RR425 - Modelling the PSA1 indicator
An analysis of the under-reporting of major accidents and over-3-day injuries in RIDDOR reports was examined using various data sources, namely the RIDDOR database of reported accidents, the Labour Force Survey (LFS) and surveys carried out by the Risk Assessment (RA) section of HSL in the manufacturing and services sector. The key points emerging from the analysis were:
- The major accident rate for the manufacturing sector shows little change over the five-year period in both the RIDDOR and RA survey data.
- The reporting levels from the manufacturing sector analysis show a consistently higher reporting for major accidents against over-3-day injuries, with perhaps a trend towards convergence of the levels. This appears to contradict the basic assumption behind the original adjustment analysis, where the model was divergence of reporting levels for the two injury categories from an equal starting point.
- The comparison of overall accident rates from the RA survey and LFS datasets indicates that there is a significant amount of under-reporting within the RA survey itself.
- The trend of an increasing reportable accident rate, as seen in the RA services survey, is not observed in the RIDDOR or LFS data. This suggests that these data may not be useful in determining the relationship (ie reporting level) between the RIDDOR and LFS data.
- The most appropriate adjustment methodology is proposed. A global adjustment of the major accident rate is calculated by using the ratio of major accident to over-3-day injury reporting levels as obtained from the RA survey.
- This adjustment will shift the baseline of the PSA1 indicator down, reflecting the conclusion that the major injury reporting level is higher that the over-3-day reporting level, but leaves the trend in the PSA1 indicator unchanged.
This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.
The Research Report Series are produced in Adobe Acrobat. The use of the latest version of the software is recommended which is available at the Adobe website via the link on this page.
Assistance in the use of Adobe Acrobat PDF files is available on our FAQs page.