By Dr Simon Mitchell
Over the last few years there has been a 'proliferation' of software programs that allow recreational divers to plan decompression from deep dives. There is much conjecture and debate over the relative merits of these various decompression planners. This is not surprising since some predict significantly shorter decompression requirements for the same combination of depth and time than others. In addition, the decompression algorithms generated by the various planners may differ not only with respect to their overall length, but also with respect to their time / depth profile. For example, some newer planners generate decompression profiles with stops that start deeper, but with shorter stops in the shallower depths. This trend towards the use of deeper decompression stops appears to be gathering a significant momentum. For divers brought up on a diet of admonishments like 'get your deco wrong and you are in big trouble', these substantial variations in the duration and nature of decompression can be very confusing.
Part 1What modern divers are witnessing is a mini-revolution in decompression planning philosophy that is gathering pace in both the recreational and technical diving communities. In this article I will try to explain the basis for this 'revolution' in non-technical terms, and its implications for divers. This is only an introduction to the most basic concepts, and those seeking details will need to look further.
The need for decompression
During a dive we accumulate the inert gas we are breathing (usually nitrogen) in dissolved form in our tissues because its solubility in fluid is increased under the increased pressures we experience at depth. The greater the depth and the longer we stay there, the greater the amount of inert gas that accumulates. During ascent, the pressure around us falls, and so does the solubility of the inert gas dissolved in our tissues. With their solubility falling, these dissolved gas molecules are physically compelled to revert into the gaseous state. Transport to the alveoli of the lungs facilitates this, but the movement of these gas molecules out of the tissues into the blood and then to the lungs ('outgassing') takes time. In some tissues this process lags behind the rate at which the water pressure around the ascending diver is falling. Under these circumstances, it is almost inevitable that there will come a point during the ascent when the pressure of dissolved gas in at least some tissues exceeds the pressure of the water surrounding the diver. The tissue is then said to be 'supersaturated'.
When some critical degree of supersaturation is reached, the gas molecules can no longer remain dissolved and bubbles will form. There is very good evidence that this process is catalysed by the constant presence of tiny micro-bubbles ('micronuclei') which act as seeds for the growth of larger inert gas bubbles when supersaturation occurs. In other words, the dissolved inert gas diffuses into these pre-formed microbubbles and makes them bigger. In theory, these micronuclei should not exist because such 'vanishingly small' bubbles are subjected to immense 'crushing pressure' generated by surface tension in their spherical walls. However, it appears that they may be stabilised by surfactants; naturally occurring detergent like molecules that reduce surface tension at a gas / fluid interface. These surfactants allow the micronuclei to persist, 'just waiting' for an inert gas supersaturation event that will allow them to grow. The resultant larger bubbles are the 'agents' that initiate DCI. All attempts at regulating our decompression (ascent) are aimed at preventing the growth of gas micronuclei into inert gas bubbles of sufficient size and number to cause DCI.
Traditional approaches to regulating decompression
The first systematic attempt to control ascent from dives in order to prevent DCI is attributed to JS Haldane, a Scottish physiologist who conducted his work in the early 1900s. He devised mathematical methods for calculating the amount of inert gas in tissue at any stage during a dive. Using appropriate adjustments to account for some tissues taking up and eliminating inert gas faster ('fast tissues') than others ('slower tissues'), Haldane was (in theory) able to track the uptake and elimination of inert gas in a variety of body tissues during compression and decompression. Most importantly, using these mathematical models he was able to predict at what depths supersaturation of inert gas would occur in a range of theoretical body tissues during decompression (ascent).
Haldane's next step was to estimate the maximum amount of supersaturation that was tolerable. To help with this estimation Haldane performed many experiments on animals (mainly goats) in which he subjected them to progressively deeper dives until he began to establish threshold exposures that caused DCI. From these experiments he was able to extrapolate the amount of supersaturation that appeared tolerable, in goats at least!!
The combination of these two vital (if somewhat approximate) pieces of information, that is, the magnitude of inert gas supersaturation in tissues during ascent and the maximum amount of supersaturation that could be tolerated, allowed Haldane to design decompression schedules that periodically stopped the ascent to give gas molecules time to leave the tissue before the maximum tolerable supersaturation was exceeded.
Given the commercial applications for his work, Haldane wanted his decompressions to be not only safe, but also efficient (as rapid as was safely possible). He reasoned that an ascent designed to achieve the maximum tolerable supersaturation (without exceeding it) would optimise the gradient for movement of the dissolved gas between tissue and lungs and achieve the best possible rate of inert gas elimination. Consequently, his decompressions were designed to maximise the supersaturation (in practical terms, by ascending as far as possible before imposing decompression stops) without exceeding the safe supersaturation limit. This strategy produced decompression tables for deep short bounce dives (typical of recreational deep diving) that were characterised by a relatively large initial ascent, followed by a series of relatively shallow stops of increasing length.
The dive tables that were designed on the basis of this work were so spectacularly effective compared to the previous regimens for controlling dives that Haldane's decompression concepts went unchallenged for decades. The Haldanian approach to dive table design was perpetrated through many evolutions and modifications. In particular, the way in which maximum safe supersaturation is defined has undergone considerable refinement, and many adjustments to tables have been made on the basis of experience with their use in the field. Nevertheless, the underlying approach has remained much the same: calculate the amount of gas in the tissues during ascent, ascend as far as possible without exceeding maximum safe supersaturation, and stop the ascent periodically to allow outgassing when the tissue gas pressure was approaching maximum safe supersaturation (called 'M-values' by more modern table designers). Most tables in use today apply this Haldanian or so-called 'neo-Haldanian' (reflecting the various modifications since Haldane's original description) approach in their design.
