Discussion of the statistics of the NEDU study on the redistribution of decompression stop time from

[tomfcrist]

"The trial was also to be concluded if a midpoint analysis after completion of approximately 188 man-dives on each dive profile found a significantly greater incidence of DCS (Fisher Exact test, one-sided α = 0.05) for the deep stops dive profile than for the shallow stops dive profile."

I love that abstract...because it implies that if a significantly greater incidence of DCS was observed in the shallow stops profile, they would have been perfectly fine with continuing the trial.

 

[Dr Simon Mitchell]

I can see you are upset, but my statements hold true.... and I will demonstrate that now below.

Upset? Incredulous more like it. Incredulous that someone with no relevant knowledge or training can come onto a pubic forum and confidently present incorrect material that they do not understand as "holding true"; and in the process risk creating misperceptions among divers who come here to obtain knowledge.

Your statements do not hold true, and you have demonstrated nothing except your own ignorance of the language and methodology of science. You have cherry-picked and cited stopping criteria that the study did not meet, and ignored the stopping criterion that the study did meet as Joseph has correctly pointed out to you.

Which part of this do you not understand?

The trial was also to be concluded if a midpoint analysis after completion of approximately 188 man-dives on each dive profile found a significantly greater incidence of DCS (Fisher Exact test, one-sided α = 0.05) for the deep stops dive profile than for the shallow stops dive profile.

This is exactly what happened.

The allowable limits of this test, were 3% to 7% pDCS.

and

Also shown, the A1 profile had a very low return, and is in danger of falling out the bottom of the limits. It is way outside its predicted range, and It was headed towards rejection.

You do not understand these limits.

The 3% limit was to prevent futility. That is, if both profiles fell below 3% then the trial was not powered (big enough) to show a difference between profiles with such a low incidence of DCS. As David said when he corrected you on RBW:

"We had stopping rules if both schedules had unexpectedly high or low risk, which were likely to result in severe DCS or an inconclusive result, respectively. We never came close to these (the figure presented in an earlier post is misinterpreted)".

The study report also clarifies this:

These rules were to limit exposure of divers to the risk of severe DCS and to limit the potentially inconclusive testing of two low-risk dive profiles.

Ask yourself: why would the authors of a study trying to show a difference between two profiles impose a stopping rule if one of their profiles was wildly successfull? This grassy-knoll accusation of yours that one profile was about to be "rejected" and so the authors had to concoct an excuse for terminating the trial "early to salvage what you can" is patently ridiculous, and once again demonstrates why participants on this forum should view anything you write about decompression science with great scepticism.

We can see, the A2 profile was right down the middle (5%). It was performing as predicted, and was a well behaved test sample.

There was no excessive injury rates present.

There would have been had the trial not been terminated in accordance with its pre-defined rules. Excess injury in this context (and from the perspective of the IRB) meant any further cases of DCS beyond those required to achieve an answer to the question. Thus, if the trial had continued after the experimental end point had been reached, that would constitute excessive injury in the eyes of the IRB.

If A1 and A2 both claim to have 5% pdcs, but in reality A1 is 1% pdcs...you are no longer comparing apples to apples. The test is invalid at that point.

Neither you nor Ross understand the relevance of the pre-trial predictions of DCS rates for the two profiles. Such predictions of risk or predicted differences between interventions (be they two decompression profiles or a drug vs placebo) are used to power a study (that is, to calcuate the number of subjects needed to demonstrate the predicted difference between the profiles to a level of statistical significance). But there is nothing that says the trial is invalid if the interventions don't match those risk predictions. Ask yourself: if we were so confident in those predictions, what would be the point in doing the study?

The fact that one or both profiles did not behave according pre-trial predictions is a legitmate outcome of the study: it does not invalidate it in any way.

Simon M

 

[Dr Simon Mitchell]

I am not a scientist,so I would like an explanation. As @rossh pointed out, A1 profile was outside of predicted range and in danger of breaching test limits. In my layman's mind, if this profile yields lower percentage of DCS injuries than predicted, that should be GOOD,right?

Hello Admikar,

Your post is a perfect example of how Ross's misinformation creates confusion in the community.

