3. Financial sustainability of health spending and efficiency

Brazil has high overall spending on health but the public share is low

Health care in Brazil is financed by a wide range of actors with the public Sistema Único de Saúde (SUS), private health insurance schemes and direct payments by private household being by far the most important payers. In 2019, Brazil allocated 9.6% of its GDP to health care (corresponding to USD 1 514 per capita when adjusted for differences in purchasing power). This share has been growing since 2000, up from 8.3%. The total share is relatively high given Brazil’s state of economic development. It is above the values seen in Chile (9.3%), Colombia (7.7%) and Mexico (5.4%), and higher than in China and Russia (Figure 3.1). It stands also above the OECD average (8.8%). That being said, the United States (16.8%), and some Western European countries such as Switzerland, Germany and France (11-12%) still allocate a substantially greater part of their GDP to health care than Brazil.

While Brazil spends more on health care than many peer countries overall, the country relies heavily on financing from the private sector, either via voluntary private health insurance or direct payments by households. In 2019, only 41% of all health spending was financed publicly1 (equivalent of 3.9% of GDP) – mainly via SUS, 30% by private health insurance (2.9% of GDP) and 25% referred to out-of-pocket payments (2.4% of GDP).2 While the spending share of private health insurance has increased significantly since 2000 (up from 1.7% of GDP), the share of public spending has grown only moderately (up from 3.5% in 2000). The proportion of household out-of-pocket spending in GDP (2.4% in 2019) is smaller than in 2000 (3.0%) but above the value seen in 2015 (2.2%), suggesting an increasing need by patients to finance health care good and services themselves in the most recent years.

In 2019, public spending on health as share of GDP in Brazil was on a similar level than in South Africa (3.6%), and higher than in Russia, China and Mexico (around 3%). However, it was below the proportions recorded in Colombia and Chile (5.7% and 6%, respectively). In most OECD countries, public health spending accounts for a significantly higher share of their GDP than in Brazil (Figure 3.2). In 2019, the average stood around 6.6% and was above 9% in France, Sweden, Germany and Japan. The share of voluntary private health insurance in GDP on the other hand is exceptionally high in Brazil, only topped by South Africa. Australia, Canada and Slovenia are OECD countries with important voluntary private health insurance markets where they account for shares between 1-1.5% of GDP.3 With a proportion of 25% out-of-pocket payments in total health spending,4 Brazil relies less on households’ ability to pay directly for health care goods and services than some other medium-income countries such as Chile (33%) or Mexico (42%), but much more than Colombia (15%). Financial protection is also much more comprehensive in the OECD as a whole – only 20% of all health spending originate directly from households.

Figure 3.1. Health spending in Brazil and selected countries, 2000-19

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; WHO Global Heallth Expenditure Database (2020[2]), https://apps.who.int/nha/database; Brazilian Ministry of Health, 2021.

Figure 3.2. Composition of health spending by key financing schemes, 2019 or latest year, Brazil and selected countries

Note: Compulsory private health insurance as is exists in Chile and Colombia are included in the category government/compulsory schemes. VHI stands for voluntary (private) health insurance. OOP stands for direct out-of-pocket payments. The category “Other or unknown” includes financing from NGO, employers, rest of the world and also spending that cannot be allocated to a financing scheme.

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; WHO Global Heallth Expenditure Database (2020[2]), https://apps.who.int/nha/database; Brazilian Ministry of Health.

Brazil spends less on inpatient care but more on outpatient care and administration than OECD countries

Analysing the composition of health spending from the perspective of the type of service reveals some particularities of the Brazilian health system. Such an analysis shows that Brazil allocates a lower proportion of health spending on inpatient care and long-term care and a higher proportion on outpatient care than on average across the OECD (Figure 3.3). The share of spending on administration (6% vs. 3%) is also higher than the OECD average. However, the fact that around 4% of health spending cannot be allocated to any service in Brazil affects the international comparability of these figures to a certain extent.

Figure 3.3. Composition of all health spending by type of service, Brazil and selected OECD countries, 2019

Note: Inpatient care refers to curative-rehabilitative inpatient care and day care. Outpatient care refers to curative-rehabilitative outpatient care and home-based care, and ancillary services. Medical goods include pharmaceuticals. Administration refers to health system and financing administration and governance.

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; Brazilian Ministry of Health (2020).

Public coverage of pharmaceutical costs in Brazil is much below the OECD average

Combining information of payers of health care with the type of services can give an insight into the generosity of the publicly financed benefit basket of a country but also about public spending priorities. As to be expected, as a general rule, health services in Brazil are less comprehensively covered by public schemes than in most OECD countries (Figure 3.4). In Brazil, only around 47% of all inpatient costs are covered by public schemes compared to 58% for outpatient care. Across the OECD, public inpatient coverage is usually more comprehensive than for outpatient care (87% vs 78% on average). Brazil has a relatively high share of dental care costs covered (33%) – exceeding the OECD average (31%). On the other hand, only 9% of all retail pharmaceutical spending is financed by public schemes in Brazil. This share is far below any other OECD country (average 58%) and suggests a substantial gap in effective coverage.

Figure 3.4. Extent of coverage in Brazil and selected OECD countries, 2019

Government and compulsory insurance spending as proportion of total health spending by type of care

Note: Outpatient care includes general outpatient and specialist curative care, other outpatient curative care but excludes dental care.

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; Brazilian Ministry of Health (2021).

Public scheme – Sistema Único de Saúde

The move towards universal health care coverage in Brazil began with the adoption of the Federal Constitution in 1988 stating that health care is a fundamental right and duty of the state and the subsequent establishment of the Sistema Único de Saúde (SUS) – the publicly funded health system. The key principles of SUS as laid out in Articles 196 to 198 of the constitution are universality, comprehensiveness, equity, decentralisation and social participation. Based on these principles, everyone in Brazil is entitled to comprehensive health services provided under SUS regardless of socio-economic circumstances or the ability to pay. The federation, the 26 states (plus the federal district) and the 5 570 municipalities have shared responsibilities for financing and management of SUS and the delivery of public services. The participation of the civil society in determining the strategic direction of SUS is enshrined in the constitution and is operationalised via regular health conferences and health councils on the federal, state and municipality level (see Chapter 2).

Services financed by SUS are delivered by a wide mix of private and public providers. In primary care, services are predominantly provided by publicly employed staff working in multi-disciplinary primary care teams. Outpatient specialist care is available in public clinics or hospital outpatient departments. Inpatient services are provided by public hospitals but private hospitals are also contracted by municipalities or states to provide services under SUS. In total, there are 345 000 health facilities registered in Brazil serving public and private patients.

The scope of health services offered under SUS is comprehensive without cost sharing for patients but there may be out-of-pocket payments for some prescribed pharmaceuticals obtained via the “farmacia popular” programme and for any other goods and services not included in the benefit package or obtained from providers outside the SUS network. While the entire population is entitled to receive services under the SUS, a non-negligible part of the population struggles with unmet health needs. Based on result of the Health Interview Survey 2019, nearly 14% of those who sought health care two weeks prior to the survey did not receive it (IBGE, 2020[3]). A particular problem is related to accessing specialist services resulting in long waiting times and unmet need (Castro et al., 2019[4]). A higher proportion (around 25%) forgo their right to treatment under SUS altogether and instead use private health insurance to obtain services or pay themselves. They mainly expect better access and higher quality from the private sector. However, for very costly and complex treatment such as organ transplant, HIV/AIDS treatment or access to high cost medication (not covered by private insurance) these patients frequently return to SUS to obtain these services.