Problems with the Haldanian approach to regulating decompression.
An implied assumption of the Haldanian approach to decompression modelling is that the decompression strategy essentially prevents bubble formation. It assumes that by imposing decompression stops to prevent tissue inert gas supersaturation from exceeding the maximum tolerable limit, bubble formation has been prevented. This would be an understandable assumption for Haldane to have made. Remember, he deduced his maximum tolerable supersaturations from the thresholds required to produce DCI in goats. It was logical that he might consider that the appearance of disease implied bubbles, and no disease implied no bubbles. He had no bubble detection technology available to him at the time.
The advent of Doppler ultrasound technology in the 1960s resulted in a mass of evidence that bubble formation did occur during many (if not most) decompressions controlled according to Haldanian decompression procedures. Despite this, the vast majority of these decompressions did not result in DCI. This was an extraordinarily interesting finding for several reasons. First, it helped explain why DCI occasionally occurred after dives that did not violate the prescribed decompression schedule. Obviously, any small upward variation in the number or size of these normally 'silent' bubbles might be enough to precipitate symptoms. Second, it also highlighted a potentially significant problem with the calculation of the latter phases of Haldanian decompression schedules. Specifically, once bubbles form, the subsequent outgassing of inert gas from tissues is slower than predicted by the techniques Haldane used to calculate tissue inert gas pressures. This helped explain why a great deal of empirical tinkering with the predicted schedules had been necessary over the years, and why long shallow stops seemed to be necessary after the initially large ascent.
If this all sounds excessively negative, I should pause to note that Haldanian and neo-Haldanian decompression regimens with their various adjustments and 'tweakings' have served us very well over the years, and continue to do so. However, these issues do set the scene for the discussion of alternative strategies for controlling decompression.
An alternative focus: prevention of bubble formation.
At around the time Doppler technology was first detecting bubbles during 'Haldanian decompressions', the pioneers of another approach to decompression were encouraging a rethink of our strategy. People like Brian Hills were suggesting that decompression strategies should focus more on prevention of bubble formation rather than on promoting outgassing by maximising tissue inert gas supersaturation within assumed 'safe' limits. These were diametrically opposed philosophies since Doppler had demonstrated that the maximisation of 'safe' supersaturation to speed outgassing very often (if not invariably) resulted in bubble formation.
The approach advocated by Hills and others (sometimes referred to as a 'bubble model' for decompression), focussed on prevention of bubble formation by limiting tissue supersaturation to a much lower level than allowed by Haldane. It was argued that supersaturation should be limited to such an extent that it avoided significant inward diffusion of inert gas into the 'gaseous micronuclei' that we discussed earlier. In practical terms, this meant that initial ascents were smaller and that decompression stops began deeper. Almost paradoxically, this did not result in longer decompressions. Indeed, the predictions arising from this approach were that the time 'cost' in having to begin stops earlier was more than compensated by avoiding the problem of bubble formation slowing the removal of inert gas during the shallow stops; that is, the shallow stops could actually be shorter. (Please note that it's not quite as simple as this, but this explanation gives a basic conceptual framework). Anyway, the result was earlier (deep) stops, shorter shallower stops, and perhaps shorter overall decompression times when profiles were compared to those generated by Haldanian models.
This approach did not really catch on in the 1970s when it was first discussed, but interest has been revived more recently as the demands of extreme technical divers for shorter (but safe) decompression strategies have become more pressing. For example, some of the extreme cave penetrations conducted by the Woodville Karst Plain Project (WKPP) have involved more than 12 hours of decompression alone, and would have been nigh on impossible using conventional Haldanian algorithms. These divers and others have begun using decompression profiles that are more typical of those generated by bubble models, and several technical diving decompression packages based on bubble models have recently been released. The most visible of these is the highly innovative 'Reduced Gradient Bubble Model' (RGBM) designed by Bruce Wienke and now included in several technical and non-technical recreational diving software packages and dive computers.
I believe that these developments, led by innovators like Wienke, are very exciting. However, our excitement must be tempered with a little caution. Although some advocates of new algorithms frequently make claims of 'testing in the field' the reality is that there is no prospectively accumulated database of dives of known outcome using bubble model algorithms. Simply put, this means that we really don't know for sure how safe or otherwise the novel models like the RGBM are. It has to be said that things look very hopeful, but more experience is required in representative populations of divers performing many dives before the concept can be considered truly established. This does not mean I would not use the RGBM (I occasionally do in fact). But as an informed user I understand that this model is in the early stages of being proven. Unfortunately, uncritical readers of material released by organisations like the WKPP often misinterpret the reports of extreme dives being conducted with 'short' decompressions, and assume that they represent proof or validation of the decompression strategies. The enthusiastic manner in which some authors from these groups present their material can be seen to exacerbate the potential for such misunderstanding. It is important to understand that the activities of highly selected populations of focussed elite technical divers cannot be generalised to the broader diving population.
Despite these cautionary words, it seems likely to me that the use of these novel decompression algorithms based on 'bubble models' will be properly validated in the fullness of time, and that their use will become more widespread. The potential benefits for all deep divers and technical divers are obvious.