Accidents at work (ESAW, 2008 onwards) (hsw_acc_work)

Reference Metadata in Euro SDMX Metadata Structure (ESMS)

Compiling agency: Eurostat, the statistical office of the European Union

Eurostat metadata
Reference metadata
1. Contact
2. Metadata update
3. Statistical presentation
4. Unit of measure
5. Reference Period
6. Institutional Mandate
7. Confidentiality
8. Release policy
9. Frequency of dissemination
10. Accessibility and clarity
11. Quality management
12. Relevance
13. Accuracy
14. Timeliness and punctuality
15. Coherence and comparability
16. Cost and Burden
17. Data revision
18. Statistical processing
19. Comment
Related Metadata
Annexes (including footnotes)
National metadata

For any question on data and metadata, please contact: EUROPEAN STATISTICAL DATA SUPPORT


1. Contact Top
1.1. Contact organisation

Eurostat, the statistical office of the European Union

1.2. Contact organisation unit

Unit ESTAT.F5: Education, health and social protection

1.5. Contact mail address

European Commission, Eurostat
2920 Luxembourg

2. Metadata update Top
2.1. Metadata last certified 31/08/2017
2.2. Metadata last posted 31/08/2017
2.3. Metadata last update 31/08/2017

3. Statistical presentation Top
3.1. Data description

Harmonised data on accidents at work are collected in the framework of the administrative data collection 'European Statistics on Accidents at Work (ESAW)', on the basis of a methodology developed first in 1990.

An accident at work is defined as 'a discrete occurrence in the course of work which leads to physical or mental harm'. The data include only fatal and non-fatal accidents involving more than 3 calendar days of absence from work. If the accident does not lead to the death of the victim it is called a 'non-fatal' (or 'serious') accident.

A fatal accident at work is defined as an accident which leads to the death of a victim within one year of the accident.

The variables collected on accidents at work include:

  • Economic activity of the employer and size of the enterprise
  • Employment status, occupation, age, sex and nationality of victim
  • Geographical location, date and time of the accident
  • Type of injury, body part injured and the severity of the accident (number of full calendar days during which the victim is unfit for work excluding the day of the accident, permanent incapacity or death within one year of the accident).
  • Variables on causes and circumstances of the accident: workstation, working environment, working process, specific physical activity, material agent of the specific physical activity, deviation and material agent of deviation, contact - mode of injury and material agent of contact - mode of injury.

The national ESAW sources are the declarations of accidents at work, either to the accident insurance of the national social security system, a private insurance for accidents at work or to other relevant national authorities (labour inspection etc.). As an exception, accident data for the Netherlands are based on survey data.

On the Eurostat website, ESAW data are disseminated in two sections: 'Main Indicators' and 'Details by economic sector (NACE Rev2, 2008 onwards)'. Depending on the table, data are broken down by: economic activity (NACE 'main sectors' (1 digit code) or more detailed NACE divisions (2 digit codes)); the occupation of the victim (ISCO-08 code); country; severity of the accident, sex, age, employment status, size of the enterprise, body part injured and type of injury.

The data is presented in form of numbers, percentages, incidence rates and standardised incidence rates of non-fatal and fatal accidents at work, either for EU aggregates, countries or certain breakdowns by dimensions such as age, sex etc.

  • Numbers correspond to a simple count of all non-fatal and fatal accidents for the entirety or certain breakdowns of the data;
  • Percentages represent shares of breakdowns;
  • The incidence rate of non-fatal or fatal accidents at work is the number of serious or fatal accidents per 100,000 persons in employment;
  • The standardised incidence rates of non-fatal or fatal accidents at work aim to eliminate differences in the structures of countries' economies (see section 18.6 Adjustment for more details).

The incidence rate indicates the relative importance of non-fatal or fatal accidents at work in the working population. For both types of accidents at work the numerator is the number of accidents that occurred during the year. The denominator is the reference population (i.e. the number of persons in employment) expressed in 100,000 persons.

The reference population (or number of persons in employment) related to the national ESAW reporting system is provided by the Member States, either from administrative sources related to accidents at work or from the EU Labour Force Survey (LFS).

