Background
Public health facilities in many low-and-middle-income countries (LMICs) often have limited human resources and provide inadequate access to health care for the population. These limitations are associated with inadequate improvements in health outcomes in recent decades. Governments seeking to move towards Universal Health Coverage (UHC) are increasingly recognizing that they need to engage non-state providers (NSPs) to address gaps in their health systems. NSPs include all health care providers outside government health facilities [
1], including private-for-profit providers, private-not-for-profit providers, and informal providers such as traditional healers [
2].
One common approach to engaging NSPs has been contracting with them to deliver primary health care services to a specified population on behalf of the government [
3]. Typically, a formal contract is established between the government and one or more NSPs that stipulates the responsibilities of all parties involved in the contract, the type(s) of health care services to be provided, how the contract will be financed, and accountability and performance monitoring mechanisms.
In LMICs, the private for-profit and non-profit sectors represent important and often well-resourced providers of health care services. Governments are motivated to contract with these NSPs both to utilize all available resources to increase coverage of health services to the population and to improve the effectiveness and efficiency of services through fostering competition [
2]. Contracting-out has also been encouraged by a range of external factors, including the need to quickly scale up vertical health programs, concerns about the quality of available health care services and the lack of adequate health care personnel in the public sector [
2,
3].
The implementation of contracting-out may, however, be hampered in many LMICs by several factors, including high administrative costs and lack of sufficient providers for meaningful competition in rural areas. As elsewhere, existing vested interests among the parties involved in bidding on and awarding contracts may present other challenges to unbiased assessment and management of contracts [
4]. Furthermore, contracting-out may result in further fragmentation of the health system, particularly in countries where monitoring is weak [
3].
In Tanzania, NSPs of health services include faith-based organizations (FBOs), non-governmental organizations (NGOs), private for-profit providers and informal providers [
5]. This paper focuses on faith-based providers, the most prominent group in terms of total infrastructure, number of staff, and geographic reach.
The private not-for-profit sector—of which the faith-based facilities make up the overwhelming majority—is the second largest provider of health services in the country [
5]. The FBO sector owns 23.3% of health infrastructure, while the state owns 60%. However, 41.1% of hospitals are owned by FBOs while 40% are owned by the state, making faith-based NSPs the largest providers of hospital services in the country [
5].
The Tanzanian government has a long history of providing subsidies to FBOs to serve areas without public health facilities. In 1992 the government formally negotiated agreements [
6] to provide bed and staff grants to hospitals managed by FBOs. In districts without a government hospital, the government designated FBO hospitals to serve as District Designated Hospitals (DDHs). In these districts, the government provided operational support to hospitals owned by FBOs.
Since the introduction of the health sector reforms agenda in the 1990s, the concept of partnerships between government and NSPs for health services delivery has continued to gain importance. In 2005 the government revised the 1992 agreements. With this reform, district officials were empowered to contract with NSPs, with contracts to be signed at district level rather than by the Ministry of Health as had previously been the case. In 2007, reforms continued with the introduction of a new type of operational contract known as the Service Agreement (SA). This reform signaled the transition to a formal system backed up by solid legal frameworks and marked the end of basing contracts mainly on informal trust-based relationships [
6].
Studies on contracting-out in LMICs have reported various, sometimes conflicting, experiences and evidence [
7]. For example, in South Africa and Zimbabwe, contracted NSPs reportedly provided health care services of the same or higher quality at lower cost [
7]. However, no significant performance differences were found between contracted and public providers in Ghana and Tanzania [
7]. One review, which focused on the effectiveness of the contracting-out interventions in reaching poor and marginalized groups in low-and-middle-income countries, underscored a lack of robust evidence [
8]. Another review, however, concluded that these interventions could be effective and should be scaled up with more robust evaluation [
9]. Other reviews suggested that while contracting-out has improved access to health services, its effects on other performance areas—such as quality of services, efficiency and equity—remain inadequately understood [
10,
11].
While at least some research has been done on impact, there is a paucity of knowledge on the actual processes underlying development and implementation of contracting-out interventions and the contextual factors that influence their performance. This paper reports on the design and implementation of SAs between local governments and NSPs for the provision of primary health care services in Tanzania. After examining the roles of actors, the policy process, the context and policy content that influenced how the SAs were designed and implemented, it discusses lessons regarding design and implementation of contracting out policy that may be useful learning for other countries. This study thus complements and explores in more depth the findings of a recent study on stakeholders’ perceptions regarding the Service Agreements (SA) [
12].
