A wide range of workers are subject to decompression (a reduction in ambient pressure) as part of their working routine. These include divers who themselves are drawn from a wide range of occupations, caisson workers, tunnellers, hyperbaric chamber workers (usually nurses), aviators and astronauts. Decompression of these individuals can and does precipitate a variety of decompression disorders. While most of the disorders are well understood, others are not and in some instances, and despite treatment, injured workers can become disabled. The decompression disorders are the subject of active research.

Mechanism of Decompression Injury

Principles of gas uptake and release

Decompression may injure the hyperbaric worker via one of two primary mechanisms. The first is the consequence of inert gas uptake during the hyperbaric exposure and bubble formation in tissues during and after the subsequent decompression. It is generally assumed that the metabolic gases, oxygen and carbon dioxide, do not contribute to bubble formation. This is almost certainly a false assumption, but the consequent error is small and such an assumption will be made here.

During the compression (increase in ambient pressure) of the worker and throughout their time under pressure, inspired and arterial inert gas tensions will be increased relative to those experienced at normal atmospheric pressure—the inert gas(es) will then be taken up into tissues until an equilibrium of inspired, arterial and tissue inert gas tensions is established. Equilibrium times will vary from less than 30 minutes to more than a day depending upon the type of tissue and gas involved, and, in particular, will vary according to:

• the blood supply to the tissue
• the solubility of the inert gas in blood and in the tissue
• the diffusion of the inert gas through blood and into the tissue
• the temperature of the tissue
• the local tissue work-loads
• the local tissue carbon dioxide tension.

The subsequent decompression of the hyperbaric worker to normal atmospheric pressure will clearly reverse this process, gas will be released from tissues and will eventually be expired. The rate of this release is determined by the factors listed above, except, for as yet poorly understood reasons, it appears to be slower than the uptake. Gas elimination will be slower still if bubbles form. The factors that influence the formation of bubbles are well established qualitatively, but not quantitatively. For a bubble to form the bubble energy must be sufficient to overcome ambient pressure, surface tension pressure and elastic tissue pressures. The disparity between theoretical predictions (of surface tension and critical bubble volumes for bubble growth) and actual observation of bubble formation is explained variously by arguing that bubbles form in tissue (blood vessel) surface defects and/or on the basis of small short-lived bubbles (nuclei) that are continually formed in the body (e.g., between tissue planes or in areas of cavitation). The conditions that must exist before gas comes out of solution are also poorly defined—although it is likely that bubbles form whenever tissue gas tensions exceed ambient pressure. Once formed, bubbles provoke injury (see below) and become increasingly stable as a consequence of coalescence and recruitment of surfactants to the bubble surface. It may be possible for bubbles to form without decompression by changing the inert gas that the hyperbaric worker is breathing. This effect is probably small and those workers that have had a sudden onset of a decompression illness after a change in inspired inert gas almost certainly already had “stable” bubbles in their tissues.

It follows that to introduce a safe working practice a decompression programme (schedule) should be employed to avoid bubble formation. This will require modelling of the following:

• the uptake of the inert gas(es) during the compression and the hyperbaric exposure
• the elimination of the inert gas(es) during and after the decompression
• the conditions for bubble formation.

It is reasonable to state that to date no completely satisfactory model of decompression kinetics and dynamics has been produced and that hyperbaric workers now rely on programmes that have been established essentially by trial and error.

Effect of Boyle’s Law on barotrauma

The second primary mechanism by which decompression can cause injury is the process of barotrauma. The barotraumata can arise from compression or decompression. In compression barotrauma, the air spaces in the body that are surrounded by soft tissue, and hence are subject to increasing ambient pressure (Pascal’s principle), will be reduced in volume (as reasonably predicted by Boyles’ law: doubling of ambient pressure will cause gas volumes to be halved). The compressed gas is displaced by fluid in a predictable sequence:

• The elastic tissues move (tympanic membrane, round and oval windows, mask material, clothing, rib cage, diaphragm).
• Blood is pooled in the high compliance vessels (essentially veins).
• Once the limits of compliance of blood vessels are reached, there is an extravasation of fluid (oedema) and then blood (haemorrhage) into the surrounding soft tissues.
• Once the limits of compliance of the surrounding soft tissues are reached, there is a shift of fluid and then blood into the air space itself.

This sequence can be interrupted at any time by an ingress of additional gas into the space (e.g., into the middle ear on performing a valsalva manoeuvre) and will stop when gas volume and tissue pressure are in equilibrium.

