The District Health System in South Africa: Progress made and next steps

July 2001

Table of Contents

Acronyms

1. Introduction
2. Legacies of the Past
3. What is meant by Decentralisation and the DHS
4. Progress in DHS Implementation in South Africa
5. It's never too late to learn: Key lessons from international experiences
6. Health Minmec Decisions and Critical sext steps
7. References

Contact Details

ACRONYMS

1. INTRODUCTION

In the 1990s South Africa is one of the few countries in the world where wholesale transformation of the health system has begun with a clear political commitment to, inter alia, ensure equity in resource allocation, restructure the health system according to a 'district health system' (DHS) and deliver health care according to the principles of the primary health care (PHC) approach.

This article attempts to describe the various forms of decentralisation and how it is being implemented in the health system. In addition, the paper will explain why the DHS was adopted and what progress has been made to date in its implementation. The paper ends by proposing next steps on the path to the establishment of well functioning health districts.

2. LEGACIES OF THE PAST

The Government of National Unity elected in 1994 inherited a highly fragmented and bureaucratic system that provided health services in a discriminatory manner (see for example Ntsaluba and Pillay, 1998). Services for whites were better than those for blacks, those in the rural areas were significantly worse off in terms of access to services compared to their urban counterparts. Expenditure on tertiary services were prioritised above PHC services.

In order to address the problems within the health sector the Department of Health developed policies on a wide range of issues that are contained in the White Paper for the Transformation of the Health Sector in South Africa released in April 1997. The White Paper lays out the vision of the Department and the Ministry of Health. Some of the issues covered by the White Paper range from the mission and goals of the Department, the structure of the national health system, to the role of non-governmental organisations and year 2000 health goals, objectives and indicators of the country. The White Paper presents what needs to be done to correct the ills of the health system and proposes how the Department intends to go about the process of reconstruction.

A significant departure from the past is the decision to create a unified but decentralised national health system based on the DHS model. One of the main reasons for this is the belief that this system is deemed to be the most appropriate vehicle for the delivery of PHC. In addition, the decision to decentralise the delivery of health care is consistent with the overall policy to decentralise government.

The next section will address issues with respect to decentralisation.

3. WHAT IS MEANT BY DECENTRALISATION AND THE DHS

3.1 Decentralisation and health sector reform

The Government of National Unity has adopted decentralisation as the model for both governance and management. Decentralised governance is embodied in the Constitution in the form of the powers and functions of the three spheres of government. The powers and functions of the local sphere of government bears testimony to the importance of this sphere in particular.

In trying to understand what the concept 'decentralisation' means a definition is required. In general terms the concept implies the shift of power, authority and functions away from the centre. It is seen as a mechanism to achieve the following: greater equity and efficiency; greater involvement of and responsiveness to communities; the reduction in the size of the bureaucracy far removed from the communities being served; and greater coordination between social sectors. The World Bank views the decentralisation of public health services as potentially the most important force for improving efficiency and responding to local health conditions and demands (World Bank, 1993).

According to Bossert (1996) decentralisation can take many forms. One set of typologies is the following:

• deconcentration;
• devolution;
• delegation; and
• privatisation.

Deconcentration is defined by Bossert as 'shifting power from the central offices to peripheral offices of the same administrative structure' (p. 147). In the South African case the establishment of provincial regional and district offices for health is an example of deconcentration. Powers are delegated to the peripheral unit to be semi-autonomous, but the peripheral unit is bound to the centre by a common bureaucracy. According to Smith (1979) decentralised personnel are typically full-time career officials, appointed, promoted, remunerated, controlled and deployed by the bureaucratic means applicable to all members of the organisation. Deconcentration emphasises policy cohesion with central planning, control and allocation of resources.

Devolution, on the other hand is the shifting of power and responsibility to separate administrative structures but that are still within the public sector. It often implies 'the transfer of functions or decision-making authority to legally incorporated local governments, such as states, provinces, districts or municipalities' (Rondinelli 1983). As such it is dependent upon the existence of sub-national levels of government. In South Africa, the existence of provincial and local government bodies with responsibilities to provide and manage health services is an example of devolution.

Delegation represents the shifting of responsibility to semi-autonomous 'agencies' ? Which may vary from parastatals, functional development authorities or special project implementation units. The key distinction is that these agencies

 'operate free of central government regulations concerning personnel, recruitment, contracting, budgeting, procurement and other matters, and that it acts as an agent for the state in performing prescribed functions with the ultimate responsibility for them remaining with the central government' (Rondinelli 1983).

