If I'm writing a risk assessment and I've identified a hazard eg. coliding with a motor vehicle whilst riding your bike. How do I rate it's severity? do I think of the worst that could happen eg. a bus running over your head and you dying of severe head injuries; or the statistically most likely outcome, more minor injuries, maybe a broken bone? Or should I list both of these scenarios seperately and rate their likelehood individually?

Any help welcomed!

Originally Posted by: GrahamC

  our organisation provides cycle training to individuals and groups on public roads, this is a hazard participants face.

Well, maybe your risk assessment can cver different scenarios separately, and consider if additional controls could reduce the risks, though I can't think what additional controls are available.

But I suppose severity varies according to these scenarios: cycle path through a park, cycle path in a 30mph street, riding along a 30mph street with no cycle path,  riding along a 40mph street with no cycle path,  riding along a 60mph country road with no cycle path......

The controls are always the same? Helmet, knee pads, training in traffic awareness, hmmm... what else is there?

John

Personally when looking at severity I always look at the worst case scenario- I don't bring likelihood into it at all. Then I look at the risk with existing controls and if that is not adequate look at additional controls to bring the risk down.

One thing you might do is to split the risk assessment across several scenarios. So instead of one risk assessment for cycling, you have different ones for different road conditions or off road cycling vs road cycling etc. Whether you think that is worth it is down to you. You could end up with half a dozen identical RAs, which is a bit pointless.  

I would try and take some of the scenario bit out of the eqation- consider whether you could sustain minor, major or fatal injuries, or none at all.

Sometimes we cannot reduce the severity of an event. You cannot control everything in this kind of scenario, but you can take steps to reduce the liklihood of it occuring instead i.e use of cycle paths, only riding in daylight hours, consider undertaking activity outside of peak times roads are in use. It won't mean the injuries would be any less severe were an accident to occur sadly, but may make it less likely to happen instead.

Likelihood and severity are two very different concepts.

Originally Posted by: Graham

Knee pads? whatever next, hel;met yes knee pads no. What are the ages of the people are they kids just learning or are they adults who want to have a better understanding. If they are adults I would suggest this is money for old rope.

Originally Posted by: JohnW

But I suppose severity varies according to these scenarios: cycle path through a park, cycle path in a 30mph street, riding along a 30mph street with no cycle path,  riding along a 40mph street with no cycle path,  riding along a 60mph country road with no cycle path......

The controls are always the same? Helmet, knee pads, training in traffic awareness, hmmm... what else is there?

Lots of interesting inferences in that posting.  Why is PPE the first control that came to mind?  We'd never accept it in the workplace, but as soon as it's cycling everyone’s first thought is that PPE for the victim is the most appropriate control.

Also, 'cycle path' doesn't necessarily make cyclists lives safer.  The accident frequency is, I believe, higher on cycle paths, but I don't know of any work that looks at aggregate severity.  The mechanism (ie, why cycle paths are worse) seems to be that the highway designers of the UK seem singularly clueless when it comes to integrating cycle paths into junctions.  That is, cycle paths and lanes probably make the 'just riding along' bit safer, but they often make the 'negotiating junctions' more difficult and more unclear (for all users - cyclists don't know if cars will give way, cars don't know if they are supposed to give way, everyone needs to be looking in twice as many different directions as once).  Since the majority of 'accidents' are at junctions anyway, making the junctions worse (and more numerous - another favourite trick of teh highways designers) has a bigger effect than making the in-between stretches better.  But that's largely speculation.

It probably is true that riding in the park leads to lesser severity injuries than riding on the street.  That's the underlying flaw in lots of the helmet effectiveness studies.  It does seem to be true that a helmeted rider in hospital (on average) has less severe head injuries, but it's also true that a helmeted rider is more likely to have been riding in the park.  In American accident data (where lots of the research originates) it's also true that poor cyclists are less likely to have been wearing a helmet and poor cyclists are also less likely to present at a hospital for relatively minor injuries.  Concluding that it was the helmet that made the injury less severe is potentially tenuous if you automatically discount that the helmet made a park appear for you to cycle in and also made you rich.

