Republic of South Africa Systematic Country Diagnostic An Incomplete Transition: Overcoming the Legacy of Exclusion in South Africa Background note A Brief Profile of the Status of Health and the Health System in South Africa Paolo Belli, Thulani Matsebula, Melusi Ndhlalambi, Maria Ngarachu A Brief Profile of the Status of Health and the Health System in South Africa Paolo Belli, Thulani Matsebula, Melusi Ndhlalambi, Maria Ngarachu1 Poor and Unequal Health Outcomes South Africans suffer from a high burden of preventable illness and premature death, despite high expenditure on health. Over the last two decades some significant improvements have occurred in maternal and child health2 and in reducing the HIV burden. However, health indicators remain poor when compared to other countries at similar levels of development. Life expectancy (LE) is approximately 62 years, lower than average for lower middle income countries (LE equal to 70 years). Since 1990, South Africa has fallen in rankings across the board (see tables 1 and 2 below). For example, over the same period (1990 to 2015) that India reduced its age standardized death rate from 1078 to 784 (per 100 000 people) – a 27% reduction, South Africa’s increased its from 936 to 1097 (a 17% increase). Table 1: South Africa compared to countries with similar GNI per capital (ATLAS method; South Africa GNIPC median ($6080), other countries range from= $4930 (Jamaica) to $6470 (Belarus) Country Age-standardized death rate Age-standardized YLL* rate Life expectancy at birth Health-adjusted life (per 100,000) (per 100,000) expectancy at birth** 1990 2015 1990 2015 1990 2015 1990 2015 Rate Rank Rate Rank Rate Rank Rate Rank Rate Rank Rank Rate Rate Rank Rate Rank Belarus 1,023 9 1,018 11 24,406 6 22,805 11 70.5 8 71.0 11 61.9 8 62.5 11 Botswana 1,491 14 1,710 15 42,233 13 51,768 15 62.3 13 58.7 15 55.0 13 51.5 15 Cuba 773 2 661 2 17,396 1 12,305 2 74.7 2 78.2 2 65.8 1 68.6 2 Iran, 1,116 10 789 8 34,824 11 17,123 7 66.0 10 74.6 9 57.0 11 64.9 9 Islamic Rep. Peru 939 7 536 1 29,911 9 11,663 1 68.7 9 79.5 1 60.1 9 69.6 1 South 1,479 13 1,458 14 42,817 14 44,187 14 62.0 14 61.3 14 54.0 14 52.9 14 Africa Thailand 928 6 790 9 24,161 5 17,115 6 71.0 7 74.7 8 62.3 6 65.9 7 Serbia 917 5 725 5 20,532 3 14,351 4 72.5 3 76.2 4 64.0 3 67.1 4 Namibia 1,531 15 1,256 13 44,126 15 38,017 13 61.7 15 64.2 13 53.8 15 55.5 13 Source: Institute for Health Metrics and Evaluation. Global Burden of Disease, 2015. https://vizhub.healthdata.org/gbd-compare/ Accessed 7 June, 2017 *Years of life are lost (YLL) take into account the age at which deaths occur by giving greater weight to deaths at younger age and lower weight to deaths at older age ** Average number of years that a person can expect to live in "full health" by taking into account years lived in less than full health due to disease. 1 Authors are World Bank Staff 2 Under five mortality rates decreased from 59 to 42 per 1,000 live births over the last two decades (SADHS 1998 and 2016), while infant mortality decreased from 45 to 35 per 1,000 live births over the same period; both the percentage of births delivered by a skilled provider (84% in 1998 and 97% in 2016) and the percentage delivered in a health facility (83% in 1998 and 96% in 2016) have significantly increased. Table 2: South Africa Health Outcomes compared with BRICS countries Country Age-standardized death rate Age-standardized YLL rate Life expectancy at birth Health-adjusted life (per 100,000) (per 100,000) expectancy at birth 1990 2015 1990 2015 1990 2015 1990 2015 Rate Rank Rate Rank Rate Rank Rate Rank Rate Rank Rank Rate Rate Rank Rate Rank Brazil 1,102 2 786 2 30,080 2 18,229 2 67.9 2 74.4 2 59.4 3 64.8 2 Russia 1,090 1 957 3 26,591 1 23,147 3 69.2 1 71.0 3 60.3 1 61.9 3 India 1,671 5 1,159 4 53,588 5 31,303 4 58.2 5 67.2 4 50.0 5 58.1 4 China 1,221 3 711 1 32,393 3 14,707 1 66.7 3 76.2 1 59.6 2 68.0 1 South 1,479 4 1,458 5 42,817 4 44,187 5 62.0 4 61.3 5 54.0 4 52.9 5 Africa Source: Institute for Health Metrics and Evaluation. Global Burden of Disease, 2015. https://vizhub.healthdata.org/gbd-compare/ Accessed 7 June, 2017 South Africans spend on health over 8.5 percent3 of their GDP, approximately half of which is public and half private expenditures. Other higher middle income countries, such as Mexico, Chile or China, which spend a comparable amount on health per capita (approximately 600 USD per capita), have achieved much better health outcomes (life expectancy above 70). In terms of disease burden, South Africa is affected by the “quadruple burden of disease�, the simultaneous occurrence of (i) infectious diseases (particularly HIV-AIDS and TB), as well as (ii) non-communicable diseases. In addition, there is still (iii) a heavy burden of perinatal and maternal disorders and (iv) trauma. As Table 3 shows, first cause of death in South Africa is tuberculosis, mostly associated with the HIV/AIDS epidemic, affecting over 7.2 million South Africans. In addition, the list of top 10 leading causes of death in South Africa is increasingly dominated by non-communicable illnesses that have conventionally been classified as diseases of affluence, such as diabetes, hypertension and cardiovascular diseases, but that, in fact, increasingly affect poor and disadvantaged groups, due to changes in diet and lifestyle. Bradshaw et al. (2007) estimated that HIV accounts for 39 percent of the total number of years of life lost (YLL) in South Africa; trauma (violence and road accidents) for 10.5 percent; tuberculosis for 4.7 percent and diarrhoeal disease for 4.2 percent of YLL. Part of the high burden of disease and poor health outcomes in the South Africa population can be explained by the extremely high exposure to risk factors, including violence, unsafe sex, unhealthy diet and alcohol and tobacco consumption (with 6.3% of women and 30.4% of men smoking tobacco daily), as Figure 1 shows. 