Incident Report 2005

[ric]

I noticed the 2005 incident report is available at:
http://www.bsac.org/techserv/increp05/intro.htm

I'm wondering if anyone knows if it's possible to normalise the findings so that the statistics show the number of incidents in relation to the total diving population? A % graph would then show whether or not the number of incident, fatalities, etc was proportionally lower or greater than in previous years.

Even with this kind of analysis it wouldn't change the fact that loss of buoyancy is a very disturbing causal factor in many other types of incident. Irrespective of "normalisation" it's very worrying.

Ric

[resting rifleman]

i dont think it is possible to collate this info short of everyone who dives, logging every dive that they carry out! and informing a "body" of this! so realistically that would be the c/g, and it seems to me that this would need legislation (it has to be paid for!) and thats a big can of worms!:eek:

[Michelle Haywood]

Ric

It's a good suggestion and one that has been made before. There was a suggestion earlier in the year that a random selection of branches would be asked to log all their dives in order to do the sort of analysis that you suggest (or at least have a stab at it). However, I don't know what happened because that information doesn't seem to have appeared.

Even then I would be cautious about extrapolating from it. Take the number of incidents reported for example. If you look at the year on year trend and put a line across it (see the Incident Reports for 2004), there appears to be a very alarming trend. In reality, the first few years have incomplete data sets due to a 'take up' period for the reporting forms. That's why in this year's report it appears much more stable (and reassuring) that the number of incidents reported isn't spiralling out of control. It has occurred to me to look at what the predictions would be if you just took the last 5 years for example - it's on my 'to-do' list when I have a spare half hour.

There are inherent problems in collecting data under any self reporting scheme. Whereas HQ and Brian Cummings do a great job of following up the fatalities, it's been suggested by Brian (at the DOC on Sat) that maybe only 50% of the incidents requiring recompression are actually reported. This means that the errors in trying to put too many statistics onto the reports are just too large to have any real confidence. The best you can hope for is spotting trends...but would buoyancy be the real concern if we knew what happened to the other incidents requiring deco but for which we have no data?

I do agree with your point that there are possibly more sophisticated ways to analyse the data. It's something that's been bugging me for a couple of years now.

Michelle

[Fiona]

I do agree with your point that there are possibly more sophisticated ways to analyse the data. It's something that's been bugging me for a couple of years now.

Michelle

Do we really want to be over analysing the data which is collated. I am not sure I want to see more. I have been reading the incident reports for quite a few years now and I have found them an interesting learning tool.

Lets not forget these are collated from all sorts of agencies and trying to work out incidents per 1000 or even 10000 dives wouldn't work unless you knew how many dives in total there were from all agencies and this just won't happen.

[Michelle Haywood]

Fiona

Points taken, but because we collect data at all people will look at it, including the Health & Safety Executive and the Maritime & Coastguard Agency. I attended a presentation by the Coastguard Commander for Dorset last night, where I learnt that the Coastguard have a ministerial target to reduce the number of deaths and serious incidents from diving. As always there is a budget to go with the target, and the current approach is to use education not legislation. Diving casualties cost approximately 5 times more than any other casualty that the Coastguard deals with.

In light of the predicted increase in incidents that came out of the 2004 Incident Report, several meetings took place involving the MCA, BSAC, PADI etc to try and prevent the number of incidents spiralling. This is all very positive, but in my view a bit of a waste of resources, because you only had to allow for the 'take up' period in the early years of the incident reports, not include that data in the total number of incidents, and actually the number of incidents per year is much more stable. Far from hitting really high numbers of incidents in the future, the trend would tend to be a gradual and fairly small rise.

It's the wrong use of statistics to draw messages out of the data that we have that can actually be dangerous to our sport. I agree that the IRs have a lot to teach us, the danger is how that information is presented because it's the headline graphs that external (legislative) bodies look at.

Michelle

[ric]

Some interesting points made here. I had overlooked the other factor, the number of dives taking place annually, and not just the number of divers, regardless of agency.

I agree that over-analysing data is a a waste of time, but I wasn't thinking that we should do the data to death. Keeping it really simple: do we have a ball-park figure for, say, the number of divers 10 years ago compared to today? Or some idea of the number of dives undertaken annually compared with 10 years ago. If, crudely, the number of divers has doubled or we do twice as many dives now compared to 10 years ago, then a similar figure for DCI 10 years ago compared to now would show that the incidence had roughly halved.

I'm not after a statement of significance at the 5% level... just wondering (and hoping) that things are getting better, not worse.

[Michelle Haywood]

I think the nearest we can get is membership numbers, and BSAC's have been declining. I suspect PADI's numbers would be commercially sensitive info, and then you have the issue of people crossing training agencies...

If you normalise for membership, then I suspect nothing much has changed statistically in the last 6-7 years, given the limitations of the data you have to go on.

I guess the take home message would be don't panic, but don't get complacent either. However, it's not the most stunning of conclusions considering the resources needed to get that far!:rolleyes:

All the thoughts I've had about whether you could have some reference point seem to end in too many confounding factors eg if you counted number of fills then all the dives done by rebreather divers and those with their own compressor wouldn't appear (plus the burden on commercial operators to count it and for no real immediate gain to themselves). Or how about asking the coastguard to lift the info from their voyage plans that everyone calls in? Well, apart from the fact that quite a few clubs still don't call in voyage plans, some coastguards don't record the data anyway, and then you've excluded the inland and shore dives. Or a sample of clubs from which you could get a 'average number of dives per member'. It's possibly the most do-able, but clubs that are likely to volunteer will be those that are actively involved in BSAC and (I'm just guessing) possibly a little safer. The trick would be to also get hold of clubs that don't train, but just go diving, only I think they probably don't engage with either HQ or regional teams. This introduces a huge ascertainment bias.

There are no easy answers on this.

Michelle

[Nigel Hewitt]

I suspect PADI's numbers would be commercially sensitive info, and then you have the issue of people crossing training agencies...Since you don't ever leave PADI's lists unless you are a proffesional I don't see how their numbers can go down. You can die and still be on the list.

I'm on PADI, BSAC the SAA not to mention TDI and IANTD so I am probably at least three UK divers.

[Michelle Haywood]

So you're a typical example of why membership numbers are probably not going to be much use.

I love it when someone proves me right!:D