Hi everyone

sorry for asking a daft question here but I am trying to work out our AIR for the last 5 years, my question is - the number of accidents is this Riddor reportable accidents only or is this all accidents including minor accidents?

the same question please for AFR - should this be calculate dwith all accidents in a year or just RIDDOR reportable ones?

Originally Posted by: peter gotch

Robert - you can keep whatever statistics you want.

If in the UK, you have to record some information as a requirement of RIDDOR. That doesn't equate to working out AIRs or AFRs.

Might be that you are ISO accredited and have signed up to keep up statistical data.

Might be that for your client(s) you might have to keep certain data.

But, in broad terms your choice as to what severity you look at in stats and whether to align to RIDDOR without or without non-RIDDOR data, or some other parameter e.g. OSHA classifications.

What do you want to use the data for?

Thanks for replying, ive been asked for our last 3 yrs AIR in a PQQ I am completing and we havnt kept a record of this, but I tied myself up in knots earlier trying to work it out.

I am going to work it out uisng all accidents in the year (inc Riddors) divided by a "man-days" figure rather than number of employees, that way I can compare some sites against others (not all our project have employees on them , some are purely ran by subcontractors so If i didnt use man-days the site with no employees on it could have 10 accidents but return an AFR of zero)

Man-days seems like the only measure that will work across any project of any size, with our employees on it or not, example 2 electricians for 5 days on our project will equal 10 man-days

There is nothing wrong with writing Not Applicable, Not Recorded or just placing a strike out line in a box.

There does not have to be an answer to every question.

As to the PQQ if you want to sit and work out a value then ideally for consistency in communication you should always use the same method the client has adopted otherwise you are comparing apples with 6mm hex bolts which may be detrimental for giving an excessivley high or ridiculously low figure compared to what would be anticipated within the industry.

The clients who actually place weight on such statistics will tell you which method of calculation to use - the rest have created a box that needs a mark.

OK, Robert, it is the dreaded Pre Qualification Questinnaire where most prudent clients should have ditched most questions if a potential supplier has relevant certifications.

BUT, many haven't and continue to do their own thing. I got particularly annoyed with one very large company whose project managers all thought they knew better so the PQQ changed time and time again.

If the Client has not spelled out the criteria (as previously explained by others) then you need to decide 

(a) what units you want to use.

(b) whether you only want to report rates for your own staff

(c) whether you want to also report rates for your subbies either separately or combined with the numbers for your own workers.

AND make sure that YOUR terms of reference are clear.

If you don't you could end up looking bad when you are not!

As example compare the following two big contractors:

1. Management contractor whose workforce spend almost all their time in the office (even if that's a site office) occasionally walking around a potentially dangerous site AND who only report the numbers for their own employees - AIR or AFR should be very low - which looks good but might mask reality.

2. Main contractor who takes H&S very seriously and wants to know about everything that happens on their site including what is going on with the high risk trades - steel erectors, demolition, roofwork etc - so they keep comprehensive stats - AIR or AFR is likely to be MUCH higher

If the Client cannot clearly see the difference Contractor No 2 looks worse - actually likely (though not necessarily) to be better than Contractor No 1.

Contractor No 1 might actually have all the supply chain data readily accessible but if the question isn't sufficiently precise the people in marketing will quite often want to present the numbers that LOOK best.

Most of the time, a sensible client will realise that actually it is usually virtually impossible to read much into accident statistics - much more value in finding out what their supply chain partners do in terms of learning from incidents.