Decompression Sickness: Recompression Treatment

The aim of recompression treatment is to produce the following1:

An immediate reduction in bubble size, which will

  • Cause the surface tension acting on the bubble to increase and perhaps collapse the bubble.
  • Increase the surface area of the bubble relative to its volume, thus enhancing gas diffusion out of the bubble.
  • Reduce the length of any intravascular gas column and hence allow perfusion pressure to push bubbles out of the tissues into the venous circulation.
  • Reduce the compressive effect of bubbles on adjacent tissues.
  • Reduce the bubble-tissue and bubble-blood interfaces and the secondary inflammatory reactions that follow.

An increase in the diffusion gradient of gas out of the bubble (see Figure 10.3), which will

  • Relieve ischaemia and hypoxia.
  • Restore normal tissue function.

There are three factors to consider after deciding a diver needs recompression therapy:

  1. The pressure (depth) required for therapy.
  2. The gas mixture to be used.
  3. The rate of decompression.

Depth of recompression

In deciding the depth of recompression, three different approaches are possible:

  1. Recompress to a pressure (depth) dependent on the depth and duration of the original dive.
  2. Recompress to a depth that produces clinical relief of symptoms and then tailor the gas mixtures for decompression from that depth.
  3. Recompress to a predetermined fixed depth, i.e. according to standard tables of recompression therapy.

Only the third is in common practice these days, but each has some logic to it under certain circumstances.


This technique uses the principle that if a dive to 4 ATA produces DCS, then recompression to 4 ATA should relieve the symptoms. This approach was best typified by the now defunct concept of treating aviator DCS merely by descent to ground level.

This is not a particularly satisfactory technique because it is designed to cope with the total quantity of gas dissolved in the body during the original dive, irrespective of its distribution. Because DCS is the clinical manifestation of a gas bubble lodged in a vulnerable area, it is necessary to recompress to reduce the size of that particular bubble, irrespective of the total quantity of inert gas dissolved in the body. There is also evidence that once a bubble is formed, even compression to very high pressures may not completely eliminate that bubble, thus leaving bubble nuclei to re-expand on decompression.

The one advantage of this approach, apart from its simplicity, is evident when a diver develops DCS very soon after surfacing from a deep dive. Under these conditions, a prompt return to the original depth will ensure that there is no tissue-to-bubble pressure gradient that could cause bubble growth at a lesser depth.


This empirical approach was used first in underwater recompressions to reduce the depth exposure as far as possible. It is still applicable, both underwater and in chambers.

The freedom to be able to choose any depth to achieve an acceptable clinical result and then select an appropriate breathing gas is invaluable in serious cases (Figure 13.1).

A severely injured diver with decompression sickness who is treated in a recompression chamber at 18 metres.
Figure 13.1 A severely injured diver with decompression sickness who is treated in a recompression chamber at 18 metres. Note the provision of both physiological monitoring and mechanical ventilation while under pressure. This is rarely required but available in many modern facilities. (Photograph by M. Bennett.)

As traditionally used for deep and saturation exposures, the patient is recompressed to the depth at which all major symptoms disappear, plus one additional atmosphere, based on the assumption that the additional pressure would result in a reduction bubble size and allow increased surface tension to promote bubble resolution.

This approach is not without difficulty because it requires a sound knowledge of saturation decompression and the ability to mix and administer different gas mixtures. Nitrogen narcosis and central nervous system and pulmonary oxygen toxicity become important considerations. A recompression chamber (RCC) should have a mixed gas capability to support such operations.


This approach uses standard recompression tables roughly divided into the following groups:

  • Air tables.
  • Oxygen tables.
  • Oxygen tables with deep excursions.
  • Helium-oxygen (heliox) tables.
  • Saturation decompression.

Many such tables have been published by both commercial and military organizations around the world. Both air and heliox tables start at 30 and 50 metres, oxygen tables start at 9 to 18 metres, deep excursions are commonly to 30 metres and saturation treatments depend on the depth of symptom relief.