Such an arrangement is sometimes referred to a system of indirect administration.

Bossert and others also view privatisation as a form of decentralisation. However, Collins and Green (1994) suggest that it is confusing and inappropriate to suggest that privatisation is a form of decentralisation as it infers a

'transfer within a particular sector or organisation and not between the public and the private sectors'.

Here the contractual relationship between the private and public sectors is the focus of attention. In our context the proposed accredited provider system and the District Surgeon system is an example of decentralisation using resources from the private sector. The use of the private sector to deliver long term and specialised hospital care is another example.

The decentralisation of the health system has already begun, is occurring in many different ways and is not without its problems (see Pillay, 1995 for a fuller critique of decentralisation). Whilst we are using deconcentration and devolution to strengthen the public sector South Africa is also thinking of creative ways of using the resources in the private sector to generate a more coherent and useful public-private mix within the health system. However, a more coherent position towards the private sector is required. Such a position, the development of which is beyond the scope of this paper, must take into consideration factors such as:

• optimal resource utilisation;
• equity;
• ethical considerations; and
• the commodification of health.

As a new democracy for which nation building is an important objective, it is important that in decentralising that we do not lose the national coherence that we are seeking. The challenge, as has been articulated in the White Paper on the Transformation of the Health System, is to build a unitary national health system that is also decentralised.

Unity and national coherence in the national health system is important to ensure that we are able to achieve, over time, a degree of equity in health care delivery within and between provinces, and secondly to promote efficiencies that are gained through economies of scale. In addition, there are health functions that can only be applied or managed at a national level but which require the involvement of provincial and sub-provincial levels of administration and management for their implementation. Examples of such functions include the implementation of national policies and programmes such as the primary school nutrition programme (PSNP), the development of appropriate policies for the management and use of quartenary services like heart transplants which cannot be performed in each of the nine provinces. This should ideally be done in a way that integrates national functions with those of provinces and lower down.

Thus there is a need within a decentralised system to move away from a bi-polar approach that sees power and authority merely shifting between two ends of a centre-periphery spectrum, to one that sees power and authority being appropriately shared in a non-polarised system consisting of different levels of government and administration that can ensure national coherence, efficiency and equity with the delivery of health care. In other words, a well functioning decentralised health system must not be seen in terms of the centre versus the periphery, but in terms of a system that allows the centre and the periphery to work together in a way that allows the potential benefits of a decentralised system to be realised.

The Constitution spells out the powers and functions of the three spheres of government that form the bedrock for the division of functions within the national health system. Thus the national level has the power to make national legislation, set norms and standards, relate to international organisations and the Ministries of Health of other countries, monitor the delivery of services and take over this function when a province is incapable of providing services, and providing services which, because of economies of scale or financial constraints cannot be provided at provincial level. The provinces are charged with planning, regulating and providing health services with the exception of municipal health services. Local government or municipalities are responsible for the rendering of municipal health services.

An important condition for making a decentralised system of governance and administration to work effectively is the acknowledgement that there is no ideal or perfect system. Regardless of what kind of structure of decentralisation and government is adopted, there will always be an overlap of functions and responsibilities between different parts of the system. The structure of the system can only hope to help define some of the boundaries and rules by which the different actors and groups within the system are expected to work together and collaborate to achieve the multiplicity of health aims and objectives (see McCoy, Buch and Palmer 2000).

3.2 Rationale for and principles underlying DHS development in South Africa

Unger and Criel (1995) note that the

 ...district concept derives from two rationales:... the implementation of the PHC strategy, requiring a decentralised management, (and) the organisation of integrated systems which implies that one single team manages simultaneously the district hospital and the network of dispensaries' (p. 125).

In terms of the developments post Alma Ata there was a clear recognition that unless one creates a coherent vehicle to manage the delivery of PHC the objectives set at Alma Ata would not be met. This recognition resulted in the development of the DHS concept that has been promoted by the World Health Organisation (WHO).

Tarimo (p. 4, 1991) defines a DHS as follows:

'A DHS based on PHC is a more or less self-contained segment of the national health system. It comprises first and foremost a well-defined population living within a clearly delineated administrative and geographic area. It includes all the relevant health care activities in the area, whether governmental or otherwise'.

