Rank: Guest 
|
Posted By Jonathan Breeze
I've just prepared a report on accident stats for the last 12 months and have been asked the following by my boss:
"Why do you multiply the number of accidents by 1000 when calculating the Incidence Rate?"
To which I replied:
"Now that's a very good question, it's a standard formula that is used across industry, but I'll go and find out why."
To which he responded:
"I'm only asking because to the layman it looks like we've had more accidents than we actually have."
Two text books later and I'm still none the wiser.
So I'm opening the question to the floor - "Why do we multiply the number of accidents by 1000 when calculating the Incidence Rate?"
|
|
|
|
|
|
Rank: Guest
|
Posted By Heather Aston
So it's not a number with lots of decimal points? Easier for we mere mortals to grasp - and compare - whole numbers.
Maybe....
Heather
|
|
|
|
|
|
Rank: Guest
|
Posted By ken mosley
The simple answer is because it is an accident rate (ratio), so it requires rating against a number. By multiplying your number of reportable accidents by 1000 and dividing by the number of persons employed you will arrive at an acident incidence rate of x per 1000 persons employed. The multiplication factor is usually 100,000, in order that a comparison can be made with the annual stats produced by HSE.
Hope this helps.
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Okay, so far I have the following reasons:
1) It's the industry norm to calculate it this way.
2) Enables you to deal in whole numbers rather than decimals. (Thanks Heather)
3) We are looking at the general trend (downward I'm pleased to report) rather than the specific numbers.
Any more?
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Thanks for that point Ken,
We're talking all reported injuries, not fatal or major injuries, so I do not think the 100000 figure is appropriate as it would cause confusion.
The formula I'm using is:
Accident Incidence Rate = Total Accidents x 1000 / Mean No. of Employees During Period
|
|
|
|
|
|
Rank: Guest
|
Posted By ken mosley
Jonathon,
It doesn't have to be reportable accidents it can be whatever you wish, so long as you state what the rate is. For example, an AIR of 650 reportable accidents per 100,000 persons employed per year. Divide that by 100 to give 6.5 per 1000. Because HSE do not publish AFRs the most requested stats are AIRs.
The 'popularity' of AIRs varies for different industries. In my own industry (construction), AIRs are invariably requested in pre-qualifications to make a comparison with those in the annual report of HSE on construction.
|
|
|
|
|
|
Rank: Guest
|
Posted By Clive Lowery
When calculating frequency rates a multiplier (usually 100,000, which roughly equates to a working lifetime) is used in to provide a suitable value.
If it was calculated per hour worked, the figure would be so low as to be meaningless. Using a multiplier makes the final figure larger and therefore more understandable.
However the multiplier should always be stated and allowances should be made when comparing statistics using other multipliers.
(Abbreviated from the RRC General Certificate Notes)
Regards
Clive
PS My MD asked exactly the same question!
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Ken & Clive,
I note you both use the 100000 figure, but the texts prepared by Holt ("Principles..." 5th Ed.) and Ridley & Channing (Safety At Work 5th Ed.) both use 1000 when calculating Incidence Rate and 100000 when calculating Frequency Rate.
Why is this? (P'raps Allan could comment if he's around?)
Clive, I take on board your point about clearly stating which multiplier is used.
|
|
|
|
|
|
Rank: Guest
|
Posted By ken mosley
Because the HSE use 100,000
|
|
|
|
|
|
Rank: Guest
|
Posted By ken mosley
Sorry Jonathan, I forgot the other reason 100,000 is because it is recommended in HSG65 (appendix 6) "Accident Incidence and frequency rates". Though for frequency rates it uses one million.
Ken
|
|
|
|
|
|
Rank: Guest
|
Posted By Dave Wilson
Its originally from the HSE 'cost of accidents at work' I think, I have a copy somewhere as I think it is now out of print.
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Ken & Dave,
Thanks for those nuggets - I have both documents available at work.
I'll take a look tomorrow.
It does beg another question though - why do the 2 texts I've quoted use 1000 rather than 100,000?
|
|
|
|
|
|
Rank: Guest
|
Posted By Heather Aston
Jon
I still go back to the first answer.
Why do we use a multiplier at all? Why not just a straight ratio? Because it enables those sites who don't have many accidents to deal with whole numbers.
Why use 1000 for incidence rates and 100,000 for frequency? Leaving aside who uses what which is an obvioous reason now to choose to do the same thing for comparisions. Why did HSE et al pick that figure in the first place? Frqeuency rates need a larger multiplier than incidence rates becayse most sites have less accidents per hour than they do per employee.
Why 1000 and 100,000? Because the maths is easy - just move the decimal place.
BTW the US uses 200,000 for its frequency rate just to complicate the issue!
Heather
|
|
|
|
|
|
Rank: Guest
|
Posted By Pete Walker
There is a logical answer to all this:
Incident Rates are calculated using 1000 so we have a common rate between the many different sizes of businesses. The resulting rate shows how many people have been injured per 1000 employees - so everyone is comparing apples with apples.
Frequency Rtaes use 100,000 hrs to represent a persons working lifetime. So your rate shows how many times you can expect to be injured during your working life.
Hope this helps?
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Thanks Pete, That may answer part of the conundrum, but...
I've just taken a look at HS(G) 65 as previously suggested.