Discussion
This study explores the contracting strategy used to engage NSPs in the Tanzanian government’s efforts to move towards universal health coverage. Most of the existing literature on contracting-out focuses on assessing impact, rarely describing specific design and implementation features in detail. The study adds new knowledge on the processes by which NSPs were engaged in the context of a resource-poor setting. The context in which contracting-out is implemented and the design features of the intervention greatly influence its chances of success [
10,
11]. The lessons learned in this study regarding contracting policy design and implementation could be relevant to future efforts in Tanzania as well as to other countries implementing contractual agreements between governments and NSPs to improve primary health care services. The paper also provides some reflections on the use of the Walt and Gilson framework.
Building contracting-out into existing policy and practice
Previous involvement of NSPs in the national health system was mainly founded on mutual knowledge and personal, trust-based relationships. The introduction of service agreements as a mechanism through which the Tanzanian government engaged FBOs in providing primary health care services added formality to the process of contracting out. Official Acts, policies and guidelines [
25‐
28] were established that institutionalized and standardized what the agreements covered and how services operated. These legal and policy frameworks and structures facilitated effective contracts to assure the provision of primary health care services according to government standards.
While the formality of the new service agreement offered more guidance and guaranteed more accountability, relationships and trust among different actors at the national and district levels remained influential. These extended beyond the parties named in the SAs. For instance, trusting relationships existed or were built among: key stakeholders in the Ministry of Health who were responsible for policy guidelines and quality assurance; staff at PO-RALG who were responsible for policies at the district level; development partners providing financial and technical support; and the government stakeholders leading the process. Likewise, existing trust between the public and the NSPs in districts were central to encouraging people to seek health services at the institutions.
In Tanzania (as in many other developing countries), development partners actively influence policy design and implementation processes [
29]. The study’s findings revealed that international partners have played a significant role in placing contracting-out on the HSR agenda, as well as in the SA policy design and implementation. This has long-term implications. The government of Tanzania remains heavily dependent on donor funding for health care expenditures, including financing for SAs. Studies elsewhere have indicated that while the support provided by DPs is significant and highly appreciated, it can create problems related to sustainability of the relevant policies and interventions [
30‐
32].
Further, dependence on donor financing and technical support leaves domestic policy processes open to external influence. This can result in a negotiated set of priorities that reflect technical, political, and economic considerations defined more by the interests of donors than domestic needs [
33,
34]. Concerns about the impact of donor dependence can be alleviated when the central government builds sufficient internal technical and financial capacity to meaningfully participate in negotiations and to support district authorities as they establish and finance contractual agreements with NSPs.
Implementing contracting-out policy
With policy guidance and technical capacity in place, implementation becomes the next challenge. Our study revealed that district leaders did hold real authority when negotiating contractual agreements with NSPs. This was a significant difference from earlier models of contracting-out in Tanzania reported in other studies. In these earlier models, contractual agreements were made centrally by the Ministry of Health and district-level authorities were left out of the choice of NSPs and contract negotiation processes [
6].
Financial management remained a problem with SAs. While the districts now had both the mandate and the power to make contractual agreements with the NSPs, they still had little power over the financing of the contracts, nor could they finance them directly with their own resources. Districts depended on the central government to provide financing for the SAs through basket funding from donors. Insufficient and untimely payments negatively affected the implementation of the contractual agreements. We found wide agreement among our respondents that contracted FBOs were compelled to compensate for financing gaps through their own or other external resources; these continued to become increasingly limited. FBOs in Tanzania reported facing growing difficulties resulting from decreased external financial support. Financial management difficulties and gaps had serious negative effects for the faith-based NSPs. This finding corroborates assessments of contracting-out experiences in Cameroon and Chad, as well as other experiences from Tanzania [
6]. For example, in Cameroon it was reported that the Ministry of Health did not fulfil commitments on subsidies, allocation of staff, and official recognition of hospitals as district hospitals, despite repeated requests from NSPs [
6]. Likewise, a recent study in other districts in Tanzania reported significant shortages and delays in disbursements of funds from the central government to NSPs [
35]. Recent studies on decentralization in Tanzania have indicated that transferring decision-making powers without fiscal power can lead to sub-optimal outcomes [
36,
37].
The inclusion of the non-state sector in budgeting and planning processes at all levels is fundamental for strong PPP relationships. Participation by NSPs leads to more efficient and effective use of available resources, especially in district-level annual health plans. However, the private sector allocation in CCHPs remained a constant 25%, without accounting for variations in available providers or level of need in a given district. More comprehensive planning and mapping of resources throughout the sector—both thematically and geographically—could facilitate improved equity in resource distribution. Moreover, the limited capacity of district governments to make timely payments to contracted NSPs may drive deterioration of the relationships between the government and NSPs [
12]. The government and the contracted NSPs must maintain continuous dialogue to ensure clear expectations of roles and responsibilities. Ongoing dialogue would also allow the parties to quickly address and resolve any misunderstandings that occur during the implementation of the SAs.
