Ministère des solidarités et de la Santé - DREES (Direction de la recherche, des études, de l’évaluation et des statistiques)

Ministère des solidarités et de la Santé - DREES (Direction de la recherche, des études, de l’évaluation et des statistiques) - BACS (Bureau de l'analyse des comptes sociaux)

Directorate of Research Analysis, Studies, Evaluation and Statistics (DREES)
18, Place des cinq martyrs du lycée Buffon, 75014 Paris, FRANCE

Health care expenditure quantifies the economic resources dedicated to health functions, excluding capital investment. Healthcare expenditure concerns itself primarily with healthcare goods and services that are consumed by resident units, irrespective of where that consumption takes place (it may be in the rest of the world) or who is paying for it. As such, exports of healthcare goods and services (to non-resident units) are excluded, whereas imports of healthcare goods and services for final use are included.
Health care expenditure data provide information on expenditure in the functionally defined area of health distinct by provider category (e.g. hospitals, general practitioners), function category (e.g. services of curative care, rehabilitative care, clinical laboratory, patient transport, prescribed medicines) and financing scheme (e.g. social security, private insurance company, household). For the collection of the data on health care expenditure the System of Health Accounts (SHA) and its related set of International Classification for the Health Accounts (ICHA) is used. SHA sets out an integrated system of comprehensive and internationally comparable accounts and provides a uniform framework of basic accounting rules and a set of standard tables for reporting health expenditure data. The System of Health Accounts - SHA 2011 is a statistical reference manual giving a comprehensive description of the financial flows in health care.

It provides a set of revised classifications of health care functions, providers of health care goods and services and financing schemes. The SHA is currently used as a basis for a joint data collection by OECD, Eurostat and WHO on health care expenditure. The manual sets out in more detail the boundaries, the definitions and the concepts of health accounting – responding to health care systems around the globe with very different organisational and financing arrangements.

Accounting period: Health expenditure and financing data pertain to the calendar year (1 January to 31 December).
The time coverage of this Quality report is 2014 to 2016 reference years.

Healthcare expenditure is recorded in relation to the international classification for health accounts (ICHA) defining:

• healthcare expenditure by financing schemes (ICHA-HF) — which classifies the types of financing arrangements through which people obtain health services; health care financing schemes include direct payments by households for services and goods and third-party financing arrangements;
• healthcare expenditure by function (ICHA-HC) — which details the split in healthcare expenditure following the purpose of healthcare activities — such as, curative care, rehabilitative care, long-term care, or preventive care;
• healthcare expenditure by provider (ICHA-HP) — which classifies units contributing to the provision of healthcare goods and services — such as hospitals, residential facilities, ambulatory health care services, ancillary services or retailers of medical goods.

1. Household individual consumption on health, including the collective consumption with two exceptions:
i. Occupational health care (intermediate consumption within establishments) minus an estimated share of occupational health in health providers’ and other medical industries net administration;
ii.“Remunerated” unpaid household production in the form of transfer payments (social benefits in cash) for home care of sick, disabled and elderly persons provided by family members.
SHA 2011 Manual recommends following the standard System of National Account (SNA) rules for drawing the production boundary of health care services, albeit with two notable exceptions:
- Occupational health care is included in the national totals of health care spending. In SNA, this item is recorded as ancillary services and part of intermediate production of enterprises and
- Part of the cash transfers to private households for care givers of home care for the sick and disabled are treated as the paid household production of health care.
2. Health care financing schemes: HF1 Government schemes and compulsory contributory health care financing schemes; HF2 -voluntary health care payment schemes;
HF3 - Household out-of-pocket payment; HF4 - rest of the world financing schemes.
3. NACE rev. 2, section Q, human health and social work activities.

SHA concept is the consumption of health care goods and services.
Health care statistics describe the process of providing and financing health care in countries by referring to health care goods and services, its providers and financing. For the collection of the data on health care expenditure the System of Health Accounts (SHA) and its related set of International Classification for the Health Accounts (ICHA) is used. SHA is a tri-axial system in which the financing, provision and consumption dimensions are covered by the ICHA (International Classification for Health Accounts): Health Care Functions (HC), Health Care Providers (HP), Health Care Financing Schemes (HF).
Data are presented in 3 summary (one-dimensional) tables and 3 cross-classification tables (2-dimensional tables).

Summary tables provide data on:

• Current expenditure by provider (ICHA-HP)
• Current expenditure by function (ICHA-HC)
• Current expenditure by financing scheme (ICHA-HF)

Cross-classification tables refer to:

• HC x HP: Health care expenditure by function and provider: data on which type of health care goods and services are supplied by which health care provider;
• HC x HF: Health care expenditure by function and by financing scheme: data on how are the different types of services and goods financed;
• HP x HF: Health care expenditure by provider and by financing scheme: data on from which health care provider and under which particular financing scheme are the services and goods purchased.