The WHO views the DHS as a vehicle for the delivery of integrated health care (WHO Technical Report Series, 1996). This is an important consideration given the Department of Health's policy decision, reflected in the White Paper on the Transformation of the Health System in South Africa, that service delivery must be both integrated and comprehensive (p. 14).

The White Paper also notes that the establishment of the DHS is a key health sector reform strategy that is also based on the Reconstruction and Development Programme (RDP):

'The health system will focus on districts as the major locus of implementation, and emphasise the primary health care (PHC) approach' (p. 12).

There is national consensus on the principles underlying the establishment of the DHS and what the DHS should strive for. These include: overcoming fragmentation; equity; provision of comprehensive services; effectiveness; efficiency; quality; improved access to services; local accountability and community participation; decentralisation; developmental and intersectoral approach; and sustainability (White Paper, 1997, p. 28).

The role of the DHS within the National Health System (NHS) is also spelled out in the White Paper:

'This level of the health care system should be responsible for the overall management and control of its health budget, and the provision and/or purchase of a full range of comprehensive primary health care services within its area of jurisdiction. Effective referral networks and systems will be ensured through co-operation with the other health districts. All services will be rendered in collaboration with other governmental, non-governmental and private structures' (p. 30).

The following aspects of the role of the DHS in South Africa should be emphasised:

• delivery of comprehensive and integrated services up to and including district hospital services;
• decentralised management responsibility, authority and accountability;
• the planning and management of services delivered at district level;
• the need for effective referral mechanisms within and between districts and levels of care;
• the need to deliver care in the most efficient and effective manner possible;
• the option of purchasing services; and
• the importance of utilising all district resources effectively, whether public, private or non-government organisation (NGO).

4. PROGRESS IN DHS IMPLEMENTATION IN SOUTH AFRICA

4.1 South African experiences pre 1994

There have been several attempts at establishing a decentralised health service in the late 1980s and early 1990s in South Africa. This section will document a few of these in an attempt to seek lessons for the current initiative.

Attempts were made in the late 1980s and early 1990s to create a health district in Bushbuckridge. Some of the results of this work reflect on the importance of creating viable partnerships with local government (Tollman et al., 1993). The importance of working closely with local government parallels the experiences of the Philippines as noted above.

Barron and Fisher (1993) reported on their work in Khayelitsha. They identified three prerequisites for district development viz., commitment and support from the national, provincial and local authorities, control over financial and personnel resources and authority over decision-making. The challenges identified by them included:

• the fragmentation of services;
• inter-professional communication problems;
• attitude of personnel with respect to community involvement;
• lack of planning and evaluation skills; and
• poor referral systems.

The creation of the health ward system in the former KwaZulu homeland was an example of deconcentration. In this system hospitals were made the node of operations for a catchment area with the medical superintendent of the hospital acting as the health ward manager. The advantage of such a model was that it integrated the local hospital with the services rendered by the clinic with the hospital and its resources supporting PHC services. One disadvantage of this system is that the service had the tendency to become hospicentric with an emphasis on the hospital and its services. Practically this would mean that the needs of the hospitals (for transport, personnel and drugs for example) would take precedence over that of the clinics.

One of the former apartheid administrations, the Department of Health and Welfare (House of Representatives) attempted to give local authorities more responsibility with respect to health service delivery (Frankish, 1993). Contracts with forty-three local governments were signed in the late 1980s and early 1990s to render comprehensive health care. The local authorities ranged from small rural management boards to regional services councils. Some local authorities also rendered school health and geriatric services.

The lessons learnt from the above experience include:

• local authorities were eager to take responsibility for rendering comprehensive PHC services provided they were given support and the necessary funds;
• additional services can be provided cheaply if local authority facilities and infrastructure was used;
• because of greater accountability and less bureaucracy action is faster; and
• local government has much flexibility as they are not bound by the central government rules and can therefore adapt to local conditions more easily.

4.2 Implementing the DHS: achievements and challenges, 1994-2000

4.2.1 Policy development process

As stated earlier the rationale for adopting the DHS was included in the African National Congress (A NC) Health Plan and the RDP. What was needed as well, however, was a more fully development policy and implementation strategy. This was drafted by a team of officials from the nine newly established provinces under the leadership of the national Department of Health. The policy document entitled 'A policy for the development of the district health system for South Africa' was released for public comment at the end of 1995.

