Introduction 

Until 1971, Uganda had one of the best health care systems in sub-Saharan Africa, which used to provide comprehensive reproductive health (RH) care. More than two decades of unrest seriously eroded health services. Uganda started experiencing the effects of the major socio-political reforms that were ushered in by the National Resistance Movement in 1986. 

 In the past 10 years, various attempts have been made to improve health service provision in the country. Increased availability of resources for health care provision from both the central and local government, donor community as well as from local and international NGOs have led to significant improvements in the utilisation and coverage of some services such as infant immunisation, antenatal care (ANC) and curative services. Despite improvements, health indicators are still a cause for concern. At 42.6 years, the life expectancy is one of the lowest in the world.² 

 Spurred by the focus on improvement in quality of care for health services, a pilot project to reduce maternal mortality, Rural Extended Services and Care for Ultimate Emergency Relief (RESCUER) was implemented in Iganga district of Uganda. Based on this successful experience, it is currently being expanded. 

 Improving Quality of Care in Health Services 

 A three-year National Health Plan drawn up in 1993 reoriented the focus of the Ministry of Health (MOH) towards primary health care under a decentralised system which shifted provision of health services from the central government to local districts. 

 The Government Health Care Service Delivery System in Uganda comprises (a) the first level or primary health care (PHC) which includes the community-level traditional birth attendants (TBAs), health centres 1 and 2, and dispensary-maternity units; (b) secondary level comprising health centre with beds and district hospitals; and (c) tertiary level with referral and national level hospitals. In addition, NGOs and the private sector have many facilities. Under PHC, the facilities are being standardised by government so that each sub-county (about 20,000 population) should have one health centre with eight beds; a county (about 80,000 population) should have a health centre with 24 beds; and each district should have at least one government hospital with 100 beds. The above standard is not yet in place and the situation may therefore vary. 

 To strengthen the capacity of local districts in provision of high quality RH services, and to address the problem of poor management of hospitals and health services, the MOH prioritised improving quality of care. 

 Pilot Project 

 The Government in a pilot project in two districts of Bundibugyo and Kabarole in 1992 pursued quality improvement in health services by strengthening systems and building infrastructure. The process started with defining catchment areas for health service provision and ensuring effective supervision in the catchment areas. Planning workshops were held to develop strategic workplans for each district including development of job descriptions for key health workers. Reactivation of information systems and encouragement of use of the information by those who collect it was intensified. Improvement in supply system by use of inventory and standard equipment lists was made. Special attention to both clinical and administrative problems was undertaken and integration of quality assurance (QA) into district level training was done to make sure that the key concept of quality improvement in health care delivery were properly understood. 

 For each district, a list of problems that affected proper implementation of health programmes was made by selected health worker representatives at workshops held in one of the pilot districts. For Kabarole district, the priority problems selected were: (a) absence of management information system (MIS) for cost-sharing programme; and (b) mismanagement of immunisation programme due to shortages or overstocking of vaccines in health centres and problems of bottled gas in gas cylinders for cold chain maintenance refrigerators. For Bundibugyo District, the priority problems were: (a) water protection scheme was protecting less than desired number of springs; and (b) collection of monthly returns from health units were poor. 

 For each problem, the following problem solving process was used: (a) problem statement; (b) root cause of the problem; (c) standards to be applied in solving the problem; (d) the major objectives to be achieved; (d) strategies to meet the objectives; (e) activities; (f) indicators; (g) the team for this problem; and (h) workplan for solving the problem. 

 The activities undertaken to address the priority problems resulted in improved communication, supervision, training and management of health programmes. As awareness of quality increased in the pilot districts, the problem solving methods were adopted and there was some integration of services. 

 Some of the lessons learnt from this pilot experience were: critical mass of active people in the district is necessary to create change, roles need to be defined more clearly and one person to be in charge of the programme needs to be identified, and additional training is needed. There was also a temptation to focus on systems improvement rather than on work processes and there was inability to ‘own’ the entire process by the district health team. 

 National Quality Assurance Programme 

 A National Quality Assurance Programme (NQAP) was subsequently launched in 1994. The primary objective of the NQAP was to manage the decentralised health services and optimise the scarce resources in order to realise the mission of the MOH. In particular, districts were experiencing difficulties managing health services under decentralisation and required management support. At the same time, there was need for co-ordination and integration of programme activities at the headquarters of MOH in order to support district health services. Discussion with the World Bank mission towards the end of 1993 ended with the agreement that a QA management support approach would best address these needs, especially following the QA pilot project discussed above. The World Bank also agreed to provide the required funds for the programme. 