3.2. Classification system

The following classifications are used in ESAW:

  • NACE Rev. 2: 2nd revision of the Statistical Classification of Economic Activities in the European Community managed by Eurostat. For ESAW data NACE Rev. 2 is used from reference year 2008 onwards;
  • ISCO-08: International Standard Classification of Occupations managed by the International Labour Organisation (ILO) (revision of 2008). ISCO-08 should be used from reference year 2011 onwards (not all countries use it yet);
  • NUTS 2013: Nomenclature of territorial units for statistics. The current NUTS 2013 classification is valid from 1 January 2015, though many datasets will still include other NUTS versions.
  • ICSE-93: International Classification by Status in Employment (revision of 1993)

For full details on these classifications, please see our nomenclature server Ramon.

3.3. Coverage - sector

All economic sectors according to NACE Rev. 2 should in principle be covered from reference year 2013 onwards. Before the reference year 2013, there was no obligation to cover sectors outside the so-called 'common branches' A and C – N. However, most Member States covered those non-common sectors C and O – U before on a voluntary basis.

The NACE sectors are (all branches, at one digit level):
















P          EDUCATION






Data delivery for sectors T and U is voluntary

Some sectors and professions are subject to confidentiality rules. According to the ESAW implementing Regulation 349/2011 (annex II), Member States deliver the following NACE Rev. 2 divisions of sector O on a voluntary basis only:

84.22   Defence activities

84.23   Justice and judicial activities

84.24   Public order and safety activities

84.25   Fire services activities

Section '3.6. Statistical population' below gives information about professions (occupations) subject to confidentiality exceptions.

On the Eurostat website, the data tables published in section 'Main indicators' (hsw_mi) cover the 13 common branches A, C-N only.

In the section 'Details by economic activity (NACE Rev. 2, 2008 onwards) (hsw_n2) some data are presented at 1 and 2 digit level of NACE Rev.2 classification.

The ESAW implementing Regulation (see section 5.2 Quality assessment) requires that Member States send data on accidents of 'employees'. Data on accidents of self-employed, family workers and other employment types are voluntary.

The metadata annex to this file contains, besides other information, a detailed overview about which Member State covers which economic sectors and employment types.

3.4. Statistical concepts and definitions

European Statistics on Accidents at Work (ESAW) are based on case-by-case data for accidents at work resulting in more than 3 days' absence from work, permanent incapacity or death of the victim.

An accident at work is 'a discrete occurrence in the course of work which leads to physical or mental harm'.

  • This includes all accidents in the course of work, whether they happen inside or outside the premises of the employer, on the premises of another employer, in public places or during transport (including road traffic accidents or accidents in any other mean of transportation) and at home (such as during teleworking). It also includes cases of acute poisoning and wilful acts of other persons;
  • It excludes: accidents on the way to or from work (commuting accidents); deliberate self-inflicted injuries; occurrences caused solely by a medical condition (such as heart attack or stroke) that occurred during work, i.e. which were not (at least partially) caused by the occupational activity of the victim;; accidents to members of the public, for example family members of a worker who are not working; and occupational diseases.

A fatal accident at work is defined as an accident which leads to the death of a victim within one year of the accident. In practice the notification of an accident as fatal ranges from national registration procedures where the accident is registered as fatal when the victim dies during the same day (Netherlands) or within 30 days after the accident (Germany) to cases where no time limits are laid down (Belgium, Greece, France except for deaths occurring after the recognition of a permanent disability, Italy, Luxemburg, Austria, Sweden and Norway). For the other Member States the time limit is one year, except for Spain where the limit is 1,5 years after the date of the accident.

In a typical fatal accident at work, the death occurs within a few days after the day of the accident and the limitation to the day of the accident would result already in a significant underestimation of such very severe accidents.

The ESAW methodology is in accordance with the ILO (International Labour Office) "Resolution concerning statistics of occupational injuries (resulting from occupational accidents)" adopted by the Sixteenth International Conference of Labour Statisticians in October 1998.

