C

I'm not aware of the 1000 hours but to benchmark to others you will be better of with either HSE (100,000) or OSHA (200,000) to give you the accident frequency rate. The calculation is;

Total number of injuries x 100,000 (or if OSHA 200,000)/ number of hours worked by all employees = accident frequency rate

I have often been asked to calculate and supply to customers RIDDOR Accident Frequency rate as Number of RIDDOR accidents x 100,000 divided by the number of hours worked (for all people).

It was once explained to me that the 100,000 was about the number of hours a single person worked in their life time ( 49 years x 50 weeks per year x 40 hours per week = 98,000). This of course could be bull !, but seemed reasonable enough. Note the government has other ideas about how long you should work, though it probably improves the maths.

I know organisations like the BCSA (British Construction Steelwork Association) use the above formula. So as said above check what others in your industry, customers use.

You may find your HR dept have accurate info. If not it depends how accurate you want to be. As hours worked is the number you divide by you want it to be as large as possible, so you may want to take account of overtime etc. I would suggest that people don't actually work 52 weeks a year ( if they have an accident on holiday that's their problem). So take off two, three or even four weeks as appropriate to your org. You do not need to multiply the hours by 100,000, just the accidents. It also depends what period you want your stats over ( I used to have a 12 month rolling average), but could be worked out monthly.

So RIDDOR AFR if you had 2 RIDDOR accidents x 100,000 gives you 200,000 divided by (2,100 employees x 37 hours a week x 48 weeks per year) you get 200,000 / 3,729,600 = 0.0536

Simples :0)

imwaldra, I don’t suppose you would be so kind as to give me the name of the IOSH guidance you referred to. I tried to search for it, but drew a blank. I was struggling to get my mind around how this actually works, and I am always keen to review alternative ideas.

By the way, we used to use the actual hours as it was requested by the MD. It turned out that it made a significant difference to the AFR and was seen as commercially attractive. It was not difficult for us as we had an electronic clock in/out system, so the info was only a few mouse clicks away. I guess this may not be so easy for others.

Chris,
See 'Reporting Performance', p4, in the section on 'minimal reports'. I agree it doesn't leap out at you - apologies for not providing the detail in my first posting.

Thanks Smitch and imwaldra

I’m not sure I agree with the report. It states you should use frequency rates opposed to actual numbers, I agree with this saying one person had a serious injury is meaningless unless it is expressed in terms of how many people there are or amount of work done by those people. But by its very nature a frequency rate is a ratio of accidents / hours worked, so will be the same number for 10 people, 100 people or 1000 people, isn’t it.

So you employ 10 people and you have one serious injury that would be (1 x 100,000) / 20,000 =5 ( I have used the reports 100 people work 200,000 hrs so 1 person works 2,000 per year).

If you pro rata it up from 10 employees to 100 ( x by 10) you get (1 x 10 x100,000)/ 200,000 = 5

Or is it suggesting you only increase the hours ! so you get 0.5, I would not be comfortable with that.

I have to admit I find the comment in the report “very few have systems for measuring total hours worked” hard to accept. Personally I have found people not paid their overtime get a little peeved, so Company’s must have a system for recording this and would have thought it would be on either some fancy accounts package or a simple spread sheet for smaller companies. If nothing else the tax man wants to know how much everyone is paid, so there must be records of hours worked. My view would therefore be this information is more readily available than ever before. Yes some hours will not be recorded by managers who often don’t get paid overtime (or at least I did not).

If in the above example each of the 10 employees worked 2 hours a week overtime throughout the year that adds an additional 1,000 hours so you would get (1x100,000)/21,000= 4.76 this is a 5% reduction and not cheating just recording accurately.

I know the company I worked for the statistics were very important for new business, I also know for others it is not the same, so each to their own.

Also the report suggests reporting fatalities per 10,000 workers, again this should be a ratio, So 1 fatality with 10 workers would still = 5 AFR for fatalities again unless it is suggesting either
a) (1 x1000 x 100,000)/20,000,000 = 5 (no difference) or
b) (1x 100,000)/20,000,000 = 0.005 !!!!

Perhaps I’m reading it incorrectly or they should have put some worked examples to clarify what they mean. Because either what they are suggesting is pointless or they are suggesting ways to water down your statistics (which I’m sure they are not).

Am I the only person who is misreading the report? Or is there a flaw in the logic I used.

Sorry if I have hijacked the original thread, but I think clarity is worthwhile.