Search

Transcutaneous PO2 and PCO2 measurements and estimates of skin respiration were monitored at different levels of inspired PO2 in 20 healthy adults during the first 4 days of the tuberculin reaction, a convenient model of acute inflammation. Hyperoxia at 1 and 2 ATA significantly increased transcutaneous PO2 levels in undisturbed and in inflamed skin but did not fully correct the relative hypoxia at the site of inflammation. Hypercapnia was reduced with O2 breathing at 2 ATA. The apparent rate of O2 consumption at the reaction site was raised during hyperoxia, most prominently at 2 ATA. The most intense reactions showed a central relative slowing of laser-Doppler blood flow indicative of microcirculatory impairment. The extent of the relative hypoxia and hypercapnia was greatest in these strongest reactions. The density of lymphocytes and monocytes in biopsies of 48-h reactions was loosely related to the corresponding transcutaneous PO2 measurements. The present study provides evidence that diffusion barriers, in addition to increased local respiration, can contribute to the apparent hypoxia and hypercapnia of this inflammatory model.

The most serious problems resulting from a change in ambient pressure are pulmonary barotrauma with air embolism and decompression sickness. The small differential pressures used in ventilators at atmospheric pressure may tear lung tissue and, in diving, deaths have occurred from the expansion of pulmonary gas on an ascent of less than two metres. The bubbles of respired gas that enter the systemic circulation often occlude cerebral arteries and may cause infarction. In decompression sickness, bubbles form in the tissues from supersaturation of the nitrogen or helium absorbed under pressure. Joint pain--the 'bends'--is associated with gas in particular connective tissue. Serious decompression sickness results from the entry of microbubbles into the systemic veins. Large numbers of bubbles trapped in the lung cause an acute respiratory syndrome known as 'chokes'. If the lung filter is overwhelmed, or microbubbles pass into the systemic arteries through an atrial septal defect, they may open the blood-brain barrier, affecting brain and spinal cord function. Untreated, demyelination with relative preservation of axons may occur, the pathological hallmarks of multiple sclerosis. Gas bubble disease requires urgent compression in a hyperbaric chamber and the use of high partial pressures of oxygen.

SYSTEMIC HYPERBARIC OXYGEN THERAPY HYPOCHLORITE SOLUTIONS.

The investigation of commercial diving accidents has indicated that the danger of anoxia, from the inhalation of gases not containing oxygen, is not fully recognized. The problem is more common in a variety of general industrial situations and is an occasional cause of death in anaesthesia. It is a particular hazard with inert gases, which, because they are recognized to be non-toxic, give a false sense of security. The pathological findings consist of pulmonary oedema and petechial haemorrhages, mainly in the brain, lungs and myocardium. Whenever possible, a minimum oxygen content should be included in all gases liable to be respired, but where this is not possible, oxygen analysers and alarms should be provided. Where a general hazard exists, personnel must be warned of the danger.

We evaluated effects of duodenojejunal (DJ) feeding on gastric pH and selected gastrointestinal hormones in 13 randomly selected patients in an intensive care unit (ICU). To obtain baseline values for gastric pH, a nasogastric (NG) tube was placed in each patient and gastric pH was measured every 30 minutes for 2 hours. To obtain control values, a Dobbhoff tube was placed fluoroscopically and 0.45 percent saline solution (NaCl), 75 ml, was infused for 1 hour and gastric pH was measured again; the previously placed NG tube was left in position. Then, by randomization, either 0.45 percent NaCl (pH = 5) was continued (n = 6) or a high-nitrogen, isotonic, enteral feeding solution (Osmolite HN, pH = 6.4) (n = 7) was infused, both at 75 ml/h. Gastric pH was noted hourly for 96 hours; antacid (Maalox TC, 15-ml aliquots) was given by NG tube when the pH was 4 or less. After 96 hours, the infusion was stopped and gastric pH was noted for 4 additional hours. Before and during initial saline solution infusion; after 24, 48, 72, and 96 hours of continuous infusion; and 4 hours after stopping the infusion, peripheral venous blood was obtained for measurement of plasma gastric inhibitory polypeptide (GIP) and serum gastrin. Data were analyzed by ANOVA (RMD), Fishers' exact test, and the unpaired t-test. Groups did not differ demographically. Throughout the infusion, gastric pH tended to be higher with the enteral feeding solution than with saline solution, but this was significant only at 24 hours. Less antacid was required with the enteral feeding solution at 24 and 48 hours than with saline solution. Plasma GIP levels were significantly higher with the enteral feeding solution than with saline solution during most of the infusion. Serum gastrin levels did not differ between the groups. In this cohort, infusion of the enteral feeding solution tended to maintain a gastric pH of more than 4 and was associated with increased plasma GIP levels, which may inhibit gastric acid secretion. Early enteral feeding may benefit certain ICU patients.