These tables clearly state the gas mixture to be used (usually oxygen or heliox), the periods to be spent at each depth and the rate of decompression. Standard air, oxygen and gas mixture tables for recompression are presented in Appendix B and Appendix C.

Standard recompression tables

A prudent diving physician will advise non-experts to adhere to strict treatment guidelines, as depicted in the manuals, but retain his or her own flexibility to treat patients who did not respond or in facilities that may not be appropriate to apply the guidelines. For this reason, this text covers more than one set of treatment techniques. The standard tables are as follows.

AIR TABLES (US NAVY 1A, 2A, 3 AND 4 AND ROYAL NAVY 52, 53, 54, 55, 71, 72 AND 73)

These tables are used increasingly rarely. The US Navy (USN) first published their air recompression Tables 1 to 4 in 1945, and these were the standard treatment tables for more than 20 years. What is remarkable about these tables is that they were never tested on subjects with DCS. The test subjects made normal dives, which required decompression, but were then subjected to a treatment table instead. The treatment tables were deemed satisfactory if the subjects did not surface with DCS after one of the treatment tables!

During the 1960s, reported failure rates of up to 50 per cent on the air tables resulted in the development of the shallow oxygen tables.

Many experienced clinicians today are reluctant to use air tables because of the problems they bring and their variable benefits. The difficulties with air tables include logistical problems with prolonged decompression, aggravation of symptoms during ascent, nitrogen narcosis and DCS in the attendants and respiratory distress caused by the increased density of air under pressure (particularly if pulmonary involvement is already present). The results are often not adequate unless the symptoms are mild, recent and not the result of gross omitted decompression.

However, if oxygen is unavailable and the RCC can logistically support a treatment table lasting up to 40 hours, air tables may be considered.


The introduction of standard oxygen tables using 100 per cent oxygen interspersed with short air breathing ‘oxygen breaks’ gives more flexibility and improved results (Figure 13.2). These tables are shorter, needing only 2 to 5 hours, require compression only to 18 metres (2.8 ATA) and are therefore logistically achievable. Divers Alert Network (DAN) reports that about 80 per cent of all DCS is treated with these tables. The physiological advantages are in the speed of bubble resolution and increased oxygenation of tissues.

The US Navy Treatment Table 6 (or Royal Navy Table 62). Depths are expressed here as metres of sea water (msw). An initial period at 18 msw (60 feet) breathing 100 per cent oxygen with air breaks is followed by a slow ascent to 9 msw (30 feet).
The US Navy Treatment Table 6 (or Royal Navy Table 62). Depths are expressed here as metres of sea water (msw). An initial period at 18 msw (60 feet) breathing 100 per cent oxygen with air breaks is followed by a slow ascent to 9 msw (30 feet).

Disadvantages include the less immediate reduction in bubble size (i.e. to less than half the volume reduction achieved with the 50-metre standard air tables), increased fire hazard, oxygen toxicity and the occasional intolerance of a distressed patient to a mask. Although the pressure gradient of nitrogen in the intravascular bubble-to-blood interface is increased with oxygen breathing, if the diver has previously dived in excess of 18 metres, there could well be a gas pressure gradient from tissue-to-extravascular bubble during the early phase of recompression.

With the foregoing qualifications, the use of oxygen tables has received worldwide acceptance as the starting point for all standard recompression therapy. Although some authors have reported significant failures with these tables, the success rate is higher than with the air tables that preceded them, and no better-performing alternative has been identified. Success rates for complete resolution of symptoms vary with both severity and the time to recompression, but they are often reported at 80 to 90 per cent.

Whenever oxygen is used, attention must be paid to oxygen toxicity. Unless one is following established safe protocols, it is suggested that the oxygen parameters should not exceed those likely to result in neurological or pulmonary toxicity (see Chapter 17). In cases of potential death or disability, exceeding these parameters may be acceptable.


