Statement of administrative sources

Background

Official statistics published by the Health and Safety Executive are based on two main sources - data gathered from statistical surveys, and data extracted from administrative or management systems.

By using data which is already available within administrative or management systems, rather than collecting data afresh, we are able to limit the overall burden placed on data providers, and also minimise data collection costs. In addition, the information we extract from such systems often has the advantage of being more timely than statistical data and, when compared with data from surveys (and particularly sample surveys), can also deliver data with a greater breadth of coverage.

The UK Statistics Authority actively encourages public bodies to exploit administrative and management sources for statistical purposes. However the Authority recognises that the statistical advantages of such arrangements can only be fully realised if statisticians have appropriate access to such systems; if statistical purposes are reflected in the design, management, and development of such systems; and if adequate safeguards are put in place to ensure the professional integrity of any official statistics derived from them.

The Authority's main requirements are set out in the third Protocol attached to their Code of Practice for Official Statistics. Amongst other things, this Protocol requires all producers of official statistics to publish, in consultation with the National Statistician, a 'Statement of Administrative Sources' which lists:

The following tables describe the administrative/management sources which the Health and Safety Executive currently uses to produce official statistics, or which have the potential to be so used, differentiating between:

1. Statistical usage of our own organisation's administrative or management sources

Name/Title of Administrative Data Source Name of overarching Administrative System Main administrative purpose of this source/system Geospatial Coverage Title(s) of all Statistical Products derived from this Source
HSE Enforcement Data COIN To record and monitor HSE enforcement action against employers resulting from breaches of health and safety legislation. GB; England; Scotland; Wales; Local authority (lowest level of coverage). Enforcement Notices issued, and Prosecutions Taken, by HSE each year.
Reported injuries and Dangerous Occurrences RIDDOR To record reports from employers of injuries, certain diseases and dangerous occurrences as required by the RIDDOR regulations GB; England; Scotland; Wales; Region; Local authority Reported injuries
Dangerous Occurrences
Blood Lead levels CLAWD (Control of Lead At Work) Collection and storage of annual blood lead data collected by Appointed Doctors from workers as required by the CLAW directive Great Britain Medical Surveillance of Blood Lead Levels in British Workers
ESQCR reports Electrical Incidents Database (EID) To collect statutory reports from Electricity suppliers in respect of the Electrical Safety and Quality Control Regulations (ESQCR) Great Britain  

2. Statistical usage of other organisations' administrative or management sources

Name/Title of Administrative Data Source Name of overarching Administrative System Name of Organisation responsible for this system/source Main administrative purpose of this system/source Geospatial Coverage Title(s) of all Statistical Products derived from this Source
Local authority Enforcement Data Each LA has their own system to record this information. LAs themselves, with advice/guidance from HSE. To record, LA enforcement action against employers resulting from breaches of health and safety legislation. Local authority (but can be aggregated to GB) Enforcement Notices issued by LAs each year.
Industrial Injuries Disablement Benefit (IIDB) IIDB Department for Work and Pensions To manage and pay IIDB GB; England; Scotland; Wales; Region; Number of IIDB claimants by prescribed industrial disease
Death certificates Death certificates Office for National Statistics (England and Wales);
National Records of Scotland
To record deaths of British citizens GB; England; Scotland; Wales;
Region; Local authority
Annual deaths from Mesothelioma;
Annual deaths from Asbestosis

3. Other administrative sources in HSE with the potential to be used for statistical purposes

Name/Title of Administrative Data Source Name of overarching Administrative System Main administrative purpose of this system/source Geospatial Coverage
Hydrocarbon release data HCR To provide voluntary information on Hydrocarbon releases in the offshore industry which supplements the statutory data provided under the RIDDOR regulation Offshore operators, GB
Notification of new construction projects (F10) COIN To alert HSE to new  construction projects which are underway and are expected to last more than 30 days or involve more than 500 man hours GB
Medical surveillance /biological monitoring data   To monitor compliance with specified Health and Safety legislation in certain industries or for particular at risk workers. GB (HSE-enforced premises only)
Central Index of Dosimetric Information CIDI To monitor compliance with regulation 20 of the Ionising Radiation Regulations 1999 GB (classified workers only)
Health and Safety Inspection activity COIN To record HSE inspector contact with duty holders GB; England; Scotland; Wales;
Region; Local authority (HSE-enforced industries only)

4. Detailed information about the Health and Safety Executive's governance arrangements for its own administrative or management sources

4.1 Enforcement data

Arrangements for providing statistical staff (whether inside or outside the organisation) with access to administrative or management sources for statistical purposes

HSE statistical staff have full access to HSE's corporate system 'COIN'. The managers of this system provide statisticians with data and interpretation on request.