 Thus, the objectives of NQAP were to: 

1. Increase awareness among health workers and community for improving quality of services;  
2. Develop and disseminate standards which are effective, acceptable and affordable;  
3. Identify priority areas for process improvement;  
4. Promote problem solving through the QA management approach;  
5. Encourage interactions between health providers and community to improve quality.  

The NQAP began with the establishment of a QA Unit composed of three medical officers. Following a QA workshop for Makerere University Medical School lecturers, senior staff of the MOH as well as Mulago National and Teaching Hospital, a 25-member National QA Committee was established, composed of senior headquarter personnel, managers of all national disease control programmes, and representatives from medical schools and central hospitals. This committee assists the QA Unit in implementing the NQAP and integrating its activities within the MOH. The principal focus of the QA programme is on improving the quality of district level health services using quality management methods to identify and solve common service-related problems. This approach induces the development and dissemination of standards or guidelines, determination of the needs of patients and their families, strengthening of communication between health care providers and users, and the use of data to identify quality gaps. 

Organisational Structure at District Level 

Quality awareness workshops introduced quality management methods to district health teams and to district administrative and political leadership. During these workshops, participants selected clinical or administrative problems from their own districts to be addressed over the next six months using quality improvement methods. They then developed workplans to collect the necessary data, formed and applied solutions, and measured resulting changes. Six months later, the district teams met again to compare results of their problem solving activities and to identify another round of problems to be addressed. At the end of the first year, a general meeting was held for all district health teams to share the lessons they had learnt. 

An important part of the programme is Continuous Quality Improvement (CQI) support visits to districts conducted by teams of two or three persons from the National QA Committee. During these visits, the CQI team works with the district health teams to solve problems related to the district workplans and to other administrative or clinical problems district managers are experiencing. At monthly meetings of the national committee, teams discuss their findings and arrange further technical assistance from specific national programmes, if needed. The programme is funded by a loan provided by the World Bank to support the District Health Services Project. The MOH supports the programmes through provision of technical personnel, vehicles and office accommodation. 

The NQAP receives technical support from the Johns Hopkins University. The University helps develop the content of training materials and assists in training, development methods of evaluation, external review of the programme process and documentation of achievements. 

Programme Results 

The NQAP has brought about a re-awakening among the health workers on the importance of quality of care in the health care delivery system. 

An important achievement at the central level has been the strengthening of interaction among disease control programmes, facilitated by monthly National QA Committee meetings and district visits by committee members. The meetings and visits have helped Ministry headquarter staff to appreciate the need for integrated district services. 

Two examples of problems detected by the visiting QA teams were inconsistent drug procurement procedures and difficulties in obtaining health funds channelled through local government. In response, the MOH and Ministry of Local Government quickly established procedures and guidelines which eliminated the problem. 

Maternal Mortality in Uganda 

The WHO defines maternal mortality as, "the death of a woman while pregnant, or within 42 days of the termination of pregnancy, irrespective of the duration or site of pregnancy, from any cause related to, or aggravated by the pregnancy or its management." In Uganda, like in other developing countries, the reasons that women die in pregnancy and child birth are multi-factorial, but can be classified into three main groups: access to health care; the quality of care at a given health facility as determined by personnel, equipment, supplies, and the referral system; and individual characteristics predisposing the mother to high-risk pregnancies. 

Among Ugandan women of reproductive age, maternal mortality is one of the major causes of death and is a major concern to the government and to all health care providers. Currently, the maternal mortality ratio for the country is estimated to be 506 per 100,000 live births.³ The figure varies greatly for different areas of the country, being highest for districts in the North and Northeastern part of the country. 