3 Average for lower middle income countries is 5.8 percent of GDP and for upper middle income countries is 6.3 percent. Table 3: Ten leading underlying natural causes of death, 2013-2015 Causes of death (based on 2013 2014 2015 ICD 10 Rank Number % Rank Number % Rank Number % Tuberculosis 1 41 904 8,8 1 39 495 8,3 1 33 063 7,2 Diabetes Mellitus 5 23 133 4,9 3 23 966 5,0 2 25 070 5,4 Cerebrovascular Diseases 4 23 158 4,9 2 24 131 5,1 3 22 879 5,0 Other forms of heart 6 22 189 4,7 4 22 928 4,8 4 22 215 4,8 disease Human 3 23 825 5,0 6 22 729 4,8 5 21 926 4,8 immunodeficiency virus disease Influenza and pneumonia 2 24 345 5,1 5 22 813 4,8 6 20 570 4,5 Hypertensive diseases 7 17 104 3,6 7 18 319 3,9 7 19 443 4,2 Other viral diseases 9 14 104 3,0 9 14 508 3,1 8 16 097 3,5 Chronic lower respiratory 10 12 384 2,6 10 12 690 2,7 9 12 667 2,8 diseases Ischaemic heart diseases … … … … … … 10 12 239 2,7 Intestinal infectious 8 16 163 3,4 8 14 795 3,1 … … … diseases Other natural causes 207 523 43,6 207 593 43,7 202 840 44,1 Non-natural causes 49 681 10,4 50 692 10,7 51 227 11,1 All causes 475 510 100 474 659 100 460 236 100 Source: Stats SA, 2017 Another, related explanatory factor of the poor health outcomes is the stark inequalities that exist dependent on socio-economic status (SES). South Africa has the starkest inequality in the world, as measured by the Gini coefficient, and socio-economic inequalities contribute to poor health outcomes, while inequalities in health in turn exacerbate socio-economic inequalities. Inequalities in health status start before birth, and are carried through to the early years. Per the South Africa Demographic Health Survey (SADHS), 2016, 27% of children under 5 are considered short for their age or stunted (below two SD median height for age), and 10% are severely stunted (below 3 SD).4 Some of these inequalities are a legacy of the apartheid era: for example, in 2005 infant mortality rates (IMR) ranged from 18/1 000 live births for white children to 74/1 000 for black children. However, health inequalities are increasingly associated more with socio-economic status, area of residence, and level of education, rather than simply race. For example, IMR is estimated at 27/1,000 in the Western Cape and 70/1 000 in the Eastern Cape (Nannan N, Dorrington R, Laubscher R, et al. Under-five Mortality Statistics in South Africa. Cape Town: South African Medical Research Council, 2012). Limpopo, Eastern Cape and North Western Province, the poorest provinces, are also those with the worst health outcomes. 4 Child stunting decreases with increasing wealth quintiles, from 36% among children in the lowest wealth quintile, to 24% among those in the middle wealth quintile, and to 13% among children in the highest wealth quintile (SADHS, 2016) Figure 1: South Africa, Top 10 causes of DALYs with key risk factors, 2015 Source: Institute for Health Metrics and Evaluation, 2017 DALYs = Disability Adjusted Life Years The sum of years of potential life lost due to premature mortality and the years of productive life lost due to disability Stunting and poor health compromise children’s ability to learn, and impair their earning potential throughout their lives. Even in adulthood, poor health outcomes are disproportionally concentrated among the poor. The poor are disproportionably affected by the HIV-AIDS and tuberculosis epidemics, by worse maternal mortality, but are also increasingly affected by non-communicable diseases. For example, according to the SADHS 2016, prevalence of high blood pressure, one of the main risk factors for cardiovascular diseases, is 75.8 percent among women with no education and 37 percent among women with a secondary and tertiary degree. Bradshaw et al. (2007) found that deaths from asthma in men were four times higher in the Eastern Cape than the Western Cape, and in a study analyzing age-standardized mortality rates in Cape Town (Groenewald al at 2008), it was found that people living in the poor sub districts of Khayelitsha had mortality rates of 856.4 deaths per 100, 000 attributable to non-communicable diseases, compared to rates between 450 to 500 per 100, 000 for people living in the wealthier suburbs. A disease which can be classified as a disease “almost exclusively� affecting the poor is tuberculosis, which, as Table 3 above showed, is currently killer number one in South Africa. Ataguba et al (2011) estimate that the