The following 15 phase I and II variables have to be sent by Member States to Eurostat from reference year 2011 onwards on an annual basis:

  • 1. Case number
  • 2. Economic activity of the employer (NACE)
  • 3. Occupation of Victim (ISCO)
  • 4. Age of Victim
  • 5. Sex of Victim
  • 6. Type of Injury
  • 7. Part of Body Injured
  • 8. Geographical Location of the Accident
  • 9. Date of the Accident
  • 10. Time of the Accident (optional)
  • 11. Size of the Enterprise (optional)
  • 12. Nationality of the Victim (optional)
  • 13. Employment Status of the Victim
  • 14. Days Lost (severity)
  • 15. Weight ESAW collection

In addition, three of the following nine phase III variables on 'causes and circumstances of the accident' have to be sent annually to Eurostat from reference year 2013 onwards:

16. - 18. Workstation, Working Environment, Working Process

19. - 20. Specific Physical Activity, Material Agent associated with the of Specific Physical Activity

21. - 22. Deviation, Material Agent associated with the Deviation

23. - 24. Contact - mode of injury, Material Agent associated with the Contact - Mode of injury.

Finally, the weight on Causes and Circumstances has to be sent if the Member State applies an additional sampling for the encoding of the ESAW Phase III variables on causes and circumstances. If not applicable the default value is 1.

The definition of the variables is stated in the Commission Regulation (EU) No 349/2011 and further specified in the ESAW methodology.

3.5. Statistical unit

Microdata are collected for each accident. If a person is a victim of more than one accident during the reference year, several cases are reported (one for each accident).

See 4. Unit of measure for the measurement units used.

3.6. Statistical population

In principle all accidents at work should be covered that fulfil the definition of '3.4. Statistical concepts and definitions'.

In terms of employment types covered, Member States are required to report on 'employees'. The other employment types (self-employed, family members, students and others) are voluntary.

In addition to NACE sector O (see section 3.3. Coverage – sector), some professions (occupations) are also subject to national confidentiality rules and delivered on a voluntary basis as mentioned in annex II of the ESAW implementing Regulation 349/2011:

For ISCO – 08:

- 0 Armed forces occupations

- 3351 Customs and border inspectors

- 3355 Police inspectors and detectives

- 541 Protective services workers

            a. 5411 Fire-fighters

            b. 5412 Police officers

            c. 5413 Prison guards

            d. 5414 Security guards

            e. 5419 Protective services workers not elsewhere classified

3.7. Reference area

Data are available for all EU-Member States and EFTA countries (Iceland (from 2012), Norway and Switzerland). Croatian and EU-28 data are available from 2010 only.

3.8. Coverage - Time

Data are available since reference year 1994 for all EU-15 Member States, from 1995 also for Norway.

From reference year 2008, data for all EU-27 Member States and EU aggregates are available. Croatian data are included since 2010.

Data for Switzerland was added from 2004 and Icelandic data from 2012.

3.9. Base period

Not applicable.

4. Unit of measure Top

The following measurement units are used in ESAW data:

- Numbers of accidents

- Percentages of accidents (in relation to different totals and breakdowns)

- Incidence rates of accidents: number of accidents per 100,000 workers

- Standardised incidence rates: number of accidents per 100,000 workers adjusted for the relative sizes of economic sectors at EU level (see section 18.6 Adjustment for more details).

5. Reference Period Top

The calendar year during which the accidents were reported to have taken place (reference year).

6. Institutional Mandate Top
6.1. Institutional Mandate - legal acts and other agreements

The implementing Regulation under which Member States report ESAW data to Eurostat is Commission Regulation (EU) No 349/2011 of 11 April 2011 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work, as regards statistics on accidents at work.

The afore mentioned Regulation 349/2011 implements the framework Regulation EC N° 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work.

Commission Decision 2011/231 of 11 April 2011 grants derogations to certain Member States with respect to the transmission of statistics pursuant to Commission Regulation (EU) No 349/2011 of 11 April 2011 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work, as regards statistics on accidents at work.

The current strategy of the EU concerning health and safety at work is outlined in the following EU Commission Communication: EU Strategic Framework on Health and Safety at Work 2014-2020.

The health and safety at work strategy of the EU from 2007 until 2012 was: Community strategy 2007-2012 on health and safety at work.

The health and safety strategy of the EU from 2002 and 2006 was: Commission Communication COM (2002) 118 Final and Council resolution 2002/C 161/01 on a new Community strategy for safety and health at work 2002-2006.

6.2. Institutional Mandate - data sharing

Not applicable.