According to studies done, the main clinical causes of maternal death in Uganda are: sepsis, haemorrhage, abortion, ruptured uterus, hypertensive diseases of pregnancy and anaemia.⁴ The risk factors associated with maternal mortality were found to include: 

Time is a critical dimension in the management of obstetric complications. If mothers in labour receive prompt, adequate, and effective obstetric attention, few mothers will die. Delays in getting proper attention can occur at three points, namely: 

Heightened Concern with Maternal Mortality 

In the Department of Maternal and Child Health and Family Planning (MCH/FP) the QA visits revealed unsatisfactory and high-risk deliveries in both hospitals and health centres. The visits supported earlier findings from the Safe Motherhood needs assessment conducted in 14 districts in 1995. In response, the Department of MCH/FP developed a curriculum to train midwives in Life Saving Skills throughout the country. The curriculum for training of trainers of TBAs was also revised to address the problems of unsafe delivery in the community. Currently, each hospital in Uganda has at least two to five midwives trained in Life Saving Skills using the new curriculum. 

In 1993, l17 of the 126 women referred to the Jinja district’s 30 rural health units died after arrival at Jinja Hospital. Principal causes were found to be haemorrhage (ante-partum and post-partum), ruptured uterus and post-partum sepsis. In addition, it was found that many PHC workers did not routinely identify high-risk pregnancies and some could not identify patients in need of urgent referral. Hospital staff and the district health team began an education campaign for outlying units which stressed the early recognition and prompt referral of high-risk pregnancies and obstetric complications. Hospital record keeping was strengthened with a monthly outcome record of patients referred for obstetric complications. The Director of District Health Services arranged dissemination of this information and follow-up of problems identified. 

In the subsequent 12 months, the number of maternal deaths was reduced from 17 deaths (13.5%) to eight deaths (2.9%), and the number of women referred for obstetric complications increased from 126 to 274. Maternity ward staff feel that the education campaign has resulted in earlier referral of many complications. Both hospital and district personnel recognise that these rates are still unacceptable, and a programme of continuing education and support during supervision visits has been put in place to continue improvements. 

Furthermore, the MOH has established a Safe Motherhood programme to address specifically the problems that contribute to high maternal deaths and extreme suffering among women of reproductive age. 

Rural Extended Services and Care for Ultimate Emergency Relief (RESCUER) 

Although the above efforts showed that maternal mortality can be reduced, their effect was limited because they addressed only a limited number of cases/factors affecting maternal mortality. Any attempt to reduce maternal mortality, therefore, must have as its core, improving women’s access to prompt and effective treatment of obstetric complications. The program would need to work with the community to improve recognition of complications, to convince people that medical care is needed, to facilitate access to services and to improve the quality and responsiveness of RH services. 

Such a system was set up on a pilot basis in Iganga district in 1995 with funding from UNFPA. It involved a collaborative effort among UNFPA, the MOH and Iganga district administration. The innovative referral system called Rural Extended Services and Care for Ultimate Emergency Relief (RESCUER), addresses the three problems of communication, transportation and quality emergency care and services that constrains referrals in many parts of Uganda. 

As a first step, a feasibility study was carried out in the district. After receiving the draft report, a meeting was held between UNFPA, the Population Secretariat, the MOH and the Safe Motherhood Initiative to discuss the findings. The MOH felt that the Initiative was innovative, timely and feasible. The importance of community involvement and contribution to the success of the project was recognised. 

The meeting established an agreement for a referral system to be put in place on a pilot basis for one and half years. UNFPA was asked to develop and approve a project as a component of the RH project within the MOH. Small-scale surveys would be done to assess the effectiveness of the system. Subsequently, other meetings were held with district officials who welcomed the Initiative and pledged their support to it. 

Communications Component 

The RESCUER strategy includes a two-way radio communication system to link TBAs and Safe Motherhood Assistants (SMAs) in remote villages with health referral points. 

During the design of the pilot project, a feasibility study to look into the viability and effectiveness of VHF radio communication system was conducted. The study took into account the need for a 24-hour working system and the provision of power where none existed. The results of the study indicated that the VHF radio communication system was viable and a decision to equip the referral points with solar-powered, VHF radio base stations was made. The stations were connected on a 24-hour basis to walkie-talkies easily accessible to a cluster of TBAs in remote villages. Each walkie-talkie was shared by four to six TBAs. 

In case of an emergency, the TBA or the person attending to the delivery or to an obstetric complication is able to instantly establish contact with the identified referral point, to report on the type of emergency and to immediately receive instructions on how to handle the case and on how to reduce the risk of maternal and neonatal death. The referral point then sends transport to the village to transport the mother quickly to the health unit. The referral point can also communicate with the district hospital whenever necessary. 

