellow band’ of the growth chart. However, it must be recognized that data use is still nascent and its use in program planning and evaluation is still limited and in need of further support. Generasi facilitators face particular challenges using data. As the program is positioned as a community empowerment program, it lacks credibility as a ‘health actor,’ and facilitators often struggle to use 9 program data for advocacy with sub‐district or district health officials. In addition, while many facilitators have an interest in health, they need support to analyze data and prioritize messaging. At an even more basic level, the same is true of village heads and members of inter‐village coordination teams, who are structurally in a position to interface with the health system, but often disempowered by the feeling that individuals without a health background are ill‐equipped to talk about health problems. GeneralOpportunities
Based on the findings above, general opportunities are identified for government institutions or other programming. These include: Provide improved support to cadre posyandu. In particular, improve and increase the interaction between the cadre posyandu and the bidan with the goal of increasing capacity and empowerment to identify and address village‐level health issues. Discussions could take place after posyandu or during a routine monthly meeting, and could include analysis of village information such as village data, problems, or examples of positive deviance. They could also cover existing health information (such as ‘first 1,000 days’ material) or health policies. If village needs are more carefully identified, then cadre can potentially be more purposefully recruited and posyandu more strategically planned. This can include better use of male cadre and approaches to engage men. Improved information sharing at the village level. Outside of posyandu, there are three key areas that can be improved at the village level: ‐
‐
‐
Incentivize the village head to engage on health outcomes. This means receiving village data from the bidan, and having sufficient training to understand it, not only for use within the village, but also for raising issues to the puskesmas. Create and support opportunities for community members to share information on experiences with the puskesmas or health department. This would likely be best structured around a particular issue or event (for example, a social audit of maternal or child death), so that real experiences can be brought to bear. Engage the dukun more effectively on issues outside of childbirth, especially nutrition. Further explore the role of NGO advocacy. The study shows that NGOs can provide important support to communities, but more work is needed to determine how to make it most effective. Support ongoing policy discussions and research. Key topics could include: the desirability of using punitive approaches to service delivery; real costs of the facilities‐based approach; and emergency service provision, among others. 10 Generasi‐specificRecommendations
Recommendations for Generasi include: Improve information collection, sharing and use at the village level to support better problem identification, change expectations about health services and encourage oversight. Specific recommendations include: ‐
‐
‐
Collect data in a way that it is more readily accessible to communities so that they can understand if and why targets were met or not. Specifically, provide ‘real’ data back to communities instead of data which has already been manipulated for bonus calculations. Identify key policy areas to target and collect data that will support advocacy on those issues. These could include: collecting information on ‘at‐risk’ children to better talk about prevention of nutrition problems or gathering accountability‐related data such as bidan attendance or immunization arrival at posyandu. Support improved discussions around data and rights with respect to health during the planning process. Strengthen PNPM Generasi linkages to and position vis‐à‐vis the health department. Generasi is not always taken seriously by the health department since it is a community empowerment program, rather than a sectoral one. Its position could be improved by: ‐
‐
‐
Engaging a specialist ‘health facilitator’ with a health background who can act as a trusted liaison between the health department and Generasi; support Generasi to use data better to advocate on health issues; and facilitate opportunities for communities to share their health experiences with government, particularly the health department. Strengthening the technical health knowledge of kecamatan facilitators. This can also be supported by the ‘health facilitator.’ They can then better support village level actors. Consider coordinating the program through Bappeda, rather than the community empowerment division, so that it is in a stronger cross‐sectoral position. 11 2.BackgroundandMethodology
2.1.BackgroundandRationale
Local governments are currently responsible for around one third of all public expenditure in Indonesia, and more than half of the spending on basic services such as health, education and infrastructure. However, the increased transfers have not necessarily correlated with improvements in service delivery at the community level. The capacity of district governments to provide equitable and accessible services varies considerably across regions. This occurs both because of breakdowns in the supply side (local governments lack the skills or resources, or the incentive, to address villages’ needs) and in the demand side (villagers are not able to articulate their development needs, or are not able to communicate them effectively to government). As a result, Indonesia is lagging behind in several key Millennium Development Goals (MDGs) and Human Development Indicators (HDIs) and many Indonesians, especially in disadvantaged areas, lack access to basic services. PNPM Generasi Sehat dan Cerdas (PNPM ‘Healthy and Bright Generation’) is an incentivized community block grant program that seeks to provide a demand‐side response to some of these problems. It targets four Millennium Development Goals lagging in Indonesia: reducing poverty, maternal mortality, child mortality, and universal basic education. PNPM Generasi is designed to address a broad‐set of demand and small‐scale supply constraints to accessing health and education services. Box 1: PNPM Generasi Health Indicators 





Four prenatal care visits for pregnant women Taking iron tablets during pregnancy Delivery assisted by a trained professional Two postnatal care visits Complete childhood immunizations Adequate monthly weight increases for infants  Monthly weighing for children under 3 and biannually for children under 5  Vitamin A twice a year for children under 5 PNPM Generasi’s 3‐Wave evaluation series (conducted from 2007‐2010) demonstrated the program’s ability to increase utilization of basic services, but also noted the inability of the program to impact long‐
term health and education outcomes. Since utilization of services is only one element influencing longer term changes, it is important to better understand both the range of other factors at play, including the quality of services available and the responsiveness of services providers to the needs and preferences of rural communities. Communities often lack the entry points, capacity, and authority required to motivate service providers and local governments to make improvements, and PNPM Generasi does not currently contain an explicit design element to incentivize or assist them in doing so. This study is motivated by a desire to better understand factors influencing long‐term health indicators, and to assess what, if anything, communities themselves can do to tackle them. Part of a suite of research work, its main focus is on the village: how health actors and institutions interact within the village, and they interact (or don’t interact) with levels above them. The primary goal of this work is to inform improvements in the design of PNPM Generasi, which will be scaled up to 500 kecamatan by 2014 as part of the Government’s suite of demand‐side health and education programs. It supports objectives defined in the government’s connectivity‐based Masterplan for the Acceleration of Indonesia’s Poverty Reduction Program (MP3KI) by targeting assistance to rural 12 areas with poor health and education performance. The program will also act as the interface through which sectoral programs will be integrated into PNPM. PNPM Generasi is also expected to play a role in new multi‐sectoral programs. Bappenas and the Ministry of Health have identified PNPM Generasi as a demand‐side pillar under the Scaling Up Nutrition framework. Bappenas and the Australian government are working to design of a multi‐sectoral program aimed at improving frontline service delivery in the poorest provinces of Indonesia. 2.2.StudyObjectives
The PNPM Generasi Health Institutions Diagnostic was commissioned to: examine and map key actors and institutions in health service provision and identify opportunities and constraints for communities to engage constructively with local health providers in order to modify the incentive structures they face or provide improved services. The research employed qualitative methods to answer the following questions: 1. Who are the actors and institutions which are critical to village‐level health services delivery? 2. What are the functional relationships between these actors/institutions which influence service delivery? 3. How do these relationships influence the effectiveness of health services delivery? 4. How do regulations, incentives and resources from the district‐wide health system and Puskesmas influence the effectiveness of health services delivery at the village level? 5. What are the key factors external to the system of defined actors/institution which influence health service delivery? How do these factors influence health services delivery? 6. What relationships are currently used by communities to hold service providers accountable? What determines their effectiveness? 7. What are key incentives facing service providers and which of these incentives can be most impacted by communities? The questions above are broad, and the study bounds them in a number of ways. First, the study places more emphasis on health institutions that are linked to Generasi incentives, which are related to maternal and child health (see Box 1). This does not mean that it only examines those institutions that explicitly focus on maternal and child health, but it does mean that it starts with the issues of maternal and child health and networks out to supporting institutions, rather than canvassing of all possible health‐related actors and institutions. This approach includes a consideration of the health experiences and support networks of fathers and other influential family or community members, but through the lens of how they affect decisions around care for mothers and children. In its consideration of government programs, the study similarly keeps its focus on those related to maternal and child health. For example, it spends more time looking at the incentives created by the National Birth Insurance Program (Jampersal), which allows for all women to give birth with a skilled birth attendant for free, rather than examining programs aimed at providing general health coverage for the poor, including the national insurance scheme for poor citizens (Jamkesmas) or the various regional insurance programs that are being developed in many locations. 13 Finally, while the study is broadly conceived, it is an input for the Generasi re‐design. Therefore, particular importance was placed on understanding how Generasi influences the answers to the questions above. In order to generate examples of alternative interventions, the study also looked at interventions under the Australian Community Development and Civil Society Strengthening Scheme (ACCESS) and Australian Indonesian Partnership on Maternal and Neonatal Health (AIPMNH) programs, which both work more directly on the community‐government interface, ACCESS more from the community side, and AIPMNH more from the supply side. 2.3.Methodology
Qualitative research methods were employed for this study. Information was largely collected through semi‐structured interviews and focus group discussions with key actors at the district, sub‐district and village levels. Required respondents are listed in the table below. For community respondents, an attempt was made to draw respondents from different economic levels and different locations within the village, paying attention to outlying areas or other factors that may impact on access. Based on these interviews, researchers identified any additional respondents who were important health actors in the village. Those that were identified in the course of the research included retired heath providers who continued to be important sources of advice, or private health providers. In all, over 35 individuals were interviewed at the kabupaten level, 46 individuals at the kecamatan level and 211 at the village level, of which 132 (63%) were women. These numbers exclude large focus groups, but include small group discussions of 2‐3 people, which are common at the village level. Table 1: Study Respondents Village level 









Village head Hamlet heads (selected) Village elders (tokoh masyarakat) Village midwife (bidan) Traditional midwife (dukun) Village health post cadre (cadre posyandu) Leaders of health‐related programs, including PKK, Desa Siaga, Dasa Wisma Generasi actors, including village male and female facilitators and treasurer Women and men with babies and small children (under 5) Pregnant women Kecamatan level Kabupaten level 







Head of the health clinic Nutrition expert Midwife coordinator Mayor (camat) Inter‐village Coordination Working Body (BKAD) Generasi actors, including the kecamatan facilitator and members of kecamatan management units (UPK). 






14 Head of the Health Department Head of the Healthy Family Division, including, Mother and Child Health and Nutrition Sections Regional planning division Regional parliament (Health commission) Regional community empowerment division (Badan Pemberdayaan Masyarakat Daerah‐ BPMD) Local NGOs working on health/service delivery Generasi Kabupaten Facilitator Head of Family Prosperity Program (village women’s group) (Program Kesejahteraan Keluarga‐ PKK) Women’s Empowerment Division In addition, a review of relevant primary and secondary data was undertaken. Primary data sources of data included village and district regulations related to health, as well as district health budgets. Secondary data came in the form of project documents (ACCESS, AIPMNH, other NGOs, etc.) and a background literature review on local level health services. Researchers worked in teams of pairs made up of a senior researcher and a junior researcher. Teams generally spent one to two days at each the kabupaten and kecamatan levels, and five days at the village level. This allowed them to identify key policies and practices at the kabupaten and kecamatan levels, follow them down to the village level, and then come back to kecamatan officials with any questions or issues for clarification that developed while in the village. The amount of time spent in the village was driven largely by the number of respondents needed at that level to triangulate information about community networks (see list above), and have enough time to cover any outlying areas of the village. This approach allowed for the voices of all health actors to be heard, and allowed researchers to see where information flows or community support systems either stopped or worked effectively. Researchers were asked to keep notes of all interviews, but were not asked to complete full, coded transcripts. They followed a common outline to report research findings at the village, sub‐district and district level. This outline was developed prior to the field test, and refined with the feedback of the research team in order to best reflect village reality. It includes: the description of actors or institutions that either provided information or supported utilization of services, followed by a network analysis of how the actors relate to each other and the community in the dimensions of power, trust, incentives and linkages. In describing village networks, researchers were not asked to provide a person‐by‐person account, but to try to identify general patterns in the village. A key challenge in undertaking the network analysis was separating information from action or access to further information. The researchers were provided with basic training on existing programs and health practices so that they could make a basic assessment of the quality of information received by villagers. While determining quality of information received at the village level was not an aim of the study, it was required at a very basic level to determine how it impacted on the formation and use of networks for information and utilization. 2.4.ResearchLocationsandSiteSelection
The qualitative research took place in four provinces: West Java (pilot), Gorontalo, Nusa Tenggara Barat (NTB) and Nusa Tenggara Timor (NTT). Provinces were selected to allow both for an understanding of the role of Generasi and to examine approaches to improving citizen engagement in service delivery taken by ACCESS and AIPMNH. In each province, one kabupaten was chosen. Within that kabupaten, one village in each of two kecamatan was selected for study. The study aimed to examine one kecamatan each with high and low supply readiness (supply readiness is a composite measure of infrastructure and other supply inputs against the population of a kecamatan, based on PODES 2011 data). The study included supply‐side readiness in an attempt to see whether better supply side resources made it easier for community members to access services. In practice, the supply readiness data proved difficult to use, since in some 15 locations all Generasi locations were either low or high supply readiness. Final kecamatan and village choices were made following conversations a) at the kabupaten level with Generasi representatives, local government and NGOs (for kecamatan choices) and b) with kecamatan‐level Generasi representatives and local government (for village choices). Factors that were considered in kecamatan and village choice were supply side resources (both physical and human); achievement of Generasi targets; problem solving capacity with respect to health issues; and a topographical variation (particularly as it related to access). Research locations are detailed below. Table 2: Research Locations Province Kabupaten Kecamatan/ Village Overlap with… West Java (pilot) Sumedang Kec Wado/ Cimungkal
‐0‐ Kec Surian/ Wanajaya
Gorontalo Gorontalo Kec Batudaa Pantai/ Buhudaa
‐0‐ Kec Mootilango/ Pilomonu
Nusa Tenggara Barat Lombok Utara Kec Bayan/ Karang Bajo
ACCESS Kec Pemenang/ Pemenang Barat
Nusa Tenggara Timur Sumba Timur Kec Haharu/ Mbatapuhu
Kec Matawai La Pau/ Katikuwei 16 ACCESS and AIPMNH 3.Findings
3.1.Introduction
This report opens by placing village actors and institutions in context by sketching a basic picture of what citizens (arguably the primary health actors) understand about key maternal and child health issues and policies, and how the realities of village life impact efforts to access care for themselves and their families. This first section of findings does not claim to present an exhaustive picture of citizen understanding or decision making, but aims to trace general contours of knowledge and action. This picture of what people know and do is then overlaid with a deeper analysis of where and how citizens get information and are supported (or not) in making maternal and neo natal health decisions. Researchers were asked to map key actors and institutions, and comment on: the level of power held by the actor/ institution; the amount of trust villagers place in the actor/institution; the incentives facing the actor/institution; and the linkages that the actor/institution has with other health actors or institutions. The second section below presents this village level network analysis, as well as considering actors and institutions at the kecamatan level, and considering key actors and policies being driven from the kabupaten level. Farther away from the village, the study is less exhaustive, but aims to keep a focus on policies that influence village level networks and behaviors. The final section describes issues around data more carefully, since a number of specific issues were raised on the topic. The section considers both Generasi and general health data, examining incentives for collection and use. 3.2.Whatdocommunitymembersknowaboutmaternalandneo‐natal
health?Wheredotheyseekcare?
This section aims to trace general contours of community knowledge and action, roughly following a cycle of pregnancy through to caring for a toddler. Researchers were asked to identify key themes within the village, not count the number of women who knew a particular piece of information, or made a certain decision. As a result, findings are presented in the form of general trends (many women know…) or areas where researchers have noted variation (some women know, others do not…). Examples are provided where they either illustrate a general finding or provide an example of an important exception to the norm. Before exploring the detailed steps below, it is worth making two cross‐cutting observations about types of knowledge at the village level. First, even if an individual knows that they should or should not do something, they often do not know why. For example, while people understand that poor nutrition for mothers and children is bad, it is not clear what “bad” might entail beyond having a skinny child who is often sick. They have little or no knowledge of the impact of nutrition on other aspects of a child’s development. Second, many aspects of health information remain “technocratic,” and there seem to be a number of incentives to keep it that way. For example, information around nutrition is highly managed: menus need to come from the nutrition expert, as does most nutrition information. The provision of special 17 nutrition counseling training for some health workers makes others feel that they are not qualified to undertake it. While important, these additional requirements also seem to have the effect of pulling information and power away from communities. It means, for example, that dukun (who regularly advise women on what to eat when they are pregnant, and are generally listened to when they do) are “too stupid,” according to one health care officer, to be engaged to help share basic nutrition messages for pregnant women. It also means that while village heads and community leaders who do have thoughts about health care in their villages often don’t share them with the health department, since they feel that they do not have the technical expertise to do so. 3.2.1.IthinkI’mpregnant(ormywife/daughter/daughterinlaw,etc.ispregnant).Whatdo
Ido?