Commission Regulation 2015/359 concerns the collection of data on "current expenditure on healthcare" which is defined as the "final consumption expenditure of resident units on health care goods and services".
There is a very close relationship between the concept of "final consumption expenditure" as defined in the System of Health Accounts (SHA) and in National Account and, as a consequence, also between the underlying economic transactions as recorded in the two accounting frameworks.
In National Accounts there are two types of statistical units: institutional units and local kind-of-activity units (KAU). A local KAU groups all the parts of an institutional unit in its capacity as producer which are located in a single site. A local KAU belongs to one and only one institutional unit.
SHA uses the same two types of units for data compilation.
Local KAUs operating as providers of healthcare goods and services to resident units are statistical units in SHA.
Also transactions by institutional units are recorded in SHA, in which framework institutional units are also referred to as "financing agents". More precisely, SHA financing agents are institutional units that manage one or more financing schemes. The transactions are executed by the financing agents, according to the rules of the financing schemes.
Financing agents serve as key statistical units in producing national health accounts. While financing schemes are the key units for analysing how the consumption of health care goods and services is financed, the data concerning the relevant transactions are collected either from the financing agents that operate the different financing schemes or from the providers.
The concept of "healthcare financing schemes" in SHA is an application and extension of the concept of "social protection schemes" defined by the European System of Social PROtection Statistics (ESSPROS):  "a distinct body of rules, supported by one or more institutional units, governing the provision of social protection benefits and their financing ...". The social protection scheme is the statistical unit in ESSPROS. It is an analytical unit that allows describing the complete structure of the social protection financing system:  expenditure and receipts.
According to SHA Manual 2011, "the key concepts for describing the structure of the health care financing system are based on measuring: (a) the expenditure of health care financing schemes, under which goods and services are purchased directly from health care providers, on the one hand, and (b) the types of revenues of health care financing schemes, on the other hand.  
Commission Regulation 2015/359 limits its scope to the collection of data on the expenditure of health care financing schemes.

SHA focuses on the consumption of health care goods and services by the resident population irrespective of where this takes place. This implies the inclusion of imports (from non-resident providers) and the exclusion of exports (health care goods and services provided to non-residents).

Nonetheless, this assumption is not perfectly verified (see 6.3.1).

The data aims at providing a complete overview of expenditure on health care goods and services consumption of services and goods by the resident population on the national territory of a country.

Data are available for 2006 to 2017 reference years.

Not applicable.

Several data sources are used (as of data notification in March 2018):

-          Surveys/census: 0

-          Public administrative records: 4

-          Financial reports: 0

-          Other: 5

Public administrative records

Other

Annual.

Data are collected through the joint health accounts questionnaire (JHAQ) that countries submit to Eurostat during the annual data collection exercise. There is a voluntary deadline to send the JHAQ questionnaire for the calendar year T by the 31st of March T+2. The joint health accounts questionnaire (JHAQ) is coordinated in agreement with the World Health Organisation (WHO) and the Organization of Economic Co-operation and Development (OECD). These three international organisations are known collectively as the International Health Accounts Team (IHAT). Countries submit data to Eurostat on the basis of Commission Regulation (EU) 2015/359 of 4 March 2015 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council as regards statistics on healthcare expenditure and financing.

The 2018 JHAQ includes a number of features which allow national data correspondents to perform various quality checks before submitting the data. The embedded programmes allow the verification of:

1- Consistency of the data between tables,
This step checks if the marginal totals reported in each table of the JHAQ are consistent across all tables. For example, for each function (HC), the total across all financing schemes (HF) in the HCxHF table has to be equal to the total across all providers (HP) in the HCxHP table, i.e. the values in the column “All HF” in the HCxHF table have to be equal to the values in the column “All HP” in the HCxHP table. Any detected differences are flagged up in the corresponding row or column in the relevant tables and all inconsistencies are listed in the “Report” worksheet by variable code together with the amount by which the respective variable differs between the two compared tables. A positive value indicates that the first listed table has a higher value for the same variable, and vice versa.

2- Consistency of the data within tables,
Any detected inconsistencies are listed by variable code together with an indication of which total is not equal to the sum of its subcomponents as well as the numerical difference. A positive figure indicates that the total is greater than the reported sub-components, and vice versa.

• The presence of negative values,

Entries in the tables cannot be negative as they refer to the consumption of goods and services.
If an individual data table is checked for internal consistency, the negative values check is performed for the relevant table and then any negative values are highlighted red and crossed out.

