Health system governance following devolution: Comparing experiences of decentralisation in Kenya and Indonesia

BMJ Global Health, Volume 3, No. 5, Year 2018

Interpretation

Introduction Devolution reforms in Indonesia and Kenya have brought extensive changes to governance structures and mechanisms for financing and delivering healthcare. Community health approaches can contribute towards attaining many of devolution’s objectives, including community participation, responsiveness, accountability and improved equity. We set out to examine governance in two countries at different stages in the devolution journey: Indonesia at 15 years postdevolution and Kenya at 3 years. Methods We collected qualitative data across multiple levels of the health system in one district in Indonesia and ten counties in Kenya, through 80 interviews and six focus group discussions (FGD) in Indonesia and 269 interviews and 14 FGDs in Kenya. Qualitative data were digitally recorded, transcribed and coded before thematic framework analysis. Common themes between contexts were identified inductively and deductively, and similarities and differences critically analysed during an inter-country analysis workshop. results Following devolution both Indonesia and Kenya experienced similar challenges ensuring good governance for health. Devolution reforms transformed power relationships, increasing responsibilities at subnational levels and introducing opportunities for citizen participation. In both contexts, the impact of these mechanisms has been undermined by insufficiently clear guidance; failure to address pre-existing negative contextual norms and practices varied decision-maker values, limited priority-setting capacity and limited genuine community accountability. As a consequence, priorities in both contexts are too often placed on curative rather than preventive health services. Conclusion We recommend consideration of increased intersectoral actions that address social determinants of health, challenge negative norms and practices and place emphasis on community-based primary health services. We used qualitative methods as these allowed the inductive generation of rich data by seeking to understand the ‘why’ and ‘how’ questions about devolution and community health services within specific contexts.31 32 We selected Indonesia and Kenya from among the six countries involved in the REACHOUTi consortium’s work on the equity and effectiveness of community health workers because devolution had been highlighted as a contextual issue affecting community health worker programmes within these countries.25–27 Mixed qualitative methods including key informant interviews, in-depth interviews with close-to-communityii (CTC) providers and their supervisors and focus group discussions with community members were applied separately in each country. In Indonesia, questions probing governance and devolution were added within the topic guide used for studies conducted before and after a quality improvement intervention (studies carried out in 2014 and 2015). Meanwhile, in Kenya, a separate substudy was conducted to explore priority-setting for community health and equity following devolution (April 2015–April 2016), and these data were analysed along with data collected for the mixed methods REACHOUT quality improvement baseline (March–April 2015). The topic guides used in both countries included similar issues but different questions and probes. Other findings from the study carried out in Kenya, which explore the implications of changing priority-setting processes for community health and equity, have been published elsewhere.33 In Indonesia, all data were collected by Indonesian nationals working as part of the REACHOUT consortium in Bahasa Indonesia. Meanwhile, in Kenya national, county and some health worker level interviews were carried out by a foreign researcher (PhD student) in English language (RM). Community and some health facility level respondents in Kenya were interviewed by trained research assistants working with REACHOUT consortium in Kiswahili or Kamba (depending on respondents’ preference). All researchers were trained and experienced in using qualitative methods. None of the researchers had prior relationship with any of the research participants. Both countries have histories of highly centralised and hierarchical government, which had created expectations of national accountability and had left a legacy of patronage and mismanagement of funds.34 35 At the time of the study, both Indonesia and Kenya were lower middle-income economies, with wide income inequalities and uneven access to health services between population groups36 37 (see box 1 and table 1). Indonesia devolved administrative and operational functions to 32 provinces and 440 districts in 2001, with districts responsible for managerial and financial responsibilities for public health30 (see table 1). Indonesia has a three-tier devolved government with national, provincial and district authorities. In Indonesia, reforms were driven in response to economic crisis, the fall of the Suharto regime and strong international pressure.30 Indonesia’s devolution reforms aim to improve quality and reduce costs for public services, including health.30 Following the financial crisis in 1997, there has been an increasing transition towards privatisation contributing to reduced spending for preventive health alongside an increasing role for private insurance companies.44 Previous studies have reported that in the eagerness to reform following devolution, socioeconomic conditions (wide differences in the ability of districts to generate revenue and up to 50-fold variation in income per capita between districts) have been overlooked, with variation in ability to pay for services and hence in access to healthcare.30 Kenya has a shorter history with devolution, having adopted a two-tier devolved government (with national and county level authorities), transferring planning, management and budgeting responsibilities for a range of services, including health, from central government to 47 new subnational