While the document was positively received by most who reviewed it there were a few criticisms of the process and aspects of the content. Local government health officials were critical that they were not formally represented on the team that developed the document arguing that those that were responsible for the document did not understand local government and how it operates. A second criticism, again largely from local government health officials was against one of the three governance options listed in the draft policy document. Provision was made for one of three governance options, viz., the provincial option, the local government option and the

statutory District Health Authority. Local government officials charged that the document provides for a structure outside of government and therefore not accountable to government. Some local government officials were also critical of the proposal of creating health regions whose role would be to assist in the creation of the health districts.

Notwithstanding these criticisms the Department of Health adopted the document and included aspects of it into the White Paper on the Transformation of the Health System that was formally endorsed by Parliament in 1997 after a series of hearings on the document.

A formal study of the process and some of its outputs was conducted in 1995 as part of a WHO multi-county study. This investigation identified the following weaknesses (Gilson et al., 1995):

• regions may become an obstacle to district development;
• implementation strategies may have overlooked some critical groups;
• top-down implementation runs counter to the PHC approach;
• the linear strategy adopted is inflexible;
• there is little change in the management style of provincial and national managers;
• lack of management capacity and skills; and
• there is no monitoring and evaluation system.

4.2.2 Implementing the DHS policy

The DHS policy was implemented throughout the country from 1995 onwards. The first task that confronted provincial managers was the demarcation of health region and health district boundaries. Two fundamental criteria were used for this process: these boundaries had to be coterminous with local government boundaries and contiguous. Other issues that provinces took into consideration were: financial viability; existing health services; transport routes; sites of economic activity; and geography. By early 1999 there were 39 health regions, 174 health districts and 843 municipalities nationally.

Also by 1999 all health regions were staffed by a management team and staff to support district development and in most health districts management staff were appointed. In a few provinces managers were appointed on an acting or seconded basis. These included: KwaZulu-Natal; Gauteng; The Free State; and the Western Cape. The rationale for this was that these provinces had stronger municipalities than others and it was possible for services to be devolved to the latter at some point in time.

Despite the demarcation of boundaries and the appointment of some personnel at district level there was slow progress in integrating health systems at district level to eliminate fragmentation and duplication between health services rendered by local government and provincial Departments of Health. The integration of health workers at district level proved to be the single most important challenge for the achievement of service integration. Ideally, all health workers in a district should be employed by and report to a single health authority. Unfortunately, the resolution of the governance issue proved to be more difficult than initially anticipated given that the pace of local government restructuring was different to that of restructuring of the health services.

The creation of a single service provider per health district was achieved in most parts of the Northern Cape where the provincial Department of Health took over the rendering of all health services from the smaller municipalities. This proved relatively easy to achieve as the provincial Department of Health provided all of the funding that the municipalities used for the rendering of health services. Staff employed by these municipalities were employed by the Department of Health. However, one municipality contested the decision of the Department of Health and took the latter to the high court. The court decided in favour of the Department of Health as the latter acted in terms of the contract that was signed between the department and the municipality in terms of which the former was obliged to give the municipality a year's notice before the contract could be terminated which was done.

While structural integration was not successful functional integration was achieved in many provinces. In Northern Province, Gauteng, North West and the Eastern Cape some functional integration has been achieved by having joint district level planning structures with local government, and by the secondment of staff to local government facilities in order to provide comprehensive service provision. However, this process has also been difficult. A major barrier to such functional integration is the disparity in salaries and service conditions between health workers employed by provinces and municipalities, and between municipalities of different grades. This issue represents a bigger challenge in large cities and Metropoles, which employ the majority of municipal health workers.

Despite barriers the Albany district in the Eastern Cape province, for example, has been able to document the processes and benefits of functional integration which included:

• curative services being introduced in all municipal clinics;
• staff redeployment strategies finalised;
• duplication of services rendered by both the province and the municipality within a single clinic was rationalised; and
• all facilities,

whether they be provincially administered or run by the municipality are using the same patient held card and tick registers thus making continuity of care and data collection easier (Toomey, 2000).

Despite these achievements, many barriers to the institutionalisation of the DHS remain. These include:

• the determination of the health rendering function of municipalities;
• the transfer of resources; and
• the building of capacity of municipalities to enable them to render comprehensive health care.