USN Table 6A involves the addition of an initial period of breathing air at 50 metres followed by a standard Table 6. This table was initially introduced as a treatment of cerebral arterial gas embolism (CAGE) in submarine escape trainees (with a low nitrogen load), although later studies have not shown any benefit from this deep excursion over the shallow 18-metre oxygen tables in divers.

Most centres now recommend starting with a standard USN Table 6 at 18 metres; however, in severe cases of DCS or cases not responding, the option exists to go on to a deeper table using a mixed gas such as a Comex 30 (Figure 13.3, Plate 3). Although based on little evidence, some centres advocate initial use of the Comex 30-metre table for severe neurological DCS. Heliox mixtures (50/50) are substituted for oxygen at depths greater than 18 metres. In refractory cases, there is then the opportunity to go to 50 metres using a modified USN Table 6A, Royal Navy (RN) Table 64 or Comex 50 breathing heliox mixtures.

The Comex 30 treatment table. The maximum excursion is to 30 msw breathing 50 per cent oxygen in helium (heliox 50).
Figure 13.3 The Comex 30 treatment table. The maximum excursion is to 30 msw breathing 50 per cent oxygen in helium (heliox 50). This table is an adaptation by Dr Xavier Fructus of an older nitrox table developed by Dr Barthelemy at the French Navy diving facility (GERS) in the late 1950s. The final version of the Comex 30 heliox table appeared in the 1986 Comex medical book. Since then it has been modified by many users, so that multiple similar versions are in use around the world. The green areas on the graph represent the time spent by the patient breathing pure oxygen. The red areas on the Comex 30 table represent the patient breathing a 50/50 mix of oxygen and helium. Breaks in oxygen therapy are scheduled and are shown as the blue areas on the tables.

A 2011 workshop reviewed the evidence for the use of 30-metre heliox recompression and made some recommendations concerning the appropriate use of these more challenging treatment schedules2.


Saturation treatments are used when divers have developed DCS during or just after decompression from saturation exposures. The customary treatment tables are inappropriate because of the extreme gas loads in ‘slow’ tissues and the often excessive oxygen exposures required. In general, increased pressure is applied, and the oxygen percentages are less. These tables may also be used after recompression to depth of relief in severe cases and when other tables have failed.

Gas mixture

Recompression results in a reduction of bubble size, but it may also be associated with a further uptake of gas into the bubble. Ideally, a breathing mixture would be selected that diffuses into the bubble at a slower rate than the inert gas diffuses out, resulting in bubble shrinkage. This is one reason that the oxygen tables are preferred to the air tables because no further inert gas can be absorbed.

There is anecdotal evidence that recompression using heliox has been beneficial in treating DCS after air dives. Hyldegaard and associates3 demonstrated in a model of spinal DCS that, while breathing air, there is a steady increase in bubble size. When compression occurred while breathing oxygen, there was an initial increase and then decrease in bubble size, whereas with heliox there was a progressive shrinkage in bubble size.

The initial increase in bubble size with oxygen may be explained by the fact that at equal partial pressure differences the flux of oxygen in fat is twice that of nitrogen and four times that of helium3. This may explain the initial worsening of symptoms seen occasionally when patients with DCS are recompressed on oxygen.
There is little available evidence to suggest heliox tables are more effective than the oxygen tables, although many centres will move to a heliox table for the difficult or non-resolving acute case (Table 13.1). Heliox is also easier to breathe at depth than air – especially relevant to patients with respiratory ‘chokes’.

Comparison of the features of standard 18-metre oxygen tables and 30-metre heliox tables

In determining which therapeutic table should be selected, remember:

The natural history of the disease:

  • The longer the surface interval before symptoms, the less likely they are to worsen over time.
  • Neurological symptoms have sequelae; many others do not.

The value of pressure: Recompression will prevent bubble growth and hasten resolution.

The depth effect of recompression is less important the longer the symptoms have been present.