governments, (now known as counties) in 2013 (see table 1). Kenya’s reforms were driven by increasing frustration with inefficiencies and inequities associated with the former centralised government process and in response to growing local and international pressure following the postelection violence of 2007–2008.41 Devolution aims in Kenya seek to strengthen democracy and accountability, enhance national unity, increase community participation, improve efficiency and reduce inequities.59 Kenya’s central government have sought to address historic inequities through transfer of the equitable share funding from central government to county governments, determined by the commission for revenue allocation formula,66 which takes into account each county’s poverty level, along with the equalisation fund for formerly marginalised counties.59 Formerly marginalised counties now benefit from higher levels of funding, along with the decision space to invest in health. At the county level, there is no formal guidance to clarify how county budget should be allocated, as a result contextual factors, power dynamics and values play key roles in determining how the budget is allocated. Study sites were selected purposively. In Indonesia, selection was based on high maternal mortality (since this was the focal research area within REACHOUT study), and in Kenya, selection ensured representation of agrarian, pastoralist and urban with varied poverty levels and coverage with maternal health as well as other key health indicators (to provide diverse range of counties studied). In both contexts, participants were selected purposively, in order to identify those who could contribute valuable information (see table 2). More information about the repsondents interviewed in Kenya have been presented elsewhere (see table 4, McCollum et al).33 Respondents were contacted in advance by via telephone or introduction by a known contact, such as their supervisor or Community Health Worker (CHW). Interviews and discussions were carried out at a location convenient for the respondents, often at their place of work for health workers or at an easily accessible community meeting point. Topic guides were used to explore local politics, support from local government for community and/or maternal health, equity of health services and CTC service delivery across both countries. Interviews and FGDs typically lasted between 30 min and 90 min. Respondent demographics in Indonesia *The health system manager is the person with the highest authority at the subdistrict level. He or she is the head of the Puskesmas and is typically a physician or dentist, or someone with a public health background. DHO, District Health Office; FGD, focus group discussion; TBA, traditional birth attendant; NGO, Non Governmental Organisation; MCH, Maternal Child Health; QIC1, Quality Improvement Cycle 1 Semistructured interviews with 80 participants and 6 focus group respondents were conducted in one district in West Java Province (see table 2). Purposive sampling was used to identify respondents for in-depth interviews with health stakeholders (Pusat Kesehatan Masyarakat (puskesmas) (community health centre) and District Health Office officials); non-health stakeholders (subdistrict and village officials); healthcare providers (village midwives, kaderiii and traditional birth attendants) and community members (including women who were pregnant in the past 2 years and men). Interviews with 86 CTC providers and their supervisors and 14 focus group discussions with a further 146 community members were carried out in two counties as part of the REACHOUT quality improvement baseline study. A further 183 interviews were conducted as part of the substudy. These included 14 key informant interviews with national-level respondents and in-depth interviews with 120 county-level decision makers (technical, political, treasury, gender and children’s representatives) across 10 diverse counties. In three of these counties, data from interviews with 49 health workers (health facility and community level) were included. This paper will present findings from a secondary analysis of qualitative data from Indonesia and Kenya. It will provide an overview of similarities and differences between governance mechanisms since devolution and potential implications for community health. Data from both countries were digitally recorded, transcribed and coded using NVivo 10, before framework analysis was conducted. Common governance themes were identified both inductively and deductively, allowing meaning to emerge from the data32 through data familiarisation by reading and rereading narratives drafted from governance-related themes between REACHOUT data collected in both countries and the Kenyan substudy (see figure 1 which presents the common themes across both contexts and their inter-relationship). A data analysis workshop involving researchers from both contexts was held, during which thematic framework analysis was carried out with data compared across and between countries to search for similarities and differences in order to refine themes for further analysis. Emerging from this workshop, we identified three main themes: contextual norms, politics and power dynamics; the influence of power, leadership capacity and values on priority-setting and decision making; and community accountability and empowerment. Common themes from Indonesia and Kenya. Discussions and interviews conducted in Bahasa Indonesia, Kiswahili or Kikamba were translated to English, with a selection (approximately 10%) back-translated for quality checking. Data collection continued until saturation was reached, and data were triangulated between sources to minimise bias. All respondents gave written informed consent, which highlighted the main study objectives. All study findings, including quotations, have been appropriately anonymised, and data collected through this study was safely stored in a password-protected laptop to ensure confidentiality. This study fulfils COnsolidated criteria for REporting Qualitative research criteria for qualitative research.38