• The presence of atypical entries,

The atypical entries check provides information whether the data tables contain values in cells which are – if at all – only reported by very few countries and are thus atypical for health accounting.
If an individual data table is checked for internal consistency, the atypical entries check is performed for the relevant table. In the data tables, any cell containing an atypical entry will be highlighted for national data providers.
Should any atypical entries be identified, compilers should scrutinize in detail the transactions that led to entries in those cells and assess whether the accounting rules of SHA have been correctly applied. If they come to the conclusion that the transactions are recorded in the correct categories of the ICHA classifications, then the corresponding atypical entries represent unique – and correctly accounted for – features of the country’s health system. In this case a short description of the nature of the transactions should be included in the accompanying Metadata file under “II.3. Atypical entries”.
If, on the other hand, compilers come to the conclusion that the transactions are not recorded correctly, then they need to make adjustments in the concerned tables. In case the transactions recorded in a cell do not belong to the boundaries of SHA (e.g. they refer to intermediate consumption) the value of the respective cell should be deleted (and all cells that are affected by this change adjusted accordingly). In case the transactions are misreported and another category of the ICHA classification is more appropriate, the value of the cell should be transferred to the correct cell of the table.

3- The growth rates against the previous year and the magnitude of revisions as compared to previously submitted data. Results are grouped into three different categories:

• Breaks in series (the current questionnaire shows no data for an item that is not null in the other file)
• Newly reported (the current questionnaire contains data for an item that is empty in the other file)
• Differences (all other types of differences)

SHA data is compiled both by a bottom-up approach as well as by a top-down approach, depending on the data source. Compilation is done by financing schemes and by different health care functions/task areas. The results of the several calculations are then aggregated.

To gain the differentiation between the different SHA-dimensions (especially HC and HP) quotas and pro-rating and utilisation keys are applied on some spending items. For some spending items it is necessary to extra-/intrapolate data as there is no up-to-date data available or data is missing for certain years. For some other spending items, estimation methods have to be applied.

Several methods are normally used for estimations:

• Balancing item/Residual method: For example, if data are available from the financing side, which permit accurate estimation of the flows to a provider or function, then an acceptable estimation method is to subtract these expenditure flows from the total revenues, and derive the expenditure flows from the unmeasured financing scheme as a residual.
• Pro-rating/Utilisation key: Typically in the absence of direct spending data, a utilisation key linked to the proportion of resources used can be constructed in order to distribute e.g. aggregate provider spending across functions. For every key a fraction of total utilisation within the cost-unit is assigned: fractions in the key must add up to 100% of all care delivered by the cost-unit. Examples of utilisation keys are admissions, bed-days, contacts, staffing, etc.
• Interpolation/Extrapolation: In the absence of data for the period in question, missing values can be estimated using known data points.
• Or other.

Not applicable.

Authorities responsible for SHA data collection are working to ensure that the statistical practices used to compile national health accounts are in compliance with SHA methodological requirements and that good practices in the field are being followed, according to the methodology underlined in the SHA 2011 Manual and European Statistics Code of Practice respecting professional independence of the statistical authorities. Procedures are in place to plan and monitor the quality of the health care expenditure statistical production process.

Not available.

Users are: French Ministry of Health, academics, health organization, international organization etc.

There is no satisfaction surveys or consultations to determine user satisfaction on SHA data. Nonetheless, French health national accounts are annualy presented in a Committee ("Commission des comptes de la santé") gathering about 50 different stakeholders from all part - Ministry of Health, public authorities, health professionals, academics, social partners, industries, etc. - with many very positive feedbacks. Press relay is also very good.

See 5.3.1.

Information concerning total pharmaceutical expenditure is not available: SHA is based on French health accounts (comptes nationaux de la santé or CNS). In CNS, we are not able to isolate medication use at hospital (due to a highly aggregate data source).

No comment.

The only sampling technique is used to compute over-the-counter medicines, which represents around 1% of total health expenditure. About 3 different sampling techniques have been tested, which provided very similar results. Quality assessments indicated an accuracy of around 0,1%.

See 6.2.

See below.

SHA data are base on French health national accounts which are centered on consumption in the French territory. This means that the consumption by French residents abroad is excluded; and conversely that the consumption by non-residents on the French territory is included. Customs data are not sufficiently detailed on this specific topic to provide an accurate estimation to be used in SHA, but qualitative assessments indicate that the amounts strictly concerning health consumption should be very limited, which also limits the coverage error.

See 6.3.1.

Not applicable: there is no overlapping between the unique survey (on over-the-counter medicines) and all the administrative data.

There is almost no measurement error: the few ones are corrected through a careful automatized process (for instance: typing error in the data source reporting euros instead of million of euros in administrative data).

