Decompression Tables

[DCBC]

That is not an accurate use of statistics. For your conclusion to be true, every diver would have to dive 2,200 dives, and every dive would have to have an equal risk for DCS.

In reality, some divers are very aggressive and push the limits regularly. These divers should have a much higher incidence of DCS than the rest of the population that dives more conservatively. At the other extreme, many divers stay almost exclusively in places like the reefs at Key Largo, where it is just about impossible to be bent.

If the familiar 80/20 rule has any bearing, then we might expect 80% of the DCS hits to occur among 20% of the diving population. Given the actual number of dives people take and the fact that most do not push the limits, it would not surprise me to learn that more than 90% of the population will never experience DCS.

John,

Assuming that the statistics given by DAN are accurate, it's reasonable to assume that they include various places, at various depths and included various experience levels. The ONLY constant given is that they ALL were using dive computers.

I fail to understand the point you're trying to make... If not 1 in 2200 dives, perhaps you might enlighten me on the math you are using... Given these figures what would you say the statistics are? 52,168 recreational air dives, 23 DCS incidents.

 

[NJDiver07866]

My rule is you dive enough you will get bent, regardless of what table you use. Don't waste all your research trying to find the 'perfect' table and looking at which ones are 'conservative' or 'safer'. Don't go banannas looking at comparing NDLs and SI credit. Take this time and hydrate yourself, keep physically fit & listen to your body. Everyone has 'silent bubbles' after a dive. You dive long enough sooner or later one will say hello. This is why I think less about the table and and more about the color of my urine, what the water temp is, how much I work at depth and how I feel after the dive. Yes statistics are important, and i respect the people trying to define and improve diving by reducing DCS. But for me, I'm a chicken, I have decided that I won't leave the water unless all my tissue compartments are less than 80% boil. This is my personal line....not based on any statistics... just based on I like the number 80. And as compartments go... they are 'theoretical'... so hopefully numbskull is taken into consideration as one of the compartments because I do have that one....

 

[boulderjohn]

John,

Assuming that the statistics given by DAN are accurate, it's reasonable to assume that they include various places, at various depths and included various experience levels. The ONLY constant given is that they ALL were using dive computers.

I fail to understand the point you're trying to make... If not 1 in 2200 dives, perhaps you might enlighten me on the math you are using... Given these figures what would you say the statistics are? 52,168 recreational air dives, 23 DCS incidents.

I am going to greatly simply this process to show the point--I am aware that the statistics are more complex than I will present.

Let's start with your assumption that all divers will reach 2,200 dives in their lifetime. I don't believe that is anywhere close to reality. Where I live, divers are almost exclusively vacation divers, meaning that they will dive during one vacation a year or maybe every other year. A 22 dive year would be a good year for most of them, and they would have to put together 100 of those years to reach 2,200. I myself started diving in late middle age, which recent statistics indicates is very common. I am chalking up a lot of dives these days, but I doubt I will hit 2,200 in my lifetime. As has been noted many times in ScubaBoard, many--perhaps most--divers get certified on a whim and leave the sport before too long. If the average diver accumulates 100 dives in a lifetime, I will be surprised.

But, just to make the math easier, let's assume the every diver hits 220 lifetime dives. Let's assume that we have a representative sample of 10,000 divers. That will give us 2,200,000 total dives, and 1,000 total DCS cases. We will also assume that no one has more than one DCS case in a lifetime.

Let's also assume that we do have a normal distribution of divers in our sample. This means their diving habits should fall into a normal bell shaped curve. As I said earlier, a relatively small percentage (let's assume 20%, which is not far off from a normal distribution) will aggressively push NDLs frequently, and a relatively small percentage will never push NDLs. The largest group will be in the middle.

Given that distribution, we can expect that the 2,000 divers in the aggressive group will have most of the hits. Normal distribution would predict that they would in fact have about 80%, or 800 DCS hits. That leaves 200 hits to be distributed among the 6,000 divers in the middle group (who would get most of them) and the 2,000 divers in the cautious group (who would get very few).

Of course, that makes some assumptions that have not been proven, such as the assumption that divers who aggressively push NDLs are more likely to get DCS. I think that is a very safe assumption.


To put it most simply, it is not statistically valid to take a probability associated with one portion of a population and distribute it across the population as a whole. Over the past 7 years, a large number of Americans have been killed by enemy action in Iraq. That does not mean that every American has an equal chance of being killed by enemy action in Iraq. I feel pretty safe myself.

 

[rstofer]

Doesn't anyone find the 1 in 2200 statistic to be preposterous?