7. Confidentiality Top
7.1. Confidentiality - policy

Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

Commission Regulation (EU) No 557/2013 of 17 June 2013 implementing Regulation (EC) No 223/2009 of the European Parliament and of the Council on European Statistics as regards access to confidential data for scientific purposes and repealing Commission Regulation (EC) No 831/2002 Text with EEA relevance, 'establishes the conditions under which access to confidential data transmitted to the Commission (Eurostat) may be granted for enabling statistical analyses for scientific purposes, and the rules of cooperation between the Commission (Eurostat) and national statistical authorities in order to facilitate such access' (article 1).

7.2. Confidentiality - data treatment

Microdata of Member States are sent via secure e-Damis application.

There are currently no restrictions applied when publishing aggregated data.

8. Release policy Top
8.1. Release calendar

ESAW data is usually reported and partially released by Member State between January - September of each year (in year N+2, where N is the reference year in which the accidents took place), with a focus on end of June - July (release of most countries' data). There is no specific release calendar for ESAW data (see the Eurostat website for all published ESAW data).
See section 14 Timeliness and punctuality for more details about the timing and punctuality of ESAW data releases.

8.2. Release calendar access

Not available.

8.3. Release policy - user access

In line with the Community legal framework and the European Statistics Code of Practice Eurostat disseminates European statistics on Eurostat's website (see item 10 - 'Accessibility and clarity') respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably. The detailed arrangements are governed by the Eurostat protocol on impartial access to Eurostat data for users.

9. Frequency of dissemination Top


10. Accessibility and clarity Top
10.1. Dissemination format - News release

See Eurostat website for occasional news releases.

10.2. Dissemination format - Publications

See Eurostat website for publications on accidents at work including:

Accidents at work statistics (Statistics Explained article - updated once per year)

8.6% of workers in the EU experienced work-related health problems - Issue number 63/2009

Health and safety at work in Europe (1999-2007)

Work and health in the European Union - A statistical portrait - Data 1994-2002

Accidents at work in the EU 1998-1999

Accidents at work in the EU in 1996

10.3. Dissemination format - online database

See Eurostat website.

10.4. Dissemination format - microdata access

Not applicable.

10.5. Dissemination format - other


10.6. Documentation on methodology

Eurostat (2013): 'European Statistics on Accidents at Work (ESAW) - Summary methodology, 2013 edition'.

Previous editions:

- Eurostat (2012): 'European Statistics on Accidents at Work (ESAW) - Summary methodology, 2012 edition', ISBN 978-92-79-23133-9, ISSN 1977-0375, doi:10.2785/2509, Cat. No KS-RA-12-002-EN-N.

- DG Employment and social affairs and Eurostat (2001): 'European statistics on accidents at work (ESAW) - Methodology - 2001 edition', Health and safety at work series, co-publication with Eurostat.

10.7. Quality management - documentation

Quality reports are planned to be available from May 2016.

11. Quality management Top
11.1. Quality assurance

Quality considerations play a central role with regard to Eurostat corporate management as well in the day-to-day statistical operations.
The European Statistics Code of Practice sets the standard for developing, producing and disseminating European statistics (see for more information). It builds upon a common European Statistical System (ESS).
The Quality Assurance Framework of the European Statistical System (ESS QAF) identifies possible activities, methods and tools that can provide guidance and evidence for the implementation of the Code of Practice when developing, producing and disseminating European statistics.
Eurostat organises about each year a Working Group on European Statistics on Accidents at Work, in which statistics and its metadata are presented. In particular, various quality issues such as data coverage and reporting levels are discussed and related improvements are discussed.
Data validation measures are desribed in concept 3.4 data validation.

11.2. Quality management - assessment

In general, Eurostat aims to continuously improve the quality of ESAW data together with Member States.
ESAW is an administrative data collection which provides information on accidents at work as they are reported by employers in Member States to relevant national state authorities.
The type of data transmitted to Eurostat as well as its timing and other aspects follow strict rules. Eurostat together with Member States systematically check and validate data transmissions according to the following legal acts and further guidelines:
- Commission Regulation (EU) No 349/2011 of 11 April 2011 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work, as regards statistics on accidents at work (text with EEA relevance).
- Framework Regulation EC N° 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work.
- Commission Decision 2011/231 of 11 April 2011 granting derogations to certain Member States with respect to the transmission of statistics pursuant to Commission Regulation (EU) No 349/2011 of 11 April 2011 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work, as regards statistics on accidents at work.
In addition, the ESAW summary methodology (version of 2013), defines further details of the data to be transmitted.