AI Digest

Study Justification:

– The study aims to examine the governance structures and mechanisms for financing and delivering healthcare in Indonesia and Kenya following devolution reforms.
– It seeks to understand the impact of devolution on community health approaches and its contribution towards achieving objectives such as community participation, responsiveness, accountability, and improved equity.
– The study is important because it provides insights into the challenges and opportunities of devolution in the health system governance of both countries.

Study Highlights:

– Both Indonesia and Kenya faced similar challenges in ensuring good governance for health following devolution.
– Devolution reforms transformed power relationships and increased responsibilities at subnational levels, allowing for citizen participation.
– However, the impact of these mechanisms was undermined by insufficiently clear guidance, failure to address negative contextual norms and practices, limited priority-setting capacity, and limited community accountability.
– As a result, both countries prioritize curative health services over preventive health services.
– The study highlights the need for increased intersectoral actions, addressing social determinants of health, challenging negative norms and practices, and emphasizing community-based primary health services.

Study Recommendations:

– Consider increased intersectoral actions that address social determinants of health.
– Challenge negative norms and practices in the health system governance.
– Place emphasis on community-based primary health services.
– Improve guidance and clarity in the implementation of devolution reforms.
– Strengthen priority-setting capacity at subnational levels.
– Enhance community accountability in decision-making processes.

Key Role Players:

– National, provincial, and district authorities in Indonesia.
– County governments, national government, and commission for revenue allocation in Kenya.
– Health system managers at the subdistrict level in Indonesia.
– Health stakeholders, non-health stakeholders, healthcare providers, and community members in both countries.

Cost Items for Planning Recommendations:

– Training and capacity-building programs for key role players.
– Development and dissemination of clear guidance materials.
– Implementation of intersectoral actions addressing social determinants of health.
– Strengthening priority-setting capacity through workshops and technical support.
– Enhancing community accountability through community engagement initiatives.
– Monitoring and evaluation activities to assess the impact of the recommendations.

Strength of Evidence

The strength of evidence for this abstract is 7 out of 10.
The evidence in the abstract is based on qualitative data collected from interviews and focus group discussions in Indonesia and Kenya. The study provides a detailed description of the methods used and the contexts of the two countries. However, the abstract does not provide specific details about the sample size, selection criteria, or data analysis process. To improve the evidence, the abstract could include more information about the sample size and selection criteria, as well as provide a clear description of the data analysis process, including how themes were identified and analyzed. Additionally, the abstract could mention any limitations of the study, such as potential biases or generalizability issues.

