Print version ISSN 0042-9686
Bull World Health Organ vol.81 n.4 Genebra Jan. 2003
POLICY AND PRACTICE
A new face for private providers in developing countries: what implications for public health?
Un nouveau type de prestataires privés dans les pays en développement : quelles conséquences pour la santé publique ?
Uno nuevo tipo de proveedores de asistencia privados en los países en desarrollo: ¿implicaciones para la salud pública?
Natasha PalmerI, 1; Anne MillsI; Haroon WadeeI; Lucy GilsonI; Helen SchneiderII
I Health Economics and Financing Programme, Department of Public Health and Policy, London School of Hygiene and Tropical Medicine, London WC1E 7HT, England(email: firstname.lastname@example.org)
II Centre for Health Policy, University of the Witwatersrand, Johannesburg, South Africa
The use of private health care providers in low- and middle-income countries (LMICs) is widespread and is the subject of considerable debate. We review here a new model of private primary care provision emerging in South Africa, in which commercial companies provide standardized primary care services at relatively low cost. The structure and operation of one such company is described, and features of service delivery are compared with the most probable alternatives: a private general practitioner or a public sector clinic. In a case study of cost and quality of services, the clinics were popular with service users and run at a cost per visit comparable to public sector primary care clinics. However, their current role in tackling important public health problems was limited. The implications for public health policy of the emergence of this new model of private provider are discussed. It is argued that encouraging the use of such clinics by those who can afford to pay for them might not help to improve care available for the poorest population groups, which are an important priority for the government. Encouraging such providers to compete for government funding could, however, be desirable if the range of services presently offered, and those able to access them, could be broadened. However, the constraints to implementing such a system successfully are notable, and these are acknowledged. Even without such contractual arrangements, these companies provide an important lesson to the public sector that acceptability of services to users and low-cost service delivery are not incompatible objectives.
Keywords: Delivery of health care; Primary health care; Private sector; Public sector; Quality of health care; Comparative study; South Africa (source: MeSH, NLM).
Le recours à des dispensateurs de soins de santé privés dans les pays à revenu faible ou intermédiaire est largement répandu et fait l'objet de nombreux débats. Nous examinons ici un nouveau type de dispensation des soins de santé primaire par des prestataires privés, qui fait son apparition en Afrique du Sud, où des sociétés commerciales proposent des services standardisés de soins de santé primaires à un coût relativement bas. L'article décrit la structure et le fonctionnement d'une de ces sociétés et compare les divers aspects des services offerts avec ceux des alternatives les plus probables, à savoir un médecin généraliste du secteur privé ou un dispensaire public. Lors d'une étude de cas portant sur le coût et la qualité des services, les dispensaires privés jouissaient de la faveur des usagers et fonctionnaient avec un coût par visite comparable à celui des dispensaires de soins de santé primaires du secteur public. Le rôle actuel de ces dispensaires dans la prise en charge de problèmes de santé publique importants est cependant limité. Les conséquences de l'émergence de ce nouveau type de prestataires sur les politiques de santé publique sont examinées. Il est souligné qu'en encourageant les personnes qui en ont les moyens à utiliser de tels services, on ne contribue pas nécessairement à améliorer les soins accessibles aux groupes de population les plus pauvres, qui constituent une priorité majeure pour le gouvernement. Il pourrait cependant être souhaitable d'encourager les prestataires privés à solliciter une aide gouvernementale si l'éventail des services offerts, ainsi que la population qui y aurait accès, étaient élargis. On sait toutefois que la réussite de la mise en uvre d'un tel système se heurte à d'importantes contraintes. Mais, même en l'absence d'accords contractuels, ces sociétés démontrent clairement au secteur public que l'acceptabilité des services par les usagers et la fourniture de services à faible coût ne sont pas des objectifs incompatibles.
Mots clés: Délivrance soins; Programme soins courants; Secteur privé; Secteur public; Qualité soins; Etude comparative; Afrique du Sud (source: MeSH, INSERM).