Member States fulfill their obligations in relation to these legal acts and guidelines. This includes all classification systems. Exceptions are mentioned in the other concepts of this file (including 19. Comments).
However, the data sources in Member States are to some degree heterogenous. For example, in some Member States accidents are reported in the framework of accident insurances whereas in other Member States universal social security system apply in case of accidents. This issue is covered in section 6. on Accuracy and reliability of this file.
A particular challenge of ESAW data are the issues of (i) under-reporting of non-fatal accidents in some Member States, (ii) differences in the coverage of self-employed, family workers and other non-employee workers (voluntary data), and (iii) accuracy of reference populations (number of employed persons in economic sectors) and incidence rates (reference populations are the denominators of incidence rates). These issues are currently worked on but are not likely to be solved completely in the short term.
According to the ESAW implementing Regulation No 349/2011, only three out of nine phase III variables need to be reported to Eurostat. Although many Member States report on more than three phase III variables, the average coverage of phase III variables accross all Member States and all variables is currently about 50% only.

12. Relevance Top
12.1. Relevance - User Needs

In line with the European Union legal framework and principle 6 (impartiality and objectivity) of the Code of Practice, Eurostat develops, produces and disseminates European statistics respecting scientific independence and in an objective, professional and transparent manner in which all users are treated equitably (see for more information).

The main users of the ESAW data at EU level are:

  • Other EU organisations concerned with health and safety at work, in particular Directorate-General for Employment, social affairs and inclusion as well as the EU agencies EU-OSHA and EUROFOUND;
  • Member States authorities such as statistical authorities, labour ministries, labour inspectorates;
  • International organisations dealing with health and safety at work such as ILO;
  • Business associations, trade unions and NGOs;
  • Researchers and students;
  • Media.

Eurostat regularly meets other EU and international organisations as well as representatives of relevant Member State authorities, business associations, workers and employer associations. In addition, it communicates with other user groups such as researchers, students and the media via several channels including its user support, website and through events such as conferences and meetings. The needs and comments from these user groups are continuously taken into account when deciding about what statistics are collected and disseminated.
ESAW data users are often interested in information concerning individual economic sectors. Incidence rates are important for comparing the situation in different Member States and economic sectors. While absolute numbers of accidents for individual economic sectors (at NACE Rev.2 at 2-digit or more detailed level) reflect appropriately the accidents reported in Member States, incidence rates (number of accidents per 100,000 workers) depend also on reference populations which are sometimes not very accurate for smaller sectors.

12.2. Relevance - User Satisfaction

Eurostat carries out regular user satisfaction surveys (see including for statistical domains such as 'health'.

Information about the downloads of individual datasets, detailed tables, publications and Statistics Explained articles is regularly compiled and analysed.

All available information about user satisfaction is taken into account when developing data collections.

However, there is currently no dedicated user satisfaction survey on ESAW statistics.

12.3. Completeness

See metadata overview in annex.

13. Accuracy Top
13.1. Accuracy - overall

As an administrative data collection, the data transmitted by Member States to Eurostat are thought to exactly reflect the number of accidents notified to the relevant national authorities. Therefore, ESAW data are considered to have a high level of accuracy in relation to the absolute number of accidents notified in Member States.

Most countries' data are not any more subject to under-coverage. Exception are Greece (data from employees not insured by IKA is missing which may represent as much as 60-70% of the data) and Germany (which does not yet cover data on civil servants but plans to do so from 2019 onwards). The data of the Netherlands and Norways have seen large decreases between reference years 2013 and 2014 which are thought to be due to methodolgoical or organisational issues. In general, under-coverage means that a known part of the economy or workforce of a country is not covered by the data, for example if an economic sector or a certain professional status (e.g. civil servants) is by definition not included in the notification system.
A significant issue for the accuracy of ESAW in the Eastern part of the European Union as well as in Norway is assumed under-reporting of accidents (see other sections in this file for more information). Under-reporting means that certain accidents that should have been reported were in fact not reported, e.g. if enterprises or workers are not aware of the obligation/possibility to notify or if they are afraid of the consequences of notification such as possible state investigations and requirements to invest in health and safety. However, even in those countries which suffer from significant under-reporting it is thought that the majority of fatal accidents are reported which then give a more comparable picture for the EU.