bottom 40 percent of the population bears 65 percent of the TB disease burden compared with 17 percent for the top 40 percent. Close living conditions, particularly in informal settlements, and long waiting times at primary care clinics contribute to high risk of exposure to TB. A number of studies conducted in South Africa reported that high levels of community income inequality were independently associated with increased prevalence of tuberculosis5 (Wood, et al 2010; Harding et al, 2008; Bhorat H and Westthuizen C. 2012; Pellicer et al, 2011). Furthermore, TB disproportionably affects specific populations, such as prisoners and miners. For example, prevalence of TB among miners and ex-miners is estimated at 2500- 3000 per 100,000 people (compared to 834/100,000 incidence in the general population), more than ten times the WHO threshold of a health emergency, which is 250 per 100,000 people. 5 What is the relationship between TB and poverty? Stop TB Partnership. http://www.stoptb.org/assets/documents/events/world_tb_day/2002/1Therelationship.pdf. Poor health outcomes are also explained by health services’ inability to effectively prevent and treat diseases. International experience shows that effective health services can decisively contribute to improvements in health. Recent research has put the impact of medical care on overall health outcomes at between 44 and 57 percent (Arah et al 2006). The London School of Hygiene and Tropical Medicine studies on "health-care-amenable mortality in Europe" estimate the influence of medical care on recent health indicators improvement to be approximately half (Nolte and McKee 2008, Figueras et al 2008). A Bimodal Health System, with Stark Inequalities in Access to Quality Services A bimodal health system exists in South Africa, which serves the poor and the rich in separate health facilities, with significant differences in terms of quality and access.6 Disparities and inequality are entrenched in access to health services, reflecting the country’s colonial and apartheid past. A key cause of these disparities in access to services is the fragmentation of health financing, with poor and rich utilizing separate revenue collection and pooling mechanisms. Such fragmentation is a common feature in low- and middle-income countries (Mills et al 2012). The public sector: Approximately 48.2 percent of the country’s Total Health Expenditure (THE), and 14.2 percent of total government expenditure is tax based, and mainly finances government owned health facilities and public health programs, catering for the healthcare needs of 46.2 million people, equivalent to 84 percent of the total population. In terms of the patient payment policy, patients who are South African citizens are eligible to receive hospital care free at point of service, subject to them satisfying means test criteria contained in the Uniform Patient Fee Schedule (UPFS). Primary health care is free for all at the point of care regardless of income. The organizational structure of the South Africa public or government health sector is described in Figure 2 below. It is organized as a hierarchy, with the National Department of Health (NDoH) at the top, followed by the provincial level (nine provinces), and then the district level. In fact, the system has been characterized as a quasi-federal system of government (Mills et al 2012), whereby the provinces have their own departments of health, mirroring the institutional structure at national level. Provincial departments are headed by their own “Ministers of Health� referred to as Members of the Executive Council (MECs). They also have heads of department who are the equivalent of the Director-General at the NDoH. The roles of each level of government were spelled out in the 2004 National Health Act: the NDOH is responsible for providing strategic direction for the health system, and for running several national programs, including prevention and control of communicable diseases. Some of the priority programs (e.g. HIV/AIDS) are financed through Conditional Grants from the National Treasury, which are either channeled to the Provinces or directly utilized by the NDoH. Formulation of policy and ensuring that the same is translated into legislation is also the responsibility of the NDOH. Provincial departments of health are charged with delivering most health care services within the nine provinces, and they utilize part of the Equitable Share they receive from National Treasury to finance their programs and health facilities. Health facilities include mobile clinics, clinics, community health centres, district hospitals, tertiary hospitals, specialized hospitals and central hospitals. Hospital services and primary health services routinely consume the largest respective shares of provincial health budgets. Local governments below the Provinces are responsible for environmental health services, coordination of health services provided by community health workers and NGOs and non-facility based health promotion services (Gilson and McIntyre, 2007). Overall, the South African public health system is often described as “hospital -centric� (McIntyre