The first step to a healthy pregnancy is recognizing that you are pregnant, and for the health department, that means being checked by a bidan in your first trimester, known as ‘K1 murni.’ Most women do know that they should be checked by a bidan, and generally do know that they should do it early. Many women are checked at the posyandu, which means that they have to publicly admit that they may be pregnant. However, some choose to visit the bidan on their own, but they may have to pay. There are, however, a number of factors that push women away from going to the bidan in their first trimester, including the advice of husbands, mothers‐in‐law and mothers. There is a great deal of regional variation: in Sumba, the role of husbands and in‐laws is particularly strong, due to the continued practice of belis, or payments for brides,1 that renders women equivalent to property in some families. The expense around marriage means that men need to save for a very long time, with the result that partnership and pregnancy often precede formal marriage. This can create significant tension between families, and impact upon a woman’s medical care. In Lombok, there is still a practice of engagement by kidnapping, and many girls get pregnant quite young as a result. In some locations, these girls are very embarrassed to be checked by a bidan, in others they are not. In addition, women who are still having children when they are older seem to be slightly stigmatized, and are less willing to come to posyandu. Fear that husbands/ mothers/ mothers in law/ other villagers might find out about a pregnancy that is unwanted or unexpected for social or economic reasons can keep women out of the formal health system. In nearly all locations, health actors and community members alike recognized the difficulty in addressing socially difficult pregnancies: young mothers, old mothers, unmarried mothers, etc. Many women do remain comfortable being checked by dukun, since they feel it is more discreet and they can better understand information that they are given by these women and men. Under Jampersal, there is funding to pay for dukun who refer or bring patients to the bidan to be checked, but not all dukun are aware of this scheme, nor was it clear that all dukun receive the incentive from the bidan. Women should get a buku KIA (maternal and child health book) when they are checked, but not all women know about this book, or if they do get it, how they can use it. There was variation across locations regarding the distribution of the book. In a few locations, women had the book and had read 1
Normally a number of cattle or horses. 18 it. In others, women did remember that they had a book and some were able to produce it, but had not read it or used it to track children’s health. In some locations women did not have it at all. Even within villages, there was variation, perhaps linked to where women were checked, and if there happened to be stock on hand. In one kecamatan in Gorontalo, buku KIA are being withheld as a punishment from women who are not checked in their first trimester. While many women are aware that they should be checked, they are not always clear as to what information they should gain in the process. Many are aware that the bidan should give them a due date, but not all get it. Since there is variation in the use of buku KIA, due dates are not always written down. For women checked after their first trimester, bidan often have trouble predicting due dates. While many dates are correct, some are off by months, and some women were not told a date. Where there is increasing focus on ensuring that women give birth in facilities outside of the village (and particularly in places where they can be fined for going to these locations seven days before their due date) knowledge of this date is critically important to help families prepare. At the same time, given the complications around predicting due dates, it seems equally important to recognize that they are only approximations, even when they are made in the first trimester. Despite the numerous problems around predicting when a baby will arrive, a number of health actors interviewed were unwilling to believe that babies will arrive very far outside of their predicted timeframe. This is perhaps linked to increased pressure for women to give birth in facilities outside of the village and incentives for bidan not to help women give birth at home (and sometimes in the village at all). Whatever the reason for this belief, it makes it very hard to talk about finding solutions to assist women at the village level in the case that something does go wrong. In all Eastern Indonesia study locations, ‘chronic low energy’ (Kurang Energi Kronis – KEK), or poor maternal nutrition, remains a large problem. Some women are aware if they are KEK, and if so, aware that it can lead to complications in pregnancy. Women are dependent on the bidan to determine if they are KEK, so if a bidan does not come routinely to the village, then they will not know. There were a few older cadre posyandu who claimed that they could also determine if a woman was KEK, but this is generally not something that most cadre feel that they can cover. In some areas, there are messages around planting “nutrition gardens,” for maternal health. The nutrition message around these kitchen gardens is generally around eating vegetables for vitamins and minerals. While pregnant women might be encouraged to eat meat or eggs, there seems to be little messaging around ways to get protein or iron. In addition, nutritional support for pregnant mothers does not always match what community members will actually eat. For example, supplemental food for KEK women is generally dictated by the Puskesmas nutrition specialist. Generasi also follows their advice and in some locations provides powdered milk to pregnant women who are KEK. In some locations this is appropriate, but in others women choose not to (or are prevented from) drinking the milk: women in Lombok threw it away since they are not used to the taste and there are rumors in other locations of husbands appropriating the milk for their coffee. 19 3.2.2.I’mabouttogivebirth(orbeafatherorgrandparent).WhatdoIdo?WheredoIgo?
In all locations, there is a large, politically motivated push to reduce maternal mortality. This seems to be driven by the increasing attention given to the indicator both by the Indonesian government and donors as a result of its inclusion in the Millennium Development Goals. The national government’s approach to reducing maternal mortality is to pay for women to give birth with a trained midwife at an ‘adequate health care facility.’2 If a woman gives birth at home, the bidan will not be paid for her services. In addition, there is strong messaging around the fact that dukun are not allowed to deliver babies. Across the research locations, there was variation in the understanding of what an ‘adequate’ health care facility is for the purposed of receiving free care. In Sumba Timor and Gorontalo, women are required to go to a puskesmas or hospital if they want free service. In Lombok, village health posts (polindes) are still considered adequate. In Jawa Barat, the head of the health department indicated that a clean polindes with basic facilities would be an adequate place to give birth, but the puskesmas in at least one kecamatan had forbidden women from giving birth in a recently constructed polindes. There were no written guidelines as to what made a location ‘adequate.’ Even in locations without a clear facilities‐based approach, dukun are being threatened with jail and potential prosecution for murder if anything happens to a mother or child in a childbirth that they are assisting. In addition, there appears to be a growing phenomenon of punitive approaches to ensure that the facilities‐based approach is implemented. In some villages in Sumba Timor, a number of village actors can be fined if they do not take appropriate steps to support pregnant women (see details in the box below). In Gorontalo, women can be forced to give birth at an approved health facility by the police and the military, and in one kecamatan, their children can face consequences (refusal to issue birth certificates), if they do not give birth in an approved facility. As one team researcher wrote, “The push to achieve targets has unintentionally led to the development of rather negative attitudes towards the community, as they are perceived almost as an enemy to their attainment of millennium development goals.” Many health sector respondents stated that community members need to be “scared” or “threatened” into making the right choices, because otherwise they are too lazy or uneducated to do so. Box 2: Village Health Regulations in Sumba Timor In Sumba Timur, village regulations around maternal and child health have been implemented in 74 villages. While they address a large range of health behaviors, they are known mainly for imposing fines on women who do not give birth at a puskesmas (Rp 250.000‐ 500.000 depending on the village) and on dukun who assist women in childbirth(Rp 500.000‐ 1.000.000, depending on the village). They also include fines for: women who are not checked four times during pregnancy, or who are checked by a dukun in the course of their pregnancy; nursing mothers who do not breastfeed for 6 months (with some exceptions, both fines of Rp 250.000); husbands who fail to support their wives’ pre‐natal checks or the nutrition of wives and children by planting nutritious food, including vegetables and nuts for protein and owning at least five hens and one rooster (Rp 500.000‐ 1.000.000, depending 2
Under the Jampersal scheme, births with a trained attendant at an adequate health care facility are free. See Jampersal guidelines, available (in Indonesian) at http://www.depkes.go.id/downloads/PERATURAN_MENTERI_KESEHATAN_JUKNIS_JAMPERSAL.pdf 20 on the village); cadre posyandu who do not do their jobs (Rp 250.000), and bidan who do not do their jobs (Rp 250.000).3 These threats unfortunately do not mean that the quality of the facilities is good, even where huge efforts are being made to improve. Both Sumba Timur and Gorontalo suffer from staff shortages, and more rural locations have trouble retaining staff, as was the case in locations visited in West Java. Though a structured review of facilities was not conducted, locations had important shortcomings. For example, in Sumba Timur, while a main driver behind the facilities based approach is the need to prevent women from bleeding to death in child birth, transfusions can only be carried out at the hospital in the district capital. Even there, there is no blood bank, so women must bring their own donors. In addition, threats do not help many women overcome the very real barriers that they still face in giving birth. There are a number of reasons for this, as illustrated in the case in the box below. Box 3: Health Realities in Sumba Timor Katikuwei is located about two hours from the nearest puskesmas in the enclave of a national forest. Staffing of health workers is limited in Kecamatan Matawai La Pau, as it is across the kabupaten. Until recently, there was a doctor at the puskesmas, but her one‐year contract has just ended and she will be returning to Java. It is not clear when a new doctor will come. Other staff dedicated to the puskesmas include two bidan and a nurse. In Katikuwei, there has not been a bidan or nurse at the village level since 2009. A bidan has been assigned to the village and was supposed to arrive at the beginning of the year. However, as of the end of February, she has not arrived, despite the efforts of the (very proactive) Camat to push on the head of the health department.4 Because of staff shortages and poor roads, a bidan has not visited the village since November, four months ago. Mrs. K believes she has been pregnant since December, but she has not been examined by a bidan. The nurse generally comes every month, but he does not check pregnant women. (He doesn’t really do examinations at all, but does bring immunizations, and provide basic medicines if people know what they need or can describe symptoms.) She heard about a new village regulation that imposes a fine on her if she does not give birth at the puskesmas, but she doesn’t see how it can be applied since she has never been checked and has never been given a due date. Even if she were to be checked, she could only guess when to go to the puskesmas. The bidan who do visit the village change a lot – some are more helpful than others. For example, some of her friends who have been pregnant are given a date, and have a book where it is written down, but others just get told “it’s a few months away,” depending on the bidan that they talk to. The dates are usually around the right time (never exact though), but sometimes they are several months off. Mrs. K knows that she should be examined by a bidan in the course of her pregnancy. So, she goes to posyandu hoping that the bidan will come. She hasn’t gone every time, since it is a waste of time if the bidan isn’t there. If she knew about the need to get “K1 murni,” (or the fact that she could be fined Rp 250.000 under the new village regulation for not being examined four times in the course of her pregnancy) Mrs. K might try harder to go to the puskesmas to be examined. However, even understanding the importance of the examinations, she probably couldn’t afford to go to the puskesmas. In the rainy season, it would cost her Rp 100.000 to take a motorcycle taxi to the puskesmas, far beyond her means. She could also take the truck that goes twice a week to Waingapu (via 3
It is not clear how this will interact with UU 53/ 2010 on discipline of PNS. Under that law, generally only the health department can apply any kind of sanction to a bidan. 4
When the Camat asked that she be disciplined, the head of the health department replied that her salary would be withheld until she arrived (when she would then get all her back pay). As it turns out, the bidan assigned to this village has had at least one, and perhaps two disciplinary problems in other locations. In the first case, she was not present in her village when a woman needed assistance in childbirth. This contributed to delays in a problematic pregnancy which ended with the death of a child. She was relocated. 21 the kecamatan). It costs considerably less (Rp 20.000), but only goes on Tuesdays and Saturdays, so she would need to find somewhere to stay until she could come back. Since she should be checked, she has gone to the dukun who lives in her neighborhood. The dukun is an older lady, and she learned about childbirth from her mother. She can explain everything that Mrs. K should do be sure that the baby is healthy and the birth is fast and smooth. After talking to her, Mrs. K knows that she should rest in the afternoons, and try to eat a little more. The dukun has massaged her, and says that the baby is in the right place. Her neighbor, the head of BPD, went to a meeting about that village regulation and says that women who don’t get checked could be at risk. His own wife is heavily pregnant and he is very concerned. He even asked the nurse if the puskesmas is “putting the women of this village at risk by not checking them?” The nurse just said they were short of staff. His wife, Mrs. J, has additional concerns. She knows that under the new village regulation, her family could be fined Rp 500.000 if she does not give birth at the puskesmas. She could go and stay with relatives near the puskesmas, but might be embarrassed if she and her husband have to stay too long. She has heard that there is now a ‘waiting house’ to stay in just behind the puskesmas, but if she stays there she will have to make sure that she can take enough rice with her for herself, her husband, her mother‐in‐law and her two small children. Someone will need to keep an eye on the chickens and feed the pig. There is money from Generasi to help her with expenses, but it will only come after she has given birth, so she needs to make sure that she can get enough before she leaves. To do that, she will have to borrow from family members. Luckily, they know she can slowly pay it back, so they will give it to her. Poorer villagers can struggle to raise this money. Even if she decides to go down to the puskesmas around her due date, she needs to be ready in case the baby comes early. Luckily, even though the village is far and the roads bad, the puskesmas has an ambulance that can make it to the main road of the village. Unluckily, there is no phone signal in the village. Most people go to the preacher, who has a phone and is willing to drive into the forest to call the puskesmas. If her husband is at home, she will ask him to walk down the hill, across the river and back up to the main road and the preacher’s house. Usually, the preacher can reach the puskesmas, and they always say that they will send the ambulance. However, sometimes it comes and sometimes it doesn’t. As long as it doesn’t rain and her contractions are not too severe, she will walk out to the village market herself to wait for the ambulance. If there are any problems, however, she will try to get help from the dukun that lives near her, since it will be too late to get help from the bidan, and she is too afraid to get on a motorbike when her labor has started. She has heard that there are also fines for the dukun (actually Rp 1.000.000) if they help women give birth. This is making some dukun nervous, but many are still willing to help. At least they have some experience. It would be scary to give birth alone with just her husband to help, as her cousin did a few months ago. The case of “Mrs. K” combines the experiences of a number of women in Katikuwei village, highlighting particular challenges in a village that does not have a bidan and in a kabupaten that faces staff shortages. However, even in West Java, transportation is an issue for pregnant women. Women in less rural areas are able to wait until they feel contractions before they go to the puskesmas, but this can also be complicated over rough roads on the back of a motorbike, even for a short distance. There is considerable variation in the amount of information that pregnant women receive from the bidan. Some of this information, such as the due date, can have a huge impact on what happens at the moment of childbirth. Even when it is correct, babies do consistently arrive early, so both women and the health systems have to be prepared to respond to a variety of situations. In all locations, the Desa Siaga program is designed to help families save money for childbirth and ensure that they have community support to get to the health facility when needed. With limited exceptions, this program does not work well. Even where awareness was highest (a village in Sumba Timur), people were unwilling to implement systems of transport that they had theoretically established, usually due to bad roads or fear that something would happen to the pregnant woman if she was on a motorbike. 22 Systems for identifying risk (green, yellow or red flags in front of houses of pregnant women based on risk levels) did not seem to change community responses in a discernible way though they may contribute slightly to a better understanding by a woman of her own risk level. While there are many similarities between locations, there are also some key differences in what individuals know and do when they need to give birth. In Lombok Utara, access is less of a problem, there are generally sufficient staff, and many privately practicing bidan, who can also access Jampersal funds for childbirth. The result is considerably greater choice for women and discussions between women about who the better bidan are so that they can select based on quality of care. 3.2.3.Iamthemother(father/grandparent)ofababy/toddler.WhatdoIdo?
Many mothers know that their babies should be checked by a bidan as soon as possible after birth. This initial check is combined with an initial immunization, which many mothers are also aware of, cited by some as an advantage of giving birth at the puskesmas. Once mothers and children return home, further checks, on‐going weighing and immunizations are meant to be conducted at the posyandu. The study did not examine knowledge about exclusive breastfeeding, but did find policies in place in Lombok Utara to try to prevent bidan from selling formula. Posyandu attendance varies greatly, though most women are aware of posyandu (as long as it is functioning). There are a range of reasons why families do or do not take their children to posyandu, including distance (including obstacles like rivers or steep hills which might be difficult or dangerous to navigate with small children) and time. It is generally women who take children to posyandu, though if they need to work, then children may be taken by grandparents or in exceptional cases, fathers. Though Generasi often spends money on food at the posyandu to get women and children to come, and sometimes supports women to go to cooking classes so that they can prepare more nutritious food, information about what constitutes nutritious food remains limited. Some of the reason for this may be cultural. In Sumba Timur, for example, many mothers and children suffer from undernourishment. Interventions on nutrition have focused on planting vegetables, but the consumption of protein (including plant protein) remains low. While families might have chickens or pigs, and even cows or water buffalo, large animals are reserved for adat and smaller animals for guests or ‘insurance.’ If meat if consumed, men get priority on the nice bits. In Lombok Utara, many women consider instant noodles nutritious, especially if they are served with an egg. This notion is cultivated by advertising and in a few villages also supported by the provision of instant noodles as supplemental food at the posyandu (by Generasi). The amount of information that people receive at posyandu also varies, though in general ‘counseling’ is fairly limited. Women often know how much their children weighed at the last posyandu, but don’t always know what that means in terms of their health (e.g. are they underweight or healthy). Only if children fall below the red line are parents informed that there is a problem. In those cases, supplemental feeding may be provided through Generasi or under routine government funding, though there can be disbursement problems in both systems. There is no discussion with parents whose children are in the ‘yellow,’ or whose weight continues to fluctuate. Many cadre posyandu are aware 23 that children in the ‘yellow’ are having a problem, but because it is not reported to the health department, bidan (and therefore cadre) do not focus on these children. If a child does fall below the red line and is reported to the health department (and Generasi), most families know that they may get supplemental food. They do not generally know that their child should have an additional check by a nutrition specialist and perhaps a doctor to figure out why the child is not gaining weight. In very serious cases where children are hospitalized they do get a medical examination, but it does not seem to be happening before that. Though Generasi provides considerable amounts of supplemental feeding, and is drawing more attention to undernourished children, there does not seem to be any discussion of underlying illnesses in the village planning process. This is true also for fairly basic problems like de‐worming. 3.3.Whataretherolesofactors,institutionsandpoliciesfromthevillageto
thedistrictlevel?
This section presents a network analysis of health actors and institutions. It focuses primarily at the village and sub‐district level, but also includes a consideration of policies and structural factors at the district level that have an impact below. In each village, researchers mapped health actors and institutions, including descriptions of: their role in providing health services or information in the village; the amount of power that they command; the amount of trust that they enjoy; and how they link to other key actors within the village or broader health system. This mapping is presented in a general form below. 3.3.1.Whataretherolesofvillageactors,institutionsandpolicies?
