Is there data on how close to NDL undeserved hits occur?

[dmaziuk]

I don't think precise is the word: IMO the problem is those metrics are derived from the model that is derived from the statistics that was at some point collected from goats and burly caisson workers. The model promises X DCS cases per Y dives, and if the statistics was good, that's what you should get -- regardless of IIS or SurfGF (or what TC the GF is in as @inquisit pointed out above).

So the way I see it, what you're really looking at here is not the level of added conservatism "at which the data begins to show an increase in risk". It's whether adding conservatism to the model would appreciably reduce that promised X in Y ratio.

 

[rsingler]

Take a look at this simultaneous discussion:
Post in thread Question - Any information on increased nitrogen uptake by a panicking diver?
(Just disregard the discussion on the title topic)

 

[dmaziuk]

I've never been a goat but I strongly suspect many of them felt stressed when experimented upon. So excess of stress hormones should also have been well represented in the original samples. I.e. panic's covered, no need to lower your GF.

 

[scubadada]

Yes, we can do much better with regard to the risk of DCS than how close we were to the NDL, under or over.

Surfacing GF is likely a much better parameter to follow. This is becoming more readily available but is far, far from universal. Integral supersaturation for individual tissues or all tissues, with appropriate weighting, may turn out to be a good marker of decompression stress in a variety of dives. I see this as a work in progress, not readily available to any of us.

 

[broncobowsher]

I don't think precise is the word: IMO the problem is those metrics are derived from the model that is derived from the statistics that was at some point collected from goats and burly caisson workers. The model promises X DCS cases per Y dives, and if the statistics was good, that's what you should get -- regardless of IIS or SurfGF (or what TC the GF is in as @inquisit pointed out above).

So the way I see it, what you're really looking at here is not the level of added conservatism "at which the data begins to show an increase in risk". It's whether adding conservatism to the model would appreciably reduce that promised X in Y ratio.

You forgot to add in the 19 year old volentolds is excellent physical health that did the navy studies, unlike the slightly overweight middle aged that are using the data.

 

[Angelo Farina]

To put it a different way can we even crudely say how much "riding the NDL" contributes to recreational undeserved Type 2 DCS?

I always tought that "riding the NDL" is a bad habit coming from the concept that recreational divers should not do dives with planned deco.
I have always taught my students that it is actually safer to plan for a deco dive with square profile, always considering the max depth and the total time from immersion to when the diver reaches the deeper planned stop (usually 6 meters for normal dives with a 15-liters tank at 232 bars, perhaps 9 if diving deeper than 40m).
At that point you look at your computer, and in most cases it will say that no deco is required.
And, despite this, the diver still makes all the planned deco stops.
For me this is the safe way.
Of course it requires to be deco-trained and properly equipped.

 

[dmaziuk]

We're not diving Navy tables anymore though. DSAT was developed for less-fit people and it's about the least conservative of the models in common (edit: ) recreational use today. See e.g. Baker's comparison:
https://www.shearwater.com/wp-content/uploads/2019/05/understanding_m-values.pdf

 

[scubadada]

You forgot to add in the 19 year old volentolds is excellent physical health that did the navy studies, unlike the slightly overweight middle aged that are using the data.

The USN tables have not been used for quite a while. I used them from 1970-1980 after my certification by LA County. Considerably more liberal than PADI RDP but not aimed at repetitive dives

 

[Kendall Raine]

Yes, we can do much better with regard to the risk of DCS than how close we were to the NDL, under or over.

Surfacing GF is likely a much better parameter to follow. This is becoming more readily available but is far, far from universal. Integral supersaturation for individual tissues or all tissues, with appropriate weighting, may turn out to be a good marker of decompression stress in a variety of dives. I see this as a work in progress, not readily available to any of us.

Sorry to jump back on an older thread, but would you explain what you mean by "with appropriate weighting?" Are you suggesting weighting indivdual halt times? As I understand how ISS is computed there is no "weighting" of respective half time super saturation integrals. The mB/time metric is simply the sum. Is that not right?

 

[Duke Dive Medicine]

Thanks for the response. Maybe it would have to be broken into Type 1 and Type 2 DCS. I agree that Type 1 would be very difficult to get data on.

However, Type 2 involving chamber rides should be doable.

My understanding is the chamber people want to see the patient's computer. I would assume they then document this data in the chart.

In my short diving career I personally know 2 people with Type 2 DCS needing multiple chamber rides who had "undeserved" hits. If I know 2 people there must be many more.

Recently an experienced SBer mentioned he now dives no closer to NDL than 20 minutes or something to exercise extreme caution.

So it might be helpful for the diving community if data could be developed to say only 1% of Type 2 DCS involving chamber rides were in dives that the diver stayed 20 minutes away from NDL.

To be fair even my sample of 2 people had an added complexity that would have to be taken into account. One had a workup showing no underlying risk factor and one had a large PFO (since closed and back diving).

There's (nearly) an infinite combination of dive profiles and divers, not to mention the option to change settings like gradient factors on computers, which means that it would be difficult to standardize the data to analyze it, if we did have access to it (more on that below). '20 minutes away from NDL' can mean a lot of different things. About the closest we can come is the research on printed dive tables. We can also do probabilistic modeling for theoretical printed-table-based dives based on data derived from known military dives and outcomes (DCS vs non-DCS). To my knowledge, most computer algorithms have at least some roots that probabilistic modeling.

Anecdotally, we find that a lot of divers we see with DCS tend to have dove their computers to the edge of the algorithm, i.e. straight to whatever the computer says is the NDL, but again, that's hard to standardize given that computers track dives in real-time and many of them have parameters that can be adjusted. We do have data that suggest that in general, the deeper the dive and the longer the bottom time, the higher the probability of DCS, which is counterintuitive - one would think that the tables would control for pDCS with deeper and longer dives, but that was not the case when we looked at it.

From my own experience, divers with suspected DCS almost never bring their computers to the chamber, usually for fear of being judged. That pretty much squashes the idea of doing research on it. If they do bring their computer, the chamber facilities typically don't have the software or patch cables to retrieve and display the profiles. Even if we could retrieve the profiles, that would not influence clinical decision-making for treatment of DCS, which is based strictly on symptoms.

We tend to avoid the words "deserved" and "undeserved" nowadays and instead use "explained" and "unexplained". Depending on the profile, your example of the diver with DCS and a large PFO could be considered an explained hit.

<edit: neglected to check the date on this thread, lots of responses, including my own, that already stated the above!>

Best regards,
DDM