Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.
Working Environment Authority, Analytics and Behavior, Postbox 1228, DK - 0900 Copenhagen C.
1.6. Contact email address
Confidential because of GDPR
1.7. Contact phone number
Confidential because of GDPR
1.8. Contact fax number
Confidential because of GDPR
2.1. Metadata last certified
28 June 2024
2.2. Metadata last posted
28 June 2024
2.3. Metadata last update
28 June 2024
3.1. Data description
Denmark submits data on accidents at work with more than 3 calendar days of expected absence from work to Eurostat. Upload is performed annually by the end of June according to the Regulation (formerly 'gentleman agreement') in the *.CSV format.
The data sets contain information about the victim, the injury, the company, the accident and the causes and circumstances of the accident. The variables included are:
Victim:
Occupation of the victim (ISCO)
Age of the victim
Sex of the victim
Employment status
Nationality
Injury:
Type of injury (see NOTE)
Part of body injured
Days lost
Company:
Economic activity of the employer (NACE)
Accident:
Date of the accident
Time of the accident
Geographical location (NUTS)
Causes and circumstances:
Work station (partially)
Working environment (partially)
Deviation (from 17th of June 2017 and onwards)*
Contact - Mode of injury
Material agent - contact (Not included in 2018-data and onwards)*
Other:
Road Trafic Accident (RTA Yes/No) - Facultative
Weight (correction for underreporting)
Data include both fatal and non-fatal accidents.
*) As from 17th of June 2017 a new notification system was introduced. In the old system the variable Material agent - contact was registrered. In the new system Deviation is registrered in stead.
NB: In 2020, 2021 and 2022 a lot of COVID-19 cases were reported as accidents at work. There is no specific registration of these cases in the Danish classifications available. Most cases were registered in the parameter 'Type of injury' (INJ) as 'Acute infection' (072), 'Other poisonings and infections' (079) or unknown (00) for the reference year 2020. For the purpose of the year 2021 ESAW data delivery, the number of COVID-19 related cases has been estimated on the basis of text mining searching for text strings in the text fields of the notifications. The following text strings have been used: 'covi', 'corona', 'pcr', cowid', 'cv', 'smit', 'positiv', '19'. For these identified cases the variable INJ has been coded as '073' (Covid-19) according to the ESTAT/ESAW recommendation on how to code occupational COVID-19 cases.
3.2. Classification system
All variables are delivered according to the ESAW methodology.
3.3. Coverage - sector
In principle all economic sectors are covered by the Danish notification scheme and hence by the ESAW-data.
Exceptions are:
Defence activities (NACE, Rev 2. 8422) are not covered for military personnel (armed forces) defined by ISCO-08 codes 01, 02 and 03, while performing combat, training, exersize, but accidents during civil activities are.
Denmark is only providing provisionary information about company size (optional variable).
3.3.2. Additional comments on coverage of Economic sectors (NACE Rev.2)
We have no additional comments on coverage of Economic sectors.
3.4. Statistical concepts and definitions
For description of the variables, see point 3.1 Data description (above).
3.4.1. Coverage of ESAW variables – annual update expected
Please check the Annex 1, table 3.4.1.
3.4.2. Coverage of ESAW variables - day lost- annual update expected
Please check the Annex 1, table 3.4.2.
3.4.3. Coverage of ESAW variables- weight - annual update expected
Please check the Annex 1, table 3.4.3.
3.4.4. Additional comments on coverage of ESAW variables
We have no additional comments on coverage of ESAW variables.
3.5. Statistical unit
Statistical unit is the 'Number of reported accidents at work' (N).
Reportable accidents in Denmark are accidents with one day or more of absence from work (days lost), wheras Eurostat publishes only accidents with more than 3 days lost.
Denmark submits only cases with more than 3 days lost together with cases where Severity is unknown.
3.6. Statistical population
The Danish ESAW data refers to all employed persons, except military personel (see 3.3 Sector coverage) but including:
students (trainees) occupied with 'work-like' activities
family members when performing 'work-like' activities
Self-employed, if the accident happens with technical equipment (machines, tools etc.)
3.6.1. Coverage of Employment status - annual update expected
Please check the Annex 1, table 3.6.1.
