Health accounts are a vital tool to obtain comprehensive and internationally comparable data on health expenditure and financing, which is key information to assess the performance of health systems. After several pilot implementations over the past decade, Brazil is now committed to institutionalising the annual production of health accounts at a more granular level. This will unlock the potential of health accounts to better inform health policy decision making. This report reviews efforts to institutionalise health accounts in Brazil aligned with international standards. It analyses the comprehensiveness of reporting, the data sources used, and methodologies applied to estimate health spending. It also provides recommendations on how to best institutionalise the regular production of health accounts and disseminate the results for greater policy impact and suggests how the current work could be further expanded in the future. Finally, the report analyses the first health expenditure data for Brazil generated with this new methodology in comparison to health spending in OECD countries.
Executive Summary
The systematic monitoring of the financial resources dedicated to health is crucial in any health system performance assessment and nearly all OECD countries have institutionalised the annual production of national health accounts as a key instrument to track health spending. Brazil has a long history in producing health satellite accounts to measure the health sector’s contribution to the overall economy and has also piloted the implementation of national health accounts over the past decade to provide more focused and detailed insights into health expenditure and financing. Yet, while two significant reports have been published, these previous studies have not resulted in a full institutionalisation of health accounts in Brazil.
In 2023, the Ministry of Health of Brazil decided to re‑engage in this work with the intention of fully institutionalising the annual production of health accounts to unlock its analytical potential. The OECD Secretariat is supporting the ministry in this endeavour and -with this report- has reviewed the new initiative to institutionalise health accounts in Brazil to ensure that it is line with international standards. The report also provides recommendations on how to best institutionalise the regular production of health accounts and disseminate its results, and indicates how the current work could be possibly expanded in the future.
In terms of the overall assessment the OECD concludes that:
- The proposed methodology used to calculate health spending in Brazil is sufficiently aligned with international standards and practices to ensure international comparability.
- All key financing schemes have been identified and for each of these appropriate distribution keys have been generated to allow an allocation of health spending to the various health services.
In addition, it is found that the level of detail that Brazil is able to provide from a service perspective generally meets international reporting requirements and, in some cases, goes beyond what several OECD countries can currently report. However, some data gaps exist and – looking towards future health accounts production rounds – it is recommended that Brazil should continue to explore how they could be filled:
- For some of the government financing schemes spending data could not be identified.
- Some types of financing schemes albeit of lesser importance in Brazil are missing entirely.
Finally, regarding the estimation techniques utilised, while there is confidence in the overall methodology, in some cases it could warrant further review or refinement. Two important areas are noted:
- The estimation methodology of overall out-of-pocket spending and the allocation to functional level.
- The production of the distribution keys for spending by SUS appears to be methodologically sound in principle but there should be further investigation to assess the extent to which the underlying data source (SIGTAP) accurately reflects the prices for the various products and services, and update as required.
One essential component of the new initiative to institutionalise health accounts is to comprehensively document the production process, the data sources used, and the methodologies applied to derive health spending estimates. It is felt that the new health accounts manual, which the Brazilian health accounts team has drafted when producing new data for year 2022, is an important step towards a comprehensive documentation of this work. In general, the structure and content of the Brazilian manual is considered robust, providing the key methodological information for people who want to get a better understanding of how results are derived. In areas advising further improvements, there may be scope to provide additional detail on how high-level “control totals” and distribution keys are derived in practice, including some specific examples.
Putting the new health accounts results for 2022 into an international context reveals a number of important insights and specificities of the Brazilian health system:
- With 45% in total health spending, the financing share of compulsory schemes in Brazil is much lower than in OECD countries (75%). At the same time, the role of voluntary health insurance (27% of total spending) is much more pronounced than in any OECD country. On average, this scheme accounts only around 5% of overall health spending across the OECD. The share of out-of-pocket spending in Brazil (27%) is above the OECD average (19%).
- The share of health spending allocated to outpatient curative care is higher in Brazil than across the OECD, while the share of inpatient services is lower. The extensive use of laboratory and imaging services in Brazil (representing 6% of total health expenditure each) are contributing to the high outpatient share. Brazil dedicated around 19% of its financial resources to basic healthcare services, a proxy measure used to compare spending primary healthcare internationally. This was similar to the OECD average. Conversely, Brazil only dedicated 22% of overall health expenditure to inpatient care in hospitals. Moreover, the overall proportion of total health expenditure dedicated to health system administration in Brazil (6%) is above most other OECD countries, reflecting somewhat the known complexities of managing SUS but also the management costs of the large private health sector in the country.