Originally Posted by: andybz

On the subject of cycling, the main thing missing (in my opinion) is the benefit.  There will always be risks, many of which cannot be fully controlled.  Without stating the benefits it will be difficult to justify. 

Very true.  There used to be an official BMA statement (policy briefing or something) that basically said that cycling was, on average, good for you whether or not you wear a helmet.  I don't know if it is still current. On average cycling extends your healthy life, however you go about it.

That's also why I run up stairs not holding the handrail (depite the stuff about 'mature safety culture' sometimes trotted out) - not only is it fun, it's good for me (in my opinion, though not that of some killjoys).

However, there's nothing (or very little meaningful, anyway) in most company health and safety systems or procedures about promoting or encouraging good stuff - it's nearly all about preventing bad stuff, isn't it?

Originally Posted by: Ian Bell2

H&S people need to grow a pair and concentrate on significant risks.

That depends on whether the H&S person's job description is to protect the health and safety of the workforce, or to protect the company from the risks arising from health and safety matters.  In the latter case, small minded auditors, claims companies and accreditation schemes are all significant risks.

Originally Posted by: jwk

The old version of MS had some very useful graphs, which admittedly were explained very badly.

It's all about frequency, so in answer to your query, is it worst case or most likely, the answer is; it's the most likely worst case.

So, to take an example, I have known two people almost die as a result of a cut to a finger. One had a heart attack on the operating table when their abcessed finger was being drained, one suffered major blood-loss when she tore the little artery in her little finger. Do I thenrate a cut finger as severe? After all, I have known two people who nearly died. Of course I don't; almost all cut fingers are minor inconveniences, so a graph of outcomes for a cut finger would peak at the left, with a very long tail ending in a miniscule number of deaths.

On the other hand a graph of outcomes for falls from a height would peak somewhere towards the right; a few people would get up and walk away from a 20 metre fall (or even very much higher given the right freak circumstances), but a large proportion of the incidents would result in quite severe injury or death.

So would your graph of outcomes be skewed left or right? That's how you determine severity,

John

This is very useful. Thank you.

It's hard to find outcome data from cycle + motorised vehicle collisions. But for reported outcomes of all on road cycling accidents, for the type of riding we do, the percentage of fatal/serious injury is about 0.3%. Which might indicate that this is not the most appropriate severity to state on the risk assessment.

Is there a point to the matrix risk assessment when using numbers an example,

10X10 risk matrix,

Severity

1 No injury, 3 minor injury, 5 Lost time, 8 Severe, 10 Fatality

Likelihood

1 remote, 2 unlikely, 5 likely, 8 Very Likely, 10 Certain.

I asked why not make it a 5X5 matrix if that is what they want to use and was told we can't because it won't work. When I asked why I was told that if people see the numbers high then they will take more notice.

Is that not an eduacation thing, because the numbers shouldn't matter it should be the control measures that matter and employees being educated in understanding the assessment.

I am not saying it doesn'e have a value but if you score things to high just to get the message across then again the function of the R/A is pointless.

Just reading some of the postings on here and I am one of them, who are the R/A's for are they for managers to cover backsides or are they for the users. If it is for the users then writing them for disco's sports days etc, are relevant so that participants no the control measures as well as the risks.

You could class them 'young users'

I don't use a matrix system for all assessments, I don't on fire, COSHH, Manual Handling, DSE, First Aid and some others.

When my first youngster was born, somebody "helpfully" bought us a "Baby Book" listing all manner of ailments & their potential consequences. If you looked up "sniffle", you were given a list of likely outcomes, which included "Death" & for some reason, your eyes were always drawn to this!

Fast forward a number of years to my career in H&S & nothing appears to have changed. I carried out some work for an organisation providing football in the community as a means of keeping kids off the streets & their Risk Matrix using numbers, concluded that the most likely outcome was "Death"!

When I pointed this out, the response was that of course that wasn't what was meant, so the figures would be altered!

I accept that in some industries a numerical Risk Matrix may be useful, indeed necessary, but I find it much easier to stick to the HSE's guidance & keep it simple.

Ensure sanity by sticking with the most statistically likely outcome. Avoid scoring matrix if you can, otherwise restrict its use to prioritising outstanding actions by the employer. Also consider giving priority to hazards and controls affecting higher populations.