The establishment and the roll-out of SUS is generally considered a success story in extending health care coverage to disadvantaged population groups that previously did not have access to health care services and SUS therefore contributes to reducing inequalities in access to care and improving health outcomes. Part of this was achieved by putting a focus on reorganisation and strengthening of primary care (Couttolenc and Dmytraczenko, 2013[5]). Yet, problems with care co-ordination, low quality of services and inefficiencies are frequently mentioned in the context of SUS (Couttolenc and Dmytraczenko, 2013[5]).

Private scheme – private health insurance

In 2020, around 22% of Brazilians (47 million) were covered by voluntary medical insurance to mainly duplicate health care coverage5 under the SUS, down from 25% in 2014 (Agência Nacional de Saúde Suplementar, 2020[6]). More than 700 entities and insurers offer this type of coverage (Agência Nacional de Saúde Suplementar, 2020[6]). There is a wide variety in the scope of services covered by the individual plans and in the legal modalities of the entities covering these services. Most cover ambulatory and hospital care but not pharmaceuticals or highly specialised treatment. In addition, more than 26 million Brazilians are covered by exclusively dental insurance plans6 which duplicate and supplement SUS coverage, a figure that nearly doubled in the last 10 years (Agência Nacional de Saúde Suplementar, 2020[6]).

Most health plans are employer-based group policies as a component of employment contracts where contributions are paid by employers but other individual or collective policies are also available. An important feature of the Brazilian private insurance system is that contributions are deductible from personal income subject to taxation,7 meaning that people with high income have a higher net tax gain. This is one of the reasons why the uptake of voluntary health insurance is much more prominent among richer population groups (Montaya Diaz and Sarti, 2020[7]). People with private coverage usually access private health providers including for profit and not-for-profit hospitals.

Health spending in Brazil is projected to increase substantially until 2040

Without taking into account any structural breaks due to COVID-19, the OECD health spending projection model suggests that – in the base-line scenario – health spending in Brazil will increase to 12.6% of GDP by 2040 (Figure 3.5). Compared to 2017 this is an increase of more than 3 percentage points – more pronounced than in most OECD countries. On average across the OECD, 10.8% of economic wealth will be allocated to health in 2040, 2 percentage points more compared to 2017. The projected increase in Brazil is also stronger than in Chile or Colombia where health spending is expected to reach 11.4% and 9.5% of GDP in 2040, respectively.

Rising health spending can also be expected under more favourable scenarios in Brazil. In case some ‘cost control’ policies are implemented, the health-to-GDP ratio is projected to rise to 12.4% of GDP; in case of more ‘healthy ageing’, the increase can be limited to 11.7%. If additional ‘cost pressures’ will occur, health spending is expected to grow further, reaching 13.3% by 2040. Significant increases in health spending, albeit below the OECD estimates, are also projected in other research work based on alternative projection models. Rocha et al., for example, project health spending to reach 12.2% of GDP by 2045 in their baseline scenario (Rocha, Furtado and Spinola, 2019[9]).

Figure 3.5. Health spending as share of GDP, Brazil and OECD countries, 2017-40

Note: Projections do not take into account any structural breaks due to COVID-19.

Source: OECD estimation based on Lorenzoni et al. (2019[10]), “Health Spending Projections to 2030: New results based on a revised OECD methodology”, https://dx.doi.org/10.1787/5667f23d-en.

In summary, the OECD health spending projection model has two key messages for Brazil:

• the health spending to GDP ratio in Brazil will increase further, even under the most favourable scenario;8 and

• the expected increase of this ratio will be higher in Brazil than in most OECD countries.

In light of these findings, there are basically four possible options (that could also be used in combination) to address the increasing health financing needs in Brazil, and they will be discussed in detail in the following subsections:

• Increasing public health spending without compensation from lower spending elsewhere, thus raising total government spending (Section 3.3.1).

• Prioritise health spending within the existing overall government spending envelope (Section 3.3.2).

• Rely more on private spending (Section 3.4).

• Find efficiency gains in the health sector (Section 3.5).

The introduction of SUS has been accompanied by several initiatives to strengthen primary health care

Since its inception, a strong focus of SUS was to strengthen the role of primary health care and move away from a health system that was historically very hospital-centred. In 1994, the Family Health programme was established which set standards on the composition of teams of health care professionals including general physicians, nurses, nurse assistants and community health workers providing a wide range of services, such as acute care, health promotion, chronic disease management and maternal and child health care (Castro et al., 2019[4]). These teams became the key element of primary health care in SUS but problems with staffing slowed down the expansion of this care model. To address problems in access, especially in rural areas, the “Mais Médicos” programme was adopted in 2013, which led to the temporary recruitment of around 18 000 additional doctors, mainly from abroad. While being successful in increasing access to primary health care in some disadvantaged areas (Gonçalves et al., 2019[28]), the programme is currently phased out and is replaced by the “Médicos pelo Brasil” programme. This initiative seeks to centralise recruitment for doctors for disadvantaged communities and to strengthen the professional development opportunities of these newly hired doctors.

A strategy to improve efficiency in primary health care and expand access was the ‘Previne Brasil’ initiative introduced in 2020. This initiative marked an important shift in the calculation method how federal funds are transferred to the municipalities to purchase primary care services but also increased overall federal resources for primary health care. Under SUS, the federal government has financed roughly one-third of total primary health care costs, with the remaining part being funded predominantly by municipalities (61%) (Ministério da Saúde/Fundação Oswaldo Cruz, 2018[29]).13 Prior to the ‘Previne Brasil’ strategy, the federal transfers to municipalities were composed of a fixed and variable capitation component, a voluntary pay-for-performance bonus based on the National Programme for Improving Primary Care Access and Quality (PMAQ) and payments for the direct provision of services through community health workers and “Mais Médicos” programme. The ‘Previne Brasil’ strategy recalibrated the capitation payments to better take account of differences in health care needs across municipalities, streamlined the pay-for-performance indicators and revisited the strategic actions to be financed by the federation (OECD, 2021[30]). An important change in ‘Previne Brasil’ is that capitation payments are only made for people registered with Family Health Teams which should incentivise municipalities to further engage in this delivery form. While it is too early to conduct an overall evaluation of this programme, the number of people registered with Family Health Teams has increased substantially, reaching 127 million by November 2020 (OECD, 2021[31]).

Analysing the composition of overall health spending in an international context, can help assessing whether a country makes spending on primary health care a priority. In 2019, Brazil dedicated around 16% of its financial resources to “basic health care services” – defined as general outpatient care, dental care, home-based curative care and preventive activities,14 a share similar to the OECD average (Figure 3.10).

Figure 3.10. Spending on basic health care as share of current health spending, 2019 or nearest year

Note: Basic health services combine general outpatient care, dental care, home-based curative care and preventive activities provided regardless of the setting. It can be used as a proxy for primary health care for international comparisons.

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; Brazilian Ministry of Health.