- When analysing financial coverage of healthcare costs in Brazil, one finding that stands out is the substantial difference in coverage across goods and services. While overall spending on generalist outpatient care is largely covered by compulsory schemes (85%), only around 10% of all pharmaceutical costs are covered by the public pursue. This can point to gaps in the SUS benefit basket but also to problems in effectively accessing items that are publicly covered.
A full institutionalisation of health accounts production is required to obtain this type of information on an annual basis. Based on best practices across OECD countries, a number of factors were identified that facilitate such an institutionalisation.
- Establishing a dedicated national health accounts team with sufficient staff and the appropriate skill-mix and access to adequate IT infrastructure is vital.
- Having a legal a mandate for the production of health accounts is not a requirement but can in some instances accelerate and ensure the regular and sustainable production of this data.
- Developing and maintaining good relationships with data providers and stakeholders is crucial and the establishment of an expert group to advise on the methodology to produce health accounts and on the interpretation of results has proven to be useful in many OECD countries.
- The technical production process should be standardised as much as possible and needs to be considered as an iterative process, driven by the desire to continuously improve the health accounts methodology by incorporating new data sources and information whenever available.
- Quality assurance across all steps in the production process and results and a comprehensive documentation of methods is also important.
Many of these elements are already aligned in Brazil, and the country can also leverage the strong high-level support for this stream of work.
In order for health accounts results to more efficiently inform policy making, Brazil should strengthen its dissemination strategy and develop a dissemination plan, including a suite of outputs for different users to maximise outreach and impact. This should range from media releases and accompanying data tables for a broad audience to more in-depth analyses for informed users. The organisation of high-level events to support the dissemination of results and to put them in wider health policy context raises the profile of health accounts data and increases the likelihood that the data is actually used to inform policy making. The publication of metadata and methodological reports can add to the legitimacy of the work and can be a way to promote this work among the research community.
The publication of the Brazilian national health accounts manual and the dissemination of the new 2022 results are important milestones towards a full institutionalisation of health accounts in Brazil but there remains scope to go further. First, to increase the relevance and potential use of the results, there should be efforts to improve the timeliness of the estimates. Then there is more untapped potential of analysis for Brazil and efforts should be made to explore an expansion of the current health accounts implementation status. This includes an allocation of health spending to health providers, the identification of revenue sources of the various financing schemes and investigating the feasibility to break down health spending by regions or states.
Putting Brazilian health spending data into an international context
Copy link to 5. Putting Brazilian health spending data into an international context
Abstract
The new approach to construct and institutionalise health accounts in Brazil has resulted in new health spending estimates. In this chapter, the results are put into an international context and compared with health spending in OECD countries. This comparison highlights that, health spending in Brazil is around the OECD average. when measured as a share of economic output, but much lower when measured on a per capita basis. It emphasises that the share of spending from government and compulsory insurance schemes in overall health spending is below 50% in Brazil, much lower than the OECD average. Conversely, the proportion of private insurance spending in total health spending in Brazil is higher than in all OECD countries.
5.1. How does health spending in Brazil compare internationally?
Copy link to 5.1. How does health spending in Brazil compare internationally?
Brazil has produced health spending data for the year 2022 according to the new approach outlined in Chapter 4. In this chapter, some of the high-level results are put into a wider international context.
Compared to OECD countries, per capita spending on health in Brazil is low (Figure 5.1). In 2022, Brazil spent around 1 700 USD (adjusted for difference in price levels) on health per capita, less than one‑third of the OECD average (USD 5 300). The level of health spending estimated for Brazil is similar to that seen in Mexico (USD 1 400), Colombia (USD 1 630) or Costa Rica (USD 1 770) but below that in Chile (USD 3 190) – the Latin American member countries of the OECD. At the other end of scale, the United States (USD 12 740), Switzerland (USD 8 910), Norway (USD 8 640) and Germany (USD 8 540) spent at least five times more on health than Brazil on a per capita basis.