3.6.2. Additional comments on coverage of professional status
We have no additional comments on professional status.
3.6.3. Coverage of Accidents outside premises - annual update expected
3.6.5. Additional comments on coverage of accidents outside premises
We have no additional comments on coverage of accidents outside premises.
3.7. Reference area
Danish ESAW data are available for Eurostat through eDAMIS. The data are not dissiminated elsewhere by national authorities.
The data cover all accidents on Danish territory whether on land, on the sea (on board any vessel of sea-transport or off shore installations) or in the air (on board aircrafts).
3.8. Coverage - Time
Annual data. Data cover all registrered accidents at work which have occured per full calendar year.
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
3.9. Base period
Full calendar years.
Same as metadata at EU level, eg. all reported accidents at work are included.
However, the basic unit of statistical observation in Denmark are the 'Number of reported accidents at work'. Reportable accidents in Denmark are accidents with one day or more of absence from work (days lost), wheras Eurostat publishes only accidents with more than 3 days lost.
Denmark submits only cases with more than 3 days lost together with cases where Severity is unknown.
2023: for date of the accident, eg. accidents at work which have occured in 2023.
6.1. Institutional Mandate - legal acts and other agreements
According to the legislation it is an obligation for the employer to notify any accident at work amongst his/her employees if the accident leads to absence from work for one day or more. Only accidents at work with more than 3 days lost are submitted to ESTAT.
6.2. Institutional Mandate - data sharing
Statistical confidentiality concerning personal information is guarantied by law in the Act on Processing of Personal Data (revised according to GDPR): Danish Data Protection Agency
7.1. Confidentiality - policy
Statistical confidentiality concerning personal information is guarantied by law in the Act on Processing of Personal Data (revised according to GDPR): Danish Data Protection Agency
7.2. Confidentiality - data treatment
Before submission to Eurostat and for other statistical purposes data that identify persons or companies are removed.
8.1. Release calendar
ESAW data are released for submission (tranfer/delivery) to Eurostat every year before the end of June.
8.2. Release calendar access
Eurostat is the only user of ESAW micro data, and the ESAW micro data are not disseminated elsewhere.
8.3. Release policy - user access
According to the ESAW regulation: Commission Regulation (EU) No. 349/2011.
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
10.3.1. Data tables - consultations
not available
10.4. Dissemination format - microdata access
CSV-file
10.5. Dissemination format - other
CSV-file
10.5.1. Metadata - consultations
Not available.
10.6. Documentation on methodology
ESAW methodology, 2013 edition.
10.6.1. Metadata completeness - rate
Not available
10.7. Quality management - documentation
Metadata is delivered together with the annual data transmissions to Eurostat. No national quality report is available.
11.1. Quality assurance
The WEA is committed to data quality, and quality consideration play a central role in data collection (notification scheme) as well as in the processing of statistical data. Data processing cover data cleaning (eg. Correction for duplets, and check for missing values) and anonymization in order to produce data sets for statistical use.
For fatal accidents the quality of information is assessed manually.
Some variables suffer from a large number (percentage) of unknown cases, where the notifier either have not answered the question or have chosen the 'Unkown' value. In particular this is the case for Deviation (DEV), Type of Injury (INJ) and Severity (SEV). Initiatives are conducted in order to increase the data quality through information campaigns targeted towards the companies.
11.2. Quality management - assessment
The overall quality assessment is described above.
Investigations of the degree of underreporting have been performed regularly, primarily by comparing RAW-data (national register) with data from emergency units at hospitals, but also with the LFS ad-hoc modules. In general these investigations estimate approximately 50 percent underreporting in Denmark.
Denmark delivers three (four) out of nine phase III variables:
Workstation
Working environment (partially)
Deviation (as from 17th of June 2017 and onwards)
Contact mode of injury
12.1. Relevance - User Needs
The primary user of Danish ESAW data is Eurostat.
12.2. Relevance - User Satisfaction
The primary user of Danish ESAW data is Eurostat. The WEA does not directly consult the users of European statistics. ESAW data for all member states are published by Eurostat at their web-site. Eurostat is the authority responsible for dissimination of statistical products from ESAW-data.
