Ask them about:

• The nature of each incident
• What they learned from them
• Actions they have taken in response

I once queried a client during an audit as to why they had chosen to give a contract to a roofing contractor who had fatality in the last year.  The answer was that they had clearly learned a lot of lessons from the incident, they had made a lot of changes to the way they managed H&S on site, changed how they did risk assessment, why they did them and what information they gave to the employees, they were more willing to talk about the hard questions and gave better answers than other contractors who were in fact cheaper and on paper better from a statistical point of view.   The statistics don't give you the whole picture!

Another PC ?

Well, there is more than one way of looking at this. First, a company working in a relatively high risk industry like construction 13 RIDDOR reportable incidents may not be so high, especially given over 10 million working hours achieved.

Another way to look at this is their systems and processes should be better assuming of course they have taken corrective measures post incidents. Obviously their reprting systems are good - Lol.

Seriously, I once worked for a company who had 6 fatalities in a year! Ok, that was worldwide, but it did not necessarily mean they were a bad company. Indeed, they are still one of the largest construction/rail companies in the UK.      

Unless my mathematics is incorrect, this organisation’s AFR is 0.17 for 2019, based on 10 million hours.  Without opening up a debate on the benefits of AFR etc. and in the context of a tender,  the figure is generally below a level that would normally raise an eye brow in the construction industry.  (Assuming these are all injuries and not a mixture of dangerous occurrences) 

Furthermore ‘10,000 000 hours worked’ for one organisation in one year is quite significant and I suspect the context of this may somewhat be lost.  I would assume this is nation/sector wide on multiple programmes for multiple Clients with multiple stakeholders etc. 

I suppose the point is that if I were looking at this organisation in the procurement phase, I would wish to understand the companies organisational context.  They clearly have a reporting culture.  Regarding their systems to manage adverse events, what else events could you ascertain (as others have said – lessons learned, system reviews after incidents etc.)  For me this is about context. 

The statistic alone wouldn’t mean a ‘no’ for me. 

Nivedita, trying to read anything into accident statistics is a complete minefield.

You need to explore all sorts of variables that make benchmarking one Contractor against others close to impossible.

One as already alluded to is how much underreporting (and/or underrecording) some Contractors may have.

Another is whether a Contractor does or does not include stats for its supply chain. If they do, then that may well include data for higher risk processes, such as roofing and scaffolding.

In contrast a Management Contractor who employs very few tradespersons should have very, very low stats if they do NOT include data for their supply chain.

....and there are many other variables. What is important is to consider how effective the Contractor's processess are, inclusive of those for investigating incidents and sharing lessons (as appropriate - sometimes there might be nothing to learn!)

Originally Posted by: peter gotch

Nivedita, trying to read anything into accident statistics is a complete minefield.

You need to explore all sorts of variables that make benchmarking one Contractor against others close to impossible.

One as already alluded to is how much underreporting (and/or underrecording) some Contractors may have.

Another is whether a Contractor does or does not include stats for its supply chain. If they do, then that may well include data for higher risk processes, such as roofing and scaffolding.

In contrast a Management Contractor who employs very few tradespersons should have very, very low stats if they do NOT include data for their supply chain.

....and there are many other variables. What is important is to consider how effective the Contractor's processess are, inclusive of those for investigating incidents and sharing lessons (as appropriate - sometimes there might be nothing to learn!)

Peter is spot on. It may be that the PC is including RIDDOR's for all work on their sites. For example, they may include RIDDOR's that are reported by a sub-contractor in their stats. We include subbies for our organisational stats, rightly or wrongly.