In addition, it is assumed that non-fatal accidents are reported to a higher degree in insurance based systems if they offer significant financial compensation for victims of these accidents. The national ESAW data sources are the notifications of accidents at work, either to national insurance systems for accidents at work (BE, BG, DE, EL, ES, FR, IT, LU, AT, PT, SI, FI and CH) or to relevant national authorities such as labour inspectorates in the framework of a universal social security system (CZ, DK, EE, IE, CY, LV, LT, HU, MT, NL, PL, RO, SK, SE, UK and NO). Incidence rates and standardised incidence rates (number of accidents per 100,000 workers) vary often strongly between the two main types of notification systems, the insurance based and the universal social security based systems as listed before.

Non-fatal accidents (serious accidents) at work:
Incidence rates are often only comparable between Member States of the same notification system (insurance or universal social security). In some Member States, weights are added to adjust the number of accidents, e.g. using data from Labour Force Surveys (LFS).

Fatal accidents at work:
In general, fatal accidents at work are assumed to be of higher accuracy than non-fatal accidents at work as fatal accidents are usually investigated by relevant state authorities.

13.2. Sampling error

The ESAW data collection is mainly based on data from administrative registers, and therefore there is usually no sampling involved.
However, data from Germany and the Netherlands (all variables), Malta and Spain (variables on causes and circumstances) and Switzerland (some variables) and more recently from the UK (all variables) are (partially) based on sampling. Weighting procedures are applied in those countries to correct for sampling errors.

13.3. Non-sampling error

Information about the coverage and non-coverage of economic sectors, employment types and occupations can be found in the metadata annex to this file.

General information about under-reporting errors can be found in other sections of this file. The Czech Republic, Denmark, Estonia, Ireland, Hungary, Italy, Sweden and Norway apply weights to correct for under-reporting.

14. Timeliness and punctuality Top
14.1. Timeliness

The legal requirement for Member States is to send Eurostat ESAW data until 30 June of year N+2, where N is the reference year in which the accidents took place. Thus, data for the referfence year 2014 had to be sent until 30 June 2016.

As soon as one or several Member States have sent their data they are published on the Eurostat website. In some cases this may happen already some months before the legal deadline of June.

14.2. Punctuality

Most countries transmit data before and a few days - weeks after the legal deadline of 30 June of year N+2. Eurostat then checks and validates this data and publishes most of it on its website within a few weeks (in July N+2 or before). A few countries may send data only during July - September of year N+2. All data is usually published until September - October of year N+2.

15. Coherence and comparability Top
15.1. Comparability - geographical

Data on fatal accidents have a high level of comparability between all countries.

Data on non-fatal accidents are considered to be of limited comparability across certain groups of countries. See section 13 on 'accuracy' for further details, in particular concerning the existence of two different types of accident notification systems (insurance based and universal social security system based).

Standardised incidence rates are calculated in order to enable between comparison between countries (some countries have larger high-risk sectors in terms of work accidents, for example concerning certain occupations in transport, construction, manufacturing and agriculture; see section 18.6 Adjustment for more details).

15.2. Comparability - over time

Data on the Eurostat website is divided between datasets with reference year up to 2007 (included), and datasets with reference year from 2008 onwards.

The datasets up to reference year 2007 are restricted to EU-15 Member States, Norway and Switzerland, and uses NACE Rev. 1 classification for economic activities.

The datasets from 2008 onwards include all countries of the previous datasets plus the other EU-28 Member States (Croatia from 2010 only) and Iceland (from 2012 onwards). They use NACE Rev. 2 classification of economic activities.

15.3. Coherence - cross domain

ESAW data may be compared with relevant data from the Labour Force Survey ad-hoc modules on accidents at work and other work-related health problems available from the same main data tree branch 'Health and safety at work (hsw)' for the reference years 2013, 2007 and 1999.

15.4. Coherence - internal

ESAW data on the Eurostat website since 2008 are largely coherent between the different datasets. There are some exceptions if some countries use old classification systems such as for ISCO-88 classification for occupations.