Abstract

We used qualitative methods as these allowed the inductive generation of rich data by seeking to understand the ‘why’ and ‘how’ questions about devolution and community health services within specific contexts.31 32 We selected Indonesia and Kenya from among the six countries involved in the REACHOUTi consortium’s work on the equity and effectiveness of community health workers because devolution had been highlighted as a contextual issue affecting community health worker programmes within these countries.25–27 Mixed qualitative methods including key informant interviews, in-depth interviews with close-to-communityii (CTC) providers and their supervisors and focus group discussions with community members were applied separately in each country. In Indonesia, questions probing governance and devolution were added within the topic guide used for studies conducted before and after a quality improvement intervention (studies carried out in 2014 and 2015). Meanwhile, in Kenya, a separate substudy was conducted to explore priority-setting for community health and equity following devolution (April 2015–April 2016), and these data were analysed along with data collected for the mixed methods REACHOUT quality improvement baseline (March–April 2015). The topic guides used in both countries included similar issues but different questions and probes. Other findings from the study carried out in Kenya, which explore the implications of changing priority-setting processes for community health and equity, have been published elsewhere.33 In Indonesia, all data were collected by Indonesian nationals working as part of the REACHOUT consortium in Bahasa Indonesia. Meanwhile, in Kenya national, county and some health worker level interviews were carried out by a foreign researcher (PhD student) in English language (RM). Community and some health facility level respondents in Kenya were interviewed by trained research assistants working with REACHOUT consortium in Kiswahili or Kamba (depending on respondents’ preference). All researchers were trained and experienced in using qualitative methods. None of the researchers had prior relationship with any of the research participants. Both countries have histories of highly centralised and hierarchical government, which had created expectations of national accountability and had left a legacy of patronage and mismanagement of funds.34 35 At the time of the study, both Indonesia and Kenya were lower middle-income economies, with wide income inequalities and uneven access to health services between population groups36 37 (see box 1 and table 1). Indonesia devolved administrative and operational functions to 32 provinces and 440 districts in 2001, with districts responsible for managerial and financial responsibilities for public health30 (see table 1). Indonesia has a three-tier devolved government with national, provincial and district authorities. In Indonesia, reforms were driven in response to economic crisis, the fall of the Suharto regime and strong international pressure.30 Indonesia’s devolution reforms aim to improve quality and reduce costs for public services, including health.30 Following the financial crisis in 1997, there has been an increasing transition towards privatisation contributing to reduced spending for preventive health alongside an increasing role for private insurance companies.44 Previous studies have reported that in the eagerness to reform following devolution, socioeconomic conditions (wide differences in the ability of districts to generate revenue and up to 50-fold variation in income per capita between districts) have been overlooked, with variation in ability to pay for services and hence in access to healthcare.30 Kenya has a shorter history with devolution, having adopted a two-tier devolved government (with national and county level authorities), transferring planning, management and budgeting responsibilities for a range of services, including health, from central government to 47 new subnational governments, (now known as counties) in 2013 (see table 1). Kenya’s reforms were driven by increasing frustration with inefficiencies and inequities associated with the former centralised government process and in response to growing local and international pressure following the postelection violence of 2007–2008.41 Devolution aims in Kenya seek to strengthen democracy and accountability, enhance national unity, increase community participation, improve efficiency and reduce inequities.59 Kenya’s central government have sought to address historic inequities through transfer of the equitable share funding from central government to county governments, determined by the commission for revenue allocation formula,66 which takes into account each county’s poverty level, along with the equalisation fund for formerly marginalised counties.59 Formerly marginalised counties now benefit from higher levels of funding, along with the decision space to invest in health. At the county level, there is no formal guidance to clarify how county budget should