As a first step in setting up the communication system, permission to operate ratio communication services was requested from Uganda Posts and Telecommunication Corporation and it was granted in July 1995. The National Frequency Registration Board (NFRB) established a national radio network to facilitate emergency ambulance radio contact so that the frequency assigned will be part of the national network and are on a shared basis. 

Procurement of the radio equipment was done through an open tender process. The base radios were installed and the walkie-talkies distributed to the TBAs and SMAs. A number of minor problems were experienced during the distribution of the radios: 

With regard to transport, a three-wheeled motorised vehicle (tricycle) was designed by one of the local engineering enterprises in Kampala (Katwe) to provide transport between the villages and the PHC or first level referral health units. The tricycle is affordable, low in fuel consumption and easy to maintain. Although this type of transport is widely used for general transport in South Asia, this is the first time it is being used for referral of health emergencies. 

To adopt the three-wheeler to "ambulance" standards, work was done locally, including: 

• User charges collected at the various health units 
• District Medical Officer (DMO) fuel allocation 
• Contribution from relatives 

To identify health referral points, the district was divided into 12 catchment areas including two islands situated in Lake Victoria. In each area, a health facility was identified as a potential referral point in terms of its proximity to the population, accessibility and serviceability. Of these, 10 are first level referral units and two hospitals are second level referral units. 

In order to ensure effective provision of basic obstetric care in all the referral units, the following procedures and supplies were  required of each referral unit: 

1. Performing surgery (caesarean section and laparatomy for ectopic pregnancy) and
2. Performing blood transfusions.   

An in-depth needs assessment of identified referral points was done, and weaknesses and gaps in terms of size, an qualification of personnel, equipment, logistics, drug supply and physical structure were examined. UNFPA, the MOH, the Population Secretariat and the district authorities developed a workplan to address the weaknesses and gaps and to ensure at least an acceptable minimum level of referral care and services. A list of minimal required equipment was established and some renovation work done to ensure a decent and welcoming health service facility. The 10 first level referral units were provided with basic equipment to enable them carry out their functions. Concessions were obtained from the MOH to enable the staff in these units to perform some of the procedures that had hitherto not been allowed at health centres at the first level, such as giving intravenous fluids. 

It was agreed that each first level referral unit should have at least one midwife and one nurse or nurse’s aide and that they should receive intense training in Life Saving Skills (LSS). Operational level RH training was given to nurse’s aides. Accordingly, additional midwives and nurse aide were recruited and trained. It was found necessary to train/re-train various cadres of health providers in order to ensure that quality services were offered at all levels. Training of TBAs in the basics of RH was ongoing in the district, and selected TBAs were also given training on usage of the radio equipment. 

Two radio technicians were recruited and as were drivers for the three wheelers. Attempts were made to attract health workers to the more remote health units. 

The quality of care offered at the referral units was of great concern and training of the health workers was gearedtowards improving it. In addition, a system was put in place for regular support supervision from the district level to the lower level health units and from the health units to the TBAs. TBA and health workers could also use the radio communication system to seek advice as needed. 

The referral system links up community level providers with the formal health system in a cost-effective yet functional way. When an obstetric emergency occurs in a village, the TBA or any other person attending to the woman uses a nearby radio to call for assistance from the nearest health unit. Advice to the attendant on what to do is immediately relayed over the radio. The three wheeler is dispatched from the health unit, with a midwife to fetch the patient. If a case cannot be managed at the health unit level, the hospital is called and an ambulance sent to transport the patient to the hospital. 

A district task force comprising the DMO and officials from other sectors in the district was set up to oversee the planning, implementation and monitoring of the project. To make the project possible, the district made an initial commitment of 10 million Uganda shillings. 

Problems that arose during implementation were addressed as needed. UNFPA Kampala Office and the MOH have been actively involved in monitoring the system from its conception. 

Evaluation of RESCUER 

RESCUER was launched on March 8, 1996 and on evaluation was done 6 months later in October 1996. The evaluation assessed the project’s strengths, weaknesses, viability, sustainability and its potential for replication in other districts and other African countries with similar referral problems. 

During the evaluation process, records were reviewed and interviews carried out with key informants at the community, health unit and national levels. Most of the data collected was qualitative. 

The evaluation found that: 

Availability of walkie-talkies in the community and their use by TBAs have greatly improved confidence of the patients in TBAs and government health units. Community awareness of medical emergencies has increased. However, it is necessary to carry out additional sensitisation of the community at all levels. In addition, the district should utilise other established population projects to sensitise the population about the importance of early referrals. 