Despite these achievements, performance of primary health care can be further strengthened

Comparing efficiency levels in primary care across municipalities in Brazil, the World Bank identifies possible efficiency gains of 37% (World Bank, 2017[26]). There is hence significant potential to expand service provision using the same amount of resources. Other issues that should be addressed include limited service accessibility and shortcomings in the co-ordination between primary and secondary care.

Lack of availability of doctors can lead to unmet needs and the 2019 Health Interview Survey has shed some light on this. In 2019, around 14% of all Brazilians who sought health care in the two weeks prior to the survey did not receive it (IBGE, 2020[3]). Particularly relevant for primary care is the fact that 21% of all Brazilians with diagnosed diabetes did not received medical care in the 12 months prior to the survey (IBGE, 2020[32]). Interestingly, while there is little statistical variation for unmet need of this kind across states, rurality or income (IBGE, 2020[33]), differences in quality of care appear to exist. Diabetic patients have a higher likelihood of complications (as measured in hospitalisations) in the North-East than in the South-East and this probability is systematically higher for those where household income is at the minimum wage or below than those with more than three times minimum wage (IBGE, 2020[34]).

Some of this unmet need will be related to the availability of doctors. While substantially increasing over the last decades, the number of doctors in Brazil is still lagging behind the OECD average. In 2017, around 2.2 physicians per 1 000 population were registered in Brazil (Scheffer and al., 2018[35]). This compares with 3.5 practicing physicians15 across the OECD (OECD, 2019[8]). Similar to most OECD countries, there is a wide geographic disparity in the availability of physicians in Brazil, also reflecting socio-economic differences. In municipalities with less than 5 000 inhabitants, the overall physician density is only 0.3 on average; for municipalities with more than 500 000 inhabitants the ratio stands at 4.3 (Scheffer and al., 2018[35]). Because specialists are more likely to practice in agglomerations, the unbalanced spatial distribution of health professionals is somewhat less pronounced when looking at generalists (i.e. registered doctors without specialisation) only. Nevertheless, significant differences still exit. On a state level, the generalist per 1 000 population ratio ranges from 0.4 in Maranhão in the North-Eastern region to 1.6 in Rio de Janeiro. A high geographical variation of health professionals can suggest an inefficient allocation of resources leading to unmet need and a lower quality of care in underserved regions and to overprovision of services in other parts of the country.

Experience from OECD countries informs that different strategies help redress the geographical imbalance of doctors (OECD, 2016[36]). A first approach tries to address the issue in the long-run by targeting future doctors during their medical education. This can be done by providing scholarships or preferred access to students who commit to practice in rural areas or communities for a defined period of time after graduation, as is the case of Australia, Japan, Canada, England and Chile (OECD, 2021[31]). For current professionals, there is a “stick and carrot” approach. Many OECD countries use financial incentives, either one-off or wage-related, to attract physicians to practice in rural or disadvantaged areas. In Canada and Denmark, some of these policies are in place. On the other hand, a few countries, including Germany, some provinces in Canada and Norway, restrict the place where doctors can practice. In Germany, self-employed doctors are not allowed to set up their practice and serve public patients in a catchment area that is considered ‘overserved’ based on a patient-to-doctor ratio. At any rate, addressing the regional imbalance of doctors, in particular for general practitioners, requires careful and coherent workforce planning. On a national level, this has so far been largely absent in Brazil (OECD, 2021[31]).

There could be potential to get more primary health care activity with existing resources

Interestingly, the substantial increase in the number of doctors in Brazil did not necessarily translate in a commensurate increase in primary care activity under SUS. The ratio of annual consultations for primary care registered under SUS divided by the total population stood at 2.2 in 2019, up from 2.0 in 2000 but below the value of 2014 (2.8)16 (Ministry of Health, 2020). While this figure should not be confused with the average number of consultations per person, the fact that this ratio increased by only 10% over the last two decades while the number of doctors grew by more than 50% should be analysed in more detail. It could mean that many newly graduated doctors do not want to pursue a career in primary care but it could signal some inefficiencies in care provision.

One issue that may explain the reduced growth in primary care consultation and associated problems in accessing SUS services17 is the wide-spread use of ‘dual practice’ which means that many physicians split their time between public and private patients. Recent regulation in Brazil prohibits doctors to have more than two jobs in the public sector, however, there is still the option to take on additional tasks in the lucrative private sector. A recent study finds that 51% of all doctors in Brazil work under dual-practice arrangements, 27% only in the private sector and only 22% exclusively in SUS (Miotto et al., 2018[37]). It also shows that those working solely in the public sector tend to be younger, less experienced and more likely to earn less than those in the private sector or in dual practice. Dual practice is also common in many OECD countries and can create problems to reduce service access of public patients if not properly regulated and monitored (Mueller and Socha-Dietrich, 2020[38]). Countries that have recently taken measures to increase physician’s time to treat public patients include Ireland and Israel. Part of their strategies include a pay raise for public doctors that forgo the option to work in the private sector.

A consequence of wide-spread application of dual practice can be absenteeism of doctors during regular working hours from their public jobs (Socha and Bech, 2011[39]). Recent research suggests that this may be an issue in Brazil. Analysing time use of 47 physicians working across 27 Family Health Teams in 12 municipalities indicate that they are absent from work during shifts in around a third of the time on average (dos Sanots Matsumoto, 2018[40]). This phenomenon is also a concern in some OECD countries but literature on this topic mentions the difficulty to clearly disentangle dual practice from poor overall management as the source of absenteeism (Socha and Bech, 2011[39]). That said, whether publically employed doctors perform their contractual work duties is clearly something that should be monitored closely. More analysis is needed to assess whether this is a common phenomenon in Brazil. If it is, the implementation of consistent monitoring tools could be an instrument to improve output and system performance.

Primary health care is still fragmented and problems in co-ordination of care across service levels persist

The latest Health Information Survey highlights the fragmentation of primary health care delivery. For example, around 50% of the diabetic population identified basic health units (‘unidade básica de saúde’) as their last contact to the health system, 11% used public speciality units in hospitals, 6% public urgent care units and 30% private practices or clinics (IBGE, 2020[32]). There is a clear socio-economic gradient to this: While basic health units were the most recent provider of diabetes care for around 60% or more of people from poorer household, these facilities are key health providers for only 15% or less of rich households (IBGE, 2020[41]). Difference in primary health care utilisation patterns and patient experience among elderly Brazilians was found in recent research (Macinko et al., 2018[42]). Those with private health plans (not using SUS) report fewer problems in terms of access, care co-ordination, continuity of care and provider co-ordination than those using SUS-Family Health Teams. Most problems are encountered by patients who use SUS but do not consult Family Health Teams.

The latter examples show that despite ongoing progress to further the roll-out of the Family Health Strategy, many Brazilians still use other providers as their regular point of care for chronic conditions. On the one hand, most people with private health insurance coverage do not use SUS at all and resort to private practices. Yet even for those without private coverage, a good part of them use public speciality units or emergency departments for provision of chronic services. This is inefficient since these types of conditions are best treated in primary care settings (OECD, 2020[43]). To further promote the Family Health Strategy more financial support may be needed, in particular in disadvantaged rural areas. In addition, out-of-hours options for primary care units should be further elaborated as in the “Saúde na Hora” programme, so that people with acute conditions do not need to resort to emergency care units. Moreover, cancer screening and preventive activities such as health check-ups should receive greater attention within primary care (OECD, 2021[31]). These are “good value for money” as these interventions reduce more costly treatment later.