be allocated, as a result contextual factors, power dynamics and values play key roles in determining how the budget is allocated. Study sites were selected purposively. In Indonesia, selection was based on high maternal mortality (since this was the focal research area within REACHOUT study), and in Kenya, selection ensured representation of agrarian, pastoralist and urban with varied poverty levels and coverage with maternal health as well as other key health indicators (to provide diverse range of counties studied). In both contexts, participants were selected purposively, in order to identify those who could contribute valuable information (see table 2). More information about the repsondents interviewed in Kenya have been presented elsewhere (see table 4, McCollum et al).33 Respondents were contacted in advance by via telephone or introduction by a known contact, such as their supervisor or Community Health Worker (CHW). Interviews and discussions were carried out at a location convenient for the respondents, often at their place of work for health workers or at an easily accessible community meeting point. Topic guides were used to explore local politics, support from local government for community and/or maternal health, equity of health services and CTC service delivery across both countries. Interviews and FGDs typically lasted between 30 min and 90 min. Respondent demographics in Indonesia *The health system manager is the person with the highest authority at the subdistrict level. He or she is the head of the Puskesmas and is typically a physician or dentist, or someone with a public health background. DHO, District Health Office; FGD, focus group discussion; TBA, traditional birth attendant; NGO, Non Governmental Organisation; MCH, Maternal Child Health; QIC1, Quality Improvement Cycle 1 Semistructured interviews with 80 participants and 6 focus group respondents were conducted in one district in West Java Province (see table 2). Purposive sampling was used to identify respondents for in-depth interviews with health stakeholders (Pusat Kesehatan Masyarakat (puskesmas) (community health centre) and District Health Office officials); non-health stakeholders (subdistrict and village officials); healthcare providers (village midwives, kaderiii and traditional birth attendants) and community members (including women who were pregnant in the past 2 years and men). Interviews with 86 CTC providers and their supervisors and 14 focus group discussions with a further 146 community members were carried out in two counties as part of the REACHOUT quality improvement baseline study. A further 183 interviews were conducted as part of the substudy. These included 14 key informant interviews with national-level respondents and in-depth interviews with 120 county-level decision makers (technical, political, treasury, gender and children’s representatives) across 10 diverse counties. In three of these counties, data from interviews with 49 health workers (health facility and community level) were included. This paper will present findings from a secondary analysis of qualitative data from Indonesia and Kenya. It will provide an overview of similarities and differences between governance mechanisms since devolution and potential implications for community health. Data from both countries were digitally recorded, transcribed and coded using NVivo 10, before framework analysis was conducted. Common governance themes were identified both inductively and deductively, allowing meaning to emerge from the data32 through data familiarisation by reading and rereading narratives drafted from governance-related themes between REACHOUT data collected in both countries and the Kenyan substudy (see figure 1 which presents the common themes across both contexts and their inter-relationship). A data analysis workshop involving researchers from both contexts was held, during which thematic framework analysis was carried out with data compared across and between countries to search for similarities and differences in order to refine themes for further analysis. Emerging from this workshop, we identified three main themes: contextual norms, politics and power dynamics; the influence of power, leadership capacity and values on priority-setting and decision making; and community accountability and empowerment. Common themes from Indonesia and Kenya. Discussions and interviews conducted in Bahasa Indonesia, Kiswahili or Kikamba were translated to English, with a selection (approximately 10%) back-translated for quality checking. Data collection continued until saturation was reached, and data were triangulated between sources to minimise bias. All respondents gave written informed consent, which highlighted the main study objectives. All study findings, including quotations, have been appropriately anonymised, and data collected through this study was safely stored in a password-protected laptop to ensure confidentiality. This study fulfils COnsolidated criteria for REporting Qualitative research criteria for qualitative research.38