For LA enforcement data, HSE makes annual requests to LAs for relevant data and interpretation.

Whether HSE or LA data, the statisticians always make clear the purpose for such data requests.

In addition, system owners often approach HSE statisticians for ad-hoc analyses of enforcement data to meet particular operational needs.

Arrangements for auditing the quality of the original source data

For HSE data, this is subject to scrutiny by HSEs internal auditing team, as well as the data managers, and operational staff, carrying out agreed DQ monitoring.

For LA data, when HSE receives the annual notices and prosecution data from LAs, this is scrutinised by HSE staff, and any irregularities addressed. However, given there are around 400 LAs, each with their own system, HSE are largely reliant on LAs to audit the quality of their own data.

In any event, all prosecution data (LA and HSE) is released publicly at a record-level as well as aggregate statistics, thus providing a certain level of 'self-auditing' by the general public.

Procedures for handling changes, and possible discontinuities, in the underlying source data

When HSE data is published, reference is made to the 'provisional' nature of the latest year, in that small changes are likely and will be reported the following year. No further changes are planned. For LA data, the most recent data is never changed, due to the complexities of data collection, and fewer records involved.

In practice, for many years no retrospective changes to 'finalised' data have been considered necessary.

Procedures for ensuring the security of the statistical processes which use administrative or management sources

All data extraction procedures are discussed and agreed with data system owners, and repeated each year. Such procedures have been in place for at least 30 years, so are well established and understood by all relevant parties. All data is stored on password-protected, Government Secure networks.

4.2 RIDDOR

Arrangements for providing statistical staff (whether inside or outside the organisation) with access to administrative or management sources for statistical purposes

The RIDDOR notification system used by employers enables reporting to be made predominantly online, with optional auto data transfer for larger reporters, also optional telephone reporting for serious incidents. HSE's Statistics branch take regular downloads of the operational data at the individual record level. This data is then securely transferred into SQL and SPSS copies of the datasets for statistical analysis.

Arrangements for auditing the quality of the original source data

Agreed data quality standards are in place and monitored by HSE. Further quality assurance is undertaken by HSE Statistics branch on a quarterly and annual basis, with any Data Quality compliance issues highlighted for review and action. Where data errors occur and are detected, these are corrected in the 'statistical copy' of the data, with the original dataset remaining unchanged.

Procedures for handling changes, and possible discontinuities, in the underlying source data

At first release, RIDDOR data are marked as provisional. The following year, data are finalised and no further revisions made (unless there is a methodological improvement or errors are found). Data for the previous 'provisional;' year are marked as 'revised'.

If a system or reporting/legal change occurs, we endeavour to estimate the effect of any discontinuity in our headline data series and produce a consistent back-series. If it is not possible or practical to adjust the series, a discontinuity line is presented in charts and tables and appropriate footnotes supplied.

Procedures for ensuring the security of the statistical processes which use administrative or management sources

All data is stored on password-protected, Government Secure networks.

4.3 Blood Lead (CLAWD)

Arrangements for providing statistical staff (whether inside or outside the organisation) with access to administrative or management sources for statistical purposes

Statisticians and supporting staff have access to the Appointed Doctors Database and the custom built Control of Lead at Work Database (CLAWD) where the doctor returns are stored.

Arrangements for auditing the quality of the original source data

In built validity checks in CLAWD are run automatically. A statistician supervises data entry. Queries are resolved internally or via clarification from the doctor.

Procedures for handling changes, and possible discontinuities, in the underlying source data

Changes in published statistics (if any) are noted in the following years release and further details are recorded in-house and stored with the blood lead forms. Data is analysed annually on returns received and validated

Procedures for ensuring the security of the statistical processes which use administrative or management sources

All data and outputs are stored on a password protected network

 
Updated 2021-05-10