Many organisations (NGOs and universities) have worked with the national and provincial departments of health to implement the DHS. While an exhaustive list of organisations and their areas of assistance is beyond the scope of this document a few will be listed:

• Health Systems Trust and the Initiative for Sub-district Support have worked in several health districts in all nine provinces and produced a number of publications which may be found on their website (www.hst.org.za);
• The EQUITY Project which initially focused its efforts in the Eastern Cape Province but has since expanded to become a national project has produced a range of documents (www.msh.co.za);
• The Centre for Health Policy and the Women's Health Project based in the School of Public Health at the University of the Witwatersrand (Wits);
• The Public Health Programme at the University of the Western Cape (UWC);
• The Health Information System Project based at the Universities of the Western Cape and Cape Town (UCT);
• The various schools of public health and training programmes (Medical University of South Africa (MEDUNSA), University of Pretoria, Wits, UCT, UWC).

4.2.3 Monitoring, assessing and evaluating DHS implementation

Early in the process of DHS development a decision was taken that monitoring and evaluation should form a vital part of the implementation process. This need was highlighted in a WHO commissioned study conducted in 1995 (Gilson, et. al., 1995). In 1997 the Centre for Health Policy with the support of the national Department of Health developed a set of indicators to monitor inputs, processes and outputs called “Towards Well-Functioning Health Districts in South Africa” (Gilson, et. al., 1997). This manual was presented and workshopped in almost all provinces in the hope that provinces will institutionalise monitoring and evaluation using these indicators. However, this did not happen for a number of reasons, including:

• structures were in the process of being established and people appointed to positions;
• people's time and energy were consumed with deciding what needed to be done and not much energy was available for monitoring and evaluation; and
• there was a lack of capacity to undertake this task at all levels of the system.

In 1999 the national Department of Health, with the support of all provinces, introduced a national DHS competition. The rationale for this was three fold:

• to institutionalise the use of indicators and the use of data for planning;
• to generate best practices so that these may be used by districts that were not doing well; and
• to reward health workers and communities who were working hard, often under difficult circumstances.

The experiences, successes and criticism of the national district competition, which was repeated in 2000 are described in a document published by the national Department of Health entitled “The District Health System Competition: What have we done and learned in the last two years” (Pillay and Asia, 2000). In brief, the indicators used have been institutionalised in the management processes of at least one province and use by many districts to improve their operations. In addition, there is some evidence to suggest that the morale of health workers in the districts that were nominated by provinces and those that won the national competition improved. However, there have also been criticisms of the competition as well and including:

• that those that do not win feel like losers despite efforts that they made to improve;
• that many districts have not been given feedback by the provincial reviewers; and
• that the competition is usually seen as an event and not integrated into the planning, action and review activities of districts and provinces.

Despite these criticisms, interviews with provincial managers responsible for district development conducted in early 2001 by the first and third authors revealed that all those interviewed viewed the competition (the name was changed to assessment in 2000) as an important activity. It is seen as important to formally evaluate and document progress in DHS development. While some changes to the indicators and process will undoubtedly be made a decision has been taken to repeat the exercise in 2001.

4.3 Constitutional imperatives and the DHS: Challenges for the next 5 years

As stated earlier there has been a dislocation between the restructuring of the health system and the transformation of local government. Ordinarily this would not have been remarkable. However, the need for these processes to cohere is vital given (a) the constitutional mandate of municipalities to render municipal health services; and (b) the decision that municipalities were the ideal sphere of government to take responsibility for the governance of health districts.

The national Department of Health, with the concurrence of the Departments of Finance and Provincial and Local Government have yet to define municipal health services as provided for in the constitution (see later sections for the decisions adopted by the Health Ministers and nine Members of the Executive Council (MINMEC)). While the Department of Health favours a situation where municipalities (district councils and metropolitan councils) take responsibility for rendering a comprehensive package of PHC services, the Departments of Finance and Provincial and Local Government appear to favour a narrow definition of municipal health services. The latter argue that municipalities are currently not, and would not for the short term, doing a reasonably efficient and effective job of rendering their core functions and that they should not be burdened with additional responsibility until they can demonstrate that they can perform their core functions adequately.

While the national departments sort out their differences, municipalities are undergoing a process of restructuring following the determination of new boundaries and the election of a new set of councilors in December 2000. It may be argued that this presents the country with a unique opportunity to obtain consensus on the role of local government with respect to the delivery of health services so that municipalities may plan accordingly.