Your layman's interpretation is correct, and I hope my post above where I say....

Ask yourself: why would the authors of a study trying to show a difference between two profiles impose a stopping rule if one of their profiles was wildly successfull?

....helps reinforce that.

Simon M

 

[tomfcrist]

I love how y’all are still unwilling to look at the study with a critical eye of any sort.

A2 took 63minutes to get plus travel time to get to 30’....the first stop was 70’ for 12 freakin minutes. On air...
A1 took 9 to get to 30’.

Well no kidding A2 took a higher hit rate.

Try allowing a difference in overall dive times....ie....same bottom time and raw deco times based off of widely used algorithms vice some random experimental garbage that’s only used when you have a chamber on site.

I’d bet that any current widely used algorithm would outperform either A1 or A2.

The fact that certain people refuse to see the uselessness of this trial is astonishing.

 

[Dr Simon Mitchell]

The fact that certain people refuse to see the uselessness of this trial is astonishing.

Well, those "certain people" include the world's leading team of decompression modellers and scientists.

And then there is the inconvenient fact that the only other published human studies of deep vs shallow stop approaches to ascent from decompression dives corroborate the findings of the NEDU study.

But, yes, we should probably all listen to you instead.

Simon M

PS: Go back to the original study report, read the analysis of 500,000 other potential ways of surfacing from the NEDU dive, note how none of them produced an integral supersaturation that was lower than the NEDU shallow stops profile, and then come back and tell me how any other widely used algorithm would have outperformed the NEDU shallow stops profile.

 

[tomfcrist]

Well, those "certain people" include the world's leading team of decompression modellers and scientists.

Simon M

That’s the astonishing part...that they are incapable of critical thought. Refusing to question the validity of their own findings.

Tesla was a genius too...and he thought he could transmit electricity from NY to Paris using the earth as his capacitor. He went public with it and never questioned his own beliefs.

 

[tomfcrist]

PS: Go back to the original study report, read the analysis of 500,000 other potential ways of surfacing from the NEDU dive, note how none of them produced an integral supersaturation that was lower than the NEDU shallow stops profile, and then come back and tell me how any other widely used algorithm would have outperformed the NEDU shallow stops profile.

Well you can start with the fact that all of those hypothetical profiles were still skewed to mandate a 174 minute total dive time.

 

[RyanT]

Let me take a slightly different tack on this topic. Can we complain about the NEDU study? Yes! Can we point out where it could have been done better? Sure! Can we come up with a million different ways the study could have been designed? Yes! For example how many iterations of two GF profiles could we have thought of to compare? Answer...lots! Are there hundreds of thousands of actual dive profiles out there that we can look at to see who got hit and who didn't? Yes. At the end of the day though, the NEDU study is one of the very few controlled studies available for direct comparison. I'll make my decisions on actual data every time.

We're doing this again? And it all started because Joesph made an awesome effort to explain basic statistics for those without the background in it...sigh.

 

[Gozo Diver]

I love that abstract...because it implies that if a significantly greater incidence of DCS was observed in the shallow stops profile, they would have been perfectly fine with continuing the trial.

The article I posted has a section that discusses this at some length. Rather than copying/pasting that, I refer the interested reader to the 'Midway Point Analysis' section.

On a separate note (with regards to the rest of the thread), since this is starting to somewhat veer off on a tangent in some respects, it should be clarified that the original post was a link to an explainer & discussion of the statistical analysis that was used in this study, particularly for those who might not be familiar with the mathematics - and it offered some "for" and "against" arguments. One could discuss myriad other aspects of course (and indeed this has been done in other places), but kindly note that this would go beyond the original motivation.

Joseph

 

[rossh]

Lets recap what we have so far.


1/
The DMO's were not blinded to the study, and were aware of the profiles each subject was diving, before making their assessments.


2/
There were no excessive injury rates in this test.

The A1 baseline test data, was well below the predicted range, and was just short of triggering an automatic test rejection.

The A2 test data was on course and well within the limiting boundaries.

The test was cut short because of the differences in the two data points, and potentially, to avoid the pending rejection to come.


Therefore: The often reported claim by some - "....that the IRB cut the test because of an excess injury rate".... that statement is false.

.