Users can obtain information about the number of accidents and general variables from the Danish ESAW-data delivered to Eurostat upon request.
12.3. Completeness
The primary user of Danish ESAW data is Eurostat.
12.3.1. Data completeness - rate
Danish ESAW data are in principle complete, ie. no sectors are exceptet from data (see section 2.3 - Coverage). The only exception from data are accidents for military personnel (defined by ISCO) during combat, training, excercize etc. Accidents during civil activities for military personel are covered.
12.3.1.1. Global reporting level of Economic sectors (NACE Rev.2) - annual update expected
Danish ESAW data are in principle complete, ie. no sectors are exceptet from data (see section 2.3 - Coverage). The only exception from data are accidents for military personnel (defined by ISCO) during combat, training, excercize etc. Accidents during civil activities for military personel are covered.
12.3.1.2. Additional comments on global reporting level
We have no additional comments on global reporting level
12.3.1.4. Additional comments on reporting levels by Economic sector (NACE Rev.2)
We have no additional comments on reporting levels by Economic sector
13.1. Accuracy - overall
Under-reporting:
The register on accidents at work is not complete, because some accidents which should be notified according to the rules (reportable accidents) are not notified.
Over the last 10 years it has been shown that the reporting level ranges form 15 to 73 percent depending on the economic activity (trade) of the company, the age of the victim and the severity of the accident. The wide variability of the reporting levels also reflects differences in investigation methods. Overall the reporting level is estimated to be approximately 50 percent. The latest survey showed an underreporting level of 46 percent.
For this reason a corection factor (weight) for non-fatal accidents is delivered to Eurostat along with the submission of the ESAW data.
For fatal accidents the coverage is 100 percent.
13.2. Sampling error
No additional information to the metadata on EU level available. Information corresponds to the EU metadata. Data collection is based on a notification system, and not on a survey (sample).
13.2.1. Sampling error - indicators
Not applicable - Data collection is based on a notification system, and not on a survey (sample).
13.3. Non-sampling error
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
13.3.1. Coverage error
Not applicable.
13.3.1.1. Over-coverage - rate
Not applicable.
13.3.1.2. Common units - proportion
Not applicable.
13.3.2. Measurement error
Not applicable.
13.3.3. Non response error
Not applicable.
13.3.3.1. Unit non-response - rate
Not applicable.
13.3.3.2. Item non-response - rate
Not applicable.
13.3.4. Processing error
Not applicable.
13.3.5. Model assumption error
Not applicable.
14.1. Timeliness
Danish ESAW data are submitted each year by the end of June for the reference year terminated 18 month earlier (N+2).
14.1.1. Time lag - first result
On date (N = 17 June 2025)
14.1.2. Time lag - final result
Not available.
14.2. Punctuality
Danish ESAW data are submitted each year by the end of June for the reference year terminated 18 month earlier (N+2).
14.2.1. Punctuality - delivery and publication
On date (17 June 2025)
15.1. Comparability - geographical
No additional information to the metadata on EU level available.
Danish data are in principle comparable from 1979 and onwards for accidents at work with 1 day absence or more. During such a long time span there are of course data breaks for certain variables. In 2003 an electronically notification system was introduced. Before 2003 notifications of accidents at work were made on paper forms and then coded into am database by professionel staff of the DWEA.
Major data breaks:
In 2007 Danish NUTS-classification was revised following the change in Denmark going from 14 counties and 273 municipalities to 5 Regions and 98 municipalities.
In 2018 NUTS 3 was implemented in Danish ESAW data, but it is regularly amended accoding to recent regional developments.
15.1.1. Asymmetry for mirror flow statistics - coefficient
Not applicable.
15.2. Comparability - over time
No additional information to the metadata on EU level available.
Danish data are in principle comparable from 1979 and onwards for accidents at work with 1 day absence or more. During such a long time span there are of course data breaks for certain variables. In 2003 an electronically notification system was introduced. Before 2003 notifications of accidents at work were made on paper forms and then coded into am database by professionel staff of the DWEA.
Major data breaks:
In 2007 Danish NUTS-classification was revised following the change in Denmark going from 273 municipalities tom 98.