The Department of Health has decided to reduce the number of health districts in line with the changes in the number of municipalities. Each metropolitan municipality and each district municipality would constitute a health district. This implies that there will be 48 health districts a reduction from the 174 demarcated in early 1999. It has also been proposed that local municipality boundaries may be used to designate sub-districts one or more local municipalities may therefore become sub-districts.

A few scenarios are possible with regard to the role of municipalities in the delivery of health services. The scenarios are wholly dependent on how municipal health services are defined. Firstly, municipal health services may be defined as the comprehensive package of PHC services. If this is the case a further determination needs to be made, i.e., should metropolitan councils and district councils or local councils be responsible for the delivery of these services. A further issue that would need to be resolved is how these services will be funded would they be funded from the revenues generated by municipalities from rates and taxes, or by the provinces or directly from the national fiscus?

Secondly, if municipal health services are defined narrowly, say as environmental health services and preventive and promotive health as preferred by the Departments of Finance and Provincial and Local Government who would render the remainder of the PHC services and how would integration be secured? An option could be that whilst the primary responsibility of municipalities is the funding and rendering of a narrowly defined basket of services, municipalities should be provided with the resources and support to render the remainder of basket of PHC services as well and that the relationship between the province and individual municipalities be regulated via performance agreements.

Given the differences in current health services rendering capacity of municipalities it is possible that a one size fits all strategy may not be feasible or desirable. It may be better to build on the current capacity of municipalities even if in the short term some health districts have more than one health rendering authority. However, this possibility should only be entertained if, via negotiation and joint planning, services are perceived to be seamless to the users and that there is an absence of duplication of services.

5. IT'S NEVER TOO LATE TO LEARN: KEY LESSONS FROM INTERNATIONAL EXPERIENCES

In order that South Africa may learn from the experiences of other countries the literature was explored to seek out the lessons learnt by countries that have a head start with respect to the  implementation of a DHS. In a review the WHO notes that following five important issues that need to be considered (quoted in Tollman et al., 1993):

• organisation, planning and management;
• financing and resource allocation;
• development of human resources;
• community involvement; and
• intersectoral action.

While an exhaustive literature review is beyond the scope of this paper lessons from the following countries will be described: Mexico; the Gambia; Tanzania; Sweden; Norway; the Sudan; and Philippines. The experiences of these countries will be illustrated under the following three headings:

• political commitment;
• decentralised management: process, skills and systems; and
• financing decentralised health systems.

5.1 Political commitment

Both in Mexico (Gonzalez-Block, et al., 1989) and in the Gambia (Conn et al., 1996) the major constraint to decentralisation of health services to lower levels was the reluctance by higher levels to give up control. It may be argued that this showed little political commitment if one assumes that politicians could have facilitated decentralisation on both countries. In Sweden political ambivalence was cited as one of the reasons that the process of decentralisation being slow (Anell, 1996).

Strong political commitment was illustrated in the case of the Philippines. Quimpo (p. 10, 1996) suggests that the process of decentralisation was intended to 'unclog the sclerotic arteries of a bloated central government'. With the creation of a special unit, the Local Government Assistance and Monitoring Service to facilitate the process and to act as the link between the Department of Health and local government, it took six months to devolve health functions to the 1 600 municipalities. This included transferring 99,5% (45 676) health personnel. The relationship between the Department of Health and local government is also cited as a reason for the rapid decentralisation.

5.2 Decentralised management: process, skills and systems

According to Gilson et al., (1994) a major obstacle to district development in Tanzania was the capacity of district managers working within a system which dis-empowered them:

‘District health managers often have limited motivation because of the combined effect of resource constraints, limited authority and inflexible administrative systems, while incentives to improve management, such as salary levels or opportunities for career development, are weak...health managers motivation is further undermined by their skill's weaknesses' (p. 470).

These authors identified four main weaknesses in the health administration structure in Tanzania:

• an inflexible and ineffective resource allocation process;
• an inadequate planning process;
• lack of clarity in terms of accountability within and between levels of the system; and
• lack of management capacity and understanding within the district health management teams. These weaknesses were also found in a study done by Barnett and Ndeki (1992).

Gilson et al., (1994) argue that to improve resource allocation, planning and budgeting the following should be done:

• the development of district health plans on which budget requests and resource allocation is based;
• clarify the roles of the centre, region and district in relation to planning and budgeting;
• develop the role of the region to support the management at district level; and
• enhance the motivation of district health managers to manage by providing them with greater control over resource allocation and greater authority to take action to meet local needs.