The network map below presents a stylized description of the key actors and institutions at the village level, and the relationships between them. The position in the map represents a hierarchy of sorts, with the bidan being the most central person in the village for support and information about maternal and child health. Those actors or institutions with a clear, active role in maternal and child health are darker than those which are not particularly active in service delivery or the provision of information, regardless of their mandate, though these are not without exceptions. While the village head is a powerful actor, and high in village hierarchy, he or she is not often strongly involved in health. A solid line between actors indicates a clear relationship that is related to health. Usually, this is a reporting relationship, but it is not always so clear. For example, the dukun does not technically report to anyone, but under new policies, should be working very closely with the bidan. Dashed lines represent a weak or theoretical relationship, or one that is not related to health per se. For example, the village head may appoint cadre posyandu, but not do it for health‐related reasons, or use them to get any health information. 24 Figure 1: Village and Kecamatan Health Networks There are actors that are not included in the network map above, primarily those in the ‘private’ sector. There is a huge variation in these actors, ranging from legitimate bidan in private practice to retired health care workers practicing at home to the functional equivalent of snake oil salesmen. In addition, the media and NGOs are not included, as they are often working at the district level. More detailed pictures of each of the actors are painted below. 3.3.1.1.Bidan
Role/ information provided In the eyes of the department of health, the bidan should be the main source of information regarding maternal and child health in the village. While she does often occupy an important role in providing information, she can also face challenges in having the power to be heard, as discussed below. Mothers (and expecting mothers) see the bidan most routinely at posyandu. If there is a bidan in the village, they may also see her independently at the village clinic and she may conduct outreach activities. In all locations, if posyandu attendance is low, the bidan should support the cadre posyandu to conduct sweeping. Across the villages studied, there were different levels of outreach and sweeping. In several locations, bidan did not make much of an effort to visit more remote locations outside of attending posyandu. In Lombok Utara and Gorontalo, attention by the health department and politicians to attendance rates at posyandu increased sweeping and outreach activities. In one village visited in Lombok Utara, the bidan 25 was particularly active with outreach, finding budget to hold three mobile clinics per month in more remote parts of the village. Across locations, key messages being provided by bidan were: ‐
‐
‐
The need to give birth with a health worker (a bidan) at an approved health facility. The fact that dukun are no longer allowed to assist with childbirth, other than providing very basic support to the attending bidan (boiling water, etc.) and moral support to the mother. Very basic information around nutrition, usually at the level of identifying problems (KEK or undernourished children) and providing some information as a solution. With some excellent exceptions, there is a tendency for bidan and health workers generally to take a very clear attitude that ‘they know better.’ This makes health messaging come across as either condescending or aggressive. With increasing pressure around targets (for example, zero maternal mortality), some bidan seem to treat communities as obstacles to their own performance. The response to this feeling is generally not to provide communities with more information, but to coerce them in some way into doing what they need to do. This approach both supports and is supported by growing punitive approaches within the health system, as detailed above. In addition to providing medical advice, bidan are responsible for compiling data from the village and sending it to the puskesmas. As will be discussed in more detail below, this data is generally not particularly well shared at the village level. Power Bidan occupy an complicated position in the village: they are an ‘insiders’ to the extent that they often live in the village, but are ‘outsiders’ in the sense that they are generally not from the village, are more educated, and are usually removed from local networks, family ties, etc. Their ‘outsider’ position is compounded within the structure of village government by the fact that in most locations, village heads see them as reporting to the department of health, and therefore outside of village control. There are some exceptions to this, as demonstrated by the case from Lombok Utara in the Box 7. Contributing to the insider/outsider dynamic, a bidan’s position as part of the elite or as a trusted source of advice is often complicated. Bidan can command respect due to their level of medical education. However, many bidan are young, unmarried and without children of their own. Many women are uncomfortable taking advice on their pregnancy from a woman who has never had a child herself. This puts younger bidan in a difficult situation regarding older counterparts or experienced dukun. In addition, even if they are trusted by the village government to provide advice to women in the village, they are not necessarily viewed as someone worth consulting in village decision making processes. There are both advantages and disadvantages to being an ‘outsider.’ In a number of locations, bidan were also ‘outsiders’ in the fact that they could not communicate in the local language. In rural Indonesia, many women (and some men) only speak and understand basic Indonesian, meaning that it is difficult for them to express themselves (to describe symptoms or ask questions, for example) and to 26 understand more complicated or technical language. These gaps make information sharing nearly impossible. However, in other locations, where bidan were trusted, the fact that they are slightly outside of family and other social networks give them greater ability to deal with difficult cases such as young pregnancies or those out of wedlock. It is difficult to objectively tell what makes for an ‘insider.’ One excellent bidan in Gorontalo came from the village that she served, and remained dedicated for many years. Another well‐trusted bidan in West Java came from a different village, but was able to settle in well. In Sumba, a bidan who was posted to her home village had still failed to appear after two months. In another village in Sumba, the bidan, from a different region of the same district, married a man from the village that she was posted in, but still did not speak the local language after being there for over eight years. Much seems to depend on individual attitudes around providing service and the fit of the bidan with the community. While these factors seem very important, there does not seem to be much discussion from the health department about what makes for successful engagement with communities. An enhanced focus on encouraging women to give birth with skilled health care workers can increase the power of bidan, as they are a key point of engagement with the formal system and can support access to free care during pregnancy and delivery. As the advocate of the formal system, they also have some level of power vis‐à‐vis the dukun. However, it can also make it difficult for them, since they are no longer supposed to assist with deliveries at home, and in some cases may face punitive measures for doing so. This reduces their ability to appear responsive, and may result in some women pulling away from formal care. Trust The level of trust from villagers to bidan is mixed. As noted above, if they are young and unwilling to assert themselves, they may not inspire much trust. However, there are a number of examples of very active and dedicated bidan who make an extraordinary effort to ensure that women get service, particularly during childbirth. These women are generally trusted. In areas where there are low staff numbers and bidan are more difficult to access, trust does not always translate to use for childbirth, but may still support being checked. Incentives There are a number of factors that influence bidan behavior. Financially, key components of their income include a minimal base salary, with an additional incentive if they are in a remote location, payments under Jampersal for performing pre‐natal checks and assisting with deliveries, revenue from private practice, and often revenue from the sale of goods (formula, medicine, etc.). Local policies shape these incentives, and provide the additional incentive of political attention to and supervision of particular outcomes, such as decreasing maternal mortality or increased posyandu attendance. While there is interaction between the incentives above, this study focused on those aiming to promote specific bidan behavior on maternal and child health. Nationally, this is driven by the Jampersal 27 program, though implementation has been modified in many districts. In addition, there are several local initiatives, described below. Nationally, the Jampersal system incentivizes bidan to ensure women deliver in approved facilities. Per the implementing guidelines, they receive Rp 20,000 for each of four pre‐natal checks and four ante‐
natal checks. In addition, Rp 500,000 is provided for medical services and accommodation during pregnancy. In some locations, most or all of these incentives go to the bidan, while in others, there is a greater focus on supporting delivery services more broadly by sharing the money between multiple actors and allocating some money for medical supplies. This usually depends on policies around where a woman is required to give birth. If the village clinic is acceptable, then the village bidan is usually incentivized to support women to come to the clinic, can attend to them herself, and receives payment. This is generally case in Lombok Utara. If women are required to go to the puskesmas, incentives become more complicated, as described in the case of Sumba Timor, below. Box 4: Paying for Deliveries in Sumba Timur To be eligible for a free pregnancy under the Jampersal program in Sumba Timor, women need to go the closest puskesmas. Village clinics are considered insufficient, and the district has made a significant effort to upgrade the quality of puskesmas to handle deliveries. It is also making an effort to ensure that the puskesmas are sufficiently staffed, both to provide adequate services to pregnant women and ensure ongoing learning for bidan. Given that there are significant staff shortages, they have done this in some locations by asking for village bidan to come to the clinic on a rotating basis for a certain number of days per month. Rotation ensures more equal distribution of Jampersal funds, since the majority of the money is paid to the lead bidan during delivery. If a woman is in labor, the village bidan can choose to accompany her to the puskesmas, but may not be in charge of the delivery. To support puskesmas costs, the district health department has also issued a policy that requires 25% of the delivery fee to go to the government to cover operational expenses at the puskesmas. The remaining 75% is divided between a number of actors at the puskesmas level, including the head of the puskesmas, attending nurse, other bidan, cleaning service, etc., with the main bidan receiving around 50% of the remaining 75%. This approach has resulted in some frustration among bidan. In one research location, the village bidan was part of group that protested the amount provided to the head of the puskesmas (7.5%), on the basis that he was not involved in the delivery process directly. She was also frustrated by the requirement to be at the puskesmas, since it meant that she had to bear additional travel expenses and lost a day of (revenue generating) work in her village to sit in the district. She didn’t feel that she received enough from Jampersal on the rotating basis to make it worth her while. In addition, service was not available to villagers on the days when she was not in the village, since she also functioned as the village nurse. She had started to simply ignore the schedule of her days in the district level, claiming she has too much to do in the village. Lombok Utara has set up some different incentives for bidan that are worth mentioning. The head of the puskesmas is trying to improve pre‐natal checks. He is requiring that all bidan provide more checks (nine checks as opposed to the usual four). They receive more money for doing so, but get nothing if all checks are not completed. This is making bidan more proactive in searching for “K1 murni,” in some cases working more effectively with dukun. There is, unfortunately, a downside to this approach, as it creates a disincentive to check women who are farther along in their pregnancy, since the bidan will get nothing. However, it is not uncommon for women who work in the city (Mataram) to return to their villages late in their pregnancies to give birth and have family support for the first months of motherhood. Incentives work against serving these women. 28 In addition to positive incentives, there are a number of negative incentives facing bidan. In nearly all locations, there is considerable pressure to reduce maternal mortality, and bidan may be called to explain themselves if a woman dies. Political pressures are also building around child mortality and malnutrition. As discussed in the next section, while this may influence service delivery, it may also incentivize falsification of data. An exception to the lack of accountability is visible in Lombok, where at least one village head has taken a larger role in oversight, as described in Box 7. In that location, which was peri‐urban, and a desirable place to be located, the threat from the village head that he would ask for the bidan to be replaced seemed to be credible. In general, however, sanctions against underperforming bidan remain very difficult to implement. While the village regulations in Sumba Timur contain fines against bidan who do not do their jobs, it is not clear how the application of those fines will interact with the law on civil servants (UU 53/2010). While the head of the health department in that location stated clearly that she was willing to apply the civil service law for bidan not at their posts, she did not do so when faced with a clear opportunity. While the Bupati in Gorontalo is taking clear steps to try to improve accountability, it is difficult to reach staff sitting under the line ministries (such as health). Interactions/ Networks To varying degrees (and with varying amounts of awareness), all bidan are engaged in a debate about modernization. Many of them see their role as moving mothers and families away from traditional practices into what they see as a ‘more correct’ scientifically based approach to medicine. Their (personal) approach to navigating the intersection of ‘modern’ and ‘traditional’ medicine also shapes their interactions and relations with community members and dukun. Cadre Posyandu. The main point of interaction between the bidan and the cadre is at the posyandu, making sure that basic activities are running as planned and gathering data. Outside of posyandu, there are no formal ways for the bidan and the cadre to interact, such as a monthly meeting. Routine meetings were reported in one village in Jawa Barat, but with the previous bidan. They had not been continued. Despite the fact that the bidan and the cadre are both in the village, there is no attempt to use the bidan to provide training for the cadre. This likely has to do with the fact that the bidan and the posyandu have different reporting structures, with the posyandu under the authority of the BPMD, and the bidan reporting directly to the Health Department. While the Health Department is charged with providing technical assistance and input to cadre posyandu, there is a strong feeling across government (and arguably, across society) that training has to be formal, and the idea that the bidan could support ‘on the job learning’ is not always well received. This formality seems to be normatively supported by a certain hierarchy of holding information within the Health Department. This affects bidan willingness to share information and impacts cadre willingness to ask questions. Puskesmas. The puskesmas provides primary support to the bidan and in some areas (where there is not a bidan in the village) assumes the role of the bidan. Generally, information is shared through the 29 monthly “mini lokakarya” (“mini‐workshops”). A good quality puskesmas aims to ensure that the bidan has what she needs to do her job in terms of medicine and routine immunizations, etc. Generally, relations with the bidan coordinator are important in ensuring good relations between bidan and the puskesmas. Puskesmas policies and engagement in the village can dramatically impact on some bidan behavior. For example, if the puskesmas is engaged in supporting and monitoring sweeping, the bidan will generally take it more seriously. With respect to nutrition, the bidan should coordinate with the puskesmas to ensure that undernourished children are checked after they are reported. This often does not happen due to low levels of operational funding (and often shortage of nutrition staff at the puskesmas). Dukun. The relationship between the bidan and dukun in the village can be characterized as a power struggle with the state sitting behind the bidan. In the absence of the state, the dukun and the bidan compete for patients, but now bidan could essentially report dukun and potentially start a prosecution. Despite the prevailing state attitude to dukun, bidan do adopt a range of approaches to working with them, from one that is dismissive to one that considerably values their experience and connection with the community. In the latter case, which is rare, dukun are provided with a larger share of the Jampersal incentive. Village government. The linkages between the bidan and the village government are generally weak, though with some exceptions. Even in locations where there are routine village meetings, the bidan is not invited to present data, discuss health problems/needs, or get government information. There are certain exceptions to this. In Gorontalo, village heads have ‘performance contracts’ directly with the Bupati. As part of the performance structure, village heads ask for health information in their village, which they get from the bidan. Unfortunately, at least two village heads admitted that when they do get health data from the bidan, they cannot understand it because of the technical language/ abbreviations used in reporting. In Lombok, research uncovered an excellent example of a village head committed to ensuring bidan performance in his village, as described in Box 7 under the analysis of the village head, below. Generasi Actors. In general, the interaction between the bidan and Generasi actors centers on data collection by the Generasi actors. In some villages, data is collected directly from the cadre posyandu, in some it is collected from the bidan. There is also generally coordination around treatment and care for malnourished children who are being supported by Generasi. While bidan are meant to be engaged as part of the planning process, there seemed to be variation in levels of engagement. In one location in West Java, the bidan was involved in reviewing the design for a polindes being funded by Generasi. She requested the design be changed, since it was inappropriate. It is unclear if it was changed, since her input came late in the process. It is also unclear if it will be accepted by the puskesmas as an adequate health facility where women will be allowed to give birth. PKK, Desa Siaga, BKKBN etc. In general, no other community health programs appeared to work all that well at the village level (see the exception in Box 7), and the level of interaction with the bidan is limited. 30 PKK has funding, and should be working with posyandu, but often this does not happen or does in a very nominal way that does not involve the bidan. In two villages, one in Jawa Barat and one in Sumba Timur, posyandu were used to gather money for health related initiatives, with collected money being managed by the bidan. In Jawa Barat, it was earmarked for an Islamic group (whether it could be used for health was debated), and in Sumba Timur money was used for Desa Siaga. Leaving aside the goals of the respective programs, there is some risk involved in bidan managing community money, since if it goes badly, trust in her can be eroded (as was the case in Jawa Barat). However, since the posyandu is technically a ‘community institution,’ there is space for it to be used flexibly. 3.3.1.2.CadrePosyandu
Role/ information provided The cadre posyandu are the core of community health efforts. They speak the local language, know their neighbors, and if they have information, can help to share it effectively. Their role is to conduct posyandu activities, at a minimum entailing weighing children and providing counseling to mothers and pregnant women on health and nutrition, (in reality mainly just weighing children). In addition to engaging with families that come to posyandu, they are also meant to conduct sweeping to ensure those that don’t come still receive service and/or understand that they should join posyandu activities. There remains huge variation in the quality of cadre posyandu, despite increased efforts by programs like Generasi to increase cadre training, which is filling an important gap. There seem to be a number of reasons for this. First, in some locations, there is high turnover among cadre. Many cadre are young, unmarried women. When they marry and prepare to have children of their own, they discontinue their work as cadre posyandu. In addition, cadre are generally appointed by the village head, so they can be changed at will. This can be useful if a village head is aware of underperforming cadre, but it also means that appointments can be political and cadre can change with elections. Second, training may be poor, inappropriate or impractical. Training provided by government is only for a day or two, and usually only for one or two cadre per village. Part of the weakness in training may be linked to the way it is funded, with money coming through BPMD and technical support from the health department. From most villages, the same people get sent to government trainings, so there are some cadre who are fairly knowledgeable. However, since there are no meetings between cadre, there is no way to share information between them, and knowledge remains siloed in a few key individuals. Trainings supported by programs like Generasi are generally a week long, and conducted for all cadre. While these trainings are useful, there were indications in Sumba Timur that they were not well targeted for the rage of people attending: many had low levels of education and could not understand Indonesian well enough to get much from the training. Others understood what was taught, but then were not allowed to take materials with them (buku PMS‐ for tracking the weight of children), and were not made aware where else they could find them (in buku KIA). 31 An interesting finding of the study was the presence of male cadre across all locations except Java. Male cadre seem to be chosen since they can more easily access difficult locations and carry things (mainly scales). There is recognition that there is a need to talk to fathers about maternal and child health, especially in areas where husbands have a large voice in health decision making or in nutrition in the home, but it is not clear how well male cadre posyandu are reaching out to men (vs taking traditional approaches to sweeping). An additional interesting finding was the approach to cadre taken in Lombok, as described in Box 5. Box 5: Dusun Heads as Cadre Posyandu Both villages studied in Lombok Utara took the approach of having dusun heads as the head cadre posyandu. In one village this practice was established formally by the village head, who issued a village decision letter (SK desa) setting the policy. In the other village, the dusun heads were traditionally the head cadre posyandu since posyandu were held at their houses. Many of the posyandu have now moved, but the practice still remains. A village secretary explained the importance of the dusun head as follows: The dusun head is normally better known by citizens. Because of that they are very strategic to involve in posyandu. The dusun head should not just be examples or people that just write letters and sign things, they also have to get out into the community and become leaders. If the dusun head is given a role at the posyandu, then he can directly take responsibility for health services at the posyandu. Power and Trust As volunteer community members who have been provided with additional training, cadre posyandu do not have a great deal of power within the health system or the village government system. They may have slightly higher levels of education than average in some communities, but are not generally part of the top village elite. In some locations, a few of the cadre posyandu may be drawn from the elite (or their children), and this creates a privileged group within the cadre. If this happens, the elite cadre are more likely to get sent for training. Cadre posyandu do have some power within the community, due to the fact that many of them are still considered people who will help community members. They can refer to local examples or role models that are meaningful to local citizens. The fact that they are drawn from the community can be a weakness in some moments, particularly around difficult cases, such as pregnancy out of wedlock, or young or old mothers. In these cases, they may find it difficult to escape community stereotypes to effectively help. In addition, they can have trouble reaching out to the elite. In one village in Lombok Utara, for example, the weakest posyandu was in the elite neighborhood, despite the fact that children were not without health problems. The same was true in Jawa Barat, where the grandchildren of both the Camat and the village head were in the “yellow” area, but all cadre were afraid to tell them. Incentives Financial incentives are generally quite low for cadre, though there have been some efforts to provide increased funding in certain locations, both from government and under Generasi. ‘Basic funding’ 32 hovers around Rp 1.000.000/ year, which is then divided between operational funding and incentives for cadre (usually five cadre per posyandu). This works out to around Rp 10.000/month per cadre. Both Lombok Utara and Gorontalo have a greater focus on posyandu attendance, and provide more funding to posyandu, particularly for sweeping and outreach, which is conducted in large part by the cadre posyandu. In addition, in many locations, Generasi is supplementing the honorarium of cadre to ensure they have enough money to cover costs to do sweeping and identify underserved individuals. Perhaps the strongest incentive for cadre posyandu is attention by government (generally the Bupati or head of the health department) to posyandu attendance rates. If those rates are being tracked, there is pressure on the puskesmas to increase attendance and subsequently on the cadre posyandu to conduct sweeping and outreach. In both Lombok Utara and Gorontalo, the puskesmas joins to conduct sweeping, which provides an additional incentive for cadre. Interactions/ Networks Cadre posyandu. There is generally no time for cadre to think together about what the data is showing at any given location. Cadre posyandu generally write down the weights of children, often re‐copy it to the register, and then provide the information to the bidan. They do not discuss health issues that they are seeing, nor do they talk about what advice to provide to community members. If they do, it is generally done within a posyandu, not between the multiple posyandu in a village. If one of the cadre has had training, this may be shared with other cadre and community members, but this is done on individual initiative, rather than in a systematic way, since there are no routine forums for sharing information. Bidan. As mentioned above, interactions between the bidan and the cadre posyandu are often limited to data exchange, and there is no attempt to use the bidan to provide additional training. Cadre may be uncomfortable asking the bidan questions, since she is more educated and elite than they are. Puskesmas. The puskesmas supports (formal) training and may provide additional support to cadre posyandu, particularly if there is not a bidan in the village. As mentioned above, puskesmas engagement in posyandu can support cadre. However, similar to the bidan, there is very little information provided to the cadre on a routine basis, and they are not supported to attempt any analysis of village level issues. Village government. With the exception of the villages studied in Lombok, discussed in Boxes 5 and 6, there is little routine interaction between village government and cadre posyandu. However, this may be becoming more complicated as village heads are being asked to take a more proactive role in health oversight. For example, with the passage of the village regulation on maternal and neo‐natal health in Sumba Timur, the village head is meant to track (and potentially fine) women who do not come to posyandu. Cadre posyandu are aware of this, and some suggested that that they might simply fail to provide that information to the village head (the village did not have a bidan). It was hard to tell if that approach would be considered resistance to the village head or a sort of administrative complicity with a village head who did not actually want to fine women. 33 Generasi facilitators. In some cases there is overlap between cadre posyandu and Generasi village facilitators. Cadre are generally aware of the facilitators and the program, and usually involved in Generasi planning processes. Their main interactions with Generasi include ensuring that supplemental food is provided for posyandu (where planned) and sharing data with facilitators. In some locations, cadre posyandu have also been supported by Generasi to complete more extensive training. Desa Siaga. Many cadre are also part of Desa Siaga networks, and some collect data. However, since Desa Siaga is largely inactive, the role of cadre outside of providing data is minimal. Dukun. Cadre posyandu have no formal interaction with dukun. 3.3.1.3.“KnowledgeHolders”