The process is reversed during decompression and gas volumes will increase, and if not vented to atmosphere will cause local trauma. In the lung this trauma may arise from either over-distension or from shearing between adjacent areas of lung that have significantly different compliance and hence expand at different rates.

Pathogenesis of Decompression Disorders

The decompression illnesses can be divided into the barotraumata, tissue bubble and intravascular bubble categories.

Barotraumata

During compression, any gas space may become involved in barotrauma and this is especially common in the ears. While damage to the external ear requires occlusion of the external ear canal (by plugs, a hood, or impacted wax), the tympanic membrane and middle ear is frequently damaged. This injury is more likely if the worker has upper respiratory tract pathology that causes eustachian tube dysfunction. The possible consequences are middle ear congestion (as described above) and/or tympanic membrane rupture. Ear pain and a conductive deafness are likely. Vertigo may result from an ingress of cold water into the middle ear through a ruptured tympanic membrane. Such vertigo is transient. More commonly, vertigo (and possibly also a sensorineural deafness) will result from inner ear barotrauma. During compression, inner ear damage often results from a forceful valsalva manoeuvre (that will cause a fluid wave to be transmitted to the inner ear via the cochlea duct). The inner ear damage is usually within the inner ear—round and oval window rupture is less common.

The paranasal sinuses often are similarly involved and usually because of a blocked ostium. In addition to local and referred pain, epistaxis is common and cranial nerves may be “compressed”. It is noteworthy that the facial nerve may be likewise affected by middle ear barotrauma in individuals with a perforate auditory nerve canal. Other areas that may be affected by compressive barotrauma, but less commonly, are the lungs, teeth, gut, diving mask, dry-suits and other equipment such as buoyancy compensating devices.

Decompressive barotraumata are less common than compressive barotraumata, but tend to have a more adverse outcome. The two areas primarily affected are the lungs and inner ear. The typical pathological lesion of pulmonary barotrauma has yet to be described. The mechanism has been variously ascribed to the over-inflation of alveoli either to “open up pores” or mechanically to disrupt the alveolus, or as the consequence of shearing of lung tissue due to local differential lung expansion. Maximum stress is likely at the base of alveoli and, given that many underwater workers often breathe with small tidal excursions at or near total lung capacity, the risk of barotrauma is increased in this group as lung compliance is lowest at these volumes. Gas release from damaged lung may track through the interstitium to the hilum of the lungs, mediastinum and perhaps into the subcutaneous tissues of the head and neck. This interstitial gas may cause dyspnoea, substernal pain and coughing which may be productive of a little bloodstained sputum. Gas in the head and neck is self-evident and may occasionally impair phonation. Cardiac compression is extremely rare. Gas from a barotraumatised lung may also escape into the pleural space (to cause a pneumothorax) or into the pulmonary veins (to eventually become arterial gas emboli). In general, such gas most commonly either escapes into the interstitium and pleural space or into the pulmonary veins. Concurrent obvious damage to the lung and arterial gas embolism are (fortunately) uncommon.

Autochthonous tissue bubbles

If, during decompression, a gas phase forms, this is usually, initially, in tissues. These tissue bubbles may induce tissue dysfunction via a variety of mechanisms—some of these are mechanical and others are biochemical.

In poorly compliant tissues, such as long bones, the spinal cord and tendons, bubbles may compress arteries, veins, lymphatics and sensory cells. Elsewhere, tissue bubbles may cause mechanical disruption of cells or, at a microscopic level, of myelin sheaths. The solubility of nitrogen in myelin may explain the frequent involvement of the nervous system in decompression illness amongst workers who have been breathing either air or an oxygen-nitrogen gas mixture. Bubbles in tissues may also induce a biochemical “foreign-body” response. This provokes an inflammatory response and may explain the observation that a common presentation of decompression illness is an influenza-like illness. The significance of the inflammatory response is demonstrated in animals such as rabbits, where inhibition of the response prevents the onset of decompression illness. The major features of the inflammatory response include a coagulopathy (this is particularly important in animals, but less so in humans) and the release of kinins. These chemicals cause pain and also an extravasation of fluid. Haemoconcentration also results from the direct effect of bubbles on blood vessels. The end result is a significant compromise of the microcirculation and, in general, measurement of the haematocrit correlates well with the severity of the illness. Correction of this haemoconcentration has a predictably significant benefit on outcome.