In 2008 NACE Rev. 2 was implemented in the form of "Dansk Branchekode 2007" which is a 6-digit version of NACE 2.0. For comparability of ESAW data 2008 is therefore the baseline.
In 2011 a Danish version of ISCO-08 was implemented.
In 2017 the variable 'Material agent of Contact Mode of Injury' was deleted from the Danish notification system, and the variable 'Deviation' was introduced.
In 2018 NUTS 3 was implemented in Danish ESAW data, but it is regularly amended accoding to recent regional developments.
15.2.1. Length of comparable time series
2008 and onwards.
15.3. Coherence - cross domain
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
15.3.1. Coherence - sub annual and annual statistics
Not available.
15.3.2. Coherence - National Accounts
Not available
15.4. Coherence - internal
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
There is no cost/burden analysis available.
The ESAW data are produced on the basis of data already existing according to Danish legislation due to the obligation to notify accidents at work.
Employers (companies) are obliged to notify electronically via the internet in a system called EASY. Companies can do this directly or from their (accounts) own internal registration system through an XML interface. When notifying the companies do not have to give details about the company (economic activity, name, address etc.) because these data are obtained automatically from the Central Register of Companies* managed by the Danish Business Authority. The key to this is that the user name to EASY is the company number in the Central Register of Companies.
In future revision of the notification system the cost effectiveness and the burden on respondents will be taken further into account.
17.1. Data revision - policy
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
17.2. Data revision - practice
Already submitted data are normally not revised.
A major revision took place for the reference year 2017 where a new notification system was implemented (launched). The new system have impact on the delivered ESAW data, as there are changes in some variables (see section 3.1 Data description).
17.2.1. Data revision - average size
Not applicable
18.1. Source data
A notification scheme for accidents at work has been laid down in a Statutory Order (administrative data), see Annex, according to which the employer shall notify accidents at work for their employees, if the accident has resulted in incapacity to work for one day or more, to the WEA.
Similar notification schemes exist for accidents within the sectors of transportation in air or at sea, administrered by the Danish Transport, Construction and Housing Authority and the Danish Maritime Agency, respectively.
Information (data) from these notifications is kept in a database (RAW) from which the ESAW data to Eurostat are created in accordance with the ESAW methodology with respect to standard concepts, definitions, variables and classifications.
The purpose of the Register on Accidents at Work (RAW) is to enable the authority (WEA) to perform inspections (investigation of accidents) and constitutes the statistical background for priorities and campaigns in order to contribute to the prevention of accidents at work.
The register suffers from reporting deficits (underreporting): See section 11.2 Quality management - assessment.
The national source of the reference population is Statistics Denmarks RAS register, see Annex.
18.2. Frequency of data collection
Data are collected continuously from the notifications received (electronically).
Data are processed annually for statistical purposes (fixed data), but the dynamic source data can be accessed.
18.3. Data collection
Administrative data based on a legally binding notification scheme:
The Register on Accidents at Work (RAW).
18.4. Data validation
On the basis of the national source register on accidents at work (RAW) the output to ESAW data format is coded and validated according to the ESAW methodology and the 'Guidelines for ESAW data transmission'.
18.5. Data compilation
On the basis of the national source register on accidents at work (RAW) the output to ESAW data format is coded and validated according to the ESAW methodology and the 'Guidelines for ESAW data transmission'.
Some variables are imputed from other variables:
Workstation (from Working environment, partially)
Working environment (recoded from RAW data om 'place of accident').
Weight for underreporting is calculated from the mean reporting percentage derived from comparison to statistics from emergency units at hospitals.
W = 100/(reporting percentage), ex: W = 100/54 = 1,86
For 2017 two additional weights have been introduced for variable 'Deviation' and variable 'Material agent - Contact Mode of Injury' (MA_CMI), because Deviation was implemented in the Dansih register in the middle of 2017 in stead of the Material agent.
For reference year 2018 and onwards the variable MA_CMI is no longer part of the Danish delivery of ESAW-data. The phase III variable DEV is part of the Danish ESAW-data in stead.
18.5.1. Imputation - rate
Not applicable.
18.6. Adjustment
Danish ESAW data covers all reported accidents that occured during a reference calendar year.