In every village, researchers came across certain individuals who have a long history working in health and are knowledgeable on maternal and child issues. A surprising number of these individuals are former birth control cadre (cadre KB). They were chosen by village heads in the 70s and 80s, provided training and sent to work in their villages. At that time, being selected seems to have held some prestige. Many of these individuals continued being active in health, often transitioning to become cadre posyandu, health staff or now, key Generasi actors. Some constructed a “health elite,” described below. Others continued to be active in public life in other ways: of the eight villages visited, two of the village heads, one village secretary and one preacher were former cadre KB. Some ‘knowledge holders’ are not former cadre KB, but are just individuals who take a particular interest in health. For example, Pak Nyoman, a cadre posyandu in Lombok, became a cadre in 1995. He was interested in health and wanted to learn. Though he works as a day laborer, he is an active cadre, he often gets selected to go to additional training. He has been selected as one of the best cadre in Lombok three times. Much of the knowledge that these individuals hold is extremely practical. For example, in one village in Sumba Timur where most cadre posyandu were not aware of where to find the growth chart, (in the buku KMS or buku KIA), one older cadre was able to simply explain how much a child’s weight should increase per month, based on different ages (0‐3 months, 4‐6 months, above 6 months). This information was roughly correct. Because other cadre were not aware of these rules of thumb, they did not know if the children they were weighing were doing well until the bidan had looked at the data and told them one way or the other. Since there was not a bidan in the village, that information came with delay. These actors are worth mentioning because of how they are currently not utilized, or even recognized within the village. They represent an asset to the village in terms of their health knowledge, but they do not necessarily share information, even with other cadre. Reasons for this include the lack of a forum for sharing and perhaps a desire not to stand out too much. Identification and support of these individuals may enable them to provide deeper assistance to the village. 34 3.3.1.4.Posyandu(theinstitution)andPuskesmasactorscomingtothevillagelevel
Role/ information In general, most people think of posyandu as the place they go to have their children weighed, and get immunizations and vitamins. They are generally correct. As mentioned above, counseling and information provided at most posyandu is minimal. While this study was not an in‐depth coverage of posyandu, it is worth noting a few examples of well‐running posyandu, as well as providing a general picture of some of the challenges for the posyandu. At a practical level, it is important to note that most posyandu, as a collection of mothers and children under five, are sites of barely controlled chaos. This helps to put some of the weaknesses of the institution into perspective. Providing counseling, cooking instruction or even basic information in this environment is challenging at best. In Kec Batudaa Pantai, in Gorontalo, the puskesmas is working hard to overcome many of these challenges. They are using different approaches to counseling, pulling pregnant women off in small groups and talking to them. In addition, they have a table providing malaria kits, which is noteworthy, since the provision of information about general health issues is very limited. A challenge in all locations is attendance. In a number of locations, new posyandu have recently been added to make it more convenient for villagers to attend. It is not clear if this is coming from a national policy, since it occurred across locations. In Lombok Utara, in an attempt to make posyandu more interesting for women, one dusun head has started a revolving savings scheme for those who attend. The same village head has asked the puskesmas for films that can be shown at posyandu. Unfortunately, they only have one film that everyone has seen. Power At a structural level, the posyandu is meant to be a cross‐sectoral activity at the village level. It sits under BPMD, with technical support from the health department. Recognizing that other parts of government should also be responsible for aspects of posyandu, the health department is loath to take on a bigger role or contribute too much funding and continues to insist that it be community owned. In the words of one respondent from the health department “If we do everything, the community will just be spoiled.” It is likely that these structural factors contribute to underfunding and reluctance of the health department to engage further with posyandu, despite the fact that many health care actors recognize its critical importance at the village level. 3.3.1.5.Dukun
Role/ information provided The title ‘dukun’ is given to all traditional midwives, but it hides enormous variation in the people who are assisting pregnant women. They are generally older women and men. Some dukun were formally trained by the government and provided with bidan kits, generally in the late 70s and 80s. Some have learned from parents or other relatives who have been delivering babies for years. They do something of an informal apprenticeship with the older dukun and then begin to practice themselves. Finally, some 35 are simply older family members who have been ‘pressed into service’ for family members with some success. Pregnant women turn to dukun at many points‐ before they are pregnant, they may go to the dukun for help conceiving or for advice on influencing the gender of a child. If they become pregnant, many women still prefer to go to a dukun at the beginning of their pregnancy, since they may be more comfortable talking to them. Many women still go to dukun during the course of their pregnancy for massage to help them be more comfortable and to ensure that the child is in the correct place. Even if they do not give birth with dukun, they may still take their children to dukun if they are ill. Dukun provide advice on everything from where and how to sit (not in a doorway, lest the child take long to come out: Sumba Timur) and what to eat for easy pregnancy (no fish, eggs or meat: Lombok Utara). They also have important traditional roles. In Gorontalo, it is the job of the dukun to present a child to the devils that exist everywhere. If the mother takes the child outside in its first months of life, she should be accompanied by a dukun so that the child is protected. For many of these services, dukun are paid, either in money or in kind. Fees for childbirth range from Rp 150.000‐ 300.000, and in some locations vary by the sex of the child. In addition to being “birthing” dukun (dukun beranak), many dukun also function as traditional apothecaries (dukun berobat), and can provide traditional medical solutions to a number of problems. Current government policy with regard to dukun focuses almost exclusively on reducing the role of dukun in childbirth. Under current policy, dukun can receive incentive for bringing women to the bidan for initial checks (K1 murni) and accompanying them to give birth. However, if they do more than that, they may face sanctions under local regulations. The medical establishment has made it clear that if they do assist women in delivery and either the mother or child is harmed or dies, the dukun can be prosecuted and sent to prison. In Sumba Timur they can be fined under the village regulations mentioned above. In one kecamatan in Jawa Barat, all trained dukun were required to bring the contents of their bidan kit (provided by the government in the 70s/80s) to the police for inspection, at which point all tools were seized. This action was nominally carried out under a district regulation on maternal and child health. However, the regulation does not yet have implementing instructions, which seem to have been pre‐empted in that location by sub‐district leadership. In all locations, ‘partnerships’ between dukun and bidan focus on limiting the role of the dukun, and training consists of making boundaries clear. Despite the fact that dukun are involved in more than childbirth, there is no positive engagement with them to support other aspects of pregnancy. For example, they are not provided with basic information about maternal and child nutrition, despite the advice that they provide to mothers. When asked about the possibility of taking such an approach, one respondent in Sumba Timur replied, “They are too stupid to understand nutrition.” This fairly accurately summarizes much of the government’s attitude toward dukun. 36 Power and Trust Dukun in most locations continue to hold power in a traditional way, but they are under considerable siege by government actors. Many dukun are afraid of government actors and some have reduced their roles in maternal health. However, they generally remain trusted actors in the provision of care for women. Importantly, even if they are questioned by younger women, they still have the trust of parents and in‐laws who have watched them assist in the births of generations of village children and recognize them as having other traditional powers as well. Since parents and in‐laws are important actors in health decision making, this trust is important. Regardless of generation, many people still see an experienced, older woman without training as preferable to a young women who does not have children of her own, even if she has medical training. Networks and Interactions Bidan. Dukun tend to characterize the ‘partnership’ with bidan in some form of fear. Many dukun have been told that they can be prosecuted and sent to jail if anything happens to a woman or baby that they treat. Most seem to think it would be the bidan who would report them, since she is the arm of the state charged with overseeing childbirth, and the person in most direct competition with the dukun. In one village a dukun told a bidan present in an interview directly “I am afraid of you.” In a place where little is so direct, this is a very strong statement. The other main feeling of dukun is that bidan take advantage of them. Dukun provide support during childbirth, and are busy during the birth preparing water etc. They then generally stay to take care of the mother, while the bidan returns to work or rests. The bidan then claims the Rp 500.000 fee from Jampersal (subject to local regulation, as clarified above). There is some evidence that dukun are willing to be more substantial partners if they are paid better. In Gorontalo, for example, one dukun working closely with the bidan allegedly receives Rp250.000. She assists with birth, which is led by the bidan, and then she stays with the mother for some time afterward. Other actors. Dukun interact formally/ systematically with no other actors. 3.3.1.6.“RoleModels”
Role models were important in all locations, though for different aspects of health knowledge and service delivery. Most examples had to do with seeing and hearing the positive side of giving birth with a bidan. Women are often afraid that they will be yelled at or exposed (naked in front of people, etc.) and it helps to hear positive experiences. In Gorontalo and Lombok, women were particularly happy to hear that friends who gave birth with a bidan did not have to pay. In Sumba Timur, women were impressed at how quickly women could walk and wear normal clothes. 37 Role models are also important in demanding good service and shaping expectations. In Bayan, Lombok Utara, where the village head takes a very proactive role in overseeing the bidan in the village, villagers are also more willing to simply call the bidan if they arrive during office hours and she is not there. 3.3.1.7.“Healthelite”/Privatesector
In some villages or kecamatan, there were health actors outside of the state or traditional sectors. These often included retired (or former) health actors who moved into the private sector at the village level. These people were often not doctors, though some provided medical consultations, and others sold medicines and shots. Depending on their position within the village, they are able to control health information. The strongest example of this comes from a village in Jawa Barat, as described in Box 6. In general, it is important to know the health elite, since they can drive village level service delivery. Box 6: Local Elite as Health Actors Mrs. Haji dominates all discussions around health in her village. She is a former cadre KB who went on to work at the puskesmas. She was recognized for her outstanding work as a cadre KB with a haj trip. She is also married to the former village head and now very active in local political and religious organizations. When she retired, she continued her own general practice at the village level. She appoints cadre posyandu as well as village level Generasi facilitators. A number of bidan have left in the past since Mrs. Haji retains such a grip on the health system. The current bidan has worked hard to bring her on side, since she is clearly the strongest voice of information about health in the village. The Generasi village facilitator struggled to name the indicators of the program, but Mrs. Haji quickly recited them all, and was very articulate on what the village needs were. She also keeps all records at her house. 3.3.1.8.VillageHead
Role/ information provided Village heads are increasingly being incentivized to have a greater role regarding health, but many still see the core of their role as encouragement: they should urge women to go to posyandu and to have birth with a bidan at an approved health facility. Indeed, without health knowledge but with power in the village, they can support behavior change. In theory, they should be getting information at the mini lokakarya at the puskesmas every three months, but this does not seem to happen. Village heads do not generally see themselves as empowered to or responsible for oversight of village health services generally or the bidan in particular, with the exception of one village head in Lombok Utara, as described in Box 7. As described in more detail below, some are also charged with getting health data, though obstacles remain in their ability to use it effectively. Box 7: A Proactive Village Head Kertamalip has been the village head of Karang Bajo in Bayan, Lombok Utara since 2005. He sees himself as a central part of village health delivery. In particular, he sees his role as providing information about health services in the village and socializing important health messages. These include the importance of using health facilities and not giving birth with a dukun. He also sees it as his duty to ensure that the bidan and nurse assigned to the village are at work. He asks them to tell him if there are any problems and also to report (via phone) if they are leaving the village. He calls when they are supposed to return, to be sure that they have returned to the village. He is clear that if he is not satisfied with them that he will try to replace them. He also asks the bidan to keep him informed of due dates of women in the village, so he can ensure that they get the care they need and that they are going to the 38 bidan. In addition to overseeing the bidan and nurse, the village head was asked to increase attendance at posyandu by the puskesmas. Working with the (female) village secretary, he passed a village regulation making all of the dusun heads in charge of the posyandu in their dusun, and encourages them to undertake sweeping. In villages in Sumba Timur that have passed a village regulation on maternal and neo‐natal health, the village head is nominally in charge of applying sanctions (fines) to village actors who have not fulfilled their responsibilities. However, in both of the villages visited by the study, there remained challenges. In the first, the village head did not know the content of the village law, or chose to describe the fines selectively, mentioning that there would be fines for women if they did not give birth at a facility, but no fine for the bidan if she was not at work. In the second, the obstacles to implementing the regulation were numerous, starting with the fact that the assigned bidan had not arrived in the village after nearly two months. The distance to the puskesmas further complicated matters. The village head was thinking of developing implementing guidelines before the regulation could be applied. Outside of maternal and neo‐natal health, village heads play an important role in supporting general access to health services for poorer villagers by providing a “poverty letter” (SKTM) that allows them to get free services if they do not have a Jamkesmas card. In addition, they provide information about how to use the health system, and additional support as necessary. For example, in Lombok Utara, any person referred to the hospital (which is in a different district) requires a letter from the Bupati within 48 hours of being moved to hospital or they need to cover their own expenses. Village heads play an important role in coordinating with government to ensure that villagers have the documents that they need. Power and Trust Village heads have considerable power. Depending on their actions, they can also command the trust of citizens. They can, and in some cases do, use this power to encourage certain health behaviors (usually in line with the facilities based approach), but only in rare cases are they searching for more information to share with constituents or overseeing health actors in the village. Incentives Incentives for village heads to engage on health issues vary considerably by location, as described below. In Gorontalo, the Bupati has a “performance contract” with every village head. The purpose of the contract is to ensure that they focus on achieving key outcomes at the village level, including reaching key health indicators. The “reward and punishment system” includes maternal mortality, though a number of the other indicators are difficult to measure or unknown to village heads.5 Village heads are 5
One village head had been in office for a year and a half without seeing the contract. The research team was able to get a copy of one, which included maternal mortality, but also life expectancy in the village, which is not actually measured at the village level, nor is it likely to be changed dramatically by a single village head. 39 particularly aware of problems of maternal mortality, since at the monthly government coordination meetings they are asked to boo any village head who has had the misfortune to have a woman from his village die in childbirth. In Sumba Timur, in village regulation on maternal and neo‐natal health, village heads themselves can be fined for not carrying out their responsibilities under the regulation, and also carry the brunt of implementation and enforcement activities under the law. In Lombok Utara there are not explicit incentives for village heads, but there are examples of strong engagement by the puskesmas and subdistrict government in some locations. Village heads are empowered to take action. There are rumors that the village head should sign the time card of the bidan, but the research team did not find this happening. The Bupati is engaged on health issues, particularly following a scandal around the falsification of data to understate the number of undernourished children. This raises the profile of health issues with the village heads. For similar reasons, the health department is particularly motivated in Lombok Utara, and hold a separate musrenbang‐like planning process focused exclusively on health. At the village level, this is led by the village head. While it does not incentivize any particular behavior, it may raise awareness. Networks and Interactions Puskesmas. Village heads are generally invited to the “mini‐lokakarya” meeting every three months at the puskesmas. It is not clear, however, how much information they share or receive at this meeting. As one community leader pointed out, without some level of technical knowledge, it is very hard to participate in those meetings. While village heads may be powerful in their own villages, they struggle against hierarchy and specialized knowledge. Cadre posyandu. Cadre are often appointed (and serving at the discretion of) the village head. How carefully their performance is overseen is questionable. In the village in Lombok Utara with the very active village head, oversight is conducted through random posyandu visits. In other locations, this does not seem to be happening, though village heads may be aware if a cadre is no longer able to do their work. Generasi actors/ Generasi. Village heads do not have much engagement with Generasi actors, even if they have appointed them (unofficially). They are generally invited to village planning meetings, and are aware of the program but are not always provided with much background information on the program or engaged to share messages. Dukun. The village head has limited engagement with the dukun, though when he or she does chose to encourage people not to use the dukun for giving birth, it carries weight. Kepala dusun. They have direct control over dusun heads and can ask them to be engaged in posyandu, as was the case in Lombok Utarta. They also rely on dusun heads for information and support in getting poorer villagers health care. See below. 40 3.3.1.9.DusunHead
Role / information provided The role of the dusun head in health is generally limited to assistance with accessing the health system in case of serious illness. They usually hand out Jampersal cards (and in some locations help re‐assign them) and provide information about the program. They also assist with “poverty letters” (SKTM), which are critical for poorer families to access health care if they do not have Jamkesmas. Generally, the dusun head does not see himself as having much of a role in maternal and child health, with the exceptions in Lombok noted above in Box 5. There is considerable variation in dusun head knowledge, which is dependent upon whether either the village head (in Lombok) or the bidan is active in sharing information. In one village in Gorontalo, one dusun head did not know about Jampersal (or free births generally). In a village in Sumba Timur, one was unaware of the meaning of the flags in the yards of pregnant women indicating their risk status, despite the fact that Desa Siaga was (theoretically) active and coordinating a community response to assist pregnant women get to a facility to give birth. Power / Trust Dusun heads are closer to community members, and generally well‐trusted. Since they often help with Jamkesmas related issues, they have a very basic knowledge of health bureaucracy. As evidenced by the example from Lombok, they can be important leaders in the community. Networks/ interactions Per the examples above, there are opportunities for the dusun heads to interact with posyandu (and thus cadre posyandu), but this is not generally the case. They report to the village head, but do not necessarily engage on health issues. 3.3.1.10.Localmedia
While not widespread, there were a few cases of local media that have been or can be used to support health knowledge or oversight that are worth recording. These include: ‐
‐
Two villages visited (one in Jawa Barat, the other in Lombok Utara) had local radio stations. These radio stations covered a few villages and many people tuned in for local news and to request songs for each other. In Java, there was no specific information provided on health, but in Lombok Utara, the radio station was able to engage a doctor to do a regular question and answer session. Unfortunately, the doctor moved, and requests to the puskesmas for another doctor that might be willing to talk on the air a have been unfulfilled. Neither radio station is used by Generasi, though the one in Lombok Utara has been used for oversight of PNPM Mandiri. In Lombok Utara, an episode of malpractice (a nurse asking the parking attendant to do an IV / set up a drip) was reported in the local newspaper and resulted in changes at the puskesmas. 41 3.3.1.11.NGOActors