En los países de ingresos bajos y medianos el recurso a proveedores de asistencia sanitaria privados constituye una práctica generalizada, que es objeto de amplio debate. Examinamos aquí un nuevo modelo de suministro de atención primaria privada surgido en Sudáfrica, en el que empresas comerciales prestan servicios normalizados de atención primaria a un costo relativamente bajo. Se describe la estructura y el funcionamiento de una de dichas empresas, comparando las características de la prestación de servicios con las alternativas más probables: un médico general privado o un dispensario del sector público. Un estudio práctico sobre el costo y la calidad de los servicios reveló que los dispensarios eran populares entre los usuarios de los servicios y tenían un costo por visita comparable al de los dispensarios de atención primaria del sector público. Sin embargo, su contribución al control de problemas importantes de salud pública era limitada. Se examinan las implicaciones para las políticas de salud pública de la aparición de este nuevo modelo de proveedor privado. Se alega que la promoción del uso de tales dispensarios por quienes pueden pagar por ellos quizá no ayude a mejorar la atención disponible para los grupos de población más pobres, que constituyen una prioridad importante para el gobierno. No obstante, si lograra ampliarse el abanico de servicios actualmente ofrecido y el número de personas con acceso a ellos, sería deseable quizá que se alentara a tales proveedores a competir para conseguir financiamiento del gobierno. Sin embargo, las dificultades existentes para implantar con éxito esos sistemas son considerables y reconocidas. Aun sin tales arreglos contractuales, estas empresas suponen una importante lección para el sector público, a saber, que la aceptación de los servicios por los usuarios y la prestación de servicios de bajo costo no son objetivos incompatibles.
Palabras clave: Prestación de atención de salud; Atención primaria de salud; Sector privado; Sector público; Calidad de la atención de salud; Estudio comparativo; Sudáfrica (fuente: DeCS, BIREME).
The use of private health care providers in low- and middle- income countries (LMICs) is widespread and its implications are the subject of continuing debate (16). One view is that private providers are likely to be more efficient than the public sector and hence that government should contract out services to the private sector. An alternative view is that private providers are often not superior in quality or efficiency to the public sector, and that contracts are not straightforward to design and implement. Finally, there is increasing recognition that neither public nor private providers have uniform characteristics, and that this distinction might overlook more important issues, such as the extent to which a provider uses public funds efficiently and serves the goals of public health (3).
The present paper contributes to this debate by describing a new model of private primary care provision emerging in South Africa, in which private companies provide standardized primary care services at a relatively low cost. Drawing on data from several case studies carried out across a range of South African primary care facilities, the paper compares aspects of service delivery by one such company with the most likely alternatives for patients on a low income: a private general practitioner or a public sector clinic. The final part of the paper discusses the opportunities and threats presented by this new model and its potential as a partner for the public sector in health service delivery.
Public sector primary care is free in South Africa, yet around 30% of people without medical insurance still choose to pay out of their own pocket to attend facilities in the private sector. Even in the lowest income quintile this proportion is estimated to be 20% (7). The market for private patients is lucrative and most general practitioners are in private practice (8). The high use of the private sector is partly due to the inaccessibility of public services, but in urban areas it might be because of perceptions of greater privacy (9, 10), speed of service (10), and quality of diagnosis, prescribing and counselling (1114). Such perceptions aside, other evidence suggests that general practitioners often deliver care of questionable technical quality, especially with respect to the quality of diagnosis and use of appropriate drugs, indicating that there is scope for improvement of the services on offer (15, 16). The profits to be made from out-of-pocket payment for primary care services are encouraging the formation of private companies aiming to compete with general practices by running clinics with higher standards of patient care at a lower charge. Patients attending these clinics pay a flat fee per visit (which includes drugs, laboratory tests, and X-rays) or are covered by low-cost medical insurance schemes. Currently, there are more than 100 such private, branded, primary care clinics in South Africa owned by a handful of companies (exact numbers fluctuate in a highly competitive market), and some companies have already opened clinics in the main cities of neighbouring countries. Larger companies are achieving widespread coverage and brand recognition. Some company managers have also expressed an interest in contracting with the government to deliver services on their behalf, but as yet they have had no clear response from the South African Department of Health.
The private clinic company used as an example in this paper was evaluated as part of a larger study examining the potential of a range of private providers to deliver services on behalf of the South African government. It was selected because it was the largest, most geographically dispersed, and long-established of the clinic chains active in the market at the time of the research. The company ran clinics in urban areas throughout South Africa, which were usually located close to centres of employment such as factory complexes. Clinics were open from 08:00 to 18:00 on weekdays and 08:00 to 15:00 on Saturdays, with no out-of-hours emergency service offered. Services were priced competitively at the lower end of the range charged by general practitioners and included access to popular services such as ultrasound and X-ray. The clinics all had similar, clean, bright, and modern surroundings. The company projected a strong brand image, with uniformed staff and frequent use of the company logo. In place of direct access to a doctor, patients were seen by a hierarchy of providers, starting with a 'primary care worker' (lay health care worker), followed by a primary health care nurse, and then, where necessary, a doctor.
