The Health and Safety Executive (HSE) welcomes this informed contribution to the debate on measuring progress against the government's targets for occupational health and safety. We recognise the importance of identifying criteria against which the success of the “Revitalising health and safety”2 and “Securing health together”3 strategies will be assessed, and also the difficulty of doing so.

We agree that in the area of occupational health, measuring progress against targets presents particular challenges. Three were highlighted by Coggon1:

• The difficulty of meaningfully attributing individual cases to work: and the fact that such attribution will be done differently by different people—for example the people themselves, their doctors, or their employers.

• The effect of cultural or psychosocial factors on reporting of symptoms: including the possibility of improved awareness leading to a “paradoxical” rise in symptom reporting (especially in the early years of a new strategy or campaign).

• The latent interval after exposure: so that the incidence of some diseases—such as cancers—will not be affected by actions taken during the 10 year period of the strategies (in such cases, monitoring exposures may give a more immediate measure of progress).

To these we would add the special challenges associated with measuring change over time:

• The degree of statistical uncertainty, especially in population based surveys, which may be such that the percentage target improvement lies within the likely margin of error.

• Concerns about the stability of the sources, in terms of consistency of coverage, timeliness, and frequency, and securing the flow of data in the longer term.

For all these reasons, the HSE agrees that no single method of monitoring can provide all the answers. Over the years we have worked with others to develop a range of different data sources for occupational ill health, including household surveys, reporting by specialist doctors, administrative data from compensation schemes and elsewhere, and the calculation of attributable fractions. Each of these is more appropriate for some disorders than for others, and each will have a part to play in monitoring progress.

Coggon concluded by suggesting that “the limitations of the monitoring methods that are adopted must be clearly articulated . . . before data are collected and trends are analysed”. To this end, the HSE statisticians—in consultation with stakeholders—have recently produced a “Statistical note on progress measurement”, which is now available on the HSE website (www.hse.gov.uk/hsestats/statnote.pdf).

This describes the challenges already mentioned. It explains how existing sources will be refined and new ones developed and sets out the statistical techniques which will be used to monitor progress against the targets. These will include the integration of data from the different sources, to arrive at an overall judgement about progress against the targets; we think that such an approach is novel in the field of occupational health, although it has been successfully used in other area of statistics—such as the National Accounts.

By setting out in advance the principles which will be followed when assessing whether the targets have been achieved, the statistical note is intended to avoid the danger mentioned by Coggon, of scientific judgement being obscured by “political wrangling and point scoring”.

The HSE sees this very much as a developing area. The statistical note has been subject to a process of external consultation, including a workshop in December 2000 attended by around 30 experts in the field. As we build on the foundations set out in the note we will continue to seek the involvement of stakeholders, including the scientific community. We do not underestimate the scale of the work still to be done but our overriding aim is that progress measurement for the occupational health and safety targets should be robust, credible, and transparent.