There is a small non-response error regarding complementary schemes' data: the coverage is higher than 97% in market shares. This non-response is corrected using sampling techniques.

See 6.3.3.

See 6.3.3.

The main processing error is reported in section 6.3.2.

See 6.3.4.

Not applicable: there is no modeling in the computation of French health accounts.

Not applicable.

Data are revised each year, according to our methological improvements. Concerning public hospital, data can also be revised because of national accounts base year change. Data are revised from year 2006 on, so that no breaks in time series will be introduced.

Futher methodological improvements have been made:

-In 2017, over the counter medicines consumption was revised. Now the estimates are based on more detailed data :  pharmacies' turnovers (including taxes - retail price) [IQVIA data].
- Collective contracts of complementary insurances have been reclassified in HF.1.2.2, because in most sectors, since last 1st Jannuary 2016, employers are obliged to subscribe such contracts for their employees. Previously , these contracts were reported in HC.2.1. The reclasssification was made only for the year 2016.
- The evaluation of HC.6.2, HC.6.3, HC.6.4 has been changed, leading to slight revisions of the results.
- The evaluation of HC.1.2 has been revised for the year 2014 and 2015 thanks to better data sources.
- The evaluation of HC.2.1, HC.2.2, HC.4.1, HC.4.2, HC.4.3, HC.5.1 has been revised due to the revision of the French national health accounts for the period 2006 -2015.

Health accounts are revised each year, according to methological improvements and also thanks to more recent data. Data are revised from year 2006 on, so that no breaks in time series will be introduced.

In this edition, the methodology for estimating over-the-counter medicines has been improved thanks to more detailed data regarding pharmacies’ turnovers. This translates into a slightly upward revision in 2006 and downward revisions between 2007 and 2016 (which reaches a maximum of 1.3 billion euros in 2016). Concerning public hospital, data can also be revised because of national accounts base year change or updated data by the French national statistical institute (INSEE): in this edition, public hospital data has been revised by - 430 million euros in 2016.

Member States are required to transmit their data to Eurostat in compliance with the Commission Regulation 359/2015 transmission deadlines.
Data and reference metadata for the reference year T should be transmitted to Eurostat by 30 April T+2.

Not applicable.

Definite figures for year T-3 are published in September of the year T.

See 7.2.1.

There were no deviations from deadlines in the reference period.

Not applicable at national level.

Not applicable.

There are no breaks in 2006-2017 series.

There are no breaks in 2006-2017 series.

SHA and ESSPROS are based on different concepts as SHA is based on final consumption whereas ESSPROS is based on total expenditure.

Also, e.g. in the domain of LTC SHA core variables are only focusing on health-related LTC whereas ESSPROS takes into account also the social aspects of LTC.

A full coherence between these different approaches is therefore not feasible.

Not applicable.

SHA is based on final consumption. The concept of "consumption" comes from national accounts SEC10.

No comment.

DREES website

https://drees.solidarites-sante.gouv.fr/etudes-et-statistiques/publications/panoramas-de-la-drees/article/les-depenses-de-sante-en-2018-resultats-des-comptes-de-la-sante-edition-2019

DREES website

https://drees.solidarites-sante.gouv.fr/etudes-et-statistiques/publications/panoramas-de-la-drees/article/les-depenses-de-sante-en-2018-resultats-des-comptes-de-la-sante-edition-2019

DREES website

https://drees.solidarites-sante.gouv.fr/etudes-et-statistiques/publications/panoramas-de-la-drees/article/les-depenses-de-sante-en-2018-resultats-des-comptes-de-la-sante-edition-2019

DREES website

https://drees.solidarites-sante.gouv.fr/etudes-et-statistiques/publications/panoramas-de-la-drees/article/les-depenses-de-sante-en-2018-resultats-des-comptes-de-la-sante-edition-2019

Not applicable.

Not applicable.

The results of the French national health accounts on which SHA relies are presented in « Les dépenses de santé en 2017, résultats des comptes de la santé – édition 2018 », Panoramas de la Drees-Santé, septembre 2018 (in French only).

See in particular annex 1, annex 2 and annex 5 for more details about the methodology of French national health accounts.

The results of the French national health accounts on which SHA relies are presented in « Les dépenses de santé en 2017, résultats des comptes de la santé – édition 2018 », Panoramas de la Drees-Santé, septembre 2018 (in French only).

See in particular annex 1, annex 2 and annex 5 for more details about the methodology of French national health accounts.

Not available.

Not available.

The Regulation (EC) No 223/2009 on European statistics (recitals 23-27, 31-32 and Articles 20-26) applies.

Further methodological improvements have been conducted this year, leading to revise some estimates.