Understanding stakeholders' roles in health sector reform process in Tanzania : the case of decentralizing the immunization programm

Abstract

The current need and enthusiasm for health reforms open an important arena for deeper analysis of the policy process with a view to understanding the political determinants of reforms and strengthening implementation. The studies described in this thesis analyse positions of different actors in the reform process, their actions in support or opposition of the process, and their impact on the health sector reform process. Globally and especially in developing countries health sector reforms have been implemented over long periods. Although there have been improvements in health, the remaining burden of disease in many countries is still very high. Reasons for the high burden of disease have been classified into lack of resources and poor organizational and managerial capacity. Good stewardship was needed to facilitate improvement in the performance of health systems. Stakeholders’ alignment and support was one of the most important components of good stewardship. However, stakeholder analysis had not been a common undertaking in developing countries despite the reforms that were being implemented in most of them. It was the aim of this study to answer the question: What has been the role and importance of stakeholders in supporting or opposing the health sector reform process? The study was conducted in Tanzania as one of the poorest countries in Africa, using the decentralization of the Expanded Programme on Immunization (EPI) as a case reference. The study units were the Ministry of Health Headquarters, Medical Stores Department, Expanded Programme on Immunization, national archives, regions and districts. At district level the study units were District Council, Council Health Management Team, EPI managers at regional and district levels, ward and village authorities, health facility, facility providers and households. Qualitative and quantitative methods were used to collect data from January 2000 to June 2002. Relevant data collection instruments were prepared and pre-tested. The qualitative data collection methods included document review,
\nin-depth interviews, key informants interviews and observations. Quantitative
\nmethods involved retrieval of secondary data, health facility survey and
\nhousehold surveys. Regular discussions with key informants and data
\ncollectors were held to verify the findings. Qualitative data was analysed
\nmanually. Quantitative data was captured and analysed using Epi Info version
\n6.1 and STATA version 6.0. The study involved answering five main questions. The first question was: Do
\nreforms learn from history? Analysis of the waves of health reforms prior to
\nthe current reforms from 1926 was done to answer the question. The main
\nstakeholders in the reforms were the political party in power, the government
\nand donors who supported the reforms each time. Each wave of health sector
\nreforms provided information on health provision, financing and resource
\ngeneration. Due to the political contexts, information on failures of health
\nfinancing did not provide lessons for succeeding reforms of the health sector.
\nStakeholders’ political interests opposed lessons that did not match the
\npolitical ideology at the time i.e. free public services versus privatization and
\npaying for social services. Lessons from previous health reforms were
\nselective, and did not consider health-financing needs among others. The
\nongoing health reforms needed to use information from all functional aspects
\nof the health system to provide lessons for improving the health system. The second question was: Who were the stakeholders in the current health
\nreforms and what were their interests and reactions? The main stakeholders
\nwere donors, and the government. The two had a very high support for the
\nreforms evidenced by their participation in problem identification, justification,
\nreform design, planning and implementation. The health sector reforms thus
\nhad high political support at central level. In the implementation process,
\nissues that triggered stakeholders’ reaction included sectoral versus local
\ngovernment decentralization. Another issue was the donor modality in
\nfinancing the health sector and need for adopting new financial management
\nsystems. Among the donors there was hesitancy to join the common financing
\nmodalities that included a Sector Wide Approach (SWAp) and Basket
\n Funding. As a result, there was delay in the process in order to reach better
\nconsensus.
\nThe third question was: What was the impact of stakeholders in the process of
\nreforming a vertical programme like EPI? Health Sector Reforms in EPI
\nincluded integration of generic functions, for example, vaccine procurement to
\nmedical stores department. Qualitative and quantitative data was collected
\nand analysed from the Ministry of Health, EPI management unit. This again
\nrevealed that EPI reforms were well supported by the government and donors
\ncentrally. EPI managers at both district and regional levels opposed some of
\nthe EPI reforms. They argued that coverage was falling due to the reforms.
\nHowever, there was no concrete evidence relating reforms in the EPI
\nprogramme and falling coverage. The primary aim of certain actors was to
\nmake sure that they continued receiving extra income from EPI functions. One
\nof the effects of stakeholders’ reaction was reversal of reforms (recentralization)
\nand return to the status quo. The fourth question was: What was the immediate reaction of stakeholders to
\ndecentralization at district level and how might it have affected performance of
\nEPI functions and the challenges? The immediate reaction of stakeholders
\nwas reduced cooperation between the Council Health Management Team
\n(CHMT) and the District Council who were politically supreme in the district.
\nWithin the Council Health Management Team there was inadequate
\ncommunication, which led to poor teamwork. The result of this was reduced
\nsupervisory visits to peripheral health facilities. The EPI coverage in the study
\ndistrict was 52.8 per cent, which was well below the previous national average
\n(80 per cent). A logistic regression model for EPI service quality variables on
\nchildren between 12 months and 23 months who had completed vaccination
\nwas applied. Certain EPI quality of service variables predicted significant
\nchanges in the odds ratio for completing vaccination. It was then suggested
\nthat strategies were needed to improve management skills among the CHMT
\nand District Council members. Also there was a need of hastening the
\nprocess of increasing remuneration and motivation of peripheral health
\nworkers. The fifth and final question was: What was the interest of the stakeholders
\nand prospects of increasing EPI coverage at district level? Decentralization
\nand integration of EPI functions were among the reforms at district level. The
\nanalysis revealed that active stakeholders at district level were the Ministry of
\nHealth, CHMT, EPI managers at district and regional levels and facility
\nproviders. The Ministry of Health opposed integration of EPI at district level by
\nissuing the directive that DCCOs and MCHCOs (EPI manager at district level)
\nshould resume their tasks. However, the CHMT had no option but to comply.
\nThis action reversed some of the health reforms at district level. Analysis of
\nthe importance the community attached to EPI, using willingness to pay for
\nEPI cold chain kerosene, was done. The support was low (48.7 per cent). EPI
\nservice quality variables were significantly negatively associated with odds
\nratio for willingness to pay for EPI input. Simulation with Policy Maker
\ncomputer software predicted that an increased number of stakeholders
\nthrough community participation would significantly improve the current low
\nlevel of EPI coverage. It was then proposed to do a similar analysis in other
\nvertical programmes and implement on a trial basis the results of the
\nsimulation.
\nIn conclusion, stakeholders were found to be active and influential in the
\nhealth sectors of developing countries like Tanzania but poorly considered in
\nimplementation of reforms. Stakeholders are important since some strongly
\nsupport while others oppose the reforms. The reaction of stakeholders is
\nevident through deployment or non-deployment of information depending on
\ninterest and context. This would result in poor management leading to
\ninefficiency in resource use, which would then be followed by poor quality of
\nservices, poor support by communities and consequently poor utilization of
\nhealth services. It is suggested that stakeholder analysis be conducted in
\nother vertical programmes in the process of integration. Promotion of
\nstakeholder analysis and also Policy Maker as a tool to manage stakeholders
\nwill facilitate the management of reforms in the health sector.