The study observed limited activity by NGO actors. However, two cases are worth noting, as they highlight some of the strengths and weaknesses of NGO engagement. In Gorontalo, a provincial NGO, ADRA, undertook data collection on nutrition and uncovered significantly higher numbers of malnourished children than reported to the health department. It also uncovered inconsistencies in health department data. While the discussion around the validity of data that the NGO has collected is ongoing, they have succeeded in forcing the health department to look more carefully at its own data and think about what it means. In Sumba Timor, ACCESS is supporting local NGOs to provide support to villages on issues that they identify. In one village visited where the clear priority is clean water, the NGO may have helped to negotiate with the camat for more regular delivery of water. The village uses part of its yearly budget to pay the sub‐district to deliver water, but it often does not arrive. Last year, the NGO supported the development of an agreement on water delivery. However, it was not written down, and a clear pricing system was never established. While the NGO was aiming for an MOU, they stopped their advocacy when the problem was resolved in the short term. This likely reflects the power of the NGO against the political interests in the area.6 3.3.1.12.Husbands/Mothers‐in‐Law
Because husbands and mothers‐in‐law / mothers continue to have such a large role in where women choose to go for care, they should be mentioned as actors. There is a huge variation in the amount of information that men have about pregnancy or children. Younger men seem to have a higher level of awareness, while older men generally seemed to leave their wives to sort things out. In many locations, men mentioned being afraid to go to posyandu, because it would be socially odd, but are slowly starting to attend. Some of the Generasi facilitators mentioned this, and it seems to be a positive spill over that some of them are much more comfortable assuming new roles within their own families. Mothers/ mothers‐in‐law are usually the ones that go to the puskesmas with a woman if she is giving birth. It was not clear how much they knew, though their support was critical. However, looking at the information flows in the village, it is not clear how they get information unless they are taking their grandchildren to posyandu (which a number of them do). 3.3.1.13.Religious/AdatLeaders
In general, the study did not find religious or adat leaders having a large role in supporting or preventing certain health decisions, though in some locations, health messages were being incorporated into sermons. This was also true in ‘difficult cases,’ such as young or out of wedlock pregnancy. While they may have opinions about young women getting married, they would remain silent on her medical well‐
being. 6
This finding is consistent with those of the LLI 3 study, which also noted that long term solutions to community problems are increasingly difficult to secure. Local Level Institutions Study 3: Overview Report, PNPM Support Facility, 2013. 42 3.3.1.14.PKK
While PKK can provide additional funding for cadre posyandu, and should be involved in posyandu activities, the program was generally not functioning and did not provide support to cadre. In general, PKK prioritized competitions of various sorts, and purchasing nice outfits. In some locations they did hold “posyandu jamboree,” or cooking classes but it was difficult to see any impact of these activities at the village level 3.3.1.15.CadreKB(SPMD)
Birth control cadre (cadre KB) were not particularly active in any village visited. They are worth mentioning because they also gather data on women and children, discussed below. 3.3.1.16.Generasiactors
There is considerable variation in the level of knowledge displayed by Generasi village facilitators. For example, there was mixed knowledge of the Generasi indicators: some facilitators could recite them easily and talk about solutions, while others did not know or struggled to remember them. Facilitators were strongest in explaining project processes, but did not always know about certain basic health tools. For example, in one village in Sumba Timur, cadre posyandu were not plotting child age/weight because they did not have the table in the buku KMS that they had used during training. As it turned out, a number of women in the village had buku KIA which contains the same table, and should be used for discussion in Generasi. Not using the table meant that they had to wait for information to come back from the bidan as to whether children in the village were below the red line. In the same village, they had requested a standing scale from Generasi, but did not know how to use it. Similarly, though they are taking villagers through a village planning process, it does not seem that they have the capacity (or have been supported) to help villagers think about health problems or solutions. In the villages studied, they generally did not talk about what targets had been met or not, even if they had talked about it with the FK. The main role of the facilitators outside of the planning process was to collect data. In some locations, data is collected from the posyandu directly, in others it is collected from the bidan. This was not always well coordinated with the cadre posyandu. In one location for example, the Generasi facilitator routinely took the posyandu records and failed to give them back in time for the next posyandu. In another, the facilitator would come to the cadre to collect data. While not disruptive in any way, the cadre who gave her data still didn’t know her name. In locations where data is collected from the bidan (and likely in some locations where it is not) it will generally match what the bidan reports to the puskesmas. This subtle process of making sure that there are no differences in the data ensures that there will be no conflict or problems later on. There were a number of questions around forms, and some facilitators struggled to fill them correctly. In addition, information that is provided back to the villages is very difficult to use. While they can tell if they hit a target or not, it is very difficult to tell by how much, meaning it is very difficult to assess their approach. As a result, there is very little discussion of whether or not approaches taken by the village are useful or not. This reduces the importance of the bonuses. 43 The PK (Pelaksana Kegiatan‐ in charge of implementing activities) generally had decent grasp of what the program could do, and worked hard to get around issues of disbursement and procurement. Kecamatan facilitators did generally come to the village, but the study did not go into depth as to how often or the extent of support that they provided to the village facilitators. Power/ Trust Village facilitators had little power in the locations visited. The PK, who were generally more senior had more power, and were able to influence the process in some locations. Networks and interactions Village government. Generasi actors did not actively engage the village head and in a few locations were unsure of the goals of the program. Though dusun consultations are held, facilitators do not seem to engage the dusun head either. Bidan. Generasi actors occasionally engaged the bidan, primarily around data and assistance to children. Bidan were also engaged in infrastructure discussions, though it seemed to be late in the process. While there is probably merit in giving villagers the space to talk about health without the presence of the bidan, who may intimidate them, citizens did not always have the knowledge that she may have been able to provide. Cadre posyandu. The main interaction that Generasi facilitators have with cadre is to gather data from them. Most are invited to the dukun level planning discussions, but not all. 3.3.1.17.DesaSiaga
Desa Siaga is an approach that is aimed at supporting villages to develop systems in support of pregnant women. This includes the identification of transport options, blood donors, savings, etc. With one exception, Desa Siaga was not functioning in any meaningful way in the villages in this study. The one village where it was working shows both village creativity and the limitations of the program. Box 8: A Model Desa Siaga In the village of Mbata Puhu, Kec Haharu, Sumba Timur, is known for being extremely dry. Drinking water alone is hard to come by in the dry season and animals suffer. Under the Desa Siaga program, all women should save Rp 5.000 each month so that they have enough money to cover costs related to pregnancy. In initial discussions around Desa Siaga (supported by AIPMNH), villagers said “We do not have any money, all we have is corn.” They decided to collect 12 kg of corn from each family (the equivalent of 1 kg/mo). It is monetized at a rate of Rp 3.000/kg and saved for use by those who needed. In one season, the village collected over Rp 2.500.000. Using that money, each pregnant woman is given Rp 50.000 for transport at the puskesmas. No money is provided for anyone to accompany her, or to come home. Though many people provided corn, not all those that contributed understood the details of the program, and many thought that they would only contribute once (not every year). In nearly a year, 12 women had been supported, and there is considerable money left over. However, because the whole program is designed to provide support in getting pregnant women to facilities and ensuring that they have blood donors in case of emergency, the options for using the additional money are extremely limited. While there are major problems with maternal 44 nutrition in the village (as an element of preparing for pregnancy), the issue is not covered by the program, and now villagers are struggling to agree what to do with extra money. 3.3.2.Whataretherolesofkecamatanactors,institutionsandpolicies?
The study was particularly interested in accountability relationships between communities and service providers. While there is considerable space for more time to be spent at the kecamatan level, this study provides a very basic picture of the incentives (and constraints) facing kecamatan‐level actors. Key institutions at the kecamatan level include the puskesmas, the camat, the kecamatan‐level Generasi facilitator and (potentially) the BKAD. Linkages between them are currently not particularly strong, since the puskesmas and the camat sit in different vertical reporting structures, and the BKAD is currently not well engaged by PNPM Generasi, but more active in PNPM Mandiri. The kecamatan facilitator interacts will all actors but, as will be discussed below, is not often in a position of strength, even when he or she has useful data. 3.3.2.1.Puskesmas
Role/ Information provided For the purposes of this study, the puskesmas carries out (or should carry out) several key functions: Bidan coordination. The puskesmas makes sure that information from the village level is reported upward, and shares information from the health department downward. Information sharing and planning is generally done in the monthly “mini lokakarya” (mini workshop), which all bidan are meant to attend. The study found some variation by location with regard to bidan coordination. Factors that influence bidan coordination include staffing constraints and the extent of the facility‐based approach being undertaken in the district. Both Gorontalo and Sumba Timor faced serious staffing problems, which puts additional pressure on bidan and on puskesmas management. In the worst cases, they have to make decisions about what type of care to provide. In Matawai La Pau, Sumba Timur, the number of staff was so limited that there was actually a trade‐off between having staff at the puskesmas and sending staff out to support posyandu. Community outreach activities. Most routinely, this includes support to villages implementing posyandu, by providing vaccinations (at a minimum), and in some locations additional staff support. There is significant variation by location, even within districts, which seems to be largely driven by puskesmas leadership. For example, in both Gorontalo and Lombok Utara, despite consistent and well socialized goals coming from the Bupati, the two puskesmas studied took very different approaches to outreach‐ in each location one connected well with community members and the other did not. In both kabupaten, district leadership paid attention to posyandu attendance rates, but this did not necessarily impact on puskesmas leadership. Given that political pressure seemed to be a key factor in driving results in other areas, this type of variation may warrant further examination. 45 At the stronger puskesmas, there seemed to be a real effort to respond to community needs by providing support to posyandu at times when people can come, and by making activities as interesting as possible. In Lombok, as mentioned above, this was achieved with the help of the dusun heads and a range of other activities. Under the district policy, the focus of the more successful puskesmas was on posyandu attendance, while the other was not. This pushed the staff to coordinate more with village heads and to implement a puskesmas policy that staff should not leave a posyandu until they had hit 100% attendance through sweeping. They also engaged more staff from the puskesmas to help with outreach, including even administrative staff in sweeping efforts. In Gorontalo, the driving force behind outreach seemed to be leadership and a willingness to meet community needs, particularly those in difficult to reach locations. Nutrition evaluation and support. Technically, if any children are reported to be below the red line, then they should be checked by the nutrition expert from the kecamatan. This person should gather additional information about the child (generally weight to height in addition to weight to age), and should check for any illness that may be contributing to undernourishment. Based on this further assessment, they determine if the child is malnourished, undernourished or simply small. If the child is malnourished, they also work to secure supplemental food or medical assistance. While they do not check undernourished mothers, some puskesmas also have additional food for them, distributed based on reports from bidan. There was variation both in implementation and staffing on this issue. Some locations did not have nutrition specialists, meaning that they needed to coordinate further with the kabupaten for any specialized assistance. This can take a long time. Even in locations where there were nutrition staff, they did not always come to villages to check underweight children. Often they relied on bidan to send additional information, particularly height. Both Lombok Utara and Gorontalo were taking a more aggressive approach to responding to undernourished children, though they are taking slightly different approaches. The approach in Lombok Utara is described in more detail below. Gorontalo established a ‘Therapy Food Centre’ to provide care to malnourished children. They have also provided funding for a child and one person to stay at the center for three months. This is helpful, but can be still difficult if the parent has multiple children to look after. The reduction of poor nutrition is also a target for the health department in its agreement with the Bupati. However, increased efforts to detect malnutrition are mixed, based on the motivation of the puskesmas to support posyandu sweeping activities. There are also no additional preventative programs or support focusing on nutrition. However, there are still expectations that malnutrition will decrease. Kecamatan Mootilango, funding for supplemental feeding (PMT) decreased last year and is now sufficient for only eight children. The planning was not based on data around the current health conditions in the kecamatan, but on projections of what would be required to hit MDGs. 46 Incentives Planning and funding The puskesmas completes both yearly and monthly planning. There does not appear to be a great deal of flexibility in these plans, since funding remains quite limited, and bound largely by rules governing the main source of funding coming from the national level, Bantuan Operational Kesehatan (BOK, Health Operational Assistance). Across all locations, BOK is the primary (and in some locations, sole) source of funding for puskesmas operations. To access this money, the puskesmas needs to make a monthly plan of action, which is used as a basis for disbursement. The effectiveness of the monthly planning process can make a significant difference in the puskesmas’ ability to respond to health needs. There seem to be different approaches being taken with regard as to when this money can be disbursed. For example, in Sumba Timur alone, puskesmas were previously (until end 2012) required to account for all money spent before any new money could be disbursed, which led to delays in disbursement. In 2013, the district claims that it is allowing a grace period on accounting to ensure that disbursements can be made consistently. In the locations studied, it is not clear that they have yet seen a difference. In addition, in two kecamatan (not covered by the study) puskesmas have been given ‘department’ (SKPD) status, meaning that they control their own budget without approval from the district level. One important aspect of BOK funding, however, is that it does provide a certain amount of flexibility on the operational side‐ as described above, there is variation in the quality of services provided by different puskesmas, even in a single district. While much of this boils down to leadership at the puskesmas, it is important that those leaders have some money at their disposal and be able to allocate it as they see fit. Funding for materials generally comes from or through the district level. For routine expenditures such as medicine, a yearly plan is made, and disbursements made against it. If a puskesmas knows that there are certain types of illness that increase at a particular time of the year, they can plan accordingly to ensure that they have requested enough medicine. However, they cannot do the same with supplemental food for children or mothers (PMT). PMT is not treated like a medicine, so though money is allocated for it, it is drawn down on a case‐by‐case basis only after undernourished children or mothers have been reported (and verified, in theory). Despite the fact that there is still a clear “hungry season” in some locations, puskesmas cannot prepare supplementary food for the children and mothers that they expect to be affected. There was some evidence of districts attempting to incentivize behavior, by providing additional money for transport or a first pair of clothes for a newborn born at an approved facility. These approaches seemed difficult to implement due to rules that resulted in delayed disbursement. For example, there needed to be evidence that someone came to the puskesmas before transport money or clothing could be requested. By the time it was received (generally a month later), the person had gone home. The planning and disbursement process should be further verified, since it was not traced carefully in this study. However, the study did see that it has an impact on the time that it takes to respond effectively to cases of undernourishment. Unless a child or mother is taken directly to the hospital, the 47 process generally starts with a report by the bidan to the puskesmas, generally following posyandu. Assuming that this report is received on time before the end of the month and is processed quickly, it will become an input to the Plan of Action (or associated plan), to be approved by the district at the beginning of the following month. If all systems are in order, funding can be released in under a month. If they are not, it can easily take more than three months. Most puskesmas expressed gratefulness to Generasi for providing both additional and quickly disbursable funding for PMT. In general, there seems to be the feeling that even if the puskesmas was to be more proactive in identifying nutrition problems, there would be little that they could do since their funding is so limited and the time that it takes to disburse is often so long. Targets (Millennium Development Goals and other political goals) Targets are discussed in more detail below, but it is worth making the point that while political pressure from above can drive action, it has not improved the use of information to inform (and refine) policy approaches at the kecamatan level. Staffing It is important to note that staffing was an issue in several of the locations. In both Gorontalo and Sumba the number of bidan available were insufficient to cover all villages. In addition, in Sumba there were also insufficient numbers of bidan at the kecamatan level. As noted above, this led to puskesmas policies requiring village bidan to leave their villages two days a week so that a greater number of staff could be on hand at the kecamatan level. In West Java, there were problems staffing less desirable kecamatan and villages. One of the kecamatan visited had been without a nutrition specialist for quite some time and suffered from constant turnover of the Puskesmas head. 3.3.2.2.Camat
The role of the Camat in health service delivery is extremely limited. Structurally, the Camat reports to the Bupati, and rather than having any formal oversight role vis‐à‐vis the sectoral agencies at the kecamatan level, his role is generally to coordinate, leaving technical solutions up to the various sectors. While many Camat were aware of basic health issues, they are usually not engaged in the details. In Gorontolo, there is a greater attempt to engage the Camat in the coordination of health outcomes. The effectiveness of this approach is not yet clear, as it mainly makes the existing role more explicit. There was one Camat worth noting in Kecamatan Matawai La Pau, Sumba Timur. With the assistance of AIPMNH, the Camat there was highly knowledgeable about and articulate on health realities and needs in his kecamatan.7 He was also extremely proactive in trying to ensure that bidan were posted to villages 7
Interestingly, he noted that a large part of the reason that he could be an effective health advocate in that kecamatan was because he was not from the area. As he said, “if I said these things where I am from, as soon as I walked out the door, I would be fined an animal by the ketua adat for insulting the elders.” However, where he sits outside of the adat system, he can be proactive about encouraging men to give more food to pregnant women and children, to set aside money for childbirth, to ensure that their wives are checked when they are pregnant and to send their children to posyandu. 48 in his care, going so far as to call the Head of the Health Department to find out why a posted bidan had not arrived yet. He occasionally attended posyandu and was also active in promoting health messages to in his area. In this particular case, it is worth highlighting how AIPMNH developed his capacity: through requesting that he give presentations about health in his kecamatan to his superiors. The program helped him to prepare, but he had to present. This ensured that he was both familiar with data from his kecamatan, and could actually discuss key issues. 3.3.2.3.BKAD
The Badan Koordinasi Antar Desa (BKAD‐ Inter‐village Coordination Body) is in charge of inter‐village coordination regarding PNPM. The group is often made up of respected individuals who have sufficient authority to coordinate across village heads in a kecamatan. They may be current or former village heads or simply tokoh masyarakat. In many locations, they also work to advocate on behalf of villages to secure funding for proposals that were not funded through the PNPM process. BKAD leads the inter‐village meeting for Generasi, but in general is not engaged particularly well by the program. For example, they are not invited to coordination meetings at the kecamatan level, nor are they provided with any information about village progress (for example, which villages have or have not met certain targets). There is, however, interest in being more involved with Generasi. A number of BKAD indicated that they would be interested not only in being engaged in coordination, but would be interested in using information to discuss with health and education departments. However, to do this, they would need assistance beyond simply getting the data. In particular, they need help to understand it and present it in a credible way. Despite the fact that the members are quite respected in the kecamatan, and some are invited to the three‐monthly mini workshop at the puskesmas, they are uncomfortable talking to ‘health people’ since they feel that they do not have the technical background to do so. As one BKAD described, he “just sits and watches” while children is his village get poor‐quality PMT (fried bread), because he doesn’t feel that he has enough of a nutrition background to say that it is insufficient. Instead, this particular man has started searching for recipes (and asked if the World Bank had any good ones to share). 3.3.2.4.KecamatanFacilitators
Kecamatan facilitators are of varying quality, both in terms of engagement with community actors and coordination with government actors at the kecamatan level. Regardless of their level of ability, however, they are all in a very difficult position vis‐à‐vis the kecamatan actors. On one hand, they can access money to solve health problems, which is valuable to the puskesmas and other government officials. On the other hand, they have to be very careful to present themselves as community empowerment actors, not health experts. This compromises their ability to present data or question information/ data provided by the puskesmas and effectively undermines their oversight role. In many locations, they also participate in the ‘mini workshops’ at the puskesmas, and in theory have the opportunity to get information and share feedback, however it is not clear how seriously the puskesmas will (or can) take input. In addition, while they do have spending flexibility, in some key areas they need to wait for technical input from health actors. For example, while Generasi provides funding for PMT, it often relies on the nutrition expert at the puskesmas to provide recipes and menus. If these are insufficient, they are in a difficult position. 49 On a positive note, many of the kecamatan facilitators seem genuinely interested in the content of health and education issues. However, as has been reported elsewhere, the pressures of program administration often prevent them from getting the support that they would need to be more articulate on health issues or more creative in supporting village level problem solving. Kabupaten meetings are often dedicated to administrative issues, and there is less space for providing on‐going learning for the facilitators, either externally or from each other. 3.3.3.Whataretherolesofkabupatenactors,institutionsandpolicies?