Materials

N/A

Innovations

Based on the provided description, here are some potential innovations that could improve access to maternal health:

1. Strengthening community health approaches: Community health approaches can contribute to improving access to maternal health by promoting community participation, responsiveness, accountability, and equity. Innovations could include training and empowering community health workers to provide maternal health services, establishing community-based clinics or mobile health units, and implementing community outreach programs to raise awareness and provide education on maternal health.

2. Enhancing governance and accountability: Improving governance structures and mechanisms for financing and delivering healthcare can help ensure better access to maternal health services. Innovations could involve establishing clear guidelines and policies for devolved health systems, strengthening decision-making processes, and promoting transparency and accountability in the allocation of resources for maternal health.

3. Addressing social determinants of health: Taking a holistic approach to maternal health by addressing social determinants of health can contribute to improving access. Innovations could include implementing intersectoral actions that address factors such as poverty, education, gender inequality, and access to clean water and sanitation, which can impact maternal health outcomes.

4. Promoting priority-setting for preventive health services: Shifting priorities towards preventive health services can help improve access to maternal health. Innovations could involve advocating for increased investment in preventive measures such as antenatal care, family planning, and nutrition programs, and promoting the integration of these services into primary healthcare settings.

5. Empowering communities and promoting community accountability: Empowering communities and promoting community accountability can help ensure that maternal health services are responsive to the needs of the population. Innovations could include establishing mechanisms for community participation in decision-making processes, promoting community-led monitoring and evaluation of maternal health services, and fostering partnerships between healthcare providers and community organizations.

These are just a few potential innovations that could be considered to improve access to maternal health based on the description provided. It is important to note that the specific context and needs of each country or region should be taken into account when designing and implementing these innovations.

AI Innovations Description

The recommendation to improve access to maternal health based on the description provided is to consider increased intersectoral actions that address social determinants of health, challenge negative norms and practices, and place emphasis on community-based primary health services. This recommendation is based on the findings that both Indonesia and Kenya faced challenges in ensuring good governance for health following devolution. The impact of devolution reforms on governance has been undermined by insufficiently clear guidance, failure to address pre-existing negative norms and practices, varied decision-maker values, limited priority-setting capacity, and limited genuine community accountability. As a result, priorities in both contexts are often focused on curative rather than preventive health services. Therefore, by implementing intersectoral actions that address social determinants of health, challenge negative norms and practices, and prioritize community-based primary health services, access to maternal health can be improved.

AI Innovations Methodology

Based on the provided description, it seems that the focus is on examining governance structures and mechanisms for financing and delivering healthcare in the context of devolution reforms in Indonesia and Kenya. The goal is to improve access to maternal health through community health approaches.

To improve access to maternal health, here are some potential recommendations:

1. Strengthening intersectoral actions: This involves addressing social determinants of health that impact maternal health outcomes. It includes collaborating with other sectors such as education, transportation, and housing to improve overall health and well-being of pregnant women.

2. Challenging negative norms and practices: This recommendation aims to address cultural and societal norms that may hinder access to maternal health services. It involves promoting gender equality, empowering women, and challenging harmful practices such as early marriage and female genital mutilation.

3. Emphasizing community-based primary health services: This recommendation focuses on strengthening primary healthcare services at the community level. It involves training and supporting community health workers, improving access to essential maternal health services, and promoting community engagement and participation in decision-making.

To simulate the impact of these recommendations on improving access to maternal health, a methodology could include the following steps:

1. Define the indicators: Identify key indicators that measure access to maternal health, such as the number of antenatal care visits, skilled birth attendance, and postnatal care coverage.

2. Collect baseline data: Gather data on the current status of access to maternal health services in the selected areas or communities. This can be done through surveys, interviews, or existing health records.

3. Introduce the recommendations: Implement the recommended interventions or strategies in the selected areas or communities. This could involve training community health workers, implementing awareness campaigns, or improving infrastructure for maternal health services.

4. Monitor and evaluate: Continuously monitor the implementation of the recommendations and collect data on the selected indicators. This can be done through regular data collection, surveys, or interviews with stakeholders.

5. Analyze the data: Analyze the collected data to assess the impact of the recommendations on improving access to maternal health. Compare the baseline data with the post-intervention data to identify any changes or improvements.

6. Draw conclusions and make recommendations: Based on the analysis of the data, draw conclusions about the effectiveness of the recommendations in improving access to maternal health. Identify any challenges or barriers that may have affected the outcomes. Make recommendations for further improvements or adjustments to the interventions.

It is important to note that this is a general methodology and may need to be adapted based on the specific context and resources available for the simulation.

Author

Dima Health

Statistics:

Citations: 26

Identifiers:

DOI: 10.1136/bmjgh-2018-000939

Research Areas:

Community Interventions, Disparities, Health System and Policy, Maternal Access, Maternal and Child Health, Quality of Care, Social Determinants, Workforce

Study Countries:

Kenya

Study Design:

Cross Sectional Study, Exploratory Study, Grounded Theory, Phenomenological Study, randomised-control-trial

Study Approach:

Mixed-methods, Qualitative

Participants Gender:

Male & Female