When putting health expenditure into the context of overall economic output, Brazil is around the OECD average (Figure 5.2). In 2022, the consumption of healthcare goods and services accounted for 9.4% of the Gross Domestic Product (GDP) in Brazil. This is on a similar level to the OECD average (9.2%) and above most OECD members in Latin America. In Mexico (5.7%), Costa Rica (7.2%) and Colombia (7.6%), a considerably smaller proportion of economic wealth was dedicated to healthcare. However, at 10%, Chile allocated a larger share of its GDP to health than Brazil.
Comparing the composition of health financing reveals some specificities of financing arrangements in Brazil (Figure 5.3). While across the OECD, around ¾ of overall health spending are typically financed by government or compulsory insurance schemes, this share was only 45% in Brazil in 2022. This was much lower than in any other OECD country, including those in Latin America. Closest to Brazil were Mexico and Chile, where 52% and 57% of all health spending, respectively, was borne by government schemes or compulsory health insurance. In Costa Rica (72%) and Colombia (77%), public payers played a much bigger role in health financing than in Brazil.
Another exceptional feature of the Brazilian health system is the importance of voluntary health insurance. In 2022, around 27% of all health expenditure was financed via duplicate coverage from health plans regulated by ANS. This was much higher than across the OECD as a whole (5%) and more than double the share seen in Slovenia (13%) and Canada (12%) – the two OECD countries where voluntary health insurance plays a comparatively large role. In all Latin American OECD countries, voluntary health insurance financed less than 10% of all healthcare costs.
On the other hand, the share of health spending that was borne directly by households was lower in Brazil (27%) than in a good number of OECD countries. In Mexico, Chile, Greece, Latvia and Lithuania, more than 30% of all health spending was out-of-pocket. However, across the OECD this share stood at only 19%. Households are best protected against direct costs for treatment or medical goods in Luxembourg, France and the Netherlands where they cover only 10% of the entire healthcare bill.
When analysing the composition of health spending by types of goods and services, Brazil stands out with its relatively high proportion of spending allocated to outpatient care (Figure 5.4). In 2022, nearly 40% of all financial resources in the Brazilian health sector were consumed by outpatient services, such as primary healthcare or specialist visits. By comparison, the OECD average stood at only 32%. The comparably high importance of outpatient care in Brazil can be partly explained by a very intensive utilisation of outpatient laboratory and imaging services; each of these services account for 6% of total health spending – a much higher proportion than in any OECD country. Brazil dedicated around 19% of its financial resources to “basic healthcare services” – defined as general outpatient care, dental care, home-based curative care and preventive activities, a share similar to the OECD average (this is a proxy measure used to compare spending on primary healthcare internationally). Conversely, Brazil only dedicated 22% of overall health expenditure to inpatient care in hospitals. This was below the OECD average (28%) and lower than the shares seen in a number of more hospital-centred European health systems.
Medical goods (which mainly refers to pharmaceuticals) accounted for 22% of total health expenditure in Brazil in 2022, slightly above the OECD average. Given its relatively young population and a less formalised long-term care sector, the low share of long-term care spending in Brazil (3%) is not surprising. Spending on collective services, referring to preventive care and health system administration, represented 14% of all health spending. Both the shares allocated to prevention and public health but also to administrative services were above the OECD average. Interestingly, Brazil allocated more resources on administration (6%) than most other OECD countries. This may be explained by the very complex structure of SUS, which requires a lot of co‑ordination and resource management across the three different levels of government but also by management costs of the large private health sector in the country.
Comprehensive health accounts data submissions by countries also allow to assess the generosity of the publicly financed benefit packages by looking into the financing composition of spending on individual healthcare goods and services. As seen in Figure 5.4, in 2022, only 45% of overall health spending in Brazil was financed by SUS or the compulsory insurances schemes for various groups of civil servants. However, not all healthcare services have the same level of financial protection (Figure 5.5). Nearly two‑thirds of spending on services of outpatient medical care were covered by these public schemes in Brazil; for general outpatient curative care, this share stood at 85%. On the other hand, only around a third of all dental care spending was covered by SUS or compulsory insurance. While dental coverage is on a par with many OECD countries the overall share of health spending on dental care is relatively low, possibly explained by higher levels of unmet need. The item where the largest discrepancy in coverage between Brazil and OECD countries exists is pharmaceuticals. In Brazil, only 9% of the total (retail) pharmaceutical bill is covered by public schemes, which means that households must cover a substantial part directly out-of-pocket. This is much lower than in any OECD country, where frequently more than half of pharmaceutical costs are borne by public purchasers. This indicates that public pharmaceutical coverage in Brazil may be inadequate or less effective than desired. High out-of-pocket spending on pharmaceuticals is typically a key driver for households experiencing financial hardship due to high healthcare costs (WHO, 2023[1]).