Implementation of SAs (and other PPP arrangements) required skill sets that were not necessarily available among either district councils or the NSPs. Other assessments of the SAs in Tanzania have reported that needs assessments were not conducted prior to signing agreements; they also found that monitoring and evaluation were not adequately done by the government [
35,
38]. A comparative study on contractual agreements between the government and faith-based health providers in Cameroon, Chad, Uganda and Tanzania reported similar challenges and their negative effects on the contracting experiences in these countries [
6]. Another study, conducted in Malawi, concluded that Service Level Agreements (SLAs) between the government and the FBOs were introduced too quickly, before adequate supporting structures, such as clear policies to guide implementation or arbitration committees to resolve difficulties, had been established [
39]. This resulted in growing mistrust, moral hazard, and in some cases the termination of SLAs. Building public sector capacity to work with the private sector, including developing skills to negotiate and oversee contracts with private providers, is imperative.
The success of contracts often depends on whether they create a sense of accountability in addition to formal requirements for monitoring adherence and providing information to improve services as needed. This study indicated that the lack of mechanisms for monitoring resulted in hospitals setting prices for services which exceeded those in the contract. The NSPs argued that the absence of a review mechanism for the SAs forced them to increase prices to reflect increasing costs and the changing economic context. This highlights the risks of implementing contracts for long periods without review. Studies in other settings also reported on several contracting projects that suffered as a result of poor monitoring [
40‐
43]. These findings suggest that the central government must play a role beyond overall strategic policy leadership and financing of health care. All levels of government should be required to monitor health care delivery in order to remain up-to-date with the situations faced by providers.
Using the policy triangle analysis framework
The Walt and Gilson policy analysis framework helped organize and simplify our study of a complex set of key factors (actors, processes, content and context) and their interrelationships in policy creation. The use of this framework particularly guided the study’s approach to analyzing the socio-economic, political and international contextual factors and actors that influenced the process by which the SA policy was designed and implemented. The framework also made it possible to analyze how the content of the SA policy fulfilled its objectives [
44‐
46]. The policy triangle framework is recommended to researchers seeking to understand complex policymaking and implementation processes [
44,
46]. Knowledge generated from this policy analysis may be useful to researchers and other stakeholders seeking to influence policy-making in LMICs [
12,
44]. Further, using the same framework to study multiple settings enables future cross-country or time-series analyses.
Limitations of the study
This study relied primarily on document reviews and interviews with stakeholders involved in the development and implementation of the service agreements at the district level. The study did not, however, interview any of the intended beneficiaries of the SAs to assess their experiences and perceptions of the health services provided. Secondly, the study was limited to four districts due to budget and time constraints. While efforts were made to sample districts with varying characteristics and respondents involved at different levels of decision-making, the results may not be generalizable to other districts or contexts.
Conclusion
Strengthening PPPs in primary health care is essential to achieving universal health coverage in Tanzania. Introducing service agreements as a mechanism for contracting-out public primary health care services in Tanzania successfully gave districts the mandate and power to make contractual agreements with NSPs. However, financing the contracts remained largely dependent on donor funds via central government budget support. The limited financial control held by the districts undermined effective implementation of the SAs with faith-based health NSPs. NSPs must be more fully involved in district annual health plans and in health budgeting and planning processes at all levels. Meaningful involvement of NSPs should lead to more efficient and effective use of limited available resources. Further, the central government needs to continue building its own and district-level capacity to provide technical and financial support to districts establishing contractual agreements with NSPs. Finally, continuous dialogue is needed between the various parties, including government, donors and contracted NSPs. Communication and dialogue reinforce the trust-based relationships that ensure clear expectations for each party and enable the parties to resolve misunderstandings or other disagreements that arise during the implementation of service agreements. Tanzania’s implementation of SAs in contracting-out delivery of primary health care services has already made significant contributions towards the country’s movement for universal health coverage. Lessons learned during the processes of the SA policy development and implementation can be applied to further strengthen and streamline partnerships among state and non-state actors for health.
Acknowledgements
This work was part of a larger programme of research on engaging non-state providers towards Universal Health Coverage (UHC) supported by the Alliance for Health Policy and Systems Research, World Health Organization with funding from the International Development Research Centre, Canada and the Rockefeller Foundation. We are grateful to the local government officials, district health authorities, non-state providers and other stakeholders in the study districts for participating in the study. We would like to thank Anya Guyer for her help with editing this paper.