In several OECD countries (including, for example, Chile, Portugal, Italy or Norway), strong gatekeeping systems are seen as a way to ensure that patients receive the best possible care for their conditions and to achieve greater appropriateness and co-ordination of care (OECD, 2020[43]). Further development of primary health care in Brazil could also entail to give GPs a stronger ‘gatekeeping’ role whereby they control and orient the patient’s entry into secondary care. This implies that patients have to register with primary care physician or practices and GPs control access to secondary care with a referral system.

Related to ‘gate-keeping’ is the need to strengthen the co-ordination of care across the health systems, also to address the long waiting times for visits to specialists or diagnostics. Historically, co-ordination between different care levels has been relatively weak in Brazil. To address these issues, the Federal Ministry of Health has established guidelines for the organisation of thematic Health Care Networks (“Redes de Atenção à Saúde – RAS”) in 2010 within the scope of SUS. This initiative seeks to integrate service provision through technical, logistical and management support across primary, secondary and tertiary care as well as ancillary services. At the moment, networks exist for prenatal care and childbirth, urgent care and emergencies, psychosocial care, for people with disabilities, and for people with chronic conditions, but the centrality of primary care in these networks is unclear.

Integrated care models can be successful in improving continuity of care leading to better outcomes but require information sharing across care levels and evidence based care pathways. However, the thematic integrated care network operate within the complex and rigid existing SUS planning and management environment and funding principles. This could be a barrier to a successful roll-out of health care network. Within the OECD, countries have embarked on different pathways to operationalise the integration of care (OECD, 2016[44]). In the United States and Germany, for example, providers joined into “Accountable Care Organisation” (ACO) where they are collectively responsible for patients in a defined catchment area. These ACO receive a virtual budget for a wide range of services and they are incentivised by shared savings contracts with payers to co-ordinate care across different levels efficiently. They can keep part of the savings if they keep total costs below the virtual budget if they meet pre-defined quality criteria. Other good examples for care integration exist in the Basque country of Spain and as primary health care networks in the United Kingdom (OECD, 2021[31]).

The Brazilian hospital sector has a lot of excess capacity in normal times but distribution is uneven

Overall, Brazil had approximately 474 000 hospital beds in 2019, around the value seen in 2007 but down from 501 000 in 2010. This equates with 2.3 beds per 1 000 population, around half of the OECD average (Figure 3.11). Nevertheless, the bed density is similar to some European countries such as Denmark (2.6), the United Kingdom (2.5) and Sweden (2.1) and above other countries in Latin America, such as Chile (2.0), Colombia (1.7) or Costa Rica (1.1). As in many OECD countries, in Brazil, too, there is some variation in bed availability within the country. On a state level, the density of hospital beds varies by a factor of 2, ranging between 1.5 beds per 1 000 population in the Northern state of Amapá and 2.9 beds in the Southern state of Rio Grande do Sul (Brazilian Ministry of Health, 2020).

Figure 3.11. Hospital beds per 1 000 population, 2019 or nearest year

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; Brazilian Ministry of Health.

An important difference between Brazil and OECD countries appears to be the level of activity. In 2019, there were 58 hospital discharges per 1 000 inhabitants in Brazil (Brazilian Ministry of Health, 2020), about one-third of the activity levels seen across the OECD and only one-fifth of the hospitalisation rate seen in Germany or Austria. The low number in Brazil also reflects the fact that the country has a relatively young population. Yet, there are more discharges per population in Brazil than in Colombia or Mexico.

As a result of the very low inpatient activity, the occupancy rate of hospital beds in Brazil is far below all OECD countries highlighting a lot of excess capacity in normal times (not considering peaks in demand associated with the COVID-19 pandemic). Only around 52% of hospital beds were used on average for treatment at any given day in 2019. This share has been slowly increasing recently but is still far below the OECD average of 76%. In Costa Rica, Israel, Ireland and Canada average occupancy rates are above 85%, which is broadly considered to be the limit for safe occupancy for patients in some countries (OECD, 2012[45]).

The very low occupancy rate of hospital beds in Brazil suggests a lot of scope of potential efficiency gains and has been the focus of numerous national and international efficiency analyses. Due to the large number of very small hospitals in Brazil, the 2019 World Bank report finds a staggering average inefficiency of 71%, highlighting substantial scope to expand service provision with existing resources. Based on this model, outpatient hospitals services could be increased by 140% and admissions by 79% while keeping expenditure constant if efficiency was maximised. Alternatively, the same level of results in hospitals could be achieved with 34% less spending, generating savings around BRL 12.7 billion (World Bank, 2017[26]). Unsurprisingly, the report finds that inefficiencies are much more pronounced in small municipalities and mainly driven by hospitals with less than 100 beds.

In a performance assessment of public hospitals, the Brazilian Federal Court of Accounts also found ample evidence of inefficient service provision (Tribunal de Contas da União, 2020[46]). Applying the same method as the World Bank but differentiating by hospital type, the results of this assessment indicate further that efficiency increases with hospitals size but no difference across regions. Interestingly, the report also finds that (i) public hospitals under direct administration are less efficient than those under indirect administration; (ii) hospitals under management of the states are more efficient than those managed by either municipalities or the federation; and (iii) that public hospitals managed by “social organisations” (Organizações Sociais de Saúde) have a better efficiency score than other public hospitals. For the latter finding, however, it is important to note that this may also be due to selection bias and the fact that OSS-managed hospitals are nearly three times bigger than the average public hospital (151 beds vs. 61 beds).

The way inpatient delivery is planned for SUS contributes to excess capacity

Even considering that the COVID-19 pandemic has shown that some easily mobilised inpatient reserve capacity can be advantageous under exceptional circumstances, the low efficiency in the hospital sector in Brazil has adverse effects.18 From a purely economic perspective, the simple solution to improve efficiency in the hospital sector would be to close many small hospitals. In reality, this is more complicated. In areas with very low population density, alternative treatment facilities may be hundreds of kilometres away. In addition, from a political economy perspective, mayors and municipal health secretaries have little to gain from a closure of existing health facilities as this would not be appreciated by voters. A more feasible approach could be a conversion of some small hospitals in rural areas into more intermediate health facilities that can treat urgent cases and patients with acute care needs for a limited period of time. This should be accompanied by a strengthening of transport opportunities to better equipped general hospitals in urban centres for patients requiring urgent inpatient care for a sustained period of time. This model could also be tied in with the roll-out of tele-health applications to help triage patients, distinguishing those who need immediate transport to general hospitals from those who can be treated locally.

On how to find the right balance between efficient acute care provision and accessibility in rural areas, lessons can be learned from some OECD countries with similar spatial challenges. In Canada, all provincial and territorial Medicare plans are expected to provide uniform access to hospital care also in rural and remote areas (Marchildon, Allin and Merkur, 2020[47]). They achieve this by extensive referral patterns involving medical transportation from rural and remote areas to tertiary care hospitals in urban areas. This involves a wide network of road- and air-based emergency medical services. In Australia, some states and territories have developed multi purpose health service (MPHS) as a service for flexible care. It is an amalgamation of services including acute hospital care, residential aged care, community health, home community care and other health related services and generally established in populations not large enough to support a separate hospital, residential aged care and community care services (Queensland Government, 2018[48]). An alternative concept popular in Finland is to embed an inpatient ward into primary health care facilities of municipalities (Keskimäki et al., 2019[49]). These GP-led facilities are typically equipped with 30-60 beds. In recent years, these facilities have taken a more active role in rehabilitation and cancer care. Around half the beds are used for acute and chronic care each.

