Why your accident frequency rate isn’t just a number

An accident frequency rate of 3 seems likes it’s close to zero but actually it means everyone in your organisation (on average) is going to be injured three times in their working lives.

The recent release of occupational injury and ill health statistics for 2010/11 by the HSE has attracted much debate. Yes, there was evidence of a continuing downward trend in the number of people injured or made unwell at work: 24,726 compared to 26,268 in 2009/10.

As the chair of the HSE, Judith Hackitt, rightly said, “Britain can be proud that it has one of the best health and safety records in Europe.” However, there was also an increase in the number of fatal injuries: 171 in total, up from 147 the previous year. This is clearly of concern, with Judith Hackitt adding in response, “we can never let up in our commitment to addressing the serious risks which continue to cause death and injuries in the workplace.”

The reasons for both the good and bad results have been much debated and the overall sentiment that we must always continue to improve cannot be argued with. However, the figures got me thinking about how Health & Safety statistics can so easily be ignored or misunderstood. The figures above are absolute totals and are very hard to misinterpret: cases of injury and ill health have decreased, fatalities have increased. Simple. But we very rarely deal with straightforward numbers. We always calculate an Accident Frequency Rate – and this is where the confusion often begins.

Let me just clarify one thing. This isn’t a criticism of accident frequency rates. Far from it. There is a growing tendency to focus on leading edge indicators such as closed out near miss reports, employee suggestions – and rightly so. It is important to concentrate on prevention rather than only dealing with incidents after they’ve happened. But it’s still really important to monitor the trailing edge measures, like AFR, to gauge how effective your leading edge initiatives actually are. Also, calculating a rate using a standard formula allows comparison across time, departments, sites, companies, industries, countries etc. and help us to quantify improvements and differences. Using a standard approach to accident frequency and continuing to measure it is a good thing. Being blind to what the numbers really mean though is not.

Sometimes the AFR can feel a little bit abstract. On its own, an AFR just looks like an arbitrary number. Out of context, it’s easy to ignore, simple to misunderstand and a doddle to regard as “pretty good actually”. This was proved to me recently when talking to a company director who was really enthused by a year-on-year reduction in his company’s AFR to 2.3. The corporate target is ZERO ACCIDENTS and he concluded that 2.3 was pretty close to ZERO, so they were doing OK.

The target is zero accidents and he concluded that 2.3 was close to zero

As an arbitrary figure, yes, 2.3 is close to zero. With this perception of what the AFR means it would be easy to imagine the directors perhaps easing off as far as Health and Safety is concerned and concentrating of the “more important” targets they had. But as far as I’m concerned 2.3 isn’t really close to zero at all, and so I discussed the AFR with him in the context of the numbers of people potentially affected and how. He began to see things very differently.

Behind any AFR are the real stories; the human cost. Any frequency rate is a reflection of the number of accidents that have actually occurred. And accidents more often than not result in someone actually getting injured. So instead of the AFR just being a lonely number, we began to consider it in terms of the number of people who were getting hurt. Amazing how people are more interested in the story behind a broken leg than they are in the number 2.3.

So, let’s look at the potential human cost by reviewing how we calculate an accident frequency rate and seeing what it really means. The most commonly used formula is:

AFR = No. of RIDDOR incidents x 100,000 / No. of hours worked

So the output of this basically provides a rate; a number of RIDDOR incidents per 100,000 hours worked, in this particular case 2.3. To show why this type of calculation is used, imagine a close competitor declared twice as many RIDDOR injuries. On paper this looks like a much worse record. However, this competitor has twice as many employees who work approximately twice the number of hours. Therefore, they both end up with approximately equal AFRs. They are declaring their performance on a level playing field – the number of RIDDOR incidents per 100,000 hours worked.

So, 100,000 – there’s another arbitrary, abstract number. But it isn’t. It isn’t selected just because it’s a big, round number. It’s used because it’s a round number which roughly equates to the number of hours we put in on average in a working lifetime. So our AFR of 2.3 means 2.3 RIDDOR reportable injuries per working lifetime. Each. That’s right, 2.3 each, per person, for everyone.

An AFR of 2.3 means 2.3 injuries in their working life. Each.