Intravascular bubbles

Venous bubbles may either form de-novo as gas comes out of solution or they may be released from tissues. These venous bubbles travel with blood flow to the lungs to be trapped in the pulmonary vasculature. The pulmonary circulation is a highly effective filter of bubbles because of the relatively low pulmonary artery pressure. In contrast, few bubbles are trapped for long periods in the systemic circulation because of the significantly greater systemic arterial pressure. The gas in bubbles trapped in the lung diffuses into the pulmonary air spaces from where it is exhaled. While these bubbles are trapped, however, they may cause adverse effects by either provoking an imbalance of lung perfusion and ventilation or by increasing pulmonary artery pressure and consequently right heart and central venous pressure. The increased right heart pressure can cause “right to left” shunting of blood through pulmonary shunts or intra-cardiac “anatomical defects” such that bubbles bypass the lung “filter” to become arterial gas emboli. Increases in venous pressure will impair venous return from tissues, thereby impairing the clearance of inert gas from the spinal cord; venous haemorrhagic infarction may result. Venous bubbles also react with blood vessels and blood constituents. An effect on blood vessels is to strip the surfactant lining from endothelial cells and hence to increase vascular permeability, which may be further compromised by the physical dislocation of endothelial cells. However, even in the absence of such damage, endothelial cells increase the concentration of glycoprotein receptors for polymorphonuclear leukocytes on their cell surface. This, together with a direct stimulation of white blood cells by bubbles, causes leucocyte binding to endothelial cells (reducing flow) and subsequent infiltration into and through the blood vessels (diapedesis). The infiltrating polymorphonuclear leukocytes cause future tissue injury by release of cytotoxins, oxygen free radicals and phospholipases. In blood, bubbles will not only cause the activation and accumulation of polymorphonuclear leukocytes, but also the activation of platelets, coagulation and complement, and the formation of fat emboli. While these effects have relatively minor importance in the highly compliant venous circulation, similar effects in the arteries can reduce blood flow to ischaemic levels.

Arterial bubbles (gas emboli) can arise from:

• pulmonary barotrauma causing the release of bubbles into the pulmonary veins
• bubbles  being  “forced”  through  the  pulmonary  arterioles (this process is enhanced by oxygen toxicity and by those bronchodilators that  are  also  vasodilators  such as  aminophylline)
• bubbles bypassing the lung filter through a right to left vascular channel (e.g., patent foramen ovale).

Once in the pulmonary veins, bubbles return to the left atrium, left ventricle, and then are pumped into the aorta. Bubbles in the arterial circulation will distribute according to buoyancy and blood flow in large vessels, but elsewhere with blood flow alone. This explains the predominant embolism of the brain and, in particular, the middle cerebral artery. The majority of bubbles that enter the arterial circulation will pass through into the systemic capillaries and into the veins to return to the right side of the heart (usually to be trapped in the lungs). During this transit these bubbles may cause a temporary interruption of function. If the bubbles remain trapped in the systemic circulation or are not redistributed within five to ten minutes, then this loss of function may persist. If bubbles embolise the brain stem circulation, then the event may be lethal. Fortunately, the majority of bubbles will be redistributed within minutes of first arrival in the brain and a recovery of function is usual. However, during this transit the bubbles will cause the same vascular (blood vessels and blood) reactions as described above in venous blood and veins. Consequently, a significant and progressive decline in cerebral blood flow may occur, which may reach the levels at which normal function cannot be sustained. The hyperbaric worker will, at this time, suffer a relapse or deterioration in function. In general, about two-thirds of hyperbaric workers who suffer cerebral arterial gas embolism will spontaneously recover and about one-third of these will subsequently relapse.

Clinical Presentation of Decompression Disorders

Time of onset

Occasionally, the onset of decompression illness is during the decompression. This is most commonly seen in the barotraumata of ascent, particularly involving the lungs. However, the onset of the majority of decompression illnesses occurs after decompression is complete. Decompression illnesses due to the formation of bubbles in tissues and in blood vessels usually become evident within minutes or hours after decompression. The natural history of many of these decompression illnesses is for the spontaneous resolution of symptoms. However, some will only resolve spontaneously incompletely and there is a need for treatment. There is substantial evidence that the earlier the treatment the better the outcome. The natural history of treated decompression illnesses is variable. In some cases, residual problems are seen to resolve over the following 6-12 months, while in others symptoms appear not to resolve.

Clinical manifestations

A common presentation of decompression illness is an influenza-like condition. Other frequent complaints are various sensory disorders, local pain, particularly in the limbs; and other neurologic manifestations, which may involve higher functions, special senses and motor weariness (less commonly the skin and lymphatic systems may be involved). In some groups of hyperbaric workers, the most common presentation of decompression illness is pain. This may be a discrete pain about a specific joint or joints, back pain or referred pain (when the pain is often located in the same limb as are overt neurologic deficits), or less commonly, in an acute decompression illness, vague migratory aches and pains may be noticed. Indeed, it is reasonable to state that the manifestations of the decompression illnesses are protean. Any illness in a hyperbaric worker that occurs up to 24-48 hours after a decompression should be assumed to be related to that decompression until proven otherwise.