The existence of many small hospitals leading to the described efficiency issues appears to be also the result of the current mode of hospital planning in Brazil. Law 8080/1990 gives municipalities a central role in planning and management of SUS services, in co-operation with states and health regions. Yet municipalities differ vastly in size and also in the capacity to perform complex managing tasks, such as the operation of secondary and tertiary care facilities. In a number of OECD countries, hospital planning is delegated to a higher governmental level. In Australia, the administration and performance of public hospitals is the responsibility of the states and territories as system managers. In Canada, provincial ministries are responsible for major new capital (e.g. hospitals) and some infrastructure planning (Marchildon, Allin and Merkur, 2020[47]). This is also the case in some social insurance-based health systems. In Germany, for example, hospital planning is carried by the 16 ‘Länder’ while most public hospitals are owned by municipalities. Thus, Brazil could consider moving hospital planning competencies to another government level, be it the states, “health regions” or “macroregions” (see Section 3.5.5). What needs to be in place are clear federal guidelines stipulating transparent planning processes and accountability.

There are few incentives for better performance in inpatient care delivery

The hospital landscape in Brazil is diverse. Public hospitals can be owned by the federal government, the states and municipalities and may be managed directly by system managers or employed hospital managers. The management of public hospitals may also be outsourced to private “Social Organisations” (OSS). In addition, state or municipal health managers also contract private non-profit and for-profit hospitals for service delivery under SUS. The latter two types of hospitals will also cater for private patients that either use their insurance coverage or pay directly. Overall, in 2019, public hospitals and private not-for-profit hospitals each accounted for around 38% of all hospital beds, with the remaining 24% provided by private for-profit hospitals.

In line with the diverse hospital landscape, the current form of financing specialist and inpatient activity in Brazil is very complex and differs across states and municipalities. On average between the years 2010-14, SUS financing for inpatient care was to 39% made from federal sources, to 48% from the states and to 13% from municipalities (Ministério da Saúde/Fundação Oswaldo Cruz, 2018[29]). Most of the federal resources to finance activities of medium and high complexity (Média e Alta Complexidade- MAC) are transferred to states and municipalities and the transfers are subdivided into different components: (i) fee-for-service payments based on a national price list for procedures; (ii) incentives related to national policies; (iii) and a global budget for a set of activities. Municipalities and states have to complement the federal FFS funding. In some states, additional incentive payments related to state priorities exist. Finally, there may be performance bonuses made to hospitals at the discretion of the contracting state or municipality manager.

The current mechanism to purchase hospital services does not appear to incentive improving hospital performance. The transfers from the federal government to states and municipalities are partially based on historic budgets and payment by procedure through a national fee schedule (SIGTAB) which, however, is updated only irregularly, mainly for high costs procedures in the context of ‘judicialisations’. As a result, the fee schedule sends distorted price signals and payments do not necessarily reflect resource use. Hence, an efficient resource allocation between different hospitals becomes challenging. Some hospitals may receive insufficient funding for the services they provide. Generally, FFS payments also encourage hospitals to provide more procedures than necessary.

To improve technical efficiency, many OECD countries have introduced prospective case-based payment systems based on Diagnosis-Related Groups (DRG) to pay for hospital activity (OECD, 2016[44]). For the operationalisation of this mode of payments, patients are grouped into different patient groups reflecting the resource-intensity of the treatment taking into account diagnosis and other factors – called DRG. Different weights are associated with the various DRGs reflecting average treatment costs. After discharge, the hospital receives a payment reflecting the weight of the DRG. Alternatively, DRGs can also be used for the allocation of hospital budgets. In this case, hospitals budgets are defined via the sum of DRG weights (or ‘case-mix’) reflecting the difference in complexity of cases across hospitals. Depending on how DRGs are used, they can incentivise hospital managers to reduce costs per patient, increase the revenue per patient and increase the number of patients (Quentin et al., 2011[50]). To avoid any possible skimping of care, quality of care should be monitored. For an accurate calculation of average resource use per patient, hospital cost accounting systems need to be in place. Brazil could explore options to move towards such a patient classification model and to use it to pay for service provision. There is a wealth of experience in the use of DRGs in OECD countries and how to adapt existing classifications to national circumstances. The existing hospital datasets in Brazil on procedures could serve as a good starting point; what seems to be missing so far is detailed costing information.

The provision of low value care remains an issue in Brazil

Brazil also encounters issue with wasteful clinical spending. One example of health services that are of low value are double examinations or costly procedures that may not be clinically necessary. Caesarean sections can fall into the latter category. While they can save the lives of mother and infant for medically indicated reasons, these surgical procedures can cause complications leading to disability and deaths especially if carried out in facilities that lack capacity to conduct safe surgery. Based on a systematic review, the WHO concludes that C-section rates above 10% are not associated with reductions in maternal and new-born mortality rates at the population level (WHO, 2015[51]). Brazil is one of the leading countries for these types of intervention. In 2019, birth via C-section accounted for 56% of all life births in Brazil with some significant variation across the country – ranging from 35% in the Northern state of Roraima to 67% in the Central-Western state of Goiás (Secretaria de Vigilância em Saúde, 2021[52]). The share in Brazil is above that of any OECD country and more than twice the OECD average (Figure 3.12). In many European countries and Israel, the proportion is below 20%. The high and growing number in Brazil suggests that other factors than medical indication play an important role in the decision making of delivery. This may be due to cultural differences in patient expectations, availability of midwives and nurses, medical practice but also financial incentives for health providers. There is also some international evidence that private hospitals tend to perform more caesarean sections than public hospitals (OECD, 2019[8]).

Figure 3.12. Caesarean section rates, 2019 (or nearest year)

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; Secretaria de Vigilância em Saúde (2021[52]), “Painel de Monitoramento de Nascidos Vivos”, svs.aids.gov.br/dantps/centrais-de-conteudos/paineis-de-monitoramento/natalidade/nascidos-vivos/.

Reducing the C-section rate is an objective in Brazil and included in the National Health Plan 2020-23. Some OECD countries have also taken measures to address the trend of increasing C-sections. In Australia, for example, a number of states have developed clinical guidelines and required reporting of hospital caesarean section rates, including investigation of performance against the guidelines. These measures have discouraged variations in practice and contributed to slowing down the rise in Caesarean sections (OECD, 2019[8]). A few countries including France, the United Kingdom and Korea have introduced financial incentives to reduce the use of unnecessary caesarean sections, for example by reducing the price gap or aligning the prices between normal delivery and C-section that hospitals can charge (OECD, 2014[53]). Brazil should consider these regulatory options to curb down C-section rates but could also intensify efforts to raise awareness of this issue among young women.

More generally and going beyond C-sections, encouraging patient-provider conversation about whether certain treatments are adequate can be one way to reduce low value care. This is the focus on the Choosing Wisely® campaign, an initiative led by clinicians that began in the United States in 2012 and which has since then spread to several other countries, including Brazil (OECD, 2017[24]). This initiative provides an evidence-based list of “do not do’s” across multiple medical specialties to help identify inappropriate care that is of little value to patients. The recommendations included in these lists cover many different clinical areas, for example under which conditions expensive imaging test for lower back pain are unjustified or when antibiotics should not be prescribed.