There are potentially a large number of actors engaged in health at the kabupaten level. These include: ‐
‐
‐
‐
‐
‐
‐
‐
‐
‐
‐
Health department (policy, management, budgeting) Bupati (policy) Bappeda (budgeting) DPRD (budgeting) Civil Service Body ‐ Badan Kepegewaian (staffing) Community Empowerment Unit (BPMD) (posyandu support/ management) PKK (posyandu support) Women’s Empowerment (general support, leans toward birth control) Social Department (socialization of entitlement programs) Hospitals NGOs (advocacy on health issues, limited) This section will not attempt an actor‐by‐actor analysis, since it is more complicated at the kabupaten level. However, in presenting key findings below, it will draw out some of the key roles and interactions as they relate to specific issues, as well as the various incentives that appear to be influencing kabupaten actors and institutions. Key findings include: Political priorities drive change. The strongest driver of change in all locations came from Bupati engagement on health issues. This is strongest when the Bupati’s vision is being carried forward by a health department that is able to take clear supporting actions to implement the general policy. Millennium Development Goals are increasingly important to Bupati, though not all of the goals have the same power. Reducing maternal mortality was the most important target in three out of four districts studied. Where it was a priority, the goal was not simply a reduction in maternal mortality, but the eradication of maternal death. “AKI nol” (zero maternal deaths) was almost a slogan in both Sumba Timur and Gorontalo. These goals seem to be based on the notion that if reducing something is good, then eliminating it is better. This leads to the development of a different incentive structure, however, often one in which any individual associated with a maternal death is shamed or punished. As discussed below, this can create some perverse incentives around how to treat community members and how data is reported. Child mortality was also recognized as an issue, but there is more flexibility around reducing numbers. Nutrition, particularly for children, was seen as a problem, especially in Lombok Utara, where there was a scandal around under reporting of undernourished children. 50 Policy priorities are blunt tools. Where effective, they are supported by health departments, which develop a clear set of priorities in implementation. The interaction between the health department and the Bupati was slightly different in each location, as described below. In Lombok Utara, there was a major scandal about under reporting of undernourished children in 2011. The media reported that there were considerably more cases of malnourished children in Lombok Utara than the government was reporting in Kecamatan Tanjung.8 The puskesmas reported 50 cases, but the media reported 2,900. The reports triggered an investigation. Sweeping by health officers revealed 3,232 children ‘below the red line,’ of which 459 were serious cases that had to be taken to hospital. In response, the Bupati has worked to ensure that the district will reach its goal of allocating 15% of funding to health to improve staff numbers and training, and to build a hospital.9 In addition, he has worked with the head of the health department to develop a program to respond to the problem. The head of the health department has targeted five key programs: posyandu attendance; Desa Siaga; toilet construction and ‘extraordinary events’ (such as dengue fever, etc.); containing contagious disease and disaster preparedness, and; Gerakan Sayang Ibu (Mother Friendly Movement). To ensure quality implementation, each of the five kecamatan in Lombok Utara focuses on only one program per year. Then it moves on to a new program. The head of the health department has a clear logic as to how each of the programs works to contribute to resolving the problem. However, this logic does not always trickle down to the puskesmas level, which sees that it is implementing the same programs as before, just with more focus. By forcing each puskesmas to report on a very specific set of indicators around a single main program, it seems that the expectations of the puskesmas are clearer. However, some of the programs are easier to measure (and oversee) than others. For example, posyandu attendance is quite clear and relatively easy to measure. Impact on nutrition is more difficult to see, and it is not clear if there have been improvements in the numbers of undernourished children. Across the kabupaten, the health department is also supporting efforts to identify pregnant women as early as possible. It has provided pregnancy tests and incentives to cadre posyandu for reporting pregnant women to the bidan. Lombok Utara has also taken pre‐natal care step farther by incentivizing nine (rather than four) checks to ensure the health of women and children. Sumba Timur has taken a different approach, with a stronger regulatory focus. Reform there has not been driven by any particular crisis (beyond consistently high maternal mortality numbers), but through 8
Interestingly, the investigations may have been triggered by a Generasi facilitator. Frustrated that the health department would not listen to reports that there were more undernourished children than they had reported in their own system (334 vs 50), the facilitator claimed to work with the reporter to determine how many cases there really were, coming up with 2,900 cases. 9
Currently, all citizens from Lombok Utara have to go to another kabupaten if they need to go to the hospital. For reasons that are not entirely clear, all patients seeking public care (Jamkesmas) must have their referral approved by the Bupati within 48 hours of arrival if they are to receive continuing care. This is a huge burden in terms of both time and money for most people, even with support from village heads and camats (or potentially because of their involvement, depending on the case). 51 an increasing recognition of the importance of health issues, particularly maternal and neo‐natal health, and a Bupati who seems committed to supporting reform. The process of developing a regulatory framework has been supported by AIPMNH. In 2009, the governor passed a gubernatorial regulation (PerGub) on health. In Sumba Timor, this was followed by the passage of a local regulation on “Revolusi KIA,” (Women’s and Children’s Health Revolution) passed in January 2011. The regulation was developed with support from the health department, the regional health advisory body and DPRD, with strong support from the Bupati. This regional regulation forms the legal basis for the numerous village regulations described in Box 2. Village regulations in 74 villages (all of the villages in AIPMNH kecamatan) were developed and passed in 2012. They were effective from the beginning of 2013. The process of developing the “Revolusi KIA” regulations helped to foster support and awareness for health issues. Partly as a result, nearly all kabupaten actors noted that health had considerable support from DPRD, critical in the budgeting process. The level of support from Bappeda is less clear. The “Revolusi KIA” generally promotes a facilities‐based approach to addressing maternal and neo‐natal health issues, and the health department has taken a number of steps to improve facilities and staffing (which remains critically low in Sumba Timur). These include improving or building rooms dedicated to maternal care at the puskesmas; developing a ‘sister hospital’ program to improve training for doctors, training teams of doctors and nurses for emergency maternal care (in puskesmas equipped to deal with emergencies); certification for 40 new bidan every year for five years and additional training for those bidan who do not have a full certificate (D4). Unfortunately, the teams that were trained for emergency maternal care have been largely disbanded‐ many doctors are only on two year contracts, so some have left and some of the remaining members of the teams have been assigned to other locations, splitting up teams that were trained together. In addition, while the bidan training is moving forward, the first class will only graduate in 2014. All of these issues add up to some fairly substantial remaining gaps on the facilities side, despite the increasingly punitive approach taken through the village regulations. While there is not a formal ‘reward and punishment’ system in Sumba Timur, in the one clear case of maternal death last year both the bidan and the head of the puskesmas were called to the district level after it happened. AIPMNH supported a broader audit of that case and child deaths for a more accurate picture of what happened. While this type of discussion does not yet fully drive policy making, or even address capacity gaps at the kecamatan level, it is making an impact: the Bupati attended the meeting and was engaged in the discussion throughout. While the Bupati is committed to health, there remain weaknesses in the system. For example, the head of the maternal and neo‐natal health unit in the health department is a doctor with limited managerial experience. Despite the support of AIPMNH, she admits that she didn’t really understand the planning process last year, though is much clearer about her needs for the future. There remain weaknesses in thinking about how to implement the full scope of a policy, the “Revolusi KIA” in particular. For example, though the government is aiming for significant (and often expensive) behavior changes by now requiring women to give birth at the puskesmas, there was little thought as to how they would support that change. The assumption was that most people had family near a puskesmas that they could stay with, so requiring them to do so would be fine. Only later were problems encountered and “waiting houses” proposed through Generasi. These are still funded only through Generasi. In addition, there 52 were a number of other management problems, including lack of discipline/ application of the civil service law for poorly performing bidan, and some strange rules around who is being accepted for bidan training.10 Unique to Sumba Timur (of the locations in this study), is the role of the health advisory body. It is chaired by a retired doctor and ex‐Bupati who brings considerable political clout to health care reform, and has been taking time to look at different levels of health care delivery from the kabupaten to the village level. He was also engaged in an AIPMNH health audit and provides something of a community perspective, though it is colored by his strong health background. For example, his work at the community level gives him a slightly more flexible position on the role of dukun in emergency situations, and he has been strongly in support of waiting houses after realizing problems of access. He also advocates for a much broader thinking about health, noting, for example, that there will need to be coordinated efforts with the agriculture and fisheries departments to address the nutrition problems that plague the kabupaten. Unfortunately, it does not seem that Generasi coordinates with this body in any way, despite the fact that it could be a powerful advocate for communities. In Gorontalo, Bupati is focusing on a “bureaucratic revolution” that erases the project mentality; builds a climate of collaboration and integration and erases sectoral egoism. To this end, Gorontalo public servants and institutions including local government actors (Camat, Village Heads), individual sectoral actors (Head of the Health Department, Head of the Puskesmas) and SKPD now have performance contracts. For organizations, such as Puskesmas and SKPD, the indicators are annual goals connected to their organizational roles and responsibilities, minimum service standards and Kabupaten goals. Examples of these for a Puskesmas and the Health SKPD include zero maternal mortality (AKInol) as well as targets for posyandu attendance, pre‐natal checks, decreasing malnutrition/severe malnutrition, and improved immunization rates. Individual performance contracts detail performance indicators and their targets. In the case of individuals, (Camats and Village Heads) the indicators for these contracts are based (in theory) on the person’s job description and responsibilities and Kabupaten goals. However, there was some confusion about the targets themselves. In one location, a village head who had been in place for a year had not yet seen his performance contract, though he was aware that there would be one, and that he needed to be responsible in some way for health. In the 2009 performance contract for another village head, the indicators for health were: zero maternal deaths; zero deaths of babies and children under 5; 100 percent participation in Posyandu; and raising the life expectancy of the village from 65 to 84 years. To incentivize the achievement of performance targets, the Bupati put in place a system of “rewards and punishments.” Rewards for individuals who exceed their targets include money and trips on the hajj, and institutions have the opportunity to have money added to their budget. While these rewards offer positive reinforcements to officials, the specter of punishment and humiliation also plays a rather 10
Despite the fact that the health department is supposedly recruiting SMA graduates from poorer and isolated villages, they have also put a height limitation of 150cm on all candidates. While the Bupati recognized the strangeness of the rule, the health department would not drop it, claiming it is nationally mandated. 53 significant role in making system function. Should an individual not fulfill their role or meet certain targets, the Bupati will provide warnings and then fire people.11 However, the most common form of punishment is humiliation in front of the quarterly Evaluation Meetings. At those meetings, it seems that the focus is on the achievement of maternal mortality targets and the elimination of severe malnutrition. Those village heads with a maternal death are brought to the front and booed by their peers. Despite the efforts above, maternal mortality numbers remain high. In response, in early 2013 the Bupati created a Health Task Force (G‐Gas) via Bupati Decree to further push coordination among actors critical to reducing maternal death. The decree calls for the engagement of a wider range of actors in health and makes their involvement more explicit. It names the following actors at the Kabupaten level: Regional Secretary, Kabupaten staff (Assistant 1, 2, 3), the Head of the Health Department, the Head of the Public Hospital, members of the security apparatus (the police and military‐TNI), the Department of Religion, Women’s Empowerment and Birth Control, and the Social Department. At the Kecamatan and village level the actors are: the Camat, Puskesmas as an institution as well as the Bidan Coordinator and Health Promotion Team, the Village Head, community figures, and PKK. According to the head of the puskesmas in Batudaa Pantai, the health task force has succeeded in increasing coordination among local actors on maternal mortality. For example, the Head of the puskesmas in Batudaa Pantai worked with the head of the local police and Camat to ensure a woman delivered at a health facility. The mother was high‐risk and malnourished, and refused to go to the hospital for her delivery, despite repeated efforts by the head of the puskesmas and puskesmas staff. Finally, the head of the puskesmas involved the Camat and the police, who went to her house to tell her that she needed to go to hospital. Interestingly, the woman continued to refuse, and eventually moved to her husband’s kecamatan in an attempt to be left alone. The new G‐Gas regulation also calls for further implementing regulations fining those who do not give birth with a health professional at a health facility. Regulatory structures are being put in place for the passage of village regulations similar to those in Sumba Timor. Though few villages in Gorontalo have passed these regulations so far, it is expected that they will. In one kecamatan, all village heads have signed an agreement stating that they will not provide children who are born outside of a health facility necessary documentation to allow them to get birth certificates. Planning. Planning remains very ‘top down’ and technocratic with little perspective from the community. In all locations, there was a focus on Musrenbang (the bottom up planning process) as the source of community input to health. In Lombok Utara, this has been taken to an extreme in the development of a separate Musrenbang planning process dedicated to health.12 In Sumedang, an interesting dynamic developed around the Pagu Indikatif Kecamatan (PIK‐money that the government has earmarked to 11
The Bupati boasts about having fired civil servants who were not fulfilling their responsibilities. One story of firing happened to a doctor (a specialist) who prioritized his private practice over public service. Another story involved health department staff who were selling medicines that should have been free to the public. PNS have been fired for being absent from the job 45 days in a year (not in a row, cumulative). 12
While this increases focus on health, it is not entirely clear how it links to the rest of the planning process. 54 fund proposals under Musrenbang). The health department seems to have encouraged its networks to lobby communities to propose training for bidan, and was successful at getting this proposal up through Musrenbang and funded under PIK. However, the quality of the training subsequently provided the health department seems to have left much to be desired, and it seems unlikely that front line actors are likely to try to propose activities again. Throughout the planning process, there remains a gap between community and health department priorities. Much of this difference seems to be attributable to different information about and perceptions of health problems (and solutions) at the village level. A useful example comes from Sumedang and is presented in Box 9. Box 9: Health policy, polindes construction and Generasi in Sumedang In Suriah, Sumedang, the village proposed and succeeded in getting funding for a polindes in a distant dusun through Generasi in 2011. They proposed the polindes so that women would be able to give birth more easily, and as a place that could be used for posyandu. While it is not extremely far to the puskesmas (30 minutes by motorbike), the road is very poor, making the journey very difficult for pregnant women. However, shortly after building was completed, the community was informed by the puskesmas that the polindes did not meet requirements as an “adequate health facility” for giving birth. Interviews revealed that what constituted “an adequate health facility” was not entirely clear. While there had possibly been discussion about the health facility, there was never any written agreement on what the health department would accept. There seemed to be questions about the size of the rooms in the building, electricity, water supply and what medical tools would be kept at the facility. Based on discussions with the puskesmas, medical supplies had been procured under Generasi for the facility, including some equipment for emergencies, such as oxygen. It is not clear who from the community (or the facilitator) was given the list of equipment, but again there was no written agreement that the materials being procured would make the facility “acceptable.” Interestingly, an interview with the head of the health department at the kabupaten level revealed a very different understanding of what was an “acceptable” health facility. She highlighted sufficient space, electricity and water, but said that the bidan had a kit that she could bring with her for assisting in delivery. In this case, there may have been opportunities to use the Generasi structure more effectively to channel information on health policy (in this case, what is an “acceptable facility”) from the kabupaten to the kecamatan, since Generasi actors should be coordinating at each level. While perhaps challenging due to the community empowerment positioning of the program, these types of standards can and should be discussed since the program is funding infrastructure. Funding. As mentioned above, the lion’s share of the funding for puskesmas operational expenses continues to come from the national level through BOK. Funding from district budgets is primarily used to cover staff costs. District spending on health varies from location to location, ranging from 9.4% (Rp 41.570.001.775) in Lombok Utara, to 6.75% (Rp 37.338.721.798) in Gorontalo, to 5.12% (Rp 37.915.833.500) in Sumba Timor. These numbers reflect funding specifically for health through the health department, though calculations for purposes of meeting funding goals (e.g. 15% for health), a broader basket of sources is often used that includes the health department, the social department (e.g. for socializing Jamkesmas); the community empowerment department (e.g. funding for posyandu); and women’s empowerment (generally related to birth control), among others. 55 The first point to be made about district level funding is that there is not much of it outside of what is used on staffing, though variation is increasing as some Bupatis are more aggressively managing their staff and working to move civil servants out of the system. The money that does go to the health department is then divided into many smaller pieces. Once key routine costs are taken out (for example, medicines and immunizations, perhaps supplementary food), it seems difficult to fund the myriad other programs properly. None of this is surprising in a resource constrained environment, but it does mean that without a clear set of health priorities, it seems relatively easy to spread money thinly over a number of ‘siloed’ programs. In addition, the managerial structure of certain activities supports budget fragmentation. A key example is funding and support to posyandu. Since it is supposed to be a ‘community‐owned’ institution, funding runs through the community empowerment division, with the health department only providing technical expertise. 3.4.KeyIssuesaroundtheuseofdata
3.4.1.Datacollectionatthevillagelevel
At the village level, data is required for nearly every program linked to maternal or child health. However, Desa Siaga, family planning programs, Dasa Wisma (limited coverage) and Generasi all collect their own data, as does the bidan for her routine reporting. Programs generally require facilitators or cadre to do house‐to‐house surveys. These data collection overlaps point to an underlying lack of trust in data generally, but requirements for independent collection of data by all of the programs likely contribute to the problem. The fact that many actors in the village overlap (a cadre posyandu may also be a part of Desa Siaga and Generasi, for example) can lead to either using data from another program (if it has actually been collected), giving up and getting data from the bidan or just making an educated guess. In Generasi, some KPMD collect data from the posyandu registers, while some simply go to the bidan. In some cases, there is a subtle ‘fixing’ of data to be sure that it aligns with the health system and therefore does not create any problems. There did not seem to be much sweeping by Generasi actors in study villages, either for routine data collection or service delivery. The timing of data collection can also be problematic. In theory, all posyandu are conducted in the opening weeks of the month, then data from the posyandu is compiled by the bidan and sent to the puskesmas, which then compiles again and sends to the district level. The data is also used at the puskesmas to develop the Plan of Action (POA), which is used to as a basis to receive BOK funding. The timelines are quite tight, and if the data is late from the bidan, the POA may not include all villages. In Sumba Timur, interviews conducted in the middle of one month revealed that only half of puskesmas had reported for the previous month. This has a direct impact on funding and on the provision of services in support of undernourished children, and deserves follow up examination. 3.4.2.Incentivestoreport