There are no adjustments such as for time series or seasonal adjustment.
18.6.1. Seasonal adjustment
No seasonal adjustments.
No additional information to the metadata on EU level available. Information corresponds to the EU metadata.
NB: In 2020, 2021 and 2022 a lot of COVID-19 cases were reported as accidents at work. There is no specific registration of these cases in the Danish classifications available. Most cases were registered in the parameter 'Type of injury' (INJ) as 'Acute infection' (072), 'Other poisonings and infections' (079) or unknown (00) for the reference year 2020. For the purpose of the year 2021 ESAW data delivery, the number of COVID-19 related cases has been estimated on the basis of text mining searching for text strings in the text fields of the notifications. The following text strings have been used: 'covi', 'corona', 'pcr', cowid', 'cv', 'smit', 'positiv', '19'. For these identified cases the variable INJ has been coded as '073' (Covid-19) according to the ESTAT/ESAW recommendation on how to code occupational COVID-19 cases.
Denmark submits data on accidents at work with more than 3 calendar days of expected absence from work to Eurostat. Upload is performed annually by the end of June according to the Regulation (formerly 'gentleman agreement') in the *.CSV format.
The data sets contain information about the victim, the injury, the company, the accident and the causes and circumstances of the accident. The variables included are:
Victim:
Occupation of the victim (ISCO)
Age of the victim
Sex of the victim
Employment status
Nationality
Injury:
Type of injury (see NOTE)
Part of body injured
Days lost
Company:
Economic activity of the employer (NACE)
Accident:
Date of the accident
Time of the accident
Geographical location (NUTS)
Causes and circumstances:
Work station (partially)
Working environment (partially)
Deviation (from 17th of June 2017 and onwards)*
Contact - Mode of injury
Material agent - contact (Not included in 2018-data and onwards)*
Other:
Road Trafic Accident (RTA Yes/No) - Facultative
Weight (correction for underreporting)
Data include both fatal and non-fatal accidents.
*) As from 17th of June 2017 a new notification system was introduced. In the old system the variable Material agent - contact was registrered. In the new system Deviation is registrered in stead.
NB: In 2020, 2021 and 2022 a lot of COVID-19 cases were reported as accidents at work. There is no specific registration of these cases in the Danish classifications available. Most cases were registered in the parameter 'Type of injury' (INJ) as 'Acute infection' (072), 'Other poisonings and infections' (079) or unknown (00) for the reference year 2020. For the purpose of the year 2021 ESAW data delivery, the number of COVID-19 related cases has been estimated on the basis of text mining searching for text strings in the text fields of the notifications. The following text strings have been used: 'covi', 'corona', 'pcr', cowid', 'cv', 'smit', 'positiv', '19'. For these identified cases the variable INJ has been coded as '073' (Covid-19) according to the ESTAT/ESAW recommendation on how to code occupational COVID-19 cases.
28 June 2024
For description of the variables, see point 3.1 Data description (above).
Statistical unit is the 'Number of reported accidents at work' (N).
Reportable accidents in Denmark are accidents with one day or more of absence from work (days lost), wheras Eurostat publishes only accidents with more than 3 days lost.
Denmark submits only cases with more than 3 days lost together with cases where Severity is unknown.
The Danish ESAW data refers to all employed persons, except military personel (see 3.3 Sector coverage) but including:
students (trainees) occupied with 'work-like' activities
family members when performing 'work-like' activities
Self-employed, if the accident happens with technical equipment (machines, tools etc.)
Danish ESAW data are available for Eurostat through eDAMIS. The data are not dissiminated elsewhere by national authorities.
The data cover all accidents on Danish territory whether on land, on the sea (on board any vessel of sea-transport or off shore installations) or in the air (on board aircrafts).
2023: for date of the accident, eg. accidents at work which have occured in 2023.
Under-reporting:
The register on accidents at work is not complete, because some accidents which should be notified according to the rules (reportable accidents) are not notified.
Over the last 10 years it has been shown that the reporting level ranges form 15 to 73 percent depending on the economic activity (trade) of the company, the age of the victim and the severity of the accident. The wide variability of the reporting levels also reflects differences in investigation methods. Overall the reporting level is estimated to be approximately 50 percent. The latest survey showed an underreporting level of 46 percent.