In Appendix 6 there is a calculation for Injury Incidence Rates just like Ken said. And behold the HSE multiply the figure by 100,000 (again just like Ken said).
But, the HSE formula uses the term "Number of reportable injuries".
Wheras Holt uses the term "Number of injuries" and Ridley & Channing use "Number of accidents".
To my mind these terms all imply different things:
the HSE term implies RIDDOR reportable injuries;
Holts term implies internal accidents reported that resulted in an injury; and
Ridley & Channings term implies all internal accidents reported irrespective of whether an injury was caused or not.
What's going on here? The more I investigate, the more complex it becomes.
|
|
|
|
|
|
Rank: Guest
|
Posted By ken mosley
Not wishing to confuse the issue further you can argue whether the numbers are only relevent to your employees or should you include contractors, agency workers and visitors?
Whether you use major, over 3 day reportable accidents, lost time accidents or accident book entries is not really an issue. You should ask why you want the data and what you intend to do with it. Is it to be a KPI ? is it part of a prequal ? is it for inclusion in the annual report and accounts? does it form part of your report on corporate governence (Turnbull)? or do you wish to use it as part of a continuous improvement programme. You need to decide what suits your purpose best. Alternatively ask the person who made the request the question why?
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
All fair points Ken.
It was originally for a graph to be submitted with a board report (so corporate governance I guess).
I made it clear to management that it was for employees only although the other data was available.
We went for accident book entries and intended to use the figures to compare with the previous two years data. So I was comparing like with like.
I had also submitted details of injuries by accident type with a view to focussing h&s training on problem areas.
|
|
|
|
|
|
Rank: Guest
|
Posted By Allan St.John Holt
Hi folks,
Sorry for the lag but I've been on the road for a while.
The simple reason I used the figures in 'Principles' as correctly quoted above is (sadly and simply) because they are what officialdom uses. I referred to number of injuries, Jonathan, rather than accidents, because I feel that the term 'accidents' is widely misused and misunderstood and 'injuries' is a better tag for students to use. I don't want to open a debate on that, but it's to get away from the "accident = and therefore unpredictable" school of thought.
Pete Walker's logical answer contains the best short summary to my mind. 100,000 hours is indeed a notional working life. We (at work here) use the rate of per 1000 Full Time Equivalents, which is a good way of included part-timers and casuals.
On the whole it's one of life's mysteries, and conventions. Personally I would much rather that we spent less time looking at lagging measures and a lot more time looking at means of measuring success. But, accidents are cheap and dirty stats to collect and compare, and so they appeal to the top floor.
Allan
|
|
|
|
|
|
Rank: Guest
|
Posted By Merv Newman
Allan,
being on the road so much can only cause you to be reactive instead of the much preferred proactive method.
Americans measure injuries per 200 000 hours. This was originally used to indicate a percentage of employees being injured - they have far fewer holidays than europeans and tend to work longer weeks so a year of 50 forty-hour weeks equaled 200 000 hours. An OSHA rate of 1 means that approximately 1% of the workforce has been injured.
Given the above, can any tell me how to compare OSHA rates with UK rates ?
|
|
|
|
|
|
Rank: Guest
|
Posted By Heather Aston
Merv
50 x 40 = 200,000??? Straight to the back of the maths class with you.
We use the OSHA rate as a Company world wide measure - it was the easiest for all our factories to calculate from their current records.
You cannot easily directly compare OSHA and UK (I assume you mean RIDDOR) because OSHA incidents are one shift absences and certain medical treatments. I find as a rough guide in our industry we have about three times as many OSHA incidents as RIDDOR ones.
Heather
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Thanks for your comments Allan.
I think I'll stick to your method but with lots of obvious labels and explanations for dummies... er directors.
I would love to calculate the rate per 100,000 hours as we have loads of PT staff and this figure would be very helpful.
But when I approached HR and requested the figures, I got 'the' look.
Merv - it's not relevant to me, but still a good question. Do they just half the figure or is it more complex than that?
|
|
|
|
|
|
Rank: Guest
|
Posted By Heather Aston Jon It's more complex than that :( There is a specific and deliciously complicated definition of an OSHA incident - see my last answer for brief outline. If you want to know more try www.osha.gov you'll be sorry! Heather
|
|
|
|
|
|
Rank: Guest
|
Posted By Jonathan Breeze
Heather,
As our company has no premises or employees Stateside, I think I'll give that one a miss.
|
|
|
|
|
|
Rank: Guest
|
Posted By Merv Newman
Heather,
grovelgrovelgrovel. 50x40 = 2000, the number of hours worked by one (american)person per year. I missed out the 100 multiplier which makes it a rough percentage. Can I go back to my seat now miss ?
And has the 100 000 lifetime hours been recalculated in respect of the Working Time Regulations ?
Your right, osha defintions of what and what isn't recordable are fascinating. Do I remember something about "painting a flagpole once a year" ?
Actually, I know a number of multi-national companies such as yours who use osha stats with one-shift standards for internal reporting.
Means that the poor safety person has to keep two lots of stats.
All other european countries work on one-shift and a million hours. (dunno about lichtenstein)
Trust the b***** British to do it differently.
Do you know where the three-day rule came from ?
When National Health came in there was a requirement for that ALL absences related to injuries and illnesses be certified by a doctor. BMC screamed that this would give them too much "trivial" work to do. Hence the three-day rule.
|
|
|
|
|
|
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.