16. Cost and Burden Top

At EU level, the ESAW data collectionis is currently managed by about 1 - 1,5 full-time persons. IT costs for receiving, processing and publishing ESAW data at Eurostat cannot be estimated as resources are shared with all other Eurostat statistical data domains. Costs of administrative overhead, missions and other supporting staff resources have not been assessed but may be similar as for the entire organisation (Eurostat / EU Commission). The costs for ESAW data collection and treatment at national level are not included but each country may report on it in its own quality report availble from this webpage. There are currently no other costs involved in the ESAW data collection.

17. Data revision Top
17.1. Data revision - policy

ESAW data are occasionally revised, such as when a country notifies Eurostat about changes in the data and metadata.

In 2014 - 2015 a revision of reference populations of some countries took place which impacted on some incidence rates (the reference populations is the denominator for incidence rates, the numerator is the numbers of accidents).

17.2. Data revision - practice

The practice for data revisions is similar to the practice of publishing new data: either the country sends revised data to Eurostat, which checks, validates and publishes it in agreement with the responsible national authority, or Eurostat calculates revised data and sends it to the country's authority for agreement prior to publishing.

18. Statistical processing Top
18.1. Source data

Eurostat receives ESAW data from the relevant national authority or insurance system (administrative data sources). The original national data sources are employers' declarations of accidents at work, either to (1) relevant insurance companies, (2) national social security systems or (3) labour inspectorates or similar national authorities. For non-fatal accidents in the Netherlands only survey data are available (a special module in the national labour force survey).

The number of employed persons (reference populations) are provided either by the countries (at NACE 2-digits level), or are taken from the EU Labour Force Survey (when countries are not able to provide their own reference populations).

18.2. Frequency of data collection


18.3. Data collection

ESAW data are extracted from relevant national administrative databases according to ESAW concepts, definitions, variables and classifications (see also section 18.1. 'Source data' above).

18.4. Data validation

Various automatic and manual checks are performed on ESAW data transmitted from Member States to Eurostat in order to identify possible errors and inconsistencies with the ESAW methodology. Countries are asked to check all inconsistencies and corrections and re-send data if necessary.

Validation activities at EU level include:
- completeness of data files transmitted (variable fields)
- correctness of codes for each variable
- comparing aggregated data at EU level with aggregates at national level
- coverage of variables, sectors, employment types and other breakdowns
- checking for differences in reporting levels (under-reporting)
- comparing ESAW reference populations with EU Labour Force Survey (LFS) populations
- consistency of totals and various breakdowns
- use of variable weights to correct for sampling and/or under-reporting
- checking for outliers

18.5. Data compilation

10 out of 31 countries provide weights for non-fatal accidents. Eurostat uses these weights to calculate the number of accidents and all derived indicators.

EU aggregates are calculated on the basis of available data from countries.

Standardised and non-standardised incidence rates are calculated using number of accidents and reference populations (see relevant annex to this file).

Standardised incidence rates are calculated in addition to normal incidence rates in order to eliminate the effect that some countries have larger high-risk sectors than others (such as agriculture, construction or transport). The standarisation method consists in multiplying incidence rates with weights corresponding to the shares of sectoral workforces in the total EU workforce. High-risk sectors are those in which the number of work accidents per 100,000 workers is typically higher. Examples are the sectors transport, construction, manufacturing and agriculture (in particular concerning some occupations within these sectors).

For more details, please refer to the ESAW summary methodology (see annex) and to the annex of this file.

18.6. Adjustment

In general, there are no adjustments done for ESAW data.

ESAW standardised and non-standardised incidence rates are checked to some degree for outliers. In case of unreasonable high or low values of these incidence rates, in particular for certain more detailed breakdowns of data, Eurostat may remove related incidence rates from dissemination and replace the values by appropriate codes.

As ESAW is an annual administrative data collection, there are no seasonal adjustments nor other time series adjustments done.

19. Comment Top

Commission Regulation (EU) 349/2011 specifies in its annex I the year of first transmission for each ESAW variable. In general, the first year of data transmission according to the Regulation was 2013 (for reference year 2011). However, certain deviations from this year apply:

- Data on the economic activity of the employer (NACE Rev., 4 digits) for sectors B and O-S needs to be transmitted only from 2015 onwards. Nevertheless, many countries provided such data already before (see document 'ESAW metadata 2012-13 update 2015-06-05' in the annex to this metadata file) for more details.

- Data on 'time of the accident', 'size of the enterprise' and 'nationality of the victim' are optional to report;

- Data on phase III variables were optional until 2015. Only three out of nine phase III variables need to be provided. However, most countries provide data on all variables. See document 'ESAW metadata (...)' in the annex to this metadata file for more details.