Classification

Until recently, the decompression illnesses were classified into:

• the barotraumata
• cerebral arterial gas embolism
• decompression sickness.

Decompression sickness was further subdivided into Type 1 (pain, itch, swelling and skin rashes), Type 2 (all other manifestations) and Type 3 (manifestations of both cerebral arterial gas embolism and decompression sickness) categories. This classification system arose from an analysis of the outcome of caisson workers using new decompression schedules. However, this system has had to be replaced both because it is neither discriminatory nor prognostic and because there is a low concordance in diagnosis between experienced physicians. The new classification of the decompression illnesses recognises the difficulty in distinguishing between cerebral arterial gas embolism and cerebral decompression sickness and similarly the difficulty in distinguishing Type 1 from Type 2 and Type 3 decompression sickness. All decompression illnesses are now classified as such—decompression illness, as described in table 1. This term is prefaced with a description of the nature of the illness, the progression of symptoms and a list of the organ systems in which the symptoms are manifest (no assumptions are made about the underlying pathology). For example, a diver may have acute progressive neurological decompression illness. The complete classification of the decompression illness includes a comment on the presence or absence of barotrauma and the likely inert gas loading. These latter terms are relevant to both treatment and likely fitness to return to work.

Table 1. Revised classification system of the decompression illnesses

First Aid Management

Rescue and resuscitation

Some hyperbaric workers develop a decompression illness and require to be rescued. This is particularly true for divers. This rescue may require their recovery to a stage or diving bell, or a rescue from underwater. Specific rescue techniques must be established and practised if they are to be successful. In general, divers should be rescued from the ocean in a horizontal posture (to avoid possibly lethal falls in cardiac output as the diver is re-subjected to gravity—during any dive there is a progressive loss of blood volume consequent to displacement of blood from the peripheries into the chest) and consequent diuresis and this posture should be maintained until the diver is, if necessary, in a recompression chamber.

The resuscitation of an injured diver should follow the same regimen as used in resuscitations elsewhere. Of specific note is that the resuscitation of a hypothermic individual should continue at least until the individual is rewarmed. There is no convincing evidence that resuscitation of an injured diver in the water is effective. In general, the divers’ best interests are usually served by early rescue ashore, or to a diving bell/platform.

Oxygen and fluid resuscitation

A hyperbaric worker with a decompression illness should be laid flat, to minimize the chances of bubbles distributing to the brain, but not placed in a head-down posture which probably adversely affects the outcome. The diver should be given 100% oxygen to breathe; this will require either a demand valve in a conscious diver or a sealing mask, high flow rates of oxygen and a reservoir system. If oxygen administration is to be prolonged, then airbreaks should be given to ameliorate or retard the development of pulmonary oxygen toxicity. Any diver with decompression illness should be re-hydrated. There is probably no place for oral fluids in the acute resuscitation of a severely injured worker. In general, it is difficult to administer oral fluids to someone lying flat. Oral fluids will require the administration of oxygen to be interrupted and then usually have negligible immediate effect on the blood volume. Finally, since subsequent hyperbaric oxygen treatment may cause a convulsion, it is not desirable to have any stomach contents. Ideally then, fluid resuscitation should be by the intravenous route. There is no evidence of any advantage of colloid over crystalloid solutions and the fluid of choice is probably normal saline. Solutions containing lactate should not be given to a cold diver and dextrose solutions should not be given to anyone with a brain injury (as aggravation of the injury is possible). It is essential that an accurate fluid balance be maintained as this is probably the best guide to the successful resuscitation of a hyperbaric worker with decompression illness. Bladder involvement is sufficiently common that early recourse to bladder catheterization is warranted in the absence of urinary output.

There are no drugs that are of proven benefit in the treatment of the decompression illnesses. However, there is growing support for lignocaine and this is under clinical trial. The role of lignocaine is thought to be both as a membrane stabiliser and as an inhibitor of the polymorphonuclear leukocyte accumulation and blood vessel adherence that is provoked by bubbles. It is noteworthy that one of the probable roles of hyperbaric oxygen is also to inhibit the accumulation of and adherence to blood vessels of leucocytes. Finally, there is no evidence that any benefit is derived from the use of platelet inhibitors such as aspirin or other anticoagulants. Indeed, as haemorrhage into the central nervous system is associated with severe neurological decompression illness, such medication may be contra-indicated.