An important tool to improve health care quality and reduce unnecessary care is the establishment of evidence-based clinical practice guidelines and monitoring their compliance. Recent research suggest that while important progress has been made in Brazil in the development of these guidelines in recent years, many still lack transparency and methodological rigour and do not necessarily follow international standards (Colpani et al., 2020[54]). High geographical variation in medical practices in a country can signal inefficient care provision and low adherence to clinical guidelines. To monitor variation in health care activity that may not be medically justified, many OECD countries including Australia, Canada, the United Kingdom and the United States have developed atlases (OECD, 2017[24]).

Accessing medication can be an issue in Brazil and better procurement may help

Given the size of the country, ensuring access to essential medicines at all times for everyone is a logistical challenge in Brazil. However, the way purchasing is organised may exacerbate this problem and there appears to be some room for efficiency gains. While making municipalities responsible for procurement for ‘basic’ medication (CBAF) may allow them to better respond to local population needs, it also leads to a higher degree of fragmentation of the purchasers market. With one-fifth of municipalities having fewer than 10 000 inhabitants, many do not have the capacity to adequately engage in such a complex activity and lack the size to be in a strong position to negotiate prices with manufacturers. While many ‘purchaser consortia’ among municipalities (and even states, for specialised medications) exist, pharmaceutical procurement responsibilities for SUS pharmaceuticals could be reconsidered all together. For medication in the CBAF category, procurement could be generally delegated to states or the federation. Alternatively, Brazil could develop national negotiations or public bidding processes for CBAF medicines on a federal level where municipalities would purchase directly medicines from (nationally) contracted manufactures for the (nationally) agreed price. Procuring pharmaceuticals in the largest possible quantities to achieve economies of scale – both in centralised and decentralised health systems- is among the principles of good pharmaceutical procurement of WHO (WHO, 1999[57]). These principles also highlight that the performance of local procurement should be monitored closely as decentralisation may threaten the appropriate selection of drugs, and the potential for discounts through bulk purchases, quality assurance and accountability. The Brazilian Federal Court of Account also sees decentralisation as a challenge for pharmaceutical procurement and has issued some guidelines to improve purchasing for the different levels of government (Tribunal de Contas da União, 2018[58]).

There is room for greater use of unbranded generics

Medical goods (including pharmaceuticals) are a substantial cost component of overall health spending in Brazil (20%) and can be a financial burden for both governments and people. Generally, the development of generic markets is an opportunity to increase efficiency in pharmaceutical spending. Underutilisation of generic drugs can be a substantial source of inefficiency as generics have the same therapeutic effects as branded alternatives but are typically much less expensive.

In Brazil, the development of the generic market has progressed rapidly since the Generic Medicines Law was passed in 1999. The law required tests for bioequivalence of generic products for marketing authorisation that made the requirements for market entry in Brazil relatively strict compared to those in other Latin American countries (Massard da Fonseca and Shadlen, 2017[59]). The legal change led to a decrease in the market share of originator drugs and price reductions for originator products and ‘similares’ (which already existed before 1999) in the ten years following the introduction of this law (Bertoldi et al., 2019[60]). In 2019, the generic market share in the retail market in Brazil stood at 78% by volume, well above the OECD average of 52%, while at 69% of all retail pharmaceutical sales by value, which was nearly three times the OECD average (26%) (Figure 3.13). However, it is important to note in this context that the majority of the generics used in Brazil -as in most Latin American countries- are similares.21 Similares are copies of off-patent products using a trade name instead of the name of the molecule. In Brazil, most simliares can be understood as ‘branded generics’ as they have to prove bioequivalence for renewal of registration since 2003 (this is not necessarily the case in other Latin American countries). Their prices are usually higher than those of non-branded generics. In contrast, in OECD countries, branded or unbranded generics do not make a major cost difference, mainly because health systems provide coverage for them irrespective of this classification. This also implies that price differentials between originator and generic medicines (including branded) are smaller in Latin American countries than across the OECD.

Figure 3.13. Volume and value share of generics in the retail pharmaceutical market, selected countries

Note: 1. Refers to the reimbursed pharmaceutical market.

Source: OECD (2019[8]), Health at a Glance 2019: OECD Indicators, https://dx.doi.org/10.1787/4dd50c09-en; OECD/The World Bank (2020[61]), Health at a Glance: Latin America and the Caribbean 2020, https://dx.doi.org/10.1787/6089164f-en.

While the use of unbranded generics is relatively high in Brazil compared to other Latin American countries (OECD/The World Bank, 2020[61]), there appears to be some scope to rein in pharmaceutical spending. In a comparison of prices for pharmaceuticals in six Latin American countries between 2010 and 2015, Álvarez and González found that, on an aggregate level, those seen in Brazil were higher than those in Argentina, Chile, Colombia, Mexico and Peru, for both ex-manufacturer and retail prices (Álvarez and González, 2020[62]). This ranking is confirmed when looking at prices for unbranded generics. Above-average international prices for unbranded and branded generics in Brazil were also found by Bertoldi et al. (Bertoldi et al., 2012[56]). Earlier research suggests that generic prices (branded and unbranded) in Brazil were above those seen in the United States, France and Spain and on the same level as in Canada and in the United Kingdom (Danzon and Furukawa, 2008[63]). The comparably high prices for generics may partly be explained by the way prices are set. While a generic drug entering the market needs to be at least 35% below the price of the originator product, there are no regular price updates required once authorisation is issued (Bertoldi et al., 2019[60]). As a result, the price of a generic may actually exceed that of the originator drug. Moreover, the 35% difference is not applicable to branded generics (Bertoldi et al., 2019[60]), and prescriptions for branded generics are currently exempt from substitution to unbranded generics (Massard da Fonseca and Shadlen, 2017[59]).

More regular revisions of maximum prices of generics and increasing the scope of substitution to include similares with proven bioequivalence could be ways for Brazil to rein in overall pharmaceutical spending and improve affordability among disadvantaged population groups. Expanding campaigns to educate prescribers and patients on the interchangeability of generics could also be an option since studies suggest that there is lack of confidence in the quality of generics and confusion how to differentiate between originator, similar and generic products (Massard da Fonseca and Shadlen, 2017[59]). The government tried to address this in 2014 by proposing to add the label “EQ” to the packaging of interchangeable products, also hoping that this would increase competition and thereby reduce prices. Yet this proposal was not enacted, giving in to industry concerns that the “EQ” symbol would send a message to patients to ignore branded medicines (Massard da Fonseca and Shadlen, 2017[59]).

The administration of the complex Brazilian health system comes at a cost

The complexity of the management of the SUS and the strong role of the private health insurance market in Brazil results in a very high share of resources needed to administer the health system. In total, more than 6% of current health expenditure is used for governance and health system and financing administration.23 This share is higher than in nearly all countries of the OECD and more than twice its average (Figure 3.14). The fact that Brazil has an important private health insurance market explains partly why the country ranks very highly in an overall comparison since private health insurance is typically associated with much higher administrative costs than public schemes (OECD, 2017[24]). However, what is surprising in the case of Brazil is that for government schemes the share of administrative expenses is also substantial, standing at 5.7% of total government health expenditure. Other countries with mainly tax-financed system and residence-based entitlements such as the United Kingdom, Italy or the Scandinavian countries dedicate a much smaller proportion of their resources to the administration of the health system. A reasonable hypothesis why the share in Brazil is so high is that it reflects the complexity of SUS.

