| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Swedish Defence Research Agency (FOI), Centre for Environmental Physiology, Karolinska Institute, Stockholm, Sweden
Submitted 30 April 2007 ; accepted in final form 25 October 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
3) was significantly (P < 0.001) shorter in the prebreathe group. In a second series, 30 goats breathed air at 0.2 MPa for 6 h. Fifteen minutes before escape from 2.5 MPa, animals were provided with either air (n = 10), oxygen (n = 12), or carbogen (97.5% O2 and 2.5% CO2) gas (n = 8) as breathing gas. Animals breathed a hyperoxic gas (60% O2-40% N2) during the escape. Two animals (carbogen group) suffered oxygen convulsions during the escape but recovered on surfacing. Only one case of DCS occurred (carbogen group). The initial bubble score was reduced in the oxygen group (P < 0.001). The period with bubble score of Kisman-Masurel 3 was also significantly reduced in the oxygen group (P < 0.001). Oxygen breathing before submarine escape reduces initial bubble scores, although its significance in reducing central nervous system DCS needs to be investigated further. submarine escape; decompression sickness; venous gas emboli
In previous experiments, hyperoxic gas (60% O2-40% N2) was breathed during the escapes instead of air. This caused the number of circulating bubbles to decrease more rapidly, both after deep submarine escapes and after escapes preceded by shallow air saturation (5). In the latter situation, hyperoxic escape gas also appeared to protect against late-occurring limb bends (5). However, the peak bubble score, which in these dive profiles occurred minutes after the end of decompression, was not reduced by the use of the hyperoxic gas, and the incidence, albeit low, of DCS affecting the central nervous system (CNS) did not decrease (5).
Computer modeling has suggested that the initial, high burst of bubbles that appears shortly after surfacing from a submarine escape maneuver emanates from the CNS and other so-called fast tissues (14). This early burst of bubbles is therefore considered to reflect the risk of CNS DCS. Preoxygenation has been used historically in altitude decompressions to reduce the decompression stress, as reflected by the decreased number of circulating bubbles and the lowered incidence of DCS, with the level of protection increasing with increased preoxygenation time (29). A previous experiment, in which goats breathed oxygen for 1 h during shallow air dives followed by simulated submarine escape, also showed a significant reduction in maximum bubble scores (30). However, during escape procedures from a disabled submarine, time is often limited, and, additionally, the supply of oxygen may also be too low to allow long periods of oxygen breathing for all of the escapers. The aim of the present study was to investigate the effect of a 15-min period of oxygen breathing before simulated submarine escape. It was hypothesized that this would primarily reduce the gas load of the fast tissues, in particular the CNS, while also providing a viable option for use in a disabled submarine situation.
The half-time for gas turnover in the brain is considered to be between 90 and 300 s (13). Therefore, a 15-min period of oxygen ventilation at surface (0.1 MPa) should be sufficient to wash out the majority of the brain's nitrogen stores, thereby lowering the initial inert gas load and reducing the risk of CNS DCS. In addition, because it is likely that submariners trapped in a disabled submarine will be exposed to raised ambient pressures, the effect of the same period of preescape oxygen breathing after saturation at a raised ambient pressure was also of interest. As noted previously, a pressure increase to 0.2 MPa is a likely disabled submarine scenario, and this will cause the partial pressure of the nitrogen stores in the tissues to increase, causing a higher risk of venous gas bubble evolution on ascent. However, given the short half-time of the brain tissues, a 15-min oxygen prebreathe at 0.2 MPa should still reduce the partial pressure of nitrogen and thus should provide some protection for the CNS.
However, oxygen, especially at high partial pressures, causes vasoconstriction, particularly in the CNS, and therefore may actually reduce the rate of removal of gases from the tissue (1). Another experimental group was therefore given a mixture of carbon dioxide and oxygen (carbogen gas), also for 15-min before simulated submarine escape. Carbon dioxide is a strong vasodilator of the cerebral circulation (7), and carbogen has been used previously in efforts to enhance tissue oxygenation (17). Although, there is an interaction between hyperoxia and carbon dioxide, the cerebral blood flow is increased by high carbon dioxide levels also during hyperoxia (15). Accordingly, it was reasoned that the addition of carbon dioxide might promote increased delivery of oxygen to the tissues and increase "washout" of the nitrogen accumulated in the CNS during the saturation phase of the dive. Therefore, three groups of animals pressurized to 0.2 MPa for 6 h were tested in total: a control group with no prebreathe (NPB), a group given 15 min of oxygen prebreathe (O2PB), and a final group given 15 min of carbogen prebreathe (CPB) before a simulated submarine escape from 2.5 MPa.
It was expected that venous gas emboli (VGE) would evolve after escapes from 240 msw (2.5 MPa) but that simulated escape from this depth would not cause a high risk of life-threatening DCS. The level of carbon dioxide (5 kPa carbon dioxide in oxygen) within the carbogen gas was chosen as a compromise between PCO2 levels that previously have been shown to increase denitrogenation (23) and levels that would be acceptable for the animals to breathe. In addition, in accordance with international standards, submariners should not be exposed to carbon dioxide levels above 4.5–5 kPa, before exiting a disabled submarine (25).
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
DCS is caused by inert gas bubbles forming in the tissue or bloodstream due to supersaturation, which in goats causes discernable symptoms such as joint pain, chokes, and neurological symptoms such as nystagmus and staggers. By definition, DCI refers to any disease that can occur during decompression, not only DCS but also barotrauma and arterial gas embolism, resulting in symptoms such as bloat, sinus bleeds, stroke, and sudden death. The occurrence of CNS oxygen toxicity was diagnosed on occurrence of convulsions during