Shunt-mediated decompression sickness in a compressed air worker with an atrial septal defect.

We report a compressed air worker who had diffuse cutaneous decompression sickness with pain in his left shoulder and visual disturbance characteristic of migraine aura after only his third hyperbaric exposure. The maximum pressure was 253 kPa gauge with oxygen decompression using the Swanscombe Oxygen Decompression Table. He was found to have a very large right-to-left shunt across a 9 mm atrial septal defect. He had transcatheter closure of the defect but had some residual shunting with release of a Valsalva manoeuvre. Thirty-two other tunnel workers undertook the same pressure profile and activities in the same working conditions during the maintenance of a tunnel boring machine for a total of 233 similar exposures and were unaffected. As far as we are aware this is the first report of shunt-mediated decompression sickness in a hyperbaric tunnel worker in the United Kingdom and the second case reported worldwide. These cases suggest that shunt-mediated decompression sickness should be considered to be an occupational risk in modern compressed air working. A right-to-left shunt in a compressed air worker should be managed in accordance with established clinical guidance for divers.

Immersion pulmonary oedema: a cardiological perspective.

It is postulated that immersion pulmonary oedema (IPE) occurs because of combinations of factors that each increase the hydrostatic pressure gradient between the pulmonary capillaries and the alveoli. The factors, by definition, include the effects of immersion, particularly raised central blood volume and hence cardiac filling pressures. Breathing against a negative pressure is important but the magnitude of the effect depends on the relation of the diver's lung centroid to the source of the breathing gas and the breathing characteristics of diving equipment. Other factors are cold-induced vasoconstriction, exertion and emotional stress, but variations of the responses of individuals to these stimuli are important. Hypertension is the most frequent cardiovascular disease predisposing to IPE but other medical conditions are implicated in some patients.

Cutis marmorata and cerebral arterial gas embolism.

Dr Kemper and colleagues reported that, when air was injected into the cerebral circulation of pigs, they developed a rash that looked very similar to cutis marmorata of cutaneous decompression illness (DCI) and to livido reticularis. They postulated that cutaneous DCI in divers may be centrally mediated as a result of cerebral gas embolism. It would be helpful if Kemper et al. described the distribution of the rash in their pigs. In divers, cutaneous DCI is generally confined to parts of the body with significant amounts of subcutaneous fat, such as the trunk and thighs, and the rash often crosses the midline. Colleagues and I have reported that cutaneous DCI is commonly associated with significant right-to-left shunts and particularly persistent foramen ovale (PFO). We postulated that the manifestations of shunt-related DCI, whether neurological or cutaneous, are in large part determined by peripheral amplification of embolic bubbles in those tissues that are most supersaturated with dissolved nitrogen (or other inert gas) at the time that emboli arrive. Hence we postulated that cutaneous DCI is the result of amplification of gas emboli that invade cutaneous capillaries. Dr Kemper has kindly sent me a number of the publications from his department on which their report of this skin rash in pigs is based. The aim of their experiments was to produce significant brain injury by means of cerebral air embolism. Their pigs had no tissues supersaturated with inert gas. They were ventilated with a FiO2 of 0.4 and anaesthetised with ketamine and midazolam. They were also given pancuronium and atropine, before air was injected into their cerebral circulation. If their findings in pigs and the resulting hypothesis were applicable to man, it would mean that one could get cutaneous DCI without decompression: one would only need cerebral gas embolism. During contrast echocardiography, I have produced arterial gas embolism in many hundreds of patients with right-to-left shunts and it is certain that some bubbles went into their cerebral circulations, but I have never seen and no patient has reported getting a rash. Nor am I aware of any reports of gas embolism causing a rash like cutaneous DCI without there being tissue supersaturation following some form of decompression. Kemper and colleagues injected between 0.25 and 1 ml·kg⁻¹ body weight of air into the ascending pharyngeal artery (roughly equivalent to human internal carotid artery) of pigs weighing 30-40kg. That immediately produced significant elevation of blood pressure and heart rate suggesting a 'sympathetic surge'. This is similar to the haemodynamic effects that can occur with subarachnoid haemorrhage and some other catastrophic brain injuries. That effect may have been potentiated by pre-treatment with atropine. There was also a considerable increase in intracranial pressure and major adverse effects on cerebral metabolism. Some pigs died quickly and the survivors were killed at the end of the experiment. I suspect that no pig would have survived the experiments without major neurological injury if they had not been killed. Most people with cutaneous DCI have no detectable neurological manifestations at the time that they have a rash. In those that do have neurological manifestations, it is rarely catastrophic. The increases in heart rate and blood pressure reported in the pigs are similar to the effects of a phaeochromocytoma, which can cause livido reticularis in man. Therefore, I wonder whether an alternative explanation for these observations might be that the cerebral injury in the pigs was so massive that the sympathetic surge was comparable to the effects of catecholamine release from a phaeochromocytoma and caused a rash similar to that seen in patients with a phaeochromocytoma.

Comparison of the size of persistent foramen ovale and atrial septal defects in divers with shunt-related decompression illness and in the general population.

INTRODUCTION: Decompression illness (DCI) is associated with a right-to-left shunt, such as persistent foramen ovale (PFO), atrial septal defect (ASD) and pulmonary arteriovenous malformations. About one-quarter of the population have a PFO, but considerably less than one-quarter of divers suffer DCI. Our aim was to determine whether shunt-related DCI occurs mainly or entirely in divers with the largest diameter atrial defects. METHODS: Case control comparison of diameters of atrial defects (PFO and ASD) in 200 consecutive divers who had transcatheter closure of an atrial defect following shunt-related DCI and in an historic group of 263 individuals in whom PFO diameter was measured at post-mortem examination. RESULTS: In the divers who had experienced DCI, the median atrial defect diameter was 10 mm and the mean (standard deviation) was 9.9 (3.6) mm. Among those in the general population who had a PFO, the median diameter was 5 mm and mean was 4.9 (2.6) mm. The difference between the two groups was highly significant (P < 0.0001). Of divers with shunt-related DCI, 101 (50.5%) had an atrial defect 10 mm diameter or larger, but only 1.3% of the general population studied had a PFO that was 10 mm diameter of larger. CONCLUSIONS: The risk of a diver suffering DCI is related to the size of the atrial defect rather than just the presence of a defect.

The role of persistent foramen ovale and other shunts in decompression illness.

A persistent foramen ovale (PFO) and other types of right-to-left shunts are associated with neurological, cutaneous and cardiovascular decompression illness (DCI). A right-to-left shunt is particularly likely to be implicated in causation when these types of DCI occur after dives that are not provocative. It is believed that venous nitrogen bubbles that form after decompression pass through the shunt to circumvent the lung filter and invade systemic tissues supersaturated with nitrogen (or other inert gas) and as a result there is peripheral amplification of bubble emboli in those tissues. Approximately a quarter of the population have a PFO, but only a small proportion of the population with the largest right-to-left shunts are at high risk of shunt-mediated DCI. The increased risk of DCI in people with migraine with aura is because migraine with aura is also associated with right-to-left shunts and this increased risk of DCI appears to be confi ned to those with a large PFO or other large shunt. Various ultrasound techniques can be used to detect and assess the size of right-to-left shunts by imaging the appearance of bubble contrast in the systemic circulation after intravenous injection. In divers with a history of shunt-mediated DCI, methods to reduce the risk of recurrence include cessation of diving, modification of future dives to prevent venous bubble liberation and transcatheter closure of a PFO.