Community members should pay for use of the referral system so that they possess a sense of ownership of the system. Since all the patients pay for treatment at the health units, they should also be required to pay for the transport to the health units. Although the community members will never admit that they are able to pay, if the service is available and is good, they will pay. 

Functioning of Health Units 

There has been marked change in the functioning of the health units. They are now offering improved quality of services and many more people are seeking help at these units. Not only are the units better equipped, but they have been painted and look much more attractive. The level of cleanliness in the units has also improved. It shows that with proper planning, even with few resources, districts can improve the quality of services offered in health units. 

The complimentarity of service by ambulances from hospitals to the RESCUER has improved referral system from first level health units to second level health units and has brought about a high sense of responsibility and collaboration among the health workers. The communities reported an improvement in the quality of services being offered at the referral units. This was further evidenced by the growing confidence of the communities in the health units and an increased in number of patients seeking service  from all the units. There is still, however, much that needs to be done to ensure better quality of services. 

Transport Component 

While the communication and service delivery components of the system are functioning well, the transport component needs to be revisited. The tricycles which are meant to transport patients from their homes to the first level of referral are inadequate, frequently breaking down and not suited for the rough roads in the rural areas. Their physical structure has generated a poor image in the communities and communities are not willing to contribute to their maintenance. However, ironically, the same physical structure has raised awareness of the need to have a means of transportation for referring serious cases to the hospital. The referral system has raised the demand for the ambulances of the hospital but the two available ambulances cannot adequately meet these demands as they are old and expensive to run. 

Although there are many complaints about the tricycles, the vehicles are still saving lives and making a difference.The community must manage with the means at hand until a better option is found. It should also be kept in mind that the idea of using the tricycle was not to have an "ambulance," but to have an affordable mode of transport, that is better than bicycles. 

Implementation Difficulties 

Several difficulties were experienced in implementation. First, due to transport problems and lack of personnel, the two referral points on the islands have not been able to operate as referral units. 

Second, a few untrained TBAs have developed a sense of false confidence and have tended to delay referral of high-risk cases. Even a few trained TBAs, who have been entrusted with the use of walkie-talkies, think that they are now qualified health workers, have started demanding salaries from government and have requested for "bigger" clinics in their areas. 

Third, time taken to transport an emergency case from TBA/community varies from 30 minutes to six hours due to several factors which include condition of the roads, whether it is the dry or rainy season, availability of fuel, availability of a driver, and number of emergencies to be transported. Poor remuneration of drivers has affected transport of some obstetric emergencies due to low morale. 

Impact

The RESCUER has greatly contributed to the reduction of maternal mortality in Iganga district as evidenced by the statistics presented in the following tables.

RESCUER does not have a project manager, as it is part of a larger RH programme of the MOH’s Department of MCH/FP, headed by Dr. E. M. Kaijuka.

Expansion to Other Districts

The results of the evaluation were discussed by various stakeholders and the following recommendation were made:

Several steps will be involved in the process of expanding RESCUER into new districts including:

Uganda’s Quality Assurance Programme has raised awareness of the importance of quality in health services both at central and district levels. At the central level, integration of national programmes has been promoted and there is increased awareness of the needsand capacities of district health teams. At the district level, pilot projects such the RESCUER, have been developed and implemented. Evaluations of such pilot ventures have indicated positive and encouraging results.

As shown by the RESCUER pilot experience, effective implementation of an integrated, affordable, accessible and acceptable referral system, a 24-hour two-way radio communication and continuous provision of essential obstetric care in the health units, can significantly reduce maternal deaths in developing countries.

However, to create a culture of quality requires commitment from health workers, patients and the communities. Barriers to establishing this culture include low morale among the health workers, low pay, inadequate supplies and medical equipment and poor health infrastructure.

	1995	1996	1997 (Jan to March)
Antenatal: New cases Re-attendance	228852 24029	26078 29894	4148 4079
Normal deliveries	5410	6094	1356
Birth before arrival	203	182	40
Discharged undelivered	598	718	122
Abortion	752	616	184
Still births	127	246	90
Premature births	77	99	26
Peri-natal death	28	10	19
	1995	1996	1997 (Jan-June)
Referrals	342	1028	579
Caesarean section	468	640	197
Maternal death	48	33	16