The structure of the company was strongly hierarchical with very limited scope for decision-making by individual clinics or treatment providers. At each clinic an on-site manager was closely supervised and supported by an area manager based at the company's head office. Clinic information systems were linked to the company's head office by a computer system that transmitted financial and clinical information daily. This system contained over 2000 diagnostic and treatment algorithms based on evidence from the Cochrane Collaboration Reviews. At each patient contact, health care workers entered the patient's details and symptoms to prompt assistance in diagnosis and to access a recommended treatment protocol; a link to the dispensary enabled the system to record all drugs dispensed and to check these against the treatment protocol that it had recommended. Staff who did not follow recommended treatment protocols were identified and followed up. This system enabled head office to review and audit clinical practice on a daily basis as well as to track costs closely.
This evaluation was part of a larger research project that comprised a series of case studies of different types of providers, providing in-depth information on a range of provider models and indicative data on cost and quality. The case-study approach was adopted to enable the exploration of performance levels and influences over performance for the main primary care provider models in existence within South Africa at the time of the study. In the model considered here, clinics were part of a directly and closely managed hierarchy, and so it was decided that a sample of two clinics was sufficient for the case study. The clinics were selected in consultation with the company's management as representative of typical operations. They were in separate provinces and had been operational for two to three years at the time of the evaluation.
At the head office, interviews with senior management were conducted, the management information system was inspected and cost and activity data for the case study clinics were obtained. At the clinic level, cost data were collected for the financial year 199899 from facility and head office records and by observation (for capital items). Capital items were valued in terms of replacement costs, and annualized using a discount rate of 8%, with life spans of 30 years for buildings, five years for medical equipment and ten years for other equipment. Utilization data were obtained from clinic information systems. Costs per visit and for specific types of service were calculated.
Several methods were used to assess quality of care. Exit interviews (50 per site) and focus group discussions (one female group near each facility) were used to assess service user perceptions of the quality and acceptability of the clinics' services. In the focus groups, participants were asked to describe their perceptions of the relative merits and accessibility of alternative public and private sector providers in the area, including this clinic chain. Objective assessment of the technical quality of services was conducted by record reviews of treatment of sexually transmitted infections (STIs), diabetes, and hypertension, with a target of 30 records for each tracer condition at each clinic. A structural checklist was used to assess the adequacy of resources and services available, including drug availability. Finally, as an indicator of process quality and acceptability of the service to users, waiting times (for 50 patients) and consultation times (for 30 patients) were measured at each clinic. The same methods were used to review service delivery in a sample of five public sector clinics and two private practices of general practitioners, enabling results below to be presented in the context of comparable data from alternative providers.
The cost to the provider of a visit to the private clinics is shown in Table 1 and compared with public sector clinics categorized by size, and whether there was a full-time doctor, and private general practices. The cost per visit at the private clinic chain was R 35 and R 44, and was within the range of public sector clinic costs, which varied from R 27 to R 68 per visit. General practice costs were considerably higher, averaging R 89 per visit. Despite having full-time doctors, the recurrent costs of the private clinics were only slightly above those of smaller public clinics without full-time doctors. The private clinics bore a high cost for external administration, reflecting strong management support from head office. However, clinical staff costs were kept low by using nurse practitioners as the main service provider, and drug costs were contained by using a basic company formulary, as well as by strict control of prescription practices via regular audit of dispensing patterns, using the computer system described above.
A comparison of the nature of service delivery and quality of care between the private clinics and other primary care providers not only reflects points commonly made in the publicprivate debate, but also highlights some that have received less emphasis. People using private clinic services were financially well off compared with those using other primary care models assessed in the overall study, but they were, nonetheless, among the lower income groups within the country. Acceptability of services to users of the private clinics appeared to be very high. Focus group discussions highlighted patients' feelings of being treated promptly and with respect ("they make you feel very important"), which, with some exceptions, was not echoed for public clinics. The structural quality assessment showed excellent standards in the private clinics in terms of cleanliness, space, and the availability of drugs and equipment, and comments in the focus group discussions showed that these were appreciated ("you get everything, pills, care, it's clean"). The waiting time at the private clinics was usually 1040 minutes, compared with 50 minutes to 3 hours recorded in the public clinics.