Though committed to the PHC approach the health status of the Sudanese showed little improvement (Abdel Rahim et al., 1992). In an effort to address this, the country was divided into 175 decentralised administrative units each with its own health management team. The functions of these teams included:

• responsibility for administrative, technical and financial matters including personnel matters;
• local planning and implementation;
• supervision of health workers;
• organisation of logistical support;
• improvement in coverage of health services;
• integration of health services (vertical services) and intersectoral collaboration;
• promotion of community participation; and
• establishment of an information system.

The main problems experienced in implementing the above policy, which resonate with some of the problems experienced in Tanzania, cited earlier, included:

• central level failed to explain the policy to regional and lower levels;
• absence of clear implementation strategy;
• health area management were unable to initiate and maintain activities given lack of support from the higher levels;
• the administrative separation of facilities run by the Ministry and local government was not resolved - this also made it difficult for the health care management team to supervise health workers at village level; and
• the uneven size of health areas made it difficult for a health team to effectively supervise personnel in large health areas.

Barnett and Ndeki (1992), in an effort to provide district managers with the necessary skills, in Tanzania used the District Action Research and Education (DARE) approach which was also used in Ghana. DARE is described as follows:

'The key terms : 'action research' and 'education' emphasise two central activities in the process. Action research is research conducted by people involved in a situation often used to analyse problems they themselves are experiencing, with the purpose of finding solutions to those problems and monitoring the process by which the solutions are implemented' (p. 301).

Use of the DARE by two different sets of researchers in different regions of Tanzania produced similar results (Barnett and Ndeki, 1992 and Ahmed et al., 1993). The former found that within a year district health management teams were able to identify problems and develop interventions. In addition team spirit was enhanced during the process. Ahmed et al., (1993) found similar results in both urban and rural districts. The urban district focussed on the following: increasing the availability of continuing education for health workers; establishing a library; preparing guidelines for good performance; and developing a system to recognise and reward the best workers. The plan for the rural district focussed on: improving the preparation and follow-up of supervisory activities through workshops and meetings involving members of the district health management team.

Even though decentralisation was adopted as national policy in the Gambia very little was achieved until a donor-funded management strengthening project was established in 1991 (Conn et al., 1996). Regional management teams were established and required to plan using a six-month planning cycle. Both decentralised management and accountability were promoted and management capacity was developed using a 'learning by doing' approach. The major constraint to decentralisation was the centralised control of budgets and other administrative practices.

From an analysis of the Swedish experience Anell (1996) suggests that two prerequisites are necessary for decentralisation to be effective: (i) the administrative level to which responsibility is delegated must be able to handle these new responsibilities and (ii) they must be willing to do so - incentives may be necessary to increase motivation.

Despite a campaign to explain the process of devolution to personnel in the Philippines there were many complaints: security of tenure was uncertain; salary increases and benefits were delayed; health service delivery was jeopardised in the hands of 'non-technical' administrators. The users of the service complained about the lack of drugs and the deterioration in the condition of hospitals (Quimpo, 1996).

5.3 Financing decentralised health systems

According to Gonzalez-Block et al., (1989) Mexican states that achieved decentralisation were characterised by: (i) having smaller social security infrastructure; and (ii) having greater financial decentralisation. It would appear that the extent of decentralisation is really measured by the extent of financial decentralisation - without the power to make financial decisions very little else is possible.

A second lesson from Mexico is that the ability to absorb and utilise resources is very important. Health status in the state of Guerrero declined as a consequence of the lack of capacity to cope with the new responsibilities required. In addition, the cities within the state benefited most as most of the resources went to them thus resulting in inequities in resource allocation within the state. It is possible that this resulted from their ability to absorb the resources more easily than the smaller towns and rural areas.

In Sweden some county councils are too small to manage services especially hospital services effectively (Anell, 1996). Economies of scale should therefore dictate what services are rendered in a particular administrative area.

According to Saether and Hertzberg (1996) during the 1970s Norway's PHC services were fragmented, unevenly financed and characterised by poor co-ordination and unclear delegation of responsibility. This changed with the passage of the Municipal Health Act in 1984 that gave municipalities the responsibility for PHC services and linked the public and private health care sectors. In addition, a new financing mechanism was introduced - block grants to municipalities.