It is clear that data is increasingly important to the Indonesian government. Rewards and sanctions, be they monetary, professional or social, around the achievement of certain targets (in particular zero 56 maternal deaths) is strong. However, when the sanctions are seen to outweigh the ‘cost’ of addressing the problem, data manipulation may occur. This seems particularly true with respect to data dealing with under‐ or malnutrition (numbers of children below the red line or those that had been checked by a nutritionist and found to be malnourished). Underreporting can come at any point in the reporting chain: in Sumba Timur, one bidan indicated that she did not report all BGM children in her village, since it would be seen as ‘too much.’ Her decision was partly rationalized by the definition she used of ‘below the red line,’ which was not simply the weight/age, but also other markers of poor health such as red hair and dry skin. In Gorontalo, there were mutterings that the numbers of BGM children decreased each time they were reported up a level. Interestingly, though Lombok Utara took the least aggressive approach to sanctions in the locations sampled, it was also the site of a massive scandal around underreporting of malnourished children, pointing to the strong possibility that incentives to underreport are systemic. More systemic incentives to under‐report may come through the planning and budgeting processes, which are often undertaken on the assumption that indicators will improve, rather than being built on current data or past performance. Resources are decreased in line with projections, and performance assessed on whether targets are met. This creates subtle pressure to report only as many cases as were planned for. There is not yet a culture of ‘testing’ approaches to see if they are working to reduce certain indicators or not. 3.4.3.Abilitytousedata‐Government
Government actors struggle with the data they do receive in a number of ways. First, data used often hides problems. For example, posyandu attendance data assumes attendance at an ‘ideal’ posyandu, where children are not just weighed, but key information related to maternal and neo‐natal health is imparted. Leaving aside the question as to whether the data that is being collected is valid, it is not clear that attendance can be equated with receiving expected information. However, when it is clear that women do not have the information that they need, the reasons why are rarely examined. Recognizing these types of gaps would be very challenging for the health department, since it would force a reconsideration of a number of their approaches. Second, where data show weaknesses, members of the health department struggle to secure funding or bureaucratic support to more carefully examine the problems that they face. For example, if data shows that women are not coming to posyandu, it is silent as to why they don’t arrive. If it shows that women are not using bidan, it is silent as to why they stay away. Outside of monitoring visits, government actors are often at a loss as to how to gather this information. In addition, many who might want to look more carefully at problems in delivery continue to face a hierarchical culture that is not always willing to hear that programs are not working or has the flexibility or initiative to seek alternative solutions. Finally, data is hard to use if there is not a clear way to report and act on it. This issue arose as a coordination problem with Generasi in Sumba Timor. While there was generally communication between the kecamatan facilitator and the puskesmas on which children were accessing assistance from Generasi, it was not clear that the puskesmas had a clear way to report that up to the kabupaten, which then complained that Generasi was not coordinating. It turned out that Generasi had also shared 57 information at the kabupaten level, but because there did not seem to be a way to report support from other sources (i.e. a line or two in a monthly reporting format), it was not effectively entering the government system. 3.4.4.Abilitytousedata‐Generasi
Generasi data is designed for the calculation of bonuses. At the village level, it is very difficult to understand actual progress against indicators and challenging to make an assessment as to whether the village has made good programming decisions. As a result, it is generally not shared back to villagers. At most, villagers will be told if they ‘hit their targets’ or not (though this did not seem to happen often), but if they didn’t, it is very hard to tell how far off they were. In some locations, there was a lack of clarity around how the data was used for village proposals and how it was updated when the money actually arrived. Women whose names were used in the proposals (for example, because they were pregnant and KEK) were sometimes angry if they did not receive money, even if they were no longer pregnant. Clarification of how to use data at the village level may help address this type of problem, and help communities plan more effectively. Above the village level, Generasi struggles to use data as an advocacy tool with government. There are several reasons for this. First, Generasi actors at every level are compromised by the position that they hold. They must continually repeat that they are community empowerment actors in a community empowerment program. The culture of the health department remains very strongly biased towards listening to its own, and discounting those without a sufficient technical background. Without anyone who can speak with authority from a health background to the health department, it is very difficult to find inroads to substantive discussions, even in locations where facilitators are doing an excellent job of coordination on funding. Second, while some blame can be placed on the “technical” culture of the health department, the facilitators do actually need more capacity to analyze data to pick and frame key messages. To date, this has not been a focus of their jobs, since most of the work that they care asked to do with government is coordination rather than advocacy. However, if they need to be having more substantive discussions based on the data, they will need additional support to identify and formulate their messages. Finally, it may be useful to more carefully analyze how data can be best used, and what data would be most useful to collect. Data in Generasi is geared toward incentivizing communities through bonuses, but it has potential value as an advocacy tool to government. This may be particularly useful where the community does not have a great deal of control over outcomes. For example, communities may not have a great deal of control over whether or not immunizations arrive in the village or not, especially if they can no longer top up payments to health care workers to deliver them. In this instance, it may be simply more useful to keep track if they arrive as an accountability mechanism than to hold the community responsible for ensuring a cold chain. If key issues are identified, additional data can be fairly easily collected through the Generasi structure. Potential topics could include stunting, through a collection of data on children in the “yellow zone” of the growth chart; accountability issues, through collection of data on bidan attendance or immunization 58 provision; or general maternal health issues, through a more careful tracking of data captured in the buku KIA. Similarly, if data are to be used to inform government, there are some definitional issues in the data that may need to be addressed. For example, the definition of ‘undernourished’ children in Generasi is different than that used by the health department. There also seems to be some difference in the way that the number of immunizations are counted, which departs from the way that they are recorded in the buku KIA, leading to confusion. Generasi needs to make the choice as to whether having a different definition will raise new issues with the health department or mean that the program is seen as unhelpful. 59 4.Analysis
Based on the information gathered about health information and local actors/institutions, at least four key themes emerge: ‐
‐
‐
‐
Linkages for information sharing about maternal and child health care Incentives and realities around accessing maternal and child health care The use of data Pathways to accountability‐ are there any? 4.1.Linkagesforinformationsharingaboutmaternalandchildhealthcare
The primacy of the posyandu, and by extension the importance of cadre posyandu, is immediately evident in talking about maternal and child health at the village level. In all of the villages visited, it was women’s main point of contact with the formal health system on issues of maternal nutrition, pregnancy risks, and baby and child immunizations and nutrition, among other things. While this does not mean that women only get information and care from the posyandu, or even that all women go to posyandu, it does mean the institution, and the cadre that drive the institution, have a critical role to play maternal and child health. This study highlights both the potential capacity (and by extension, the underutilization) of resources to more effectively help cadre posyandu reach out to their neighbors and better understand village level health issues. Unfortunately, linkages between the puskesmas, the bidan, and the cadre posyandu are often weak, and linkages with “non‐health” actors who could play important supporting roles on health issues are virtually non‐existent. There are several factors that contribute to weak linkages to and support for posyandu and cadre posyandu: First, the posyandu suffers from sitting in a structurally awkward position. Technically, it is under BPMD, as a community empowerment institution, with the health department only providing technical assistance on training to cadre. This results in a great deal of variation in support for cadre posyandu from the puskesmas. The study shows that if puskesmas are both interested and incentivized, they can provide more active support to posyandu. Unfortunately, as variation in both Gorontalo and Lombok Utara show, incentives to increase posyandu attendance do not always result in more active outreach efforts. At least some part of this problem can be attributed to the fact that health officials can claim, with some unfortunate legitimacy, that posyandu is not their responsibility. Structural arrangements have the additional problem of creating a gap between the bidan and cadre posyandu. While cadre provide the bidan with data, she has little responsibility for their training or support, which falls largely to the puskesmas. In no studied villages did the bidan have a routine way of 60 talking to cadre posyandu outside of the posyandu itself, which tends to be quite busy.13 As discussed below, this is also likely linked to the fact that there is little support for on‐the‐job training. Second, there is a gap between the normative conception of the role cadre posyandu and what they are actually supported to do. As part of an institution based on a concept of community empowerment, cadre posyandu theoretically play an important role in helping communities think about health issues and access information and care. In reality, the role that cadre actually play is largely limited to one of data collection and they are generally not supported to understand health issues within the village. At some level, the same is true of the bidan. The gap between what cadre and bidan are described as doing and what they actually do raises a fundamental question about community empowerment in health. If empowerment is taken seriously, bidan and cadre would need to look at the data from posyandu and have the capacity to identify key issues to address or raise to the health department. This does not happen, for a number of reasons. Primarily, there remains a top‐down technocratic approach to health care delivery in many locations that precludes support for village‐level problem solving on health. This is often coupled with the idea that many cadre are simply too uneducated to talk about complex health topics. In areas with low levels of education and literacy, there is merit to this argument. However, the study also found individuals in every village who were highly interested in and engaged with health issues, regardless of education or economic background. In addition, there is a pervasive belief that training needs to be formal (in a meeting room or hotel, with associated per diems and snacks) means that there is generally only budget to train a handful of cadre posyandu per village with state resources. Many programs are supplementing this training, which is an extremely useful starting point. However, the gains from this training diminish rapidly over time without village‐level follow‐up. Despite being the village‐level health resource, bidan are not asked to provide on‐the‐job training to cadre posyandu. While well trained cadre can more effectively support patients getting to the bidan, the bidan is more likely to see training cadre posyandu as a demand on her that is not part of her job, but takes time away from private practice or other activities. Finally, it has to be acknowledged that village‐level actors may face frustration if they identify problems that they cannot solve or, in the worst case, are penalized for even discussing. In many locations, this may be the case with malnutrition and stunting. Third, outside actors that might support a better use of health resources at the village level, primarily village heads or dusun heads, do not usually feel that they have either the authority or the expertise to advocate for better information sharing. The study uncovered initial attempts to address this problem in both Gorontalo and Lombok Utara. In Gorontalo, bidan are required to report certain health statistics to village heads. Unfortunately, many village heads do not understand the information that they are being given. In Lombok Utara, dusun heads are required to be the head of cadre posyandu. In one village with a village head who was very active on health, this works well, allowing the power of village government to come behind supporting posyandu attendance, and allowing it a better understanding of village 13
In one village in West Java, cadre posyandu did mention that the previous bidan had held a routine meeting with the cadre, but that had not been continued by the new bidan. 61 health issues in the process. In a village where the village government was less committed, it did not work as well. While there are myriad challenges in more effectively supporting more meaningful community empowerment on health issues, it is important to recognize the level of existing capacity at the village level. While linkages are currently weak, the bidan, the cadre posyandu (including those among them why are particularly motivated and knowledgeable), and the village government are all sitting at the village level. It would seem that support for capacity building, routine interaction and problem solving at the village level could help citizens access both health information and services, even in the absence of supply side reform. Some of this work has already begun: both Generasi and ACCESS are supporting additional training for cadre posyandu. District governments are increasingly recognizing the value of village head engagement on key issues. 4.2.Incentivesandrealitiesaroundaccessingmaternalandchildhealthcare
This study highlights the general finding that political attention to health outcomes matters. In three out of the four districts researched, requests by Bupati for routine reports on health data or additional engagement in health discussions are probably the strongest drivers of change in reporting and responding to health problems, particularly those related to maternal health. While all three locations took different approaches, political attention to maternal and child mortality and malnutrition has helped to support increased outreach efforts by puskesmas and more active posyandu (particularly in Lombok and Gorontalo) and significant investments in key infrastructure, including physical infrastructure, training for doctors, and increased numbers of more highly skilled bidan (in all three locations). However, the study also strongly shows that there is a fine line between motivating the health system to help families access health information and care and creating incentives to blame or penalize citizens for not taking desired actions. At its core, this trend shows a weakness in the ability of the health system to understand realities of village life and balance them effectively with health needs and responses. In particular, while a facilities‐based approach may provide some of the best options for health outcomes, it does need to be balanced against other factors such as poverty impact and emergency responsiveness. That balancing can only happen with sufficient understanding of the current obstacles facing families seeking to access care. These go significantly beyond laziness, stupidity or backwardness, to issues of information, communications systems, transportation systems and sometimes significant up‐front costs in accessing health care outside of the village. It is important to note that punitive responses documented in this study are strongest in locations (Sumba Timor and Gorontalo) where obstacles are largest. While programs like Generasi can help, there remain limitations. As an example, money for transport to give birth at the puskesmas is still only paid after it is clear that a women went to the puskesmas, so she still needs to have resources up front to cover the actual payment. At the very least it seems important to ensure that village level safety nets are not destroyed without being replaced by an effective health system. This sentiment was best expressed by a villager in Katikuwai, Sumba Timur: “Since the time of our ancestors we always had a person in every neighborhood ready to help women give birth. Now we cannot use them anymore. But they have not 62 been replaced by anyone from the health department. What are we supposed to do now?” Given the multiple challenges facing the current health system, responding to that question needs to be a joint effort between communities and service providers. At the moment, there are some programs that are acutely aware of village health realities, including Generasi and ACCESS. Generasi in particular continues to bridge the gap between health policies and health realities by identifying obstacles to community access and funding infrastructure such as waiting houses and covering transportation costs. However, the program struggles to start or support a much‐
needed dialogue between communities and the health department that would allow for the development of more sensible policy. The weakness of the program in this area seems to stem largely from the fact that it must present itself as a community empowerment program (not a health or education program) due to its institutional alignments. It currently also does not have a systematic way to access expertise in the field that could improve its legitimacy with the health department or improve its capacity support the presentation of information in a way that would be convincing to the health department. 4.3.Theuseofdata
Data collection at the village level is a case study in how a lack of trust compounds and perpetuates poor data quality. Nearly every health program at the village level collects its own data, often requiring a door‐to‐door survey, because they do not trust other sources of data. Since individuals willing to work on health (or in the networks to secure cadre‐type positions) are limited, and much of the data is available (in theory) at the bidan, multiple data requests may actually decrease collection. At the same time, there are some very valid reasons that data should be treated carefully, including the fact that increasing sanctions for poor results may push some individuals to falsify data. As noted above, while political attention to health data can help drive results, it can also create perverse incentives for data falsification. In Gorontalo, village heads are publically humiliated for maternal deaths in their village. In both Gorontalo and Sumba Utara, bidan may be fined or called up to the district office to account for their actions if a woman dies. In all locations, budget limitations around supplemental feeding often mean that not all children reported will be able to receive assistance, and there seems to be at least some self‐censoring as a result. Data quality problems aside, there are at least two more issues to be considered around data use. First, while data is increasingly being used to monitor progress on certain indicators (notably maternal mortality), use in program planning is still inconsistent. While data on under‐ and malnourishment in Lombok has prompted a different approach to managing puskesmas goals, for example, it is not clear if that approach will be evaluated against its effectiveness at reducing problems in nutrition. Nor did that data drive the development of any new types of outreach‐ existing programs were simply implemented with greater vigor and focus. Second, since data is beginning to influence policy direction (at least at the highest levels), space is opening to think more carefully about what is collected, be it by the health system, programs like Generasi or NGOs. For example, while problems of stunting are increasingly recognized, it is hard to talk 63 about the scale of the program because children in the “yellow band” of the growth chart are not reported. They are only reported when they drop “below the red line.” A discussion of at risk children would need to start earlier, but that is currently very difficult because the information is missing. In addition, as discussed above, without more credibility as a ‘health actor,’ Generasi facilitators have trouble using data as an advocacy tool with sub‐district or district health officials. In addition, while many facilitators have an interest in health, they need support to analyze data and prioritize messaging. At an even more basic level, the same is true of village heads and members of BKAD, who are structurally in a position to interface with the health system, but often disempowered by the feeling that individuals without a health background are ill‐equipped to talk about health problems. 4.4.Pathwaystoaccountability?