Decompression sickness affecting the nervous system is still a serious problem in diving, but the mechanisms involved are in dispute. Although microbubbles can be detected in the pulmonary artery on decompression using ultrasound, mammalian lungs are competent filters for microbubbles larger than 20 microns in diameter. It has been assumed that smaller bubbles released by the lungs are harmless, because there is evidence that they do not arrest in the cerebral circulation. We injected 15 +/- 5 microns diameter microbubbles in 5 ml of plasma slowly into the right carotid artery of anesthetized guinea pigs. At intervals of 1, 2, or 3 h postinjection, 2% trypan blue in 2 ml of plasma was injected into the same artery or the contralateral carotid artery. A control animal for each experiment was injected with 5 ml of plasma only, followed by the injection of dye at the same interval. After the animals were killed, the brains were examined for evidence of blood-brain barrier dysfunction. All animals at 1 h, and 9 out of 10 animals at 2 h after the injection of microbubbles, showed extravasation of the albumin-binding dye in the ipsilateral hemisphere, indicating gross blood-brain barrier dysfunction. In each of the matched controls, the barrier in the neocortex remained intact. At Hour 3 the barrier was impermeable to the trypan blue in both experimental and control animals. These experiments demonstrate that microbubbles impair the blood-brain barrier integrity to protein, causing focal edema.(ABSTRACT TRUNCATED AT 250 WORDS)

Transcutaneous measurements of pO2 and pCO2 were made on the forearm skin after intradermal injection of histamine, PGE2, and saline. The mediators, used at concentrations which induce intense hyperaemia, did not modify the steady state tcpO2/pCO2 levels measured with a sensor head temperature of 44 degrees C when breathing air or hyperbaric (2ATA) oxygen. It was deduced that gas transport is unaffected by mediator-induced conditions in the skin. The rates of fall of tcpO2 and of rise of tcpCO2 after arresting the forearm circulation by cuff occlusion of the arm were significantly less at the histamine site than at the PGE2 and saline sites. The values over the PGE2 and saline injection sites were less than those over undisturbed skin. The dynamic tests of respiratory gas exchange indicate that the skin metabolic rate is reduced at all injection sites and the greatest effect was seen with histamine. Measurement of dermal thickness after saline injection has shown that the excess interstitial fluid persists at the time of maximal hyperaemia: this is further accentuated at the histamine site through active oedema formation. Accumulation of excess interstitial fluid (persistence of aqueous injection or oedema generated by the action of mediator) separates the tissue cells. The reduction in the number of cells per unit volume is sufficient to explain the observed reduction in oxygen consumption per unit volume of skin. It is concluded that the increased diffusional distances in mediator-induced oedema are unimportant for the respiration of otherwise normal tissues, but that oedema by reducing oxygen flux may contribute appreciably to the hypoxia of inflamed tissue infiltrated with metabolically active cells.