Figure 3.14. Costs for administration as share of current health expenditure, 2019 or nearest year

Note: Costs for administration only refer to governance, health system administration and financing and exclude administration costs that occur in particular health providers such as hospitals.

Source: OECD Health Statistics (2021[1]), https://doi.org/10.1787/health-data-en; Ministry of Health of Brazil 2021.

It should be made clear that expenditure for governance and health system and financing administration are not bad per se. Stewardship and oversight are needed to make sure that services are provided safely and effectively throughout the country. However, countries with high administrative costs should carefully weigh if they obtain corresponding benefits in return. In the case of Brazil, this is not obvious. It rather seems that the joint responsibilities across all three levels of government in the management of SUS contribute to a situation characterised by unnecessary complexity, redundant tasks and a lack of clarity and accountability. Moving forward, Brazil should map out and evaluate critically all processes related to the management and planning of SUS at all levels of government. This can contribute to identifying superfluous administrative procedures, can help streamline competencies for each level of government and can be the starting point for discussing changes in governance and responsibilities across the levels of government to increase efficiency and accountability in SUS. In that context, Brazil should also carefully review data requirements for improved service delivery planning in SUS. While existing knowledge gaps that have been identified should be addressed (Chapter 4), there also appears to be disconnect in the country between data availability and data usage by planning authorities and decision-makers. Improving data exchange and communication can prevent that Brazil becomes “data rich but information poor”.

Stronger focus on regionalised planning can overcome some inefficiency barriers

A key topic when reviewing management responsibilities in SUS is the role of municipalities. They are important actors in the financing and management of SUS as well as in service delivery planning in line with the principle of “decentralisation” on which SUS is based on. Yet the size of municipalities differs vastly across the country, which affects the capacity of municipal health officers to effectively organise and manage SUS. The existence of many small municipalities raises the question whether some of the SUS functions should not better be delegated from a local to a regional level.

The need for better regional co-ordination to improve service delivery planning has been realised in Brazil with the Health Pact of 2006 and with Law 7 508 of 2011. The latter provides guidelines for the establishment of “health regions” which are composed of a varying number of (neighbouring) municipalities within states. These “health regions” are defined in the multi-annual state health plan and have to combine at least primary care, urgent and emergency care, psychosocial care, specialised outpatient and hospital care and health surveillance. To organise care of medium and high complexity the regions should organise themselves in “macroregions”.24 On paper, it appear that the installation of such health regions can be a powerful tool to improve care co-ordination and efficient service provision and also facilitate the operation of health care networks. However, as of now, it does not seem that regionalisation has shown its full potential. One problem appears to be that these entities have no decision-making power, budget authority or human resources and rely entirely on municipal and state support. There may also be reluctance among state and municipality officials to delegate financing and planning competences to the health regions for fear of loss of influence. However, the strengthening of the roles and responsibilities of health regions appears to be a vital element to improve SUS performance.

Access to health services through courts are a drain on available resources

A specific challenge to the Brazilian health system is the fact that a high number of people resort to the judicial system to access health care at the expense of SUS.25 This is the case when people who have prescribed treatment or pharmaceuticals denied to them, invoke the Federal Constitution which stipulates that “…health is a right for all and duty of the state”. These so-called “judicialisations” have been growing constantly for many years and represent a considerable drain to the scare resources of SUS. Between 2010 and 2015, federal spending based on judicial decisions increased tenfold, reaching BRL 1 billion (Tribunal de Contas da União, 2017[71]) or roughly 1% of the federal health budget, mainly related to access to high cost medicine.26 At the state level, the aggregated costs for these cases seem to be even higher (Tribunal de Contas da União, 2017[71]).

Getting access to health care services, which are not deemed cost-effective based on judicial interventions, however, is problematic, both, from an efficiency and equity perspective. It is an inefficient use of resources because it complicates priority setting in health and runs against the purpose of HTA conducted by CONITEC. As an example, based on court rulings, a substantial amount of “judicial spending” in 2014 went on the drug Soliris®, which had not been granted market authorisation at the time and where annual treatment costs are estimated to be over BRL 1 million per person (Tribunal de Contas da União, 2017[71]). In addition, the high and growing number of court cases also put a strain on the judicial resources. By the end 2018, all pending health related court cases at all court levels in Brazil (including the “judicial cases”) amounted to over 2.2 million (Vieira, 2019[72]).27 Beyond preventing an efficient use of resources based on criteria such as cost-effectiveness, “judicialisation” is also a problem from an equity perspective since those who get access to health care based on court rulings are typically those who are able to pay for legal fees or know how to navigate the legal system – mainly the rich and the better educated.

For international observers, the magnitude of the problem and the fact that judges give favourable rulings for access to experimental treatment for not authorised medication on a regular basis seems puzzling, especially since the establishment of CONITEC was an important achievement for the Brazilian health system.28 By doing so, judges become de facto co-managers of the health budget. The Ministry of Health is also concerned about the rise in ‘judicialisations’ and has taken a number of measures to help judges make an informed decision. For example, a co-operation agreement was signed with National Council of Justice (CNJ) with the objective to provide advisory services to the different competent courts in the judicial system. While the legal system can play an important role in guaranteeing universal health care in many countries, courts in OECD countries typically do not systematically bypass standardised HTA processes but rather hold HTA agencies to account in case they find flaws in the assessment processes. A change in the judicial culture in Brazil seems critical for a solution to this problem. A recent analysis of judicialised court rulings finds that the establishment of CONITEC had little impact on courts’ tendency to rule in favour of patients (Wang et al., 2020[73]). In most cases, CONITEC decisions were simple ignored and not even cited in the ruling decisions. Making the concept of cost-effectiveness in the provision of SUS more explicit could also be an option to reduce the size of the problem but critics would argue that the current legal framework is already sufficient to reign in ‘judicialisations’.29

Budget issues prevail at federal and other government levels and raise questions whether they can accommodate changing priorities

The overall resources available to finance activity of SUS is decided on the three levels of government taking account of the relevant laws and constitutional amendments which, on the one hand, define the minimum spending floor for each governmental entity but, on the other hand, limit annual federal public spending growth to inflation. On the federal level, the overall budget is generally perceived as very rigid with 93% of spending mandated by law either through earmarking, indexation or mandatory spending floors (OECD, 2020[11]).

The executed federal budget on health in 2018 stood at around BRL 121 billion, financing around 42% of all SUS expenditure (SIGA Brasil, n.d.[74]). Due to the decentralised nature of the management of SUS around two-thirds of federal health spending refers to transfers to the states and municipalities to purchase health care goods and service. Most of these transfers are directed towards municipalities. Since 2007, the federal transfers are structured around six financing blocks (primary care, medium and high complexity hospital and outpatient care, health surveillance, pharmaceutical assistance, SUS management and investment), including subcomponents, with the individual transfers being determined by population size but for some blocks, such as primary care, other factors reflecting difference in needs also play a role. While some flexibility in the use of funds has been allowed for over the years, there is an issue at state and municipality level to reallocate earmarked federal funds for other purposes (dos Santos and de Luiz, 2016[75]). The current budgeting practice is also not conducive to the envisaged roles of the health networks and it is difficult to see how ‘health regions’ or ‘macro regions’ can play a bigger part in planning in the current environment without having some sort of budget autonomy.