et al., 2007), due to the overall predominance of hospitals in the service chain, and by contrast its failure to provide 6 See, for example, Ataguba et al, 2014, or McIntyre et al, 2007. effective preventive and curative care at the primary care level, as well as limited scope of services and operating hours for primary care units. Patients generally bypass primary care and go directly to the nearest hospital as their first point of contact with the health system each time they get sick7, and primary level health services have continued to be underdeveloped and systematically underfunded (Coovadia et al 2009). These features have persisted, in spite of several attempts over the last 25 years to strengthen the primary care level. When the democratic government came into power, it had the responsibility to unite disparate health systems that existed in the then Bantustans. In 1994, the African National Congress (ANC) published its Health Plan which presented a model for a post-apartheid health system. The new trajectory placed primary health care at the core of the health system, to be implemented through a district-centered health system which was never implemented. Acute care hospitals spent over 75 percent of total recurrent public health expenditure in 1992/93; academic and other tertiary hospitals alone accounted for over 40 percent of total recurrent public health expenditure, while non-hospital primary care services accounted for only 11 percent. These proportions have not changed significantly over the last 25 years. The latest initiative to strengthen primary care has been launched since 2011, in the context of the first phase of the National Health Insurance initiative. The NDoH designed a new self-assessment tool to be used by primary care units to assess the quality of their infrastructure, management and service delivery capacity, and has invested in a selected number of PHCs to become “ideal clinics�. This policy initiative has not been rigorously assessed. The private sector: The private sector covers 8.8 million individuals, which is equivalent to approximately 16 percent of the country’s population (CMS 2016); these people belong to private medical aid schemes (84 of them existed in 2016), and the majority of private health expenditure (equivalent to 51.8 percent of total health expenditure) is collected in the form of private medical premiums. Per capita health expenditure among those who have medical aid scheme (in US terminology, private insurance) coverage is more than five times larger than among those who rely exclusively on the public sector. The unequal distribution of finances translates in a very unequal distribution of medical assets and personnel. Approximately 6 out of ten general doctors, two thirds of specialists, and 9 out of ten dentists and pharmacists practice in the private sector, and cater only for the privately insured population or the very small minority of uninsured patients who pay on an out-of-pocket basis. Medical schemes are not-for-profit mutual societies that are, in theory, owned by all members who make contributions towards such schemes. There are two types of medical schemes – open and restricted or closed. The term “open medical schemes� refers to those schemes in which any individual can enroll and make the necessary contributions. Restricted medical schemes, on the other hand, are those schemes in which individuals must meet certain qualifying criteria to join. Such criteria may include, for instance, employment in a particular industry or company, or membership of a particular professional association (Hodge et al 2012). 7 South Africa has considerable hospital capacity, albeit heavily concentrated in urban areas and at the higher levels of care. Figure 2: Structure of the Health System in South Africa National National DoH Treasury National priority programs Provincial Provincial DoH Treasury Provincial/Regional/Tertiary hospitals District Health Authority District hospitals Community Health Centres Primary Healthcare Primary Healthcare Primary Healthcare Clinics Clinics Clinics Medical schemes are required by legislation to cover prescribed minimum benefits, which include hospital- based interventions and long-term care for certain chronic diseases. The cover provided by medical schemes is characterized as substitutive, not complementary to public sector services (Ataguba and McIntrye 2012), in the sense that those with medical scheme coverage predominantly utilize private sector providers, perceived to be of higher quality and with shorter waiting lists than public providers. However, medical scheme members are also entitled to receive care in the public sector and for the public sector to be reimbursed by medical schemes for the provision of such services. In practice, though, majority of medical scheme members only access public sector facilities if they exhaust benefit entitlements in their cover. Earlier analytical work by the World Bank also pointed to the almost insurmountable backlog by public sector hospitals in processing the administrative paperwork required to claim fees for services rendered to persons covered by medical