For this reason a corection factor (weight) for non-fatal accidents is delivered to Eurostat along with the submission of the ESAW data.
For fatal accidents the coverage is 100 percent.
Same as metadata at EU level, eg. all reported accidents at work are included.
However, the basic unit of statistical observation in Denmark are the 'Number of reported accidents at work'. Reportable accidents in Denmark are accidents with one day or more of absence from work (days lost), wheras Eurostat publishes only accidents with more than 3 days lost.
Denmark submits only cases with more than 3 days lost together with cases where Severity is unknown.
On the basis of the national source register on accidents at work (RAW) the output to ESAW data format is coded and validated according to the ESAW methodology and the 'Guidelines for ESAW data transmission'.
Some variables are imputed from other variables:
Workstation (from Working environment, partially)
Working environment (recoded from RAW data om 'place of accident').
Weight for underreporting is calculated from the mean reporting percentage derived from comparison to statistics from emergency units at hospitals.
W = 100/(reporting percentage), ex: W = 100/54 = 1,86
For 2017 two additional weights have been introduced for variable 'Deviation' and variable 'Material agent - Contact Mode of Injury' (MA_CMI), because Deviation was implemented in the Dansih register in the middle of 2017 in stead of the Material agent.
For reference year 2018 and onwards the variable MA_CMI is no longer part of the Danish delivery of ESAW-data. The phase III variable DEV is part of the Danish ESAW-data in stead.
A notification scheme for accidents at work has been laid down in a Statutory Order (administrative data), see Annex, according to which the employer shall notify accidents at work for their employees, if the accident has resulted in incapacity to work for one day or more, to the WEA.
Similar notification schemes exist for accidents within the sectors of transportation in air or at sea, administrered by the Danish Transport, Construction and Housing Authority and the Danish Maritime Agency, respectively.
Information (data) from these notifications is kept in a database (RAW) from which the ESAW data to Eurostat are created in accordance with the ESAW methodology with respect to standard concepts, definitions, variables and classifications.
The purpose of the Register on Accidents at Work (RAW) is to enable the authority (WEA) to perform inspections (investigation of accidents) and constitutes the statistical background for priorities and campaigns in order to contribute to the prevention of accidents at work.
The register suffers from reporting deficits (underreporting): See section 11.2 Quality management - assessment.
The national source of the reference population is Statistics Denmarks RAS register, see Annex.
Yearly.
Danish ESAW data are submitted each year by the end of June for the reference year terminated 18 month earlier (N+2).
No additional information to the metadata on EU level available.
Danish data are in principle comparable from 1979 and onwards for accidents at work with 1 day absence or more. During such a long time span there are of course data breaks for certain variables. In 2003 an electronically notification system was introduced. Before 2003 notifications of accidents at work were made on paper forms and then coded into am database by professionel staff of the DWEA.
Major data breaks:
In 2007 Danish NUTS-classification was revised following the change in Denmark going from 14 counties and 273 municipalities to 5 Regions and 98 municipalities.
In 2018 NUTS 3 was implemented in Danish ESAW data, but it is regularly amended accoding to recent regional developments.
No additional information to the metadata on EU level available.
Danish data are in principle comparable from 1979 and onwards for accidents at work with 1 day absence or more. During such a long time span there are of course data breaks for certain variables. In 2003 an electronically notification system was introduced. Before 2003 notifications of accidents at work were made on paper forms and then coded into am database by professionel staff of the DWEA.
Major data breaks:
In 2007 Danish NUTS-classification was revised following the change in Denmark going from 273 municipalities tom 98.
In 2008 NACE Rev. 2 was implemented in the form of "Dansk Branchekode 2007" which is a 6-digit version of NACE 2.0. For comparability of ESAW data 2008 is therefore the baseline.
In 2011 a Danish version of ISCO-08 was implemented.
In 2017 the variable 'Material agent of Contact Mode of Injury' was deleted from the Danish notification system, and the variable 'Deviation' was introduced.
In 2018 NUTS 3 was implemented in Danish ESAW data, but it is regularly amended accoding to recent regional developments.
