Slow tissue on gas from stops

[Storker]

Myself and other experts in the field consider it to have a significant design flaw in that they had two profiles of different length (RD deco was longer) which would have made it impossible to conclude anything if the RD profile had proved superior.

Given the experimental design, the results from the study seem to show a seriously negative outcome for RD.

 

[Dr Simon Mitchell]

Given the experimental design, the results from the study seem to show a seriously negative outcome for RD.

Hello Storker,

Yes, you could certainly interpret it that way. This is the point that UWSojourner has made, and which may be under-appreciated by Dan. They showed a difference in one of their pre-defined outcome measures between RD and a shorter GF decompression which nevertheless also incorporated a moderate deep stop emphasis. This difference would almost certainly have been bigger if the comparator 30-85 profile had tacked on more shallow stop time to make it the same length as the RD decompression, and bigger again if a GF profile of same length but even less emphasis on deep stops had been used (see UWSojourner's integral supersaturation analysis here).

The caveat in all this is that the relationship between the changes measured and the absolute risk of DCS is unknown. But if divers want to know the truth about decompression safety then these results are confluent with other recent findings, and are almost certainly pointing in the right direction. I would also come back to a point that defenders of deep stop approaches often overlook; it is not as though the recent studies are challenging a previously proven paradigm. There has never been any validation of the deep stop approach. It was always just an attractive theory, and all the evidence that has recently emerged suggests that we probably embraced it too enthusiastically in some of our applications for decompression diving (like RD and bubble models).

Simon

 

[Dan_P]

Hi Simon,

I took note of the following remarks. I'm sure you'll forgive me for the extraction from context in these cases;

I suspect he had very little to do with it. He was not an author on the paper.

Shy of USN of course, how often do organizations line up a small armada of tech divers for the scientific community to tinker with at facilities and on boats also provided by the organizations?

I congratulate the authors on reporting the results objectively in the context of what must have been immense pressure to sanitize the outcomes to be as favourable as possible to RD.

I find that remark uncharacteristically unobjective, but you are of course free to congratulate anyone you may wish to, for whichever reason.

On another note, I don't know how extrapolatable the results of Spisni are, namely because the conclusion section of the report states they're not at all, but I appreciate the implications of UWSojourner's previous post, and am eagerly awaiting the trials to confirm if the correlation to measurements will match up accordingly.

To be honest though, I feel we've hacked through every bit of this already, and it's almost old news at this point - this was pre-RD2.0 anyway, and the vBlog that seems to have sparked the resentment, isn't getting any younger, either...

That said - and I want to be absolutely clear that I'm not being dismissive of anything here - there were more details in the Spisni trial configuration that I find problematic than just the length of deco time for the RD profile;

The RD-profile was locked to a 1:2 ratio profile (60m setpoint) even so the average depth was 48m, which is ideal for 1:1 ratio profiles (45m setpoint).
The reason it matters, is because the further you move away from the reference setpoints, the worse off you get - hence the need to have multiple setpoints in the first place.

If in the water on that dive, and not "locked" to a setpoint too deep because of a trial, even using RD1.0, I'd have spent at least 6 minutes less on decompression in the water, if following the protocol to the letter.

With RD2.0, even less, and with significantly reduced emphasis on depth during the ascend.

I actually think it's fair to say that this will almost certainly have had an impact on results, although I still think results show that the deep stop emphasis needed to be reduced (I want to be clear on that).


Best Regards,

Dan

 

[UWSojourner]

Myself and other experts in the field consider it to have a significant design flaw in that they had two profiles of different length (RD deco was longer) which would have made it impossible to conclude anything if the RD profile had proved superior.

Clearly the design flaw put the study at risk for the reason you state.

But one fortuitous side-effect of that flaw was the study provided a point estimate of the degree of inefficiency of the UTD-RD strategy. That is, the UTD-RD strategy produced inferior results with a 44% time advantage. That's a significant gap and should help dispel the idea that these issues are much ado about nothing.

 

[Dan_P]

Clearly the design flaw put the study at risk for the reason you state.

But one fortuitous side-effect of that flaw was the study provided a point estimate of the degree of inefficiency of the UTD-RD strategy. That is, the UTD-RD strategy produced inferior results with a 44% time advantage. That's a significant gap and should help dispel the idea that these issues are much ado about nothing.

Well no, not really.
The practical application of what we learned because of the design flaw is pretty much this:

If I plan to do a dive to 60m but instead adjust in-water to do a 48m (avg.) dive, but can't comfortably negotiate making that adjustment to my decompression profile ad hoc, I'd have been better off using a computer (rather, shouldn't do the dive).

If I were doing a dive such as the ones in the trial, exact same parametres but without the limitations of the design(flaw), even if I were using RD1.0 and "by the book", I'd spend significantly less time on deco than the trial used.

Let me be clear that I also read the results of the study to mean that the deep stop emphasis in RD1.0 was too great - but it is a factor that could mean we should be very careful about extrapolating on the results.

 

[UWSojourner]

Well no, not really.

Well, yes, really.

Let me be clear that I also read the results of the study to mean that the deep stop emphasis in RD1.0 was too great - but it is a factor that could mean we should be very careful about extrapolating on the results.

There was no extrapolation. If both profiles had the same run time we would know that the UTD-RD profile was inferior, but we would not have received any information about the degree of inefficiency (i.e. how much more time would be needed by the UTD-RD profile to correct the issue).