The advantages of the block grant system according to Saether and Hertzberg, 1996 were:

• easy to administer with few transaction costs;
• ensures equitable distribution of resources between municipalities and counties;
• funding is attached to clearly defined responsibilities; and
• it is easier to ensure cost control and containment at the different levels of government.

The use of global budgets in Sweden also showed its usefulness in containing costs but management by objective proved difficult to implement (Anell, 1996). Other disadvantages found in Norway included:

• block grants provide few incentives to increase productivity and efficiency;
• it is a rigid system which can lead to increased waiting times; and
• it does not adapt quickly to changes in health care needs and public preferences.

An early weakness discovered in the Philippines was the non-payment of salary increases in small local government units. Another problem has been the inequitable distribution of resources with cities receiving 3,5 times more than the provinces. This has resulted in many local government units being unable to finance the health services that they were required to provide.

6. HEALTH MINMEC DECISIONS AND CRITICAL NEXT STEPS IN DHS DEVELOPMENT

The Health MINMEC meeting on 13 February 2001 took the following decisions regarding the implementation of the DHS and the role of local government in health service delivery:

District and the Metropolitan Council areas shall be the focal point for the organisation and coordination of health services.

Provincial Departments of Health will be responsible for coordinating the planning and delivery of district health services within the District and Metropolitan areas, in collaboration with local government

Each Member of the Executive Committee (MEC) for Health shall establish a Provincial Health Authority (PHA) in her/his province by the 30 June 2001 whose function will be to advise the MEC for Health;

The PHA shall comprise the MEC for Health and the councillors responsible for health for each District or Metropolitan Council in the province;

The Head of the provincial Department of Health will establish a Provincial Health Advisory Committee (PHAC) by 30 June 2001 whose functions will be to coordinate the planning and delivery of health services and to advise the Provincial Health Authority;

The PHAC will be composed of the Heads of Health of the provincial Department of Health and each District Council and Metropolitan Council;

The MEC for Health will facilitate the establishment of District Health Authorities and community health committees within the District municipalities and Metropolitan areas, using the criteria and guidelines agreed to by the PHA, with the participation of local government;

District Health Services will be provided in every District municipality and Metropolitan  area;

Although the long-term vision is to capacitate municipalities to deliver comprehensive PHC services, in the short-term, these services will exclude services provided by district hospitals;

Municipal Health Services should be defined to include the following:

• environmental health services;
• provision of clean water and sanitation;
• prevention of infectious or communicable diseases;
• health promotion and education;
• provision of community rehabilitation services;
• treatment of minor injuries and diseases; and
• provision of essential medicines for primary care.

After conducting an audit of services provided in each municipality, the MEC for Health may delegate the delivery of PHC services to a Metropolitan or District Council, a local municipality, or a group of local municipalities, with the appropriate capacity, support and resources and this relationship will be managed through a service agreement signed between the province and the municipality, with clearly outlined performance indicators.

The MINMEC has provided clear direction with regard to what needs to be done and, in some instances, some timeframes. The task remains a large one which must be done in the context of limited financial and other resources. As expressed in numerous policy documents, the South African government is committed to the establishment of the DHS but needs to provide clear leadership in a few areas so that progress can be accelerated. There are many lessons from the international experience in establishing decentralised system and South Africa is well placed to learn from these experiences.

7. REFERENCES

Abdel Rahim IM, Elkaki BA, Ali MMA, Elsayd AH, Nalder S and Gorosh M (1992) Smaller health areas for better service. World Health Forum, 13:31-37.

Ahmed AM, Desta A, Tekle K and Mweta E (1993) Pursuing better health care delivery at district

CONTACT DETAILS

Yogan Pillay MSH/The EQUITY Project P O Box 1665 Pretoria 0001 Tel: 021-312 0663 Fax: 021-323 0847

David McCoy Health Systems Trust 504 General Building Cnr Smith and Field Streets Durban 4001 Tel: 031-307 2954 Fax: 031-304 0775

Bennett Asia Department of Health Private Bag X828 Pretoria 0001 Tel: 012-31000 Fax: 012-326 2740

ACKNOWLEDGEMENT AND DISCLAIMER

This Report was made possible through support provided by the US Agency for International Development (USAID)/South Africa under the terms of contract 674-0320-C-00-7010-00. The opinions expressed herein are those of the author(s) and do not necessarily reflect the views of the USAID or Management Sciences for Health (MSH).