Decompression controversies

[Diver0001]

Rossh brought it up again a month ago, showing how VPM B+3 was very similar to GF 40/70 ...

Actually it wasn't close at all. Take another closer look at the graphs he posted.

While you can get similar looking ascent curves using GF-low of 30 or 40, VPM still calculated a deco time in the shallow zone (60ft and shallower) that was significantly faster than the Buhlmann algorithm in both of the examples he posted. On the longer of the two dives, the difference is very pronounced. The Buhlmann algorithm would have told you after that dive that you had skipped about 30 minutes of stops if you had followed the VPM schedule.

What Ross did there, perhaps unintentionally, was to present the two graphs at very different time scales. The time scale on the second graph is compressed by a little more than 15% as compared to the first graph. They LOOK similar as posted, even when you lay them one on top of the other, but if he had printed them on the same time scale (without the compression) you would have very clearly seen at first glance that the two profiles are nothing alike at all, especially in the shallows. They are not even in the same ball park.

.... And that is *exactly* what this thread and the previous thread are about. The bubble models simply do not calculate enough time shallow--as the NEDU study seems to suggest--due to incorrect assumptions about the effect of deep ascent curves on slow tissues.

If you really wanted to get something that looked like VPM using GF then you'd probably have to pick 30/90 or 40/110 or something like that. Don't just look at the ascent curve. Look at total time spent shallow. That's where the differences are.

R..

 

[Patoux01]

I think you just fell in the sarchasm . I wasn't even being subtle as I asked "where is the mimicking?". I agree that the consensus on GFs to use does not mimick VPM.

 

[Macan]

I think we, the Tec diving community in particular, owe a debt of gratitude to Dr. Mitchell, Dr. Doolette, Bruce Parson of Shearwater, and Ross Hemingway. They are professionals with strong views on the issue and have gone public to express those views. Without their participation in the passionate exchanges and discussions at RBW and SB, I think many of us would still be doing deep stops.

I've moved from GF 20/85 to 40/70-85. The NEDU report and subsequent discussions over the past years gave me reason to make the change.

Sadly, the agencies are staying silent despite the compelling evidence. Deep Stops is still in their Tec curriculum. Some leading computer manufacturers continue to have deep stops as a feature even at the recreational level and bubble model algorithms a mainstay in their computers.

I've gone specifically to share the NEDU report with my students and advised them to avoid deep stops. I also recommended that they avoid purchasing computers with proprietary algorithms, particularly computers with bubble model algorithms and to bring up their GF-Lo while padding their final stop.

 

[dreamdive]

Now, if we can just get Bruce to display the ACTUAL GF that was dived (rather than the one planned), that would be very educational!

 

[Macan]

Now, if we can just get Bruce to display the ACTUAL GF that was dived (rather than the one planned), that would be very educational!

You mean the actual GF Lo? GF hi is shown on the GF99 field.

 

[dreamdive]

You mean the actual GF Lo? GF hi is shown on the GF99 field.

What you plan and what you dive is rarely the same. If I plan on an 80/90 what did I actually dive? A 60/70.....etc. As soon as you are ascending less than 30 ft/min and stay longer at a stop, you scew your profile off the intended.

 

[EFX]

The presentation by Simon Mitchell was interesting and informative. I was a late contributor to the "Deep Stops Increases DCS" thread where I came to the same conclusion as Ross that the NEDU study's deep stop is something very different than used by other bubble models. I don't wish to rehash the old arguments but to lay down some of my thoughts here and if you would comment on them.

Here are the numbers from the NEDU study (depth = 170 ft, BT = 30 min.):
Shallow stop deco stops (depth/time): 40/9, 30/20, 20/52, 10/93. DT = 174 (neglecting ascent time).
Deep stop deco stops (depth/time): 70/12, 60/17, 50/15, 40/18, 30/23, 20/17, 10/72. DT = 174.

Running the dive plan on my Perdix with VPM-B +2 gave this profile for the same depth and BT:
Deco stops (depth/time): 100/1, 90/2, 80/3, 70/4, 60/5, 50/5, 40/7, 30/14, 20/20, 10/34. DT = 95.

As you can see the deepest stops for VPM-B are very short compared to the Navy's deep stop schedule (1,2,3,4 min vs. 12,17,15,18 min.). This is in keeping with the idea of limiting slow tissue on-gassing and the resultant super-saturation of those tissues which led to the DCS hits as reported in the Navy study, while limiting bubble growth beyond the critical size.

My question is: why would you be so quick to agree that bubble models in general and VPM in particular are inferior to dissolved gas models in general and Buhlmann GF's in particular, when the NEDU study is testing something quite different from what bubble models are doing?

(BTW: the only gas is air in keeping with the NEDU study. I've read the arguments regarding the use of O2 for accelerated deco which excludes deeper stops. I agree with that so don't go there.)