While Brazil has so far only calculated data for 2022 according to the new methodology, it may be interesting to analyse how this compares to the health spending data produced previously as part of the first pilot implementations. Overall, the health spending level reported for 2022 is roughly in line with previous data (Figure 5.6), although caution should be exercised to not overinterpret the trend as possible methodological breaks can be difficult to quantify. Common with many OECD countries, including Costa Rica and Colombia, health spending as a share of GDP dropped in Brazil in 2022. This followed record levels reached in either 2020 or 2021 when substantial resources were mobilised to fight the spread of SARS-Cov‑2 virus, to improve pandemic preparedness and to treat COVID‑19 patients.
5.2. How does the scope of data reporting in Brazil compare internationally?
Copy link to 5.2. How does the scope of data reporting in Brazil compare internationally?
Overall, the new Brazilian health spending data submission for financing schemes and services for year 2022 is very comprehensive and appears to be in line with international reporting standards. Other than the missing estimates for spending by non-profit financing schemes (HF22) and enterprise financing schemes (HF23) –which might be presumed to play a limited role in Brazil – and the incomplete reporting of spending for the public schemes of civil servants and armed forces – there are no major reporting gaps that would impede the international comparability of the data. As has been outlined in the previous chapter, further refinements to the methodologies and calculations of some of the spending components could help to further improve the overall comparability (see Section 4.4).
In some areas, the level of detail in the Brazilian data actually goes beyond the levels of reporting possible in many OECD countries.
- For example, Brazil can separately identify the subcategories of general outpatient care, dental care and specialist outpatient care, a prerequisite for international estimates of spending on primary healthcare -a key variable to measure health system performance. A number of OECD countries, for example Italy or Portugal, are currently not in a position to do this.
- Brazil is able to report spending on (health) long-term care for all subcategories and all financing schemes. Again, a good number of OECD countries, such as Greece, the Slovak Republic or Iceland face some reporting challenges in this domain. Even if issues around the provision of long-term care in a financially sustainable way do not currently appear to be urgent for Brazil, it can be expected that the demographic transition will eventually put this topic higher on the policy agenda. Having comprehensive data on long-term care spending that would allow to monitor this spending trend over time is a key asset to inform this debate.
- Brazil provides a very detailed breakdown of preventive spending. Some OECD countries, such as Ireland or the Netherlands, are not in a position to provide spending on prevention services at this level of detail. Detailed reporting enables monitoring of resource allocation to the various preventive and public health areas. This can also highlight the prioritisation of target groups or risk factors for public health interventions.
- Brazil is able to comprehensively report spending on (retail) pharmaceuticals, including what is financed by SUS and what is borne directly out-of-pocket by households. This puts the country in an advantageous position compared to a number of Latin American OECD countries. Costa Rica and Mexico, for example, are currently unable to report pharmaceutical spending by government schemes. In these countries, these transactions are largely (or fully) allocated to curative care spending since dispensing to outpatients occurs in hospital pharmacies – and this cannot be distinguished from other hospital activity.
- Finally, the share of health spending that cannot be allocated to services in Brazil is only around 1% of total health spending. While, ideally, all spending can be categorised into the existing classification to not impede any functional analysis, the share in Brazil is considerably lower than in Chile (24%).
In summary, the refined calculation methodology of the Brazil health accounts provides a solid foundation for meaningful international comparisons. Moving forward, Brazil should explore options to revise previous years using the same methodology to provide a longer time series of health spending data (and avoid any possible methodological breaks). Moreover, producing data for the dimensions of health providers and revenues of financing schemes (see Chapter 6) would further increase the analytical power of the Brazilian health accounts. Finally, Brazil could benefit from the timely availability of many of its data sources to also accelerate the production of annual health accounts. In line with international standards, it should be technically possible for Brazil to produce detailed health spending data within 15 months and high-level preliminary estimates six months after the end of the reporting year. Increased timeliness of health accounts results would greatly increase its relevance.