schemes in their hospitals and how easy it is for someone to fall through the cracks with the hospital delivering such services for free and never getting reimbursed by the medical scheme for it. Affordability is the main barrier to medical scheme membership for poor households. As in other countries, private medical aid premiums are based on expected health costs rather than income, which tends to be positively correlated with age, but loosely correlated with income. As a result, lower income earners need to spend a significantly larger proportion of their income on premiums in order to become members of a private medical aid scheme, and most poor households cannot afford it. Premium payments account for over 6 percent of all household incomes, as compared to general tax which account for about 5 percent (Ataguba et al., 2017). Because of higher coverage of higher income groups, expenditure on medical aid schemes overall is progressive (higher income quintile households spending over 8 percent of their total incomes on medical scheme premiums, as opposed to a negligible expenditure by the two lowest income groups). It is estimated that the sub-population which voluntarily enrolls in one of the private medical schemes has an income roughly 2.7 times the average population’s income (Discovery Health, 2016), equivalent to a GDP per capita of approximately US$35,000 (PPP). However, as shown in Table 4 (Finmark Trust, 2009), among those with private health insurance coverage, spending on private medical aid is regressive. Households earning between ZAR1,500 (which was just above the poverty line in 2009) and ZAR3,500 per month (only 3.9% of the households in this income bracket were insured in 2009) spent the highest proportion (13.1%) of income to contribute to a medical aid, if they were privately insured). Table 4: Proportion of households who are members of medical aid schemes and proportion of members’ income spent on health insurance Monthly HH Income Av. % HH Income Spent on Medical HH Penetration of Medical Aid Aid for those who are MA members R850 – R1500 8.6% 1.2% R1500 – R3500 13.1% 3.9% R3500 – R7500 8.1% 22.1% R7500 – R10000 6.5% 47.9% R10000 + 3.9% 71.6% Source: Finmark Trust (2009) Tax subsidies: Note that those who can afford private coverage continue also contributing –through taxation- to the publicly financed and publicly provided health system. So, in a sense they elect to pay twice for health services, even though a certain amount of their medical scheme contributions is publicly- subsidized through a system of tax rebates, which -it can be argued- translates to a regressive state-subsidy of privately financed healthcare. This subsidy in 2017 was equal to approximately Rand 18 billion, or 11 percent of total private health expenditure. Other sources of funding for health care are small in comparison and come in the form of out-of-pocket spending (12 percent of the total health expenditure), and some financing specifically geared towards trauma-related services, such as the Compensation Fund, the Road Accident Fund, Rand Mutual Assurance and Mine Hospitals (van den Heever 2011). International assistance is significant, particularly for HIV AIDS8, but overall the country depends very little on development partners to fund its health activities. Latest available data indicates that the amount of development assistance for health is 1.8% of the total (Health Policy Project 2016). 8 South Africa receives significant support from the Global Fund to Fight AIDS, Tuberculosis and Malaria (“The Global Fund�), a nd The United States President’s Emergency Plan for AIDS Relief (“PEPFAR�) to fight HIV/AIDS and TB. Market and Government Failures in the South Africa Health System Consumption of healthcare is characterized by market failures stemming from asymmetries of information between patients and health providers, which lead to suboptimal market equilibria (Mooney & Ryan 1993). The imperfections extend to the market for health insurance where behaviors such as adverse and risk selection limit the number of people who purchase health insurance –for any given level of the premiums- to those who have higher self-perceived risk of using health services. Adverse selection is somewhat mitigated since those with medical scheme coverage mainly get it through their employment and not on an individual basis, but can be seen, for example, in the age structure of those privately insured, which tends to underrepresent youth. In turn, private health insurers counteract potential adverse selection by structuring their marketing and product design strategies to attract good risks, those who are healthy and less likely to utilize health services. Another shortcoming of medical scheme membership being for the most part an employment benefit is that for most people, when they leave employment or retire and can no longer afford to make contributions, cover is lost. It is thus possible for people to make substantial private contributions while employed, but then be left without access to private health care in their older years. In fact, South Africa has, at several points, explored the potential of