 

[UWSojourner]

My question is: why would you be so quick to agree that bubble models in general and VPM in particular are inferior to dissolved gas models in general and Buhlmann GF's in particular, when the NEDU study is testing something quite different from what bubble models are doing?

1. I don't think anyone has been "so quick" to come to the conclusions about bubble models. I remember debates on TheDecoStop about the NEDU study 10 years ago. This has been a long process.

2. The conclusion *most* are coming to is that the emphasis on deep stops typical in profiles generated by models like VPM is counterproductive in limiting deco risk.

3. Mitchell's presentation also showed a typical technical dive profile that exhibited the same pattern as the NEDU profiles -- deep stops trading lower supersaturation early in the profile for higher supersaturation upon surfacing. This pattern was clearly inferior in the NEDU study.

4. Your question implies that bubble models might be superior at lower conservatism levels, but inferior at higher conservatism. At what level of conservatism do you posit the crossover for VPM to go from good to bad? The idea is rather strange I think, and in any case has no evidence to support it. A more reasonable conclusion based on the evidence is that distributing deco time deeper to "protect the fast tissues" is likely not optimal no matter what conservatism you choose.

 

[kensuf]

Actually it wasn't close at all. Take another closer look at the graphs he posted.

While you can get similar looking ascent curves using GF-low of 30 or 40, VPM still calculated a deco time in the shallow zone (60ft and shallower) that was significantly faster than the Buhlmann algorithm in both of the examples he posted.

R..

This.

I run what some of my friends consider a fairly aggressive GF, but it's still more conservative than VPM by a long-shot. The skew is noticeable on extremely long dives. My last dive in Cathedral was more than a 7 hour run-time, 190 minutes at 150' average depth running a PO2 of 1.1 (19/35 bottom mix). Flushing with 30% nitrox at 100', and bumping up the PO2 to 1.3, VPM with a +3 conservatism setting gives the following deco schedule:

90' - 2 minutes
80' - 5 minutes
70' - 7 minutes
60' - 10 minutes
50' - 15 minutes
40' - 20 minutes
30' - 28 minutes
20' - 37 minutes
10' - 68 minutes
Total runtime: 384 minutes

My preferred, and "aggressive" Buhlman / GF combination gave this schedule:

90' - 5
80' - 7
70' - 9
60' - 13
50' - 19
40' - 25
30' - 36
20' - 48
10' - 94
Total runtime: 449 minutes

Yes, the deep stops were a little bit longer (my GF lo is set to 30), but my shallow stops are also longer.

BTW, for comparison, Buhlman on 100/100 gives this:

70' - 7
60' - 10
50' - 15
40' - 22
30' - 32
20' - 44
10' - 81
Total runtime: 403 minutes

Notice, VPM +3 has shorter shallow stops than the straight Buhlman algorithm.

In a real world application, a regular dive buddy of mine, who usually is more conservative than me on deco by about 10-20% more time, "cleared" an hour before me with his computer set to VPM on that 190 minute bottom-time dive. When he saw my computer called for another hour of deco, he decided that he didn't trust VPM +3 anymore, and he waited it out with me.

I cannot tell you if he would have been bent or not using VPM +3 because he wasn't willing to risk it. I do find it interesting that most of the people I know that are doing long dives have moved away from VPM and just accept that doing extra deco time is part of the cost of doing the bottom time.

 

[David Doolette]

Just because you had a negative test for a PFO does not necessarily mean you don't have one.

I was tested for a PFO and tested negative. I was tested again 15 years later after getting bent a number of times over a 3 year period, and tested positive. I was then tested a third time and tested negative. The surgeon confirmed the PFO when he closed it with an amplatzer device, and I haven't had any problems since.

Both negative tests were a trans-esophagael echocardiogram (TEE). The positive was a trans-thoracic echocardiogram (TTE). The TEE was supposed to be the gold standard, but I know of at least one other person that used to get bent a lot and now has a bionic heart that also tested negative on a TEE. The best my surgeon figured was that the prophyl I was on for the TEE skewed my ability to do a deep valsvalva while they had the probe jammed down my throat.

Hi Ken

Your experience is in accord with current thinking about testing for a PFO. In the past, some have thought that because TEE puts the transducer closer to the heart and provides a better picture than TTE, that TEE was the preferred method for detecting PFO. However, the consensus now is that TTE is better for detecting a PFO because it better allows manoeuvers like releasing a Valsalva or sniffing, which can provoke bubble contrast to shunt across a PFO that might not otherwise be detected. Also, if the PFO is two small to be detected by TTE with bubble contrast and provocation, it probably too small to matter with respect to DCS. The joint position statement on persistent foramen ovale (PFO) and diving of the South Pacific Underwater Medicine Society and the United Kingdom Sports Diving Medical Committee can be found here:

Diving and Hyperbaric Medicine Journal - DHM Journal, Immediate release articles

David