Ian-F Your stated aspiration - introducing risk profiling as a method of gap analysis and potentially increasing ownership of health and safety risk control - appears admirable and plausible. It also creates enormous risks of failing and of avoidable litigation, unless you base it on relevant methods of statistical analysis. For a risk profile in any specific location is unavoidably qualified by any changes to equipment, furniture, handtools, work processes and people; and any risk profile of any particular manager is qualified by these and a host of other individual characteristics including age, gender, qualifications, training, to name only the most obvious. While 'templates' for such statistical analysis are available in numerous guides to business statistics, using them is extremely risky unless you understand in some depth the conceptual basis on which they are designed. 'Research design and statistics for the safety and health professional', C A Cacha, Van Nostrand Reinhold, 1997 is a reasonably good guide to useful statistical models and templates, that you can adapt. Up-to-date versions of Ms Excel or of Minitab greatly reduce any drudgery. Yet the requirement to understand the rationale associated with every single model and template remains absolutely imperative if you are to avoid, at best, wasting your time.

We do this right across our business and I cannot understand where any of the suggested fall out might be generated. But then this is risk profiling in my terms and maybe not theirs. For each and every location, we take all of the risk assessments at that site and through a process not least driven by the outcomes of the assessments we prioritise and report the top six into a register. That is the profile for that site. Each site submits its register into its operating group. Using the same magic, we produce a group profile. All of the groups, yes you guessed it, submit into the division and then each division as an operational risk profile. Finally all of these combine and produce a strategic profile for the entire business. The registers are then sent back to the level below so that they can compare and contrast themselves against their peer sites. How else would a site know how well they are doing. Numbers of RIDDORs? I don't think so. Whilst we don't profile the manager exactly, essentially, if they raise the issue and have no means to tackle it, then it belongs at the next level. If they can tackle it and don't then trust me, this will come home to roost with them. And why not? Ok the process has been massively simplified to get into a para, but how that would introduce any threat of litigation is beyond me. What we are saying is 'please report on the issues that you have yet to crack - and if you have, what's left that is now your highest. That we should not include how well we do or don't do something within our risk assessments because it may be used against us either in criminal or civil action simply isn't in our thinking.

Kieran, sorry, are you suggesting that our methodology lacks reliability and validity? If that's the case I can assure that it does not. It has been highly successful and is now a model that many other large businesses are becoming interested in. I gave evidence to Prof L in his enquiry and my colleague to Lord Young. Our methodology won us the first case at the HSE challenge panel against a local authority ruling too. I don't want to get into a list of battle honours, but our efforts recently secured Corporate Membership of the CIEH as well; only the second business to do so. The magic to which I referred was really shorthand for the processes involved. They are not finger in the air stuff but based on sound logic and management decisions made on evidence. Exactly the sort of evidence advocated by Professor Lofstedt. Out approach is not for everyone, but having some 90,000 employees in the UK alone, it has been proven to work for us. You may disagree that this constitutes 'risk profiling' and that is fine. But what it does do is deliver actionable results across a multi site, multi disciplined business - in a pretty short time-scale. I feel that we've had 40 years of trying to find a way through all this and some people are still looking. Well, we're no longer looking, we've found our approach, prioritised, acted and now are very close to saying 'Right, that's sorted then' If people want to send another 40 years searching for their 'grails' then that's their choice, naturally.