Part of the goal of this study was to identify initial pathways to accountability, particularly those from the demand side that could be more effectively supported by Generasi. No particular pathways were ultimately identified. What was very clear, however, was the disconnect between community perceptions of their health needs and the approaches taken by the puskesmas. The ability for citizens to talk to (and be heard by), health service providers is fundamental to any future discussion about accountability. This gap may stem from a number of things: training, concepts of modernity, feelings of superiority, etc., but the result is that the health department is often reluctant to value the opinion of anyone without a health background, and does not fully explore community obstacles to access. The lack of recognition of what obstacles and costs communities face in accessing health services has a huge impact on potential pathways to accountability by denying community experience in planning, budgeting and policy‐making. With the increased push to a facilities‐based approach, this lack of understanding may also be supporting what appears to be an increasingly punitive approach to health care. While there remain significant challenges in raising community voices within the health department in a routine way, the role of NGOs and media to advocate on specific issues should not be overlooked. In all three out of the four research provinces, there were examples, however small, of media and NGO activities supporting accountability efforts. 64 5.OpportunitiesandGeneralRecommendations
In thinking about how village actors and institutions can better support health outcomes, this study focuses on how village assets can be most effectively supported and strengthened to provide better care for families and their children. 5.1.Posyandusupport
There are high expectations of the posyandu to drive or support a range of health outcomes, from behavior change to service delivery. However, support for cadre remains both limited and highly formalized, with little to no emphasis on problem solving. While programs like Generasi provide important inputs (such as more extensive training) that improve the skills that cadre bring to the village, routine support and on‐going learning remains both important and lacking. Opportunities include: Provide ongoing support and information to cadre posyandu. Concepts of training in Indonesia remain very formal: training takes place in a meeting room, with a power point, with few questions being asked of the instructor. This formality seems to become an obstacle to empowerment, since cadre are not encouraged or supported to apply lessons to the village environment or to conduct any analysis of problems in their own village. There are a number of informal ways that their capacity could be increased using existing resources, being aware of the time limitations of cadre (who are all volunteers): ‐
‐
Conduct a routine debriefing after each posyandu to get an initial sense of the data with the cadre: think about who didn’t come (and what needs to be done about it), etc. In comparison to the below monthly meeting, it has the disadvantage of probably being rushed and not allowing for cross learning between cadre posyandu. It does have the advantage of being quick and not requiring an extra trip for a meeting somewhere in the village. Hold a meeting between cadre from multiple posyandu and the bidan/ puskesmas each month. It should not be done at the posyandu, since there is not time, and should ideally include cadre from all posyandu (though in large villages this may be impractical) so that they can learn from each other and work to think about problems at the village level, not only at their own posyandu. Topics of discussion could include: o An examination of data over time for children, noting if there are trends and talking about problem cases. This could cover BGM, but could also start considering children in the “yellow zone” of the growth chart, or children whose weight continues to fluctuate. Once children have been identified, information about stunting could be provided and information for counseling parents of at risk children provided. o Technical health information. There are many sources of information available. For example, the cadre posyandu who do receive training from the health department could share what they have learned with the other cadre. There are a number of publications already available from the health department on specific contagious or seasonal illness which could be targeted to communities facing those health problems. Materials developed around “the first 1000 days” are specifically targeted to mothers and young children. 65 Information on local health policies. For example, some local governments are providing incentives for women to give birth at facilities. However, in many instances, women are not aware of these programs, so they can neither access them nor hold government accountable if they are not delivered. Cadre are in a good position to provide this information to women that come to posyandu. o Identification of village examples of positive deviance or ‘role models’ who can demonstrate positive health behavior. o Discussion of problem cases. In many locations, cadre know which women are pregnant but are not coming to posyandu, but feel uncomfortable approaching them if they are in a difficult social situation (unmarried, young, old, etc.). A discussion of what to do in individual cases may help get these women service, potentially including support from other actors (male cadre, bidan, respected community members, etc.). o Challenges: not all cadre may be able or willing to come to a monthly meeting, and not all bidan may be motivated to hold it. Incentives to encourage bidan to hold meetings of this type would need to be discussed, either through the health department or from an outside partner (potentially GSC). In locations without a bidan, there will be a particular challenge for all aspects of coordination with cadre posyandu. Work to identify “knowledge holders” in villages and find ways to support them to share knowledge with cadre. This will be considerably easier, and can potentially happen organically, if there is a routine meeting for the cadre. o
‐
Take “cadreization” more seriously. While the recommendation above would be part of this approach, more ideas may also come out of thinking more carefully about how to recruit and train new cadre. An examination of previous cadre KB approaches may be useful, since it seems to have been successful. Discussions about who in the community gets routinely overlooked in service provision or who drives family decision making around health may also lead to the selection of different cadre who can more effectively support the community. Further discussions with NGOs working on posyandu may also be useful. More interesting media at posyandu. The film that was used in Lombok was well received, and easier to play in the chaos of posyandu. More films may be useful. The use of local language will be an issue in the development/adaptation of films (as it is with information received from many bidan). However, it may also be possible for someone to help explain key messages to community members. Use opportunities to engage men in posyandu, and think about how to use them best. There were a surprising number of male cadre posyandu in research locations. It was not clear how much they are actually actively talking to other men about important issues in maternal and child health. Especially in locations where men have a lot of power in the household (large swathes of rural Indonesia), it is important to have other men willing to talk to them. To do so, there would need to be some targeted support to men, and perhaps a specific approach to attracting them as cadre. 66 Engage with kepala dusun on health and posyandu. Lombok is a special case, but it shows the potential value in engaging village apparatus in the posyandu so that they a) encourage people (including men, since kepala dusun are largely male) to come and b) have a better understanding of village health issues. 5.2.Sharinginformationatthevillagelevel(outsideofposyandu)
Engage the dukun on issues outside of childbirth, particularly nutrition. The current system is pushing dukun out of the discussion about health in an attempt to ensure that they no longer deliver children. This comes at the expense of engaging them more productively on other topics such as nutrition, where they still play a considerable role, and might benefit from positive engagement. Improve the empowerment of the village head on health issues. Village heads are in sufficiently powerful positions to say something if there is a health problem at the village level. Theoretically, they have space to speak at the three‐monthly ‘mini workshops.’ However, many feel intimidated by their lack of ‘technical’ health training. Providing them with information, both of a technical and administrative nature, may improve their ability to identify problems and play a stronger role in the oversight of village service delivery. It will likely be more effective if information is presented at a formal training (particularly if it is taught with input/ people from the health department) that lets them feel that they have a stamp of approval to be acting in an oversight role. Improve incentives for village head to understand health issues in the village. As is evident in Gorontalo and to a certain extent in Lombok, some districts are beginning to see the need for village heads to have a better grasp of health outcomes in their villages. However, this understanding is still nascent, and most village heads have little to no idea about health outcomes in the village, nor do they see it as their role to oversee the bidan in her work. Beyond learning more about approaches taken in different districts to address this problem, there a number of potential ways to incentivize village heads to be more engaged: ‐
‐
Ask village heads to talk about health indicators in their villages. This could be carried out by district governments (Bupatis) in routine meetings, potentially as a more constructive alternative to ‘naming and shaming.’ Or, it could be undertaken in a program context such as Generasi. Either way, the focus should be on helping village heads to understand health indicators and impacts of poor health at the village level. Gather information about bidan attendance from village heads. This may make them more aware that they should be tracking it themselves. This can also be done through government channels or through a program. Create and support opportunities for community members (or representatives through Generasi) to present and discuss health realities with the health department. For example, a realistic discussion about obstacles to access or to implementing a facilities‐based approach would be useful. These discussions will be initially difficult to hold due to power imbalances and preconceptions (of community member behavior by health workers, and vice versa) but may be useful if done in a structured way around a particular problem. For example, supporting community members to provide information in a social audit would be useful both for the health department and community members. 67 5.3.FurtherexploretheroleofNGOadvocacy
The locations selected for this study gave a very limited picture of the role of NGO advocacy in the health sector. The examples that were encountered indicate that NGOs can provide a useful role in helping communities develop solutions to health needs, but that they are sometimes provisional rather than long term.14 There was, however, positive use of media by multiple actors (community members, NGOs, Generasi actors) that should not be overlooked. When used effectively, media was critical in driving change (exposing health problems, etc.) and providing information to community members (community radio). When the media draws on data, it is particularly potent. These initiatives are important, and should be followed up as further outreach initiatives are developed. 5.4.Policydiscussionsandresearchopportunities
While not necessarily an identification of opportunities at the village level, there are certain issues that should continue to be elevated into policy discussions. Is a punitive approach to health care desirable and effective? The response to this question needs to include a broad consideration of effectiveness that include discussions of impact on non‐health outcomes, including poverty and inclusion. It should also take into account the human rights afforded to women under the Indonesian Constitution. What are the costs of the facilities‐based approach? Most of the work on the facilities‐based approach comes from the supply side. While there is evidence on the benefits to women in childbirth, it may be useful from a policy perspective to understand the burdens that the approach places on communities and individuals (and how best to address them). This would entail conducting more detailed research on the costs of childbirth (a reality check type study in health) and how the poorest deal with those or not. How do systems work when things ‘go wrong’? Due dates are never that accurate, even in developed countries. Putting a woman in labor on the back of a motorcycle over a bad road, even for half an hour, is not always something she will (or should) agree to, even if it is the plan under Desa Siaga. The list of potential variables and problems around childbirth goes on. It can be difficult to separate these real problems from the ‘excuses’ of community members who do not want to go to a health facility, but they are real enough that they should not be ignored. Punitive approaches for community members, bidan and dukun and incentives for bidan are increasingly shutting down village level solutions to health problems related to childbirth. Particularly where there are access issues are real and facilities weak, there should be an ongoing consideration of how women’s options in the case of emergency. 14
Interestingly, this echoes findings from the recent LLI 3 study, that villages are now more able to find solutions to problems that they face, but more frequently those solutions only last for the short term. 68 6.RecommendationsforPNPMGenerasi
While many of the general recommendations identified in the previous section can be adopted into Generasi, this section considers specific areas where Generasi can build on or improve its processes or structures to more effectively interact with or promote the accountability of health actors and institutions. Recommendations focus on three key areas: ‐
‐
‐
Improving information collection, sharing and use at the village level to support better problem identification, change expectations about health services and encourage oversight; Strengthening PNPM Generasi linkages to and position vis‐à‐vis the health department; and, Supporting and/or creating opportunities for communities to share their health experiences with government, particularly the health department. 6.1.Improveinformationcollection,sharinganduseatthevillagelevel
Collect data better. The current forms are designed with the calculation of the bonus in mind, rather than how the forms (or the information that they contain) can be used to help inform communities as to whether or not they have reached targets and help them to find solutions. Most village facilitators struggle to understand them (as did the research team) and it is nearly impossible for an average villager. In addition, because Generasi takes a different approach than the health department to recording data (and with some data definitions, discussed below), it is difficult to integrate with government systems. Potential solutions include: ‐
‐
Record data in line with buku KIA. This is a potentially useful way to talk about what is happening in the village and use existing resources to encourage learning. Track ‘real numbers’ (not bonus formula) so that communities can more easily understand the size of the gap that they face in reaching their target. Collect different data. This could be useful in targeting specific problems. To determine what additional data the program may want to collect, it needs to consider what issues it will pursue. For example: ‐
‐
Nutrition, for discussion. Generasi focuses on the number of children who drop weight over a two month period, while the health system focuses on children who have dropped below the red line. This creates some confusion. Both focus on children that are already ‘in trouble’ rather than ‘at risk.’ Collecting data on the number of children in the “yellow” category of the growth chart, and talking about it, may help identify solutions for children before they fall into more serious trouble. This may be beyond current village capacity, but Generasi will struggle to advocate about stunting without data on children who fall into an ‘at risk’ category. For accountability. If the program wants to have more serious conversations about accountability, the program can target data collection around: o Who from the kecamatan comes to posyandu (if they are supposed to); o If there are immunizations at the posyandu; o If the bidan is in the village. The last indicator could be tracked by the by village head and collected by GSC, in line with existing regulations. Asking village head for information about the bidan attendance could help 69 them understand that they are responsible for health. The approach would have the additional value of determining if the provision of data by the village head acts as an incentive for improved oversight. All of the other data could be collected by village facilitators without any additional cost, since they are supposed to attend the posyandu. Support improved data use in program discussions at the village level. Currently, the results of the village survey are an input to discussions about health, but performance against targets in previous years is not. Supporting the use of data in planning discussions may help improve village decision making. For facilitators who may struggle to understand data themselves, it is very difficult to insert data into a discussion geared toward generating village proposals. Helping them to think about how to use data is an important skill that requires training. If a health specialist is introduced into the program (see recommendations below), she or he may be able to help provide this type of support. For example, data might be better used to identify cases of positive deviance, but would require a structured approach to both identify and then discuss useful cases in a village. Provide rights and health policy information during socialization and some regulations. Generasi can help community members know what they should be expecting from government. This can cover basic checks, responses to underweight children, etc. ACCESS may have helped to change service provision through a rights discussion in Lombok, and it may be worth following up with some of their NGO partners regarding the best ways to present this information. Make better use of media. There are examples from Lombok of media use for both information (talk show with a doctor) and oversight (newspaper for complaints about service). Both of these examples can be replicated where there is the infrastructure, and facilitators need to be aware of the potential of these types of activities. 6.2.StrengtheningPNPMGenerasilinkagestoandpositionvis‐à‐visthehealth
department
Generasi currently suffers from being limited to community empowerment. Facilitators are told not to encroach on the sectors by claiming to have any type of technical skill, but then are not taken seriously in the sectors for exactly that reason. It makes it very difficult for them to hold health actors to account. In response, it is necessary to improve the capacity of facilitators at every level to understand health issues and analyze health problems. In addition to increased capacity, Generasi actors need to be able to refer to (and ask questions of) an expert to be able to increase their credibility, even as they continue their role in community empowerment. The health facilitator recommended below acts as the ‘expert’ and can also provide much of the capacity building. Engage a ‘health facilitator’ at the kabupaten level. The current kabupaten facilitators are both overwhelmed with coordination in an administrative sense, and do not have backgrounds to be taken seriously by the health department. The health department needs someone that they feel can help them, and Generasi needs more technical information from the health department. There are a few options to try to achieve improved coordination: 70 ‐
‐
A technical facilitator to sit at the health department, in the Generasi structure. Strong coordination with a different program in similar role. For example, explicit coordination with a program like AIPMNH. However they are engaged, these facilitators need to have a background in health, but with a community perspective. This may be hard to find. Their TOR would consist of a large percentage of time in the community, working with village and kecamatan facilitators, but they would also need to work to provide support to the health department. Examples of work they could undertake include: ‐
‐
‐
‐
‐
‐
Social audits together with Generasi in support of maternal/child audits to see what happens in community in key cases. Cross‐ sectoral coordination on specific issues (nutrition for example). This support could be provided to the health department but also benefit Generasi. Analysis of Generasi data before taking it to government, at both the kecamatan and kabupaten levels, combined with capacity building for facilitators. Development of approaches to holding community discussions about health using village data (as discussed above). Identification of opportunities for engagement between the health department and communities. This could come through the social audit, but also though other events, such as by inviting a few community members or BKAD to routine coordination meeting to present on a thematic issue to build confidence. Share local health policies with Generasi networks, to support information dissemination and also accountability. Strengthen technical knowledge of kecamatan facilitators. This includes: ‐
‐
‐
Support from technical person (health facilitator, above) to help problem solve at the village level. Analytical skills to better understand village problems/ needs and to build approaches for more effectively sharing information with the health department. Substantive discussions in routine meetings to help draw lessons from experience and build practical technical knowledge. Strengthen village facilitators generally, with a focus on understanding health information. Village facilitators can really only be supported by stronger kecamatan facilitators, but more technical information can already begin to be provided through existing training. Coordinate with Bappeda. The program is potentially weakened by its affiliation with the Community Empowerment Division, which is fairly weak. It may be work considering a stronger cross‐sectoral positioning with Bappeda. This would require a significant restructuring of existing relationships. 71 
Fly UP