Patterns of use differed markedly at the private clinics compared with the public sector. Key public health services such as immunization and treatment for tuberculosis were not available at the private clinics, which referred patients into the public sector or to a general practice. Patients appeared to "shop around" between different providers for different services, tending to use the public sector for treatment of chronic conditions and private clinics for curative care. In exit interviews at the private clinics, most respondents (54%) had visited another facility in the previous six months, compared with 22% of respondents at public clinics. One third (30%) of private attendees who had been to another provider in the previous six months had been attending the public sector for chronic care. This use of multiple providers was also reflected in the results of the record reviews at the private clinics. In all, 64% of diabetic patients whose records were reviewed (n = 53) and 48% of patients with hypertension (n = 58) had visited the clinic only once, suggesting deficiencies in the continuity of chronic care provision.
Other indicators of the care available at the private clinics were generally good, with the notable exception of one clinic having no condoms available. Treatment of STIs was used as a tracer and, other than the lack of condoms, technical quality of curative care appeared to be high. A total of 85% of STI patient records (n = 60) showed that the disease had been diagnosed using the syndromic approach (compared with 68% in the public clinics) and 97% had received treatment in line with the Department of Health's guidelines (80% in public clinics). Technical quality of care in the private clinics is likely to have been strengthened by the computerized treatment protocols.
Discussion: policy implications
Data on cost and quality for all clinics presented here are drawn from small samples, but nonetheless, they enable important policy issues to be identified for further review. The clinic chain's operations and service delivery were highly standardized across all clinics, so there is no reason to believe that the clinics studied were atypical. Although the two clinics in which fieldwork was carried out were chosen in consultation with the company's head office, other clinics were visited during the course of the study and the nature of their operations and layout were all extremely similar.
What are the implications of this new model of primary care for the current debate about the role of the private sector in LMICs and publicprivate partnerships, and how should the public health care system react to its arrival? The appearance of such highly organized chains within the primary care market indicate both the importance and size of this market, at least in urban areas enough to attract considerable private sector investment. It is also interesting to note how effectively these clinics use management systems to control drug and staff costs. Such findings do not support the commonly held belief that lavish use of these resources, particularly drugs, are why patients prefer the private sector. That these private clinics do not give automatic access to a doctor, or prescribe large amounts of medicine, indicates that other aspects of perceived quality, such as short waiting times and greater politeness of staff, might be important motives for private sector use.
Services offered by the clinics were mainly curative. Continuity of chronic care was weak; chronic patients had few return visits and were also seeking care in the public sector. Data on STI treatment suggested that the quality was good and in line with recommended treatment guidelines, but the lack of condoms at one facility, and the lack of availability of immunization and treatment for tuberculosis, highlighted the limited role such clinics currently play in preventing and treating key public health problems.
For patients, the implications of this new type of primary care provider can be viewed as both positive and negative. Patients were extremely satisfied with the care they received in the private clinics and there appears to be a growing taste for affordable private care among employed but uninsured workers. However, patients are paying for care that is already available at no charge in the public sector; also, they are not accessing a comprehensive service for this higher charge. They still have to resort to the public sector for out-of-hours care and certain other types of treatment, and the risk of under-treatment related to the flat fee must also be considered. Both financially and clinically, this 'dual' use of the public and private sectors is unlikely to be efficient or optimal in terms of quality of care.
For the public sector, in its current mode of operation, the arrival of this new model of private provider presents threats and opportunities. One direct threat is competition for primary health care nurses who are in short supply and who traditionally have been employed largely by the public sector. In interviews at public sector clinics, nurses often mentioned the option of leaving the increasingly over-burdened public sector for the better pay and working conditions of these new clinic chains. Other possibilities posed by this new model relate to its probable impact on the shape of the health care system. In the first place, the arrival of clinic networks and the growing taste for affordable private health care among the employed but low-income segment of the population could be an opportunity to encourage this group of people using the public sector to use this model, while the public sector concentrates on its role as regulator and providing services to the poorest. Potentially, this could remove some of the burden on the public sector, and the task of regulation might be made easier by the strong hierarchical control exercised within these clinic chains. However, the drawbacks of this approach are many. As long as these clinics are not providing a comprehensive service, it is doubtful how much of a burden they can remove from the public sector. In addition, this approach would only help the poorest if services for those who remained in a less-crowded public sector improved. However, if services did not improve, the approach could further consolidate the segmentation of private and public sectors, with the poorest services being reserved for the poorest people. The desirability of encouraging any low-income group to pay for primary care services is also debatable.
