Eight times risk of death

tomasrojo

Because of this discussion, I remembered this paper, which is an attempt to compare risk in the UK for different transport modes, based on per-hour metrics and adjusting for gender, age and other factors.

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0050606

Concerning the first two errors, the importance of using the most appropriate measure of exposure has been demonstrated in inter-country comparisons. Risk by distance travelled does not capture large differences in average speed, which enable differential mobility for drivers, cyclists, and pedestrians. As the speed differential between cars and bicycles is not great for local journeys, time-based comparison minimises the distorting impact of misleading comparisons with long distance car journeys. The average “main driver” travels 7,034 miles, plus a further 1,254 miles as a passenger; the average cyclist rides about 830 miles annually.

When all relevant ICD-10 codes are used, fatalities by time spent travelling vary within similar ranges for walking, cycling and driving. Risks for drivers were highest in youth and fell with age, while for pedestrians and cyclists, risks increased with age. For the young, especially males, cycling is safer than driving.

It also mentions a distortion by miscategorisation I wrote about earlier in the thread:

The third category of error is to include off-highway falls and children injured at play as cycling transport injuries, whilst omitting all falls of pedestrians (whether on- or off-highway). This substantially overstates cycling injuries and understates pedestrian injuries

I don't quite see what the objection to per-hour metrics is, per se. I can see why you might want per-km as well, but per-km metrics are not intrinsically better, and it's not "comparing apples with apples" to restrict yourself to per-km metrics. Put it another way: say you want to know what your chances of incurring a fatal injury is if you cycle habitually for a lifetime. Even if you use the per-km metric, you won't end up with a risk eight times higher* than a habitual motorist, because you won't end up covering the same distance. The ratio of the lifetime risks for the habitual cyclist and the habitual driver is going to be much closer the ratio of the per-hour risks. So that's one quite good reason why per-hour is valid.


(*I'm using ESRI's claim, as misreported by the Independent for convenience here. It seems wrong on several levels, not least that it appears to be a measure of the probable outcome of a collision, not the inherent risk in using the road.)

HivemindXX

daithi7 wrote: »

Thank you for proving my point so clearly!! :-)

You say your commute is 45 minutes.... ehhhh but by what mode of transport does it take 45 minutes???
And what are the traffic conditions to make it 45minutes?
And what way does the weather have to be to ensure it is 45 minutes?
And if you stop for a coffee, or a chat, or a whatever, does it still take 45 minutes!?

You see your commute time actually varies day to day according to a whole load of factors, but I'm guessing, just guessing now, that it's always the same length? As measured in units of length I.E metres!!

Actually I'm pretty sure it is you who has proved my point. You simply won't accept a fact, either because you don't understand it or because it doesn't suit your agenda. I don't particularly care why.

NeedMoreGears

It might be useful if the data could be arranged into sub-groups of distance travelled (assuming that there is enough data for it to be statistically valid). Maybe something like 5km intervals up to say 25km - I would expect that to cover most of the time and distance accumulated by cyclists and it might eliminate the "motorway effect" for cars.

The other thing I'd like to see more emphasis on is non-fatal injury. It seems to me that a lot of debate is focused on deaths only, which ignores a lot of suffering and expense and reduces the data available. On the (possibly big) assumption that there is a similar correlation between injury rates and death rates for the various transport modes, looking at injuries might allow "better" analysis, assuming some consistent definition of the appropriate injuries

CramCycle

NeedMoreGears wrote: »

assuming some consistent definition of the appropriate injuries

Anything that leaves no permanent damage and does not require a hospital admission (whether one was taken or not) should be ignored IMO

daithi7

HivemindXX wrote: »

Actually I'm pretty sure it is you who has proved my point. You simply won't accept a fact, either because you don't understand it or because it doesn't suit your agenda. I don't particularly care why.

era yeah....

Qualitymark

NeedMoreGears wrote: »

It might be useful if the data could be arranged into sub-groups of distance travelled (assuming that there is enough data for it to be statistically valid). Maybe something like 5km intervals up to say 25km - I would expect that to cover most of the time and distance accumulated by cyclists and it might eliminate the "motorway effect" for cars.

The other thing I'd like to see more emphasis on is non-fatal injury. It seems to me that a lot of debate is focused on deaths only, which ignores a lot of suffering and expense and reduces the data available. On the (possibly big) assumption that there is a similar correlation between injury rates and death rates for the various transport modes, looking at injuries might allow "better" analysis, assuming some consistent definition of the appropriate injuries

I wonder do cyclist accident statistics sway the general statistics on "you're more likely to be injured on short journeys close to home".

Agree with needmoregears that statistics on injuries should be kept. In fact, there's a need for statistics on all accidents that are what you might call reportable - the ones where someone has to go to hospital. If a basic, standard description form were filled out by all parties in these cases, it would be clearer what hurts cyclists, drivers and pedestrians.

We already have some of this - wasn't someone saying that the most common fatal motorcycle accident involves a car driver coming out of a side road to turn left or right and colliding with a motorcyclist, for example?

Doorings must be a large proportion of cycle injuries, ranging from broken fingers to cyclists thrown out across the road in front of other traffic to fractured skulls from cyclists somersaulting over the door of the car when it's slammed open in their way.

I'd suspect that manhole covers and potholes account for a lot of one-person crashes - the manhole covers because the wheels slip on them and the potholes because crashing into one wrenches the handlebars out of your hands and over you go. (And again: if I were in the Council I'd be gradually moving all infrastructure that needs manhole inspections under pavements instead of under roads: less pressure on them, easier inspection, less traffic disruption and safer for cyclists.)

Probably the main killer is trucks turning left and crushing a cyclist on their inside; I saw a bin lorry the other day with a full-length glass door; the driver told me they were standard now - let's hope this happens with all trucks soon, as well as a sensor that can beep if there's a cyclist passing on the left.

tomasrojo

I suspect you're more likely to be injured on short journeys close to home (if that is indeed the case) because that's where a very large proportion of your journeys are spent.

I'm not sure what breakdown of statistics based on distance is being suggested. What distance had been covered from the start of the journey when the injury was sustained?

The safety of motorways has been quantified relatively well, as far as I know. For example, in the UK there are about 10 crashes per 100 million kilometres travelled on motorways, compared with about 70 on urban 'A' roads. (Again this is a per-km metric, and it may be important that each km on a motorway represents a smaller unit of time than on an 'A' road, but it's hard to see how the sevenfold disparity can be bridged. The relative severity of crashes on A roads and motorways probably needs to be included though.) I think the proportion of total km travelled by car that is on motorways is also reasonably estimated. So a reasonable approximation of risk on roads other than motorways can be made, though I don't have those figures to hand.

When More or Less, the BBC Radio 4 statistics programme, looked into how dangerous cycling was, they concluded that changing your journey from a car to bike probably entailed an increase in risk equivalent to switching from driving on a motorway to driving on a suburban road. They suggested that this wasn't an increase in risk that would bother most people in the second scenario, so it perhaps shouldn't exercise them so much in the first.

Lumen

tomasrojo wrote: »

When More or Less, the BBC Radio 4 statistics programme, looked into how dangerous cycling was, they concluded that changing your journey from a car to bike probably entailed an increase in risk equivalent to switching from driving on a motorway to driving on a suburban road. They suggested that this wasn't an increase in risk that would bother most people in the second scenario, so it perhaps shouldn't exercise them so much in the first.

This is why I always try to cycle to work on the motorway. The RSA told me they were the safest roads in the country.