More budget flexibility would make it easier to pursue policy priorities. In addition, there is a wide-spread use of amendments to the federal budget by parliamentarians as a tool to increase spending in their constituency. This practice undermines spending efficiency and transparency and makes local authorities depend on members of congress (OECD, 2020[11]). These spending items are also typically not audited. The fiscal implications of these amendments have been growing in recent years, with spending commitments reaching BRL 12.9 billion in 2019 (Portal da Transparência, 2019[76]). More than half of the committed amount stem from amendments in the area of health (BRL 6.7 billion) mainly for primary care or care in hospitals.

Substantial resources are diverted from health service delivery

The intentional misuse of funds and corruption is commonplace to differing degrees around the globe in many areas of public life and Brazil is no exception. In fact, Brazil had some major corruption scandals involving prominent public figures in the last decade. In 2020, Brazil ranks 94^th of 180 countries in the perceived level of public sector corruption,30 much lower than nearly all OECD countries (Transparency International, 2020[77]). When it comes to health care, fraud and the diversion of funds can take different forms. Health providers and professionals may bill services to purchasers that they did not deliver, system managers may contract health providers at artificially high prices or divert health care goods for their own benefits and patients may offer bribes for health providers or professionals to ‘jump the queue’ for quicker treatment. This practice is problematic for a number of reasons. It is inefficient spending, frequently creates equity issues and undermines the trust of people in the public health system. In 2013, corruption in health was perceived to be more widespread in Brazil than on average across OECD countries (Figure 3.15). It was similar to the United States and Chile, but below Colombia, Greece or Lithuania (Transparency International, 2013[78]). On the other hand, it was much higher than in Denmark, Spain, Finland or Australia.

Since information on illegal activity its patchy at best, it is very difficult to quantify the problem. When looking at the population willing to engage in integrity violations, 5% of Brazilians reported having paid bribes in public clinics and health centres within 12 months in 2019, down from 7% in 2017 (Transparency International, 2019[79]). Among Latin American and Caribbean countries, this was a very low share. However, misuse of funds does not necessarily require the patient. Recent research estimated the total costs of fraud, corruption and misuse of funds in the health sector in Brazil to be around BRL 14.5 billion or 2.3% of total health spending (Instituto Ética Saúde, 2020[80]). Due to its complex nature that involves many actors at different levels of government, the misuse of fund may be more difficult to detect in SUS than in other health system. Indeed, in a 2017 report, the Federal Court of Accounts highlighted important deficiencies in the National Auditing System of SUS as an oversight instrument in municipalities and states (Tribunal de Contas da União, 2017[71]). It also noted that 55% of the federative entities that were audited did not implement SUS’ internal auditing component.

Figure 3.15. Perceived level of corruption in medical and health, 2013, Brazil and selected OECD countries

Level on a scale of 1 to 5; where 1 means not at all corrupt, 5 means extremely corrupt

Source: Transparency International (2013[78]), Global Corruption Barometer 2013”, https://www.transparency.org/en/gcb/global/global-corruption-barometer-2013.

Brazil tries to address these issues. For example, to detect misuse of public funds the Comptroller General in Brazil conducts a substantial number of special audits each year. Between 2002 and 2018, the funds associated with irregularities amounted to BRL 23.8 billion, of which 28% (BRL 6.6 billion) were under the remit of the Ministry of Health (Controladoria-Geral da União, n.d.[81]).

There are different policy levers used in OECD countries to mitigate integrity violations in health addressing the different stakeholders (OECD, 2017[24]). One response is primarily organisational in the sense that it involves assigning responsibilities for detecting or tackling integrity violations in service delivery and financing to specific institutions. Depending on the country, these can be dedicated central or government programmes or institutions, or these responsibilities are delegated to public or private payers. In some instances fraud detection is very pro-active going beyond simple audits. The Belgian National Institute for Health and Disability Insurance, for example, uses data mining to detect integrity violations in service delivery and payment. To tackle inappropriate business practices, countries’ responses are typically regulatory in nature and consist of limiting or banning certain practices. The main domains where some countries have introduced regulation seek to limit self-interested referrals by health providers and the means by which the pharmaceutical industry is allowed to promote sales – including Sunshine regulations.31 In the last 15 years, a number of countries introduced specific and comprehensive legislation, notably France, Portugal, the Slovak Republic and the United States.

More clarity about the role of the private sector can improve health system performance

Traditionally, Brazil has an important private health sector predating the inception of SUS and the Federal Constitution of 1988 warrants the continuation of private service delivery and private health insurance. Private providers play a vital role to ensure access for SUS patients, in particular for inpatient care. In addition, private organisations are contracted to manage public hospitals or other facilities, and private health insurance can ‘supplement’ SUS coverage. However, the co-existence between a public and private sector is not without friction and can be a source of inefficiency.

Private “social organisations” (OSS) have become a popular tool among some SUS managers to operate public health facilities. By contracting out the management of a health facility to a private company, this option gives more flexibility in hiring staff. Some studies highlight that public facilities under OSS management have more activity or provide services at lower costs than those under direct management (World Bank, 2017[26]), but other research cannot rule out some selection bias when finding that OSS operate more efficiently (Tribunal de Contas da União, 2020[46]). However, OSS contract management seems to be challenging for SUS manager as audit courts have frequently highlighted the insufficiency of monitoring mechanism (dos Santos and Servo, 2016[82]).

The fact that some private hospitals treat SUS patients and private patients simultaneously can lead to reduced access for public patients and associated equity issues. This happens when hospitals (or hospital doctors) receive higher payments when treating private patients. Access problems for public patients were the reason why Ireland introduced a ban on this practice and promoted the clear separation between hospitals that treat public patients from those that treat private patients (Mueller and Socha-Dietrich, 2020[38]). One way to address this issue would be to define a fee schedule that is applicable to all payers. This is the case for example in Germany, where for inpatient treatment the DRG tariffs are identical for public and private patients and applied regardless of whether they are treated in public, private for profit or non-for-profit hospitals.32

The well-developed private health insurance market in Brazil has some peculiarities which distinguishes it from those in OECD countries. It provides mainly duplicate insurance to SUS coverage but de-facto it is used by many as a substitute insurance for SUS as the range of services covered by private plans can be extremely wide. This unclear role of private health insurance is accentuated by the fact that SUS can demand under certain conditions reimbursement of costs from private insurers in case patients use SUS services instead of their private insurance coverage (“ressarcimento ao SUS”). This is unusual compared to OECD countries with a tax-financed system and duplicate insurance. Brazil could benefit from an open discussion to clarify the role private health insurance should play within the health system. As already mentioned, the generous tax subsidies are highly regressive. Moreover, the co-existence of SUS and private insurance is also problematic for care integration. So far, there is very little communication or data exchange between providers (or schemes) in case patients move between treatment under SUS and private coverage. The introduction of a unique patient identifier and interoperability health records between SUS and private insurance would facilitate the co-ordination of care.