incorporating medical schemes into national arrangements (Mills et al 2012), to limit the impact of risk selection. The Council for Medical Schemes, which is the regulatory body tasked with regulating the medical schemes industry, had prepared for the introduction of a system of risk-adjustment that was meant to address the challenges of differences in risk profiles among medical schemes in 2008 (Steenkamp, 2016). However, implementation of the risk- adjustment system was abandoned and in 2009 the government announced a new National Health Insurance (NHI) initiative. Another common source of “market failures� in health (which in fact affects both private and public sector when payment to health providers is linked to number and cost of services provided) is moral hazard. Because of better information and market power, health providers tend to over-diagnose, over-treat and over-prescribe (supply-induced demand) vis-à-vis optimal practice, whenever their income is dependent on the service rendered (retroactive payments), just as much as car mechanics tend to find fault and over-repair and substitute parts in our cars, whenever we leave them for periodical revision. This phenomenon is much stronger in health, however, because for those with insurance a third-party payer covers part of the treatment costs. In South Africa, fee-for-service based payment systems are prevalent in the private health sector, resulting in moral hazard and cost escalation. Medical schemes are accused of not being able to control costs or steer the private sector towards greater efficiency. Rather than changing the provider payment system towards prospective payments, such as capitation-based for primary care or Diagnostic-Group- based for hospital care, medical aid schemes’ reaction to control moral hazard has been to increase their members’ co-payments and deductibles, which makes patients more cost-aware but also reduces their financial protection. In the draft NHI “green� policy paper released in 2011, the National Department of Health stated the following: “Similarly to the public health system, the private sector has its own problems albeit these are of a different nature and mainly relate to the cost of services. This relates to the pricing and utilization of services. The high costs are linked to high service tariffs, provider-induced utilization of services and the continued over-servicing of patients on a fee-for-service basis (pg. 7).� The nature and extent of moral hazard in the private sector is contested, but it is a fact that expenditure by medical schemes has increased faster than the rate of inflation over the last two decades, and that private hospitals and specialists have been the main drivers of medical scheme total costs. The Council for Medical Schemes has singled-out hospital cost increases as the most important contributor to overall cost increases (CMS 2008). Even though the rate of increases in medical scheme contributions have significantly slowed down in recent years, the issue of affordability as a threat to the sustainability of the private healthcare sector continues to be raised by the National Department of Health, 2011, 2015, 2017). Table 5: Distribution of Medical Scheme Expenditure Among Different Type of Providers over time Provider Type 2005 2010 2015 Private Hospitals 35.3% 36.4% 37.7% Specialists 20.5% 22.4% 24.5% Medicines 15.8% 16.6% 15.25 General Practitioners 8.0% 6.4% 5.35 Dental 4.6% 2.9% 2.84% Other services 15.9% 15.3% 14.36 Source: Council for Medical Schemes, Annual Reports This cost escalation has been attributed not only to the moral hazard phenomenon just discussed, but also to excessive concentration in the market for hospital services. A working paper published by the OECD in 2016 reached the conclusion that private hospitals in South Africa are more expensive than in other OECD countries, in relation to the country’s income. The study blamed the high concentration and the barriers to access in the hospital industry as key contributors to escalation of hospital care prices. However, the OECD results have been contested by the industry, which claims that the OECD article reached flawed conclusions on the basis of a non-representative sample, and that in fact most of the cost escalation of privately financed health services in South Africa is due to the evolution of demand side factors (adverse selection and progressive aging of the insurance pools), and to the depreciation of the Rand, because most medicines and medical equipment used in SA are imported. Importantly, the growth of private hospitals is a relatively recent phenomenon that began in the 1980s. The growth in private hospital beds has been accompanied by growing concentration in the hospital market. Van den Heever (2012) concluded that private hospitals essentially do not face any meaningful competition in South Africa. In the absence of competition, private hospitals exist in an oligopolistic environment whereby 80 per cent of ownership is concentrated in the hands of three hospital groups, Mediclinic, Netcare and Life Healthcare. Recently, the SA Competition Commission established a Health Market Inquiry (HMI) to further investigate how the hospital market concentration may contribute to cost escalation and produce a negative impact on access and affordability of healthcare.9 Failures in the private sector coexist with failures in the public sector, the so-called government failures in the public finance literature. Several independent assessments, by scholars, think tanks, as well as assessments by the Office of Health Standards Compliance, indicate severe challenges in the public health sector in terms of fragmentation, low responsiveness, lack of performance and wasteful allocation and use of resources.10 Note that resource allocation in the public sector is input-based and rearranged every year based on historical expenditures. There is no strategic purchasing of services. All public facilities are budget units, with extremely limited autonomy or accountability for results. The Role of National Health Insurance in Addressing Sectoral Challenges Several plans to reform the health sector in South Africa have been designed but none has so far been implemented: One of the key focus areas of policies enacted over the past 23 years has been to redress the disparities in health outcomes and access to quality health services described above. The ANC Health Plan released in 1994 emphasized development of a publicly funded, publicly-provided and decentralized health system. The idea of creating a universal National Health Insurance first emerged in this document. The 9 Non-price competition involves hospitals engaging in the so-called “medical arms race� by investing in high-end medical equipment as a means to attract patients to their facilities and then engaging in strategies to induce utilization. 10 McIntyre et 2007, Van den Heever 2012 Taylor Committee, which published its report in 2002 and the Ministerial Task Team on Social Health Insurance created thereafter also proposed a radical reshaping of both the public sector and the medical schemes environment. The National Development Plan (NDP) acknowledged the role of universal health access towards inclusive growth and that access to healthcare services is one key determinants of health, alongside social determinants such as poverty and poor sanitation, lack of education, and others. Unfortunately, few of these past plans and policy pronouncements have been translated in actual implementation on the ground. Significant progress has been achieved on several fronts (for example, in the control of the HIV AIDS epidemic), but the health system is still as bimodal and polarized as it was two decades ago, as described earlier. Quality of services in the public sector has been lagging behind. The fundamental resource misallocations, and lack of incentives to produce god performance have not been addressed. In 2011 the ANC Government launched a new initiative with publication of a Green Paper on the development of National Health Insurance (NHI), which was followed four years later by the publication of a NHI White Paper in December 2015, confirmed in June 2017. The vision outlined in the Green and White Papers would see the progressive implementation of an ambitious reform plan over a period of 14 years. The first 5 years would be focused on quality improvements in the public health system and strengthening primary health care, especially in historically underserved areas, while the second and third phases would introduce deeper reforms both in financing as well service delivery and in both the public as well as the private sectors. The aim is to address some of the market and government failures described above, by: - Strengthening and modernizing provision of services in the public sector, through greater investments in primary care, health management information systems, outreach programs with greater use of community health workers, etc. - Creating a new National NHI Fund, meant to purchase services for the overall population based on a uniform set of medical benefits from semi-independent public providers (hospitals and primary care facilities) as well as from the private sector. - Progressively realigning the role of the private sector in health financing and provision, by making private aid scheme benefits complementary and not substitutive of the universal benefit package funded through the NHI Fund, and by modifying the business model for private providers, towards greater contribution to delivering the NHI universal benefits. Key dimensions of the reform, for example the role of the Provincial Governments and the NHI Fund, are yet to be specified. At the heart of the intention to introduce NHI is a need to address the stark inequalities that exist in both the financing and provision of healthcare in South Africa, which we have outlined above. The vision is to enable all South Africans—regardless of socioeconomic status—to have access to affordable, quality healthcare services, without suffering financial costs. However, this vision can only be translated into reality in stages and through progressive evolution. Rather than trying to go heads-on against the “distortions� of the past which have created a polarized system, may be the