The principle has been evaluated of detecting stationary tissue gas bubbles simply by measuring the velocity of ultrasound. Agar gel has been used to simulate tissue. At wavelengths appreciably larger than bubble diameter (at least 10:1) the ultrasound seems to "view" the gel-gas mix as though it were one medium, when the velocity of pulsed ultrasound is reduced approximately as predicted theoretically from the decreased modulus. The velocity of sound shows an appreciable (10%) decrease for only 0.65% of the gas phase as bubbles of 20-200 microns diameter and drops to about one-third of the bubble-free value for only 0.79% gas by volume--values well in excess of those theoretically predicted to elicit the symptoms of limb bends. These large changes in the velocity of sound can probably be measured with a much less expensive unit than the one used in this study and would seem to warrant further investigation as a very simple method for detecting bubbles in tissue--whether intravascular or extravascular.

The formulation of decompression procedures has generally been based on the observation that divers can be decompressed without stoppages to surface, from steady-state exposures of about twice the atmospheric pressure. Because decompression sickness rarely develops from this "no-stop decompression", it has been assumed that no gas is liberated. It is therefore assumed, in the calculation of the majority of decompression tables, that using a 2:1 decompression ratio allows the additional gas load from the hyperbaric exposure to be transported to the lungs in solution. Ultrasonic scanning and Doppler techniques have shown that this is not the case. Decompression tables must therefore be formulated so as to take into account the presence of gas, the critical diameter of circulating bubbles and the inherent unsaturation introduced by oxygen.

The neurological features of decompression sickness, which is thought to be due to gas embolism, are similar to those of multiple sclerosis (MS). This similarity suggested the re-examination of a concept, first proposed in 1882, that the demyelination in MS is due to venous thrombosis. Unfortunately, although the plaques of MS are often perivenular, thromboses are not always present. Nevertheless, vascular theories can explain the topography of the lesions in MS. Embolism is generally associated with arterial rather than venous damage, and with neuronal infarction rather than loss of myelin. However, the intra-arterial injection of a range of substances can cause venous damage and perivenous demyelination in the brain, although it does not exactly reproduce the plaques seen in man. There is also evidence in man that fat may lodge in the microcirculation of the nervous system and cause distal perivenous oedema with the loss of myelin from axons. Since acute fat embolism may produce lesions not only in the white matter of the brain, but also in the cord, the retina, the meninges, and the skin, and since all these have been described in MS, subacute fat embolism may be the cause of MS.

The initial event in decompression sickness is the separation of gas from solution because of supersaturation. If this event gives rise to immediate symptoms, recompression is remarkably effective. This end-point is characteristic of joint pain, that is, Type 1 decompression sickness. Unfortunately the onset of serious Type 2 decompression sickness may be insidious and the delay may be associated with blood-brain barrier dysfunction. Pressure is less effective in the resolution of this problem than a raised partial pressure of oxygen. Standard therapy using oxygen may be associated with worsening of symptoms and air tables with recurrence. Recompression to 4 ata and the use of a mixture of 50% oxygen and 50% helium offers a good working compromise in the treatment of both serious decompression sickness and gas embolism arising in air diving, avoiding the need for a differential diagnosis. Only oxygen or helium and oxygen mixtures should be used in the therapy of decompression sickness in helium and oxygen diving. When therapy has been delayed, intravenous fluids and steroids are important adjuncts.

Six experimental investigations of various mechanical aspects of the spinal cord are described relevant to its injury by gas deposited from solution by decompression. These show appreciable resistances to gas pockets dissipating by tracking along tissue boundaries or distending tissue, the back pressure often exceeding the probable blood perfusion pressure--particularly in the watershed zones. This leads to a simple mechanical model of spinal decompression sickness based on the vascular "waterfall" that is consistent with the pathology, the major quantitative aspects, and the symptomatology--especially the reversibility with recompression that is so difficult to explain by an embolic mechanism. The hypothesis is that autochthonous gas separating from solution in the spinal cord can reach sufficient local pressure to exceed the perfusion pressure and thus occlude blood flow.