SPR Not sure where you got the above definition for AFR/IFR? However the AFR/IFR is normally based on 100,000 working hours of a project, or factory, or whatever. Unless you are using OSHA stats in which case it is 200,000 working hours x AFR/IFR. How you arrive at your working hours total may be dependant on the type of industry or who is providing the data. It may be pragmatic to divide each worker by 40hrs in order to reach your sum, nothing wrong in that. Ray

"Neither of them derives from anything - they are arbitrary figures designed to give you numbers without too many decimal places in them so you can work with them more easily." SPR you are quite correct when you say that the 100000 is very roughly based on an average person's working lifetime as explained at this website by Judith Hacket http://www.lattitudesafety.co.uk/

There are no legislatively required formulae so arguing over what the components should be and where they are derived from is unnecessary! It doesn't really matter what your internal formulae are so long as:- 1 - everyone using them understands the components; 2 - it provides the data that the business needs in a form that can be readily used; and 3 - the business recognises that they may not be directly comparable to the HSE promoted formulae and has a mechanism for converting as required. I personally use a modified version of the HSE formulae that is designed to provide a much more incisive and focussed set of data than the very blunt HSE instruement. Frank Hallett

Frank Hallett uses the interesting expression, 'the very blunt HSE instruement (sic)' about the HSE measure for accident management but doesn't reveal evidence of the validity of any less blunt and valid instrument. Any safety professional interested in designing,developing and testing a not-too-blunt measurement instrument is well advised to draw on the principles of Pragmatist philosopher John Dewey and were refined for business use by Laurie Thomas and Sheila Harri-Augstein who published guides in 1988 and 1993. While the bespoke software they use has now become obsolete for use on PCs or Macs, it's straightforward to conduct the matrix analyses they use for graphical feedback with programs like Mathematica and MATLAB. The great advantage that repertory grids offer lies in how they can be used to negotiate consensus between managers about the accident rates relevant to their particular organisation at the stage of evolution (or decline) of their organisation, and adapted upwards as they improve. Instead of any benchmark based on averages that don't fit any individual organisation, what managers get is a range of numerical values relevant to their own situation, and justified in words of their own choices. While Frank might find repertory grids are in some ways also lacking, they can be justified on the basis of objective measurements combined with indicators of the commitment levels of those who designed them.

Perhaps the 'real world' competency system which the Director of IOSH sketched to a meeting of the London Metropolitan branch meeting in January will go some way towards effective uses of measurements in the task of protecting people at work, without enmeshing the task in clouds of irrelevant legalese. In an article, titled 'We need a revolution in the way maths is taught', in The Observer today, 23.2.2014, a world class mathematician Conrad Wolfram, explains how 'maths is the world's most successful system of problem-solving'. While the case he makes for radical transformation of the quality of maths at school level, it also applies to education and train for most professions, not least occupational safety and health. Information about programs Wolfram has invented and developed is available at www.wolfram.com; I have no vested interest in his business other than as a value-creating user of 'Mathematica 9' their flagship program.

Thank you Ray for the addition to my points. I consider that another of the major failings in the way such info is used is that it has a tendency to be the "headline" that becomes part of the corporate PR. A failure that tarnishes the corporate image will also have a direct impact on the person[s] in charge of that activity at that time - there's incentive for the manager who may have a job or bonus riding on a clean page!!. This then has a subliminal, and sometimes actual, drive to suppress, ignore or minimise data that could "destroy the image of perfection" promoted by the lack of reported failures. I have no doubt that there will be a range of responses telling me that this just doesn't happen 'cos of the legal impetus, 18001 systems, etc - well, we have enough data in the public domain and also from empirical experience that it does. To deny the possibility is to ignore the reality - rather like the suppression of data that doesn't look good! I know that there are a considerable number of organisations out there who try to do it well; that's not the point. What is to the point is how they ensure that all relevant data is given equal prominence and attention - warts and all! Frank Hallett