I told the director their AFR was unlikely to get any better if they were collectively satisfied with it. In truth, it’s more likely to start creeping back up but we agreed to assume it would remain fixed at 2.3. That means that if the injuries are shared out equally, every employee can expect to have 2.3 RIDDOR injuries during their working lifetime. But let’s also assume that not everyone is going to have their fair share of the RIDDOR incidents; the office staff, the managers, the directors themselves aren’t likely to have as many. So the remainder of the employees will have to “take a few for the team”. They’re likely to be facing around 3 each now.

The director really connected with this analogy. All of a sudden he realised, it’s the number of injuries that each person who works for him was likely to sustain during the course of their working lifetime. Injuries so serious that at the very least they’ll be unable to work for nearly a week. Or worse. Who’s going to have the dislocated shoulder? Who’ll be happy to take the broken ankle, the chemical burn, the penetrating eye injury, the amputation? Allow enough to happen, share enough around and he might be left guessing which of his staff will have to have the fatal accident.

So, I asked him this question, “If the board is satisfied with 2.3, why not tell everyone that’s what to expect in the future. Why not get it out the way and ask for volunteers to take the nasty injuries? If the AFR sticks at 2.3 it’ll happen anyway so why not plan for it?”

All of a sudden, that arbitrary number looked very different to him. He realised it really wasn’t anywhere near zero. We agreed that the target to strive for zero was there because it wasn’t just a number to boast about, it had a real meaning. It meant people weren’t going to get hurt. No-one.

So, if you’re ever facing some ambivalence about the need to improve your AFR, or the people who can help to improve it think it’s good enough or don’t really think about what the number actually means, perhaps present it to them in this context. Let it sink in, let them imagine the way all of those injuries will affect their workforce and their families. And before they move on to the “more important things” ask them, “What is more important than preventing a single person from suffering a couple of serious injuries?” Well, preventing your entire workforce from suffering a couple each – that’s what.

Statistics and quotes from: http://www.hse.gov.uk/press/2011/hse-statistics2011.htm – accessed 17/11/11.

5 thoughts on “Why your accident frequency rate isn’t just a number

  1. Target for TRC FR (Total Reportable Case Frequency Rate) should be 0.00, because, target must be highest (i.e. lowest FR).
    While benchmark is set on different various criterian e.g. average of record of last three years of the organization in question, same for the sister concerns and similar industries shouls be considered for setting benchmark FR.
    Also, benchmark should be realistic and achieavable. Too low (high FR) benchmark may encourage deteriorate safety performance hence culture, while on the othe end to high (low FR) benchmark may sometime bring in frustration !

  2. Thank you for the explanation on AFR & why 100,000 hours are used by RIDDOR, but why do companies change that number my company uses 200,000 & I have known others use 1,000,000?

    • Hi Joy

      Thanks for the feedback.

      Based on a nominal 40 hours a week, 50 weeks a year (as we use in 100,000 hour calculations), 200,000 hours is the approximate total hours worked by 100 people in a year (40 x 50 x 100). So the figure per 200,000 hours indicates the percentage of a workforce hurt each year. Just like the analogy used in my blog, this output could be put into context really effectively. For example, let’s imagine the rate per 200,000 hours is 10 – that means 10% of the workforce will suffer a RIDDOR injury in a year. When conveying this to a management meeting attended by, say, 10 people, that effectively means on average one on them should be guaranteed a RIDDOR injury per year – you could ask for a volunteer to take this year’s injury on behalf of the team!

      As for 1,000,000 hours – I’m not too sure to be honest. I know 1,000,000 man hours injury-free is an often celebrated milestone and so companies may calculate their frequency rate in this way because that is their target; zero injuries per one million hours.

      I hope this helps – and thanks again for your comments.


  3. I have always been stressing myself out about our company’s AFR target. Have we set an achievable target is the target to high or is it too low, are we better than our competitors, will the target appease our clients and so on. When I attend meetings with our clients they are more than happy to point out “why haven’t you hit your AFR target for last year” and “what action are you going to take to bring your AFR figure down to your set target”. I would always go away and feel that we had failed because we did not hit our AFR target. This coming Monday I have to give a safety briefing to one of our clients. My MD came to me today and said “do you think it is going to be a problem that we haven’t achieved last year’s AFR target” my reply was “I dare say they will have something to say about it”. After reading your article it has eventually made the penny drop. I now feel allot more confident and when I am asked next Monday about “why haven’t you reached your AFR target”, you know what my reply will be? Thank you for taking some of the pressure of my shoulders. It is a very good article.

Leave a Reply

Your email address will not be published. Required fields are marked *