An alternative form of cooperation would be for the government to contract with these companies to provide services to the whole population. These clinic chains can deliver services that are very acceptable to users and which are at a similar cost to the public sector, and they are attracted by the market expansion that contracting with government would offer. In addition, contracts with a hierarchically structured company might be easier to manage than with a series of self- employed professionals such as general practitioners, because the internal management systems and standardized procedures of the former would lessen the variation in service delivery and give greater control. In addition, contracts could secure access for the poorest section of the population and draw the clinics into the framework of public sector service delivery, offering a similar range of services, rather than allowing them to remain largely as complements to the public sector.
Typically, arguments for contracting with the private sector fall into two complementary categories, which place emphasis on different motivations to contract. The first, drawing on theoretical arguments, espouses the virtues of competition in increasing consumer choice and the responsiveness of the public sector. According to this view, contracts with the private sector should be introduced to impose competitive pressure on all publicly financed providers. The second argument emphasizes a more pragmatic approach, stating that the private sector can be contracted as a means of bringing in additional resources and capacity to an under- resourced public sector.
The policy implications of this new model depend on which motivation to contract is dominant. For the pragmatists, the short-term benefits are evident making contracts with such providers to provide services in rural areas where public service provision is inadequate. Options could include clinics run purely under contract to government or offering a mix of private and public sector services, although this risks creating an incentive to deliver poor-quality services to public patients to boost the numbers of those willing to pay private fees. Currently, however, private clinic infrastructure is concentrated in urban centres where public sector health provision is also relatively well resourced. There is less opportunity for the public sector to make use of this type of private resource in poorly resourced areas (unless they create an incentive for private providers to move into rural areas such as subsidizing clinic construction). The desirability of paying the private sector to provide services in urban areas, where the public sector has greater infrastructure for delivery, might become for those motivated by the pragmatic argument debatable.
Alternatively, policy-makers attracted by the competition argument might want to encourage these clinic chains to become competitors with the public sector for allocating public funds. This could happen in both rural and urban settings, and could potentially have positive effects on service quality, consumer choice, and responsiveness to patients. However, many real-life constraints, of market structure, information and capacity, would need to be taken into account in designing and implementing such a system. The South African experience of contracts for district hospital provision (17) and curative primary care services (18) emphasizes the crucial prerequisite of adequate public sector capacity to manage these constraints effectively, especially when dealing with agile and creative private companies, although positive experiences of contracting out have been recorded in a variety of settings (19). Scepticism among some policy makers and managers about the desirability of working with the private sector, and of their motivations, is a further potential hurdle (20).
Irrespective of what view is taken on implications for contracting out, this model can show much to the public sector about delivering acceptable primary care services at an affordable cost. Comments from focus groups indicated that staff attitudes were a key difference between the private clinics and the public sector, impacting both on the way that patients were treated and how the facility was presented and maintained. Factors underlying the differences in morale and motivation of primary care nurses working in these private clinics versus their public sector counterparts need to be understood. The overall performance of the commercial company in this study, and the fact that the company's use of resources was comparable to the public sector, points to the key role that management can play in improving service delivery. It provides an important demonstration to the public sector that the human aspect of services might be improved without greatly increasing resource use. However, two key questions remain for further investigation. First, would the advantages of the private model remain once a contract is in place to treat poorer patients who do not pay directly for the service? Second, can the human aspect of public service delivery in South Africa be improved without major changes in the incentive structure of the public bureaucracy? Experimentation with alternative arrangements would contribute valuable evidence to answering these questions. Finally, this analysis is based on a small-scale study, and policy options need to be further developed on the basis of a larger-scale investigation.
The UK's Department for International Development (DFID) supports the London School of Hygiene and Tropical Medicine's Health Economics and Financing Programme, of which three of the authors are members. This Department also supported the research reported here part of a project "New purchaser/provider relationships in primary health care: the desirability of contracts between health authorities and private providers".