But the UTD-RD strategy still failed despite a 44% time advantage. That provides information about the magnitude of inefficiency for the UTD-RD profile. You seem to be thrashing about trying to tuck UTD-RD into the cracks of verbal precision (e.g. "not unsafe", etc.).

Let's suppose someone asks, "How inefficient might UTD-RD be?". If you say, "We have no idea", you're being disingenuous.

You should say, "We don't know everything and likely never will. But in one study of a pretty benign profile UTD-RD was still inferior even given a 44% time advantage."

 

[Dan_P]

Well, yes, really.


There was no extrapolation. If both profiles had the same run time we would know that the UTD-RD profile was inferior, but we would not have received any information about the degree of inefficiency (i.e. how much more time would be needed by the UTD-RD profile to correct the issue).

But the UTD-RD strategy still failed despite a 44% time advantage. That provides information about the magnitude of inefficiency for the UTD-RD profile. You seem to be thrashing about trying to tuck UTD-RD into the cracks of verbal precision (e.g. "not unsafe", etc.).

Let's suppose someone asks, "How inefficient might UTD-RD be?". If you say, "We have no idea", you're being disingenuous.

You should say, "We don't know everything and likely never will. But in one study of a pretty benign profile UTD-RD was still inferior even given a 44% time advantage."

No, and I think you really need to try and focus on what I'm actually saying for a moment, instead of jumping to conclusions about my motivation:

What I am saying, is this:
There is a difference between "arbitrarily adding extra time", and "given a 44% time advantage".

If the RD blueprint states 36 minutes are called for, but the trial imposes doing 42 minutes, you wouldn't call those extra 6 minutes a "bonus", you'd call them a "design flaw".

That's a reasonable concern to have about the trial's design.

I'm not saying RD1.0 wasn't overemphasising deep stops - but you're laying too much into the results of the Spisni trial, given the design.

 

[Dr Simon Mitchell]

There is a difference between "arbitrarily adding extra time", and "given a 44% time advantage".

If the RD blueprint states 36 minutes are called for, but the trial imposes doing 42 minutes, you wouldn't call those extra 6 minutes a "bonus", you'd call them a "design flaw".

Dan,

This is a fascinating admission. I think we can safely assume that, if nothing else, the one role that the inventor of RD algorithm almost certainly had was to endorse the RD profile used in the trial. Yet you, (presumably) a senior instructor on that algorithm, do not agree on how it has been applied by the inventor? You are labelling part of the decompression "arbitrarily added extra time". I'm not sure what this says about the consistency with which the algorithm is applied in the wider world.

I get the sense you are clutching at straws. I just don't understand why you are so passionate about defending an approach to decompression that has NO validating research underpinning it, in the face of an emerging body of human experimental evidence that its underlying theory is wrong.

Simon M

 

[Dan_P]

Dan,

This is a fascinating admission. I think we can safely assume that, if nothing else, the one role that the inventor of RD algorithm almost certainly had was to endorse the RD profile used in the trial. Yet you, (presumably) a senior instructor on that algorithm, do not agree on how it has been applied by the inventor? You are labelling part of the decompression "arbitrarily added extra time". I'm not sure what this says about the consistency with which the algorithm is applied in the wider world.

I get the sense you are clutching at straws. I just don't understand why you are so passionate about defending an approach to decompression that has NO validating research underpinning it, in the face of an emerging body of human experimental evidence that its underlying theory is wrong.

Simon M

Hi Simon,

First, let me be clear - again - I'm not trying to say that the deep stop emphasis wasn't too great.
I've been perfectly clear on several occasions already; I acknowledge that the results of the study show the deep stop emphasis too great.
When you're saying that I'm grasping at straws, it's simply incorrect and incontextual to what I'm actually saying.

For good order, I've been open about my credentials - I am not a Technical Instructor, but I am a technical diver and certified to use RD to the depths involved in this study, and I have done similar dives in the locale where the Spisni trials were conducted.

Second, and more to the point:

There is a 1:1 setpoint at 45m.
The depth of the trial dives was 50m, with an average depth of the trial dives being 48m, and in either case, RD "per protocol" rounds up. So 48m.
You'd do 1:1 plus 5 minutes of deco for this dive - that's 30 minutes, because the bottom time was 25 minutes.
Adding a total of 6 minutes of deep stops across 36m and 24m, it's 36 total.
That's the "letter of the book", the "protocol", if you will.

You can't say that we have to assume AG got the trial design how he wanted it.
Nor exactly how the RD blueprint would "stipulate" it per protocol.
In fact, that's not the case.

You could say that AG didn't seem to make a big fuss over it.
And you could even make your own deduction that if he didn't, you may think it's because he was overconfident - but shy of that, the post you've made above is well beyond anything you have any basis for saying.
It's also beside the point entirely, obfuscating the objectivity of the conversation and quite frankly, starting to look like a poor attempt of deflecting attention from the fact that you've only just practically approved of UWSojourner's wrongful interpretation of the Spisni trial's findings.

Keep in mind, we still agree that the deep stop emphasis was too great - and I haven't even a footnote, question or remark to your general positions on the matter, which we have discussed previously and objectively.


Best Regards,

Dan

 

[leadduck]

The two profiles from the video were listed here:
UTD Decompression profile study results published

The total decompression time of RD here is 43min, right? But you say it should've been only 36min including the 6min deep stops. So, you suggest they should've done even less shallow time to get a better picture of UTD RD performance?