I have never met a diver that has had a DCS hit nor have I ever met a diver who knew a diver that was rumored to have had a hit. Now, I lead a sheltered life and, personally, don't push the envelope. Nor do I know more that a very few divers who do. The dives they do border on insane (in my limited recreational view). Yet they're still walking.

It would be very interesting to find out how many chamber rides they give annually in Monterey and Catalina (just to name two places that I know have chambers). Even with those numbers, I would expect the number to be much less than 1 in 2200. Properly, I would like to know the number of discrete cases, not the number of rides.

Of course, the number could be right...

Richard

 

[sschlesi]

John,

Assuming that the statistics given by DAN are accurate, it's reasonable to assume that they include various places, at various depths and included various experience levels. The ONLY constant given is that they ALL were using dive computers.

I fail to understand the point you're trying to make... If not 1 in 2200 dives, perhaps you might enlighten me on the math you are using... Given these figures what would you say the statistics are? 52,168 recreational air dives, 23 DCS incidents.

I have no idea how DAN collected this data, but I would expect that a dive that produced a DCS hit is MUCH more likely to be reported than an uneventful dive. It is entirely possible that dives that produce hits are substantially overrepresented for this reason.

 

[DCBC]

Doesn't anyone find the 1 in 2200 statistic to be preposterous?

I have never met a diver that has had a DCS hit nor have I ever met a diver who knew a diver that was rumored to have had a hit.
Richard

Please to meet you Richard. Now you can't repeat what you wrote; I've had DCS (Type I after adhering to the decompression profile).

 

[DCBC]

John,

Anyway you look at it, 52,168 dives were made that resulted in 23 DCS hits. ALL dives were undertaken with dive computers.

Perhaps all 23 hits were attributable to a few divers, who decided that their computers didn't require batteries, or maybe just didn't turn them on...

Personally however, I'm willing to assume that Dr Richard Vann, Director of Applied Research at the Duke University Medical Center for Hyperbaric Medicine and Environmental Physiology, as well as the Research Director for DAN, would insure (in his opinion) that these statistics represented the recreational diving public at-large and not solely a few people who were just trying to commit suicide.

 

[DCBC]

I have no idea how DAN collected this data, but I would expect that a dive that produced a DCS hit is MUCH more likely to be reported than an uneventful dive. It is entirely possible that dives that produce hits are substantially overrepresented for this reason.

It also begs the question, how many DCS Type I cases were not reported. I believe quite a few more than were.

 

[sschlesi]

It also begs the question, how many DCS Type I cases were not reported. I believe quite a few more than were.

Does anyone have a link to the complete study, Google only turned up the abstract.

I'd like to see what if anything they did in their data collection to make the data representative of the recreational diving population as a whole.

The study itself is trying evaluate "a mathematical model of bubble growth and resolution". Basically they take a profile run it through their algorithm and see what it says about the odds of getting bent. Then they compare that to the results of the actual dive and see if their algrorithm is doing a good job. Based on reading the abstract I see no reason that this study requires a broadly representative set of recreational dives, more important would be lots of dives, and accurate reporting of DCS.

It seems likely that some mild cases of DCS were not accurately reported.

Before concluding that on average 1 in 2200 computer dives result in DCS, I would want to know a lot more about the how the data is collected, how are they locating divers to report, how are they motivating people to report is (e.g. are they getting lots of reports from new divers, who stop reporting after a bit, are they getting lots of reports from divers involved in some other program e.g. REEF or ...), how is DCS evaluated etc.

 

[DCBC]

Does anyone have a link to the complete study, Google only turned up the abstract. ASA Abstracts - Search

I'd like to see what if anything they did in their data collection to make the data representative of the recreational diving population as a whole.

The study itself is trying evaluate "a mathematical model of bubble growth and resolution". Basically they take a profile run it through their algorithm and see what it says about the odds of getting bent. Then they compare that to the results of the actual dive and see if their algrorithm is doing a good job. Based on reading the abstract I see no reason that this study requires a broadly representative set of recreational dives, more important would be lots of dives, and accurate reporting of DCS.

It seems likely that some mild cases of DCS were not accurately reported.

Before concluding that on average 1 in 2200 computer dives result in DCS, I would want to know a lot more about the how the data is collected, how are they locating divers to report, how are they motivating people to report is (e.g. are they getting lots of reports from new divers, who stop reporting after a bit, are they getting lots of reports from divers involved in some other program e.g. REEF or ...), how is DCS evaluated etc.

You can probably give Richard a call at the Hyperbaric Center I'll send you his telephone number and e-mail address.

To break the ice, ask him if he's the famous guy who tested silicone breast implants?


I haven't been to Duke in years; since Pete Bennett ran the Atlantis Project there in the early 80's (chamber dives to 2300 ft). I was an Observer for one of the tests (visiting from DCIEM).