Retrieval

Retrieval of a hyperbaric worker with decompression illness to a therapeutic recompression facility should occur as soon as is possible, but must not involve any further decompression. The maximum altitude to which such a worker should be decompressed during aeromedical evacuation is 300 m above sea level. During this retrieval, the first aid and adjuvant care described above should be provided.

Recompression Treatment

Applications

The definitive treatment of most of the decompression illnesses is recompression in a chamber. The exception to this statement are the barotraumata that do not cause arterial gas embolism. The majority of aural barotrauma victims require serial audiology, nasal decongestants, analgesics and, if inner ear barotrauma is suspected, strict bed rest. It is possible however that hyperbaric oxygen (plus stellate ganglion blockade) may be an effective treatment of this latter group of patients. The other barotraumata that often require treatment are those of the lung—most of those respond well to 100% oxygen at atmospheric pressure. Occasionally, chest cannulation may be needed for a pneumothorax. For other patients, early recompression is indicated.

Mechanisms

An increase in ambient pressure will make bubbles smaller and hence less stable (by increasing surface tension pressure). These smaller bubbles will also have a greater surface area to volume for resolution by diffusion and their mechanical disruptive and compressive effects on tissue will be reduced. It is also possible that there is a threshold bubble volume that will stimulate a “foreign-body” reaction. By reducing bubble size, this effect may be reduced. Finally, reducing the volume (length) of columns of gas that are trapped in the systemic circulation will promote their redistribution to the veins. The other outcome of most recompressions is an increase in the inspired (PiO2) and arterial oxygen tension (PaO2). This will relieve hypoxia, lower interstitial fluid pressure, inhibit the activation and accumulation of polymorphonuclear leukocytes that is usually provoked by bubbles, and lower the haematocrit and hence blood viscosity.

Pressure

The ideal pressure at which to treat decompression illness is not established, although the conventional first choice is 2.8 bar absolute (60 fsw; 282 kPa), with further compression to 4 and 6 bar absolute pressure if the response of symptoms and signs is poor. Experiments in animals suggest that 2 bars absolute pressure is as effective a treatment pressure as greater compressions.

Gas(es)

Similarly, the ideal gas to be breathed during the therapeutic recompression of these injured workers is not established. Oxygen-helium mixtures may be more effective in the shrinkage of air bubbles than either air or 100% oxygen and are the subject of ongoing research. The ideal PiO2 is thought, from in vivo research, to be about 2 bar absolute pressure although it is well established, in head injured patients, that the ideal tension is lower at 1.5 bars absolute. The dose relationship with regard to oxygen and inhibition of bubble-provoked polymorphonuclear leukocyte accumulation has not yet been established.

Adjuvant care

The treatment of an injured hyperbaric worker in a recompression chamber must not be allowed to compromise his/her need for adjuvant care such as ventilation, rehydration and monitoring. To be a definitive treatment facility, a recompression chamber must have a working interface with the equipment routinely used in critical care medical units.

Follow-up treatment and investigations

Persistent and relapsing symptoms and signs of decompression illness are common and most injured workers will require repeated recompressions. These should continue until the injury is and remains corrected or at least until two successive treatments have failed to produce any sustained benefit. The basis of ongoing investigation is careful clinical neurological examination (including mental status), as available imaging or provocative investigative techniques have either an associated excessive false positive rate (EEG, bone radio-isotope scans, SPECT scans) or an associated excessive false negative rate (CT, MRI, PET, evoked response studies). One year after an episode of decompression illness, the worker should be x-rayed to determine if there is any dysbaric osteonecrosis (aseptic necrosis) of their long bones.

Outcome

The outcome after recompression therapy of decompression illness depends entirely upon the group being studied. Most hyperbaric workers (e.g., military and oil-field divers) respond well to treatment and significant residual deficits are uncommon. In contrast, many recreational divers treated for decompression illness have a subsequent poor outcome. The reasons for this difference in outcome are not established. Common sequelae of decompression illness are in order of decreasing frequency: depressed mood; problems in short-term memory; sensory symptoms such as numbness; difficulties with micturition and sexual dysfunction; and vague aches and pains.

Return to hyperbaric work

Fortunately, most hyperbaric workers are able to return to hyperbaric work after an episode of decompression illness. This should be delayed for at least a month (to allow a return to normal of the disordered physiology) and must be discouraged if the worker suffered pulmonary barotrauma or has a history of recurrent or severe inner ear barotrauma. A return to work should also be contingent upon:

• the severity of the decompression illness being commensurate with the extent of the hyperbaric exposure/decompression stress
• a good response to treatment
• no evidence of sequelae.

Back