Delay to recompression

Once a manifestation of DCS has developed, the subsequent progress may be related more to the time elapsed before recompression than to the specific treatment table selected. With prompt treatment the destructive tissue distension effects of bubbles are lessened, as are the effects of ischaemia and the chemical and cytological reactions to the bubble.

Even using air tables, Rivera4 demonstrated that if initial treatment was administered less than 30 minutes after symptoms developed, there was a 95 per cent probability of relief. This rate falls to 77 per cent if the delay exceeds 6 hours.
The delay among the dive, the development of symptoms and the presentation for treatment allows the clinician to assess the clinical importance and propose the most rational therapy.

To illustrate these principles, the approach to the patient who presents 30 minutes after a dive with an ascending paralysis is different from the patient who presents 48 hours after a dive with shoulder pain. The first patient’s neurological injury is likely to progress with time, whereas the shoulder pain will not and may undergo spontaneous resolution.

The first patient should be treated aggressively, commencing with an 18-metre oxygen table or even a 30-metre heliox table. If using the former, consideration should be given to the options of going deeper and changing gas mixtures if early or full recovery is not evident.

Patients who delay seeking medical assistance may benefit from recompression even up to 14 days after injury; however, most diving physicians would not consider deep or saturation tables for patients with these late presentations. Consider two illustrative case histories (Case Report 13.1 and Case Report 13.2).

CASE REPORT 13.1: The patient made a dive to 18 metres for 60 minutes on scuba. There had been no dives for a month before this. The isolated symptom of left shoulder pain developed 5 hours after the dive and had been present for 24 hours before the diver presented for medical treatment.

Comment: This is not only a mild case of DCS, it is not going to get significantly worse as long as the patient avoids further exposure to hypobaric or hyperbaric conditions. By the time the medical assessment was made (29 hours after the dive), the tissues will have equilibrated fully with the atmospheric pressure, and thus there will be no significant pressure gradient pushing nitrogen into the bubble. On the contrary, there will be a mild gradient in the opposite direction. The administration of 100 per cent oxygen will enhance this gradient further.

The authors’ approach to such a mild case would be to relieve the patient’s symptoms and perhaps to reduce the possibility of subsequent bone damage (although there is no clear evidence that this latter is really possible) by recompression on an oxygen treatment table – probably at 18 metres. If recompression was not feasible, then surface oxygen would be appropriate.

CASE REPORT 13.2: This patient has symptoms identical to those of the patient in Case Report 13.1, but this time the left shoulder symptoms followed 10 minutes after a 30-metre dive for 30 minutes with a rapid ascent and omitted decompression. The diver has presented for assistance immediately after the symptom developed.

Comment: This is a very different situation from Case Report 13.1. Even assuming that the left shoulder pain is musculoskeletal and not referred neurological or cardiac, the likelihood of progression from the theoretical ‘minor symptom’ to a major case of DCS is much higher.

First, because the symptom developed soon after the dive, it is likely to become worse. Second, more symptoms are likely to develop (remembering that DCS manifestations may continue to arise over the next 24 hours). Third, the tissues surrounding the bubbles may well have nitrogen supersaturation pressures of almost 4 ATA. The bubble, existing on the surface, will have a nitrogen pressure of approximately 1 ATA, as the bubble is at the same ambient pressure as the body. Under these conditions, there will be a gradient between the tissues and the bubbles, increasing the size of the latter until the tissue gas tension becomes equated with the bubble gas tension.

The therapeutic approach to this diver is to recompress him to the maximum depth at which 100 per cent oxygen can be used therapeutically, i.e. USN Table 6 at 18 metres, and if a satisfactory response is obtained and maintained, to decompress him from that depth. Recompression treatment at 18 metres will not rapidly reverse the tissue-to-bubble nitrogen gradient and, if other more serious symptoms develop, it may be necessary to recompress him deeper on heliox (or an oxygen and nitrogen mixture). The authors of this text would select a 30-metre 50 per cent heliox treatment schedule (e.g. Comex 30). At this depth there would be no tissue-to-bubble nitrogen gradient.