The authors thank other members of the 'Pricon' team who were involved in the various case studies drawn on for this paper. In particular, Di McIntyre and Edina Sinanovic of the Health Economics Unit in Cape Town, and Khethisa Taole, formerly of the Centre for Health Policy in Johannesburg, were instrumental in these studies. We would also thank the clinic company evaluated for their willingness to take part in this research.
Conflicts of interest: none declared.
1. Word Bank. World development report 1993 Investing in health. New York: Oxford University Press; 1993. [ Links ]
2. World Bank. World development report 1997 the state in a changing world. New York: Oxford University Press; 1997. [ Links ]
3. Giustu D, Criel B, De Bethune X. Viewpoint: Public versus private health care delivery: beyond the slogans. Health Policy and Planning 1997;12:193-8. [ Links ]
4. Zwi A, Brugha R, Smith E. Private health care in developing countries if it is to work, it must start from what users need. BMJ 2001;323:463-4. [ Links ]
5. Bennett S, McPake B, Mills A. Private health providers in developing countries: Serving the public interest? London: Zed Books; 1997. [ Links ]
6. Mills A, Brugha R, Hanson K, McPake B. What can be done about the private sector in low-income countries? Bulletin of the World Health Organization 2002;80:325-30. [ Links ]
7. Cornell J, Goudge J, McIntyre D, Mbatsha S. National Health Accounts The Private Sector Report. Technical Report, Health Economics Unit, University of Cape Town, 2001. [ Links ]
8. McIntyre D, Bloom G, Doherty J, Brijlal P. Health expenditure and finance in South Africa. Durban: Health Systems Trust; 1995. [ Links ]
9. Goldstein S, Price M. Utilisation of primary curative services in Diepkloof, Soweto. South African Medical Journal 1995;85:505-8. [ Links ]
10. Usdin L. Patterns of health care utilisation and factors affecting use of different providers in Alexandra township, South Africa [Dissertation]. New Orleans (LA): Tulane University; 1993. [ Links ]
11. Beattie A, Rispel L. A description of primary health care services delivered by general practitioners in the New Hanover District, KwaZuluNatal Province. Technical Report, Centre for Health Policy, Johannesburg, 1995. [ Links ]
12. Rispel L, Beatti A, Xaba M, Cabral J, Marawa N, Fonn S. A description and evaluation of primary health care services delivered by the Alexandra Health Centre and University Clinic. Technical Report, Centre for Health Policy, Johannesburg, 1995. [ Links ]
13. Rispel L, Cabral J, Marawa N, Xaba M, Fonn S. The integration of primary health care services at the district level using Upington as a case study. Technical Report, Centre for Health Policy, Johannesburg, 1995. [ Links ]
14. Rispel L, Beatti A, Xaba M, Fonn S, Cabral J, Marawa N. A description and evaluation of primary health care services delivered by Vaalmed Health Maintenance Organisation. Technical Report, Centre for Health Policy, Johannesburg, 1995. [ Links ]
15. Brugha R, Zwi A. Improving the quality of private sector delivery of public health services: challenges and strategies. Health Policy and Planning 1998;13: 107-20. [ Links ]
16. Schneider H, Blaauw D, Dartnall E, Coetzee DJ, Ballard RC. STD care in the South African private health sector. South African Medical Journal 2001;91:1516. [ Links ]
17. Broomberg J, Masobe P, Mills A. To purchase or to provide? The relative efficiency of contracting out versus direct public provision of hospital services in South Africa. Bennett S et al., editors. Private health providers in developing countries: serving the public interest? London: Zed Books; 1997. [ Links ]
18. Palmer N, Mills A. Classical versus relational approaches to understanding controls on a contract with independent GPs in South Africa. Health Economics (forthcoming). [ Links ]
19. Perrot J. Une partenariat base sur des relations contractualles: une option strategique pour ameliorer la performance des systemes de sante. Paper presented at the international colloquium "Financement des systèmes de santé dans les pays à faible revenus d'Afrique et d'Asie" France, ClermontFerrand, 30 November 1 December 2000. [ Links ]
20. Gilson L, Doherty J, McIntyre D, Thomas S, Brijlal V, Bowa C, et al. The dynamics of policy change: health care financing in South Africa 19941999. Monograph no. 66, Johannesburg, Centre for Health Policy, University of the Witwatersrand/Cape Town, Health Economics Unit, University of Cape Town, 1999. [ Links ]
1 Correspondence should be sent to Dr Palmer at this address