country could progressively transform the two systems, both public and private, as follows: Establish a new contractual relationship between purchasers (NHI Fund or Provincial Health authorities) and providers, characterized by: a) a stronger purchasing function, buying services based on costs and quality from both public and private sectors; b) greater autonomy at the facility level in the public sector, empowering for example hospitals managers to introduce innovations and service delivery improvements, with new governance (hospital Boards, whose role and prerogatives need to be defined); c) output and performance based payment system for health service providers. Allow the private sector to progressively take up a greater role in service provision, for example by seeking discounts in exchange for higher volumes, with new forms of provider payment agreements between the public and private sectors, and with explicit incentives to take up a large share of the poor. Improve accountability for results at all levels and in both the public and private sectors. This could be done also by more systematically collecting and publishing data on costs and quality of care in all facilities. References Nannan N, Dorrington R, Laubscher R, et al. Under-five Mortality Statistics in South Africa. Cape Town: South African Medical Research Council, 2012 Magennis R and van Zyl J. Making Health Insurance Work for the Low Income Market in South Africa: Cost Drivers and Strategies. Johannesburg. Finmark Trust, 2009. Groenewald P, Bradshow D, Daniels J, Matzopoulos R, Bourne D, Blease D, Zinyaktira N, Naledi NT. Cause of Death and Premature Mortality in Cape Town, 2001 – 2006. Cape Town. South African Medical Research Council, 2008. Coovadia H, Jewkes R, Barron P, Sanders D, McIntyre D. The health and health system of South Africa: Historical roots and current public health challenges. Lancet 2009. Published online August 25, 2009. DOI:10.1016/S0140-6736(09)60951-X Mayosi BM, Flisher AJ, Lalloo UG, Sitas F, Tollman TM, Bradshaw D. The burden of non-communicable diseases in South Africa. Volume 374, No. 9693, p934–947, 12 September 2009. DOI: http://dx.doi.org/10.1016/S0140-6736(09)61087-4 Patel V, Lund C, Hatherill S, Plagerson S, Corrigall J, Funk M, and Flisher A. J. (2010). Mental disorders: equity and social determinants. Equity, social determinants and public health programmes, 115. Schneider M, Bradshaw D, Steyn K, Norman R. and Laubscher R. (2009). Poverty and non-communicable diseases in South Africa. Scandinavian journal of public health, 37(2), 176-186. Bradshaw D., "Determinants of health and their trends: Primary health care: in context." South African health review 2008.1 (2008): 51-69. Bradshaw D, Norman R, and Schneider M. "A clarion call for action based on refined DALY estimates for South Africa." South African Medical Journal 97.6 (2007): 438-440. Harling G, Ehrlich R, Myer L (2008). The social epidemiology of tuberculosis in South Africa: a multilevel analysis. Soc Sci Med 66: 492–505. What is the relationship between TB and poverty? Stop TB Partnership. http://www.stoptb.org/assets/documents/events/world_tb_day/2002/1Therelationship.pdf. Accessed on 16 May 2017 Coovadia H, Jewkes R, Barron P, Sanders D, and McIntyre D. (2009). The health and health system of South Africa: historical roots of current public health challenges. The Lancet, 374(9692), 817-834. Stuckler D, Basu S, McKee M. "Health care capacity and allocations among South Africa's provinces: infrastructure–inequality traps after the end of apartheid." American journal of public health 101.1 (2011): 165-172. Wood, R., Maartens, G. and Lombard, C.J. (2000). Risk factors for developing tuberculosis in HIV-1- infected adults from communities with a low or very high incidence of tuberculosis. JAIDS- HAGERSTOWN MD, 23(1), pp.75-80. Bachmann, M. O, & Booysen, F. L. (2003). Health and economic impact of HIV/AIDS on South African households: a cohort study. BMC Public Health, 3(1), 14. Bhorat H, Westthuizen C. (2012). Poverty, Inequality and the Nature of economic growth in South Africa. Working Paper 12/151 ISBN 978-1-920055-92-9 November 2012 © University of Cape Town 2012 Pellicer, Miquel, Vimal Ranchhod, Mare Sarr, and Eva Wegner. (2011). "Inequality Traps in South Africa: An overview and research agenda." Bradshaw D., Nannam N., Laubscher, R., Gronewald, P., Joubert, J., Nojilana, B., Norman, R. Pieterse, D., Schneider, M, South Africa National Burden of Disease Study 2000: estimates of provincial mortality: summary report. Parrow, Cape Town: South African Medical Research Council, 2006. Ataguba, J.E., J. Akazili, and D. McIntyre, 2011, “Socieconomic-related health inequality in South Africa: evidence from General Household Surveys, International Journal for Equity in Health, 10: 48. McIntyre, D., Thiede M., Nkosi M., Mutyambizi V., Marianela Castillo-Riquelme, M. Gilson L., Erasmus E., Goudge J. A Critical Analysis of the Current South African Health System. SHIELD Work Package 1 Report, Cape Town, 2007. Van den Heever, A. Review of Competition in the South African Health System. Produced for the Competition Commission, Pretoria, 2012.