Until June 2016 there were several derogations for the transmission of data to Eurostat (see also section 6.1 Institutional Mandate - legal acts and other agreements) which are mentioned in Commission Decision 2011/231 of 11 April 2011 granting derogations to certain Member States with respect to the transmission of statistics pursuant to Commission Regulation (EU) No 349/2011 of 11 April 2011 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work, as regards statistics on accidents at work. They included (years below relate always to the year of data transmission, not to reference years):

- Economic sectors (NACE)
First delivery of data concerning employees of the public sector (NACE O) for Belgium, Germany (civil servants) and Greece is 2016. Germany does not yet provide data for civil servants nor 4-digit NACE codes. It is planned to do so until June 2019. the inclusion of this sector led to a significant increase in the number of accidents in Belgium.

Severity (days lost)
Dutch, German and Greek data on days lost will only be provided from 2016 onwards. The derogation for the United Kingdom ended already in 2015. Germnay is not yet sending severity data but plans to do so from June 2019 onwards

Road traffic accidents and accidents on board of any other mean of transport
Ireland provides data on road traffic accidents and accidents on any other means of transport from 2016 onwards (for reference year 2014). The derogations for the United Kingdom ended in 2015 for the reference year 2013 (except for air crews and sailors for which the derogation ended in 2016). This also concerns accidents of sector H 'Transportation and storage', for example truck drivers who had road traffic accidents. In order to compare data from all Member States, the table 'Accidents at work by sex and age, excluding road traffic accidents and accidents on board of any mean of transport in the course of work (NACE Rev. 2, A, C-N excluding H) (HSW_MI03)' was created which excludes such accidents for all countries.

Type of injury
First delivery of Greek and Dutch data on type of injury started only in 2016.

Phase III variables
France, Greece and the Netherlands provided phase III variables only from 2016 onwards.

First delivery of Dutch and German data on occupation (ISCO-08) was due in 2016.

Size of enterprise
Comparability of data by size of enterprise is limited by the fact that some countries provide data for the local unit of the enterprise only (CY, EE, ES, FR, IE, IT, NL, PL, SE, NO) whereas other countries include all units of the enterprise (AT, BE, BG, CZ, DE, HU, LT, LV, LU, MT, PT, SI, SK).

In addition to the above mentioned derogations, the following applies according Commission Decision 2011/231:

Full coverage of all employees in NACE Rev. sectors A-S from 2016. The increase in coverage between reference years 2013 and 2014 led to substantial increase in the number of accidents between these two years.

No data about employees which are not insured by the social Insurance Foundation (IKA) is transmitted until 2016. Greece is still working on the inclusion of such data envisaged currently from June 2017 onwards. At the same time, Greece removed for 204 reference year data for the first time the uniform weight for all its cases. This led to a very significant decrease of the number of accidents.

2009 data for Ireland shows a strong decrease in the number of non-fatal accidents and consequently a related decrease in incidence rates. Ireland applies corrections for under-reporting using weights based on a survey. A part of the 2009 decrease may be due to the effects of the economic crisis but another part may have been caused by chances in the survey methodology at that time.

First delivery of data part of body injured and date of the accident is 2016. The data of the Netherlands show a strong decrease between reference years 2013 and 2014 which is a result of a methodological change.


Norwegian data shows also a srong decrease between reference years 2013 and 2014. This decresae is also a result of a methodological change including a change of competences in the organisation collecting and reporting about the data.


Data from Bulgaria, Latvia, Lithuania and Romania seem to include a higher degree of under-reporting concerning non-fatal accidents at work in particular (in 2012 their ratios of fatal to non-fatal accidents is between 14 and 55 times higher than the ratio of the EU-28 average). Data from several other Member States may also be subject to under-reporting, though to a lesser extent.

Economic crisis in 2008 and afterwards
Several countries stated that decreases in accidents between 2008 and 2009 could partially be explained by a reduction of economic activity between the same two years (closure of enterprises and manufacturing sites, fewer temporary workers, etc.).

Related metadata Top

Annexes Top
Guidelines for ESAW data transmission
ESAW incidence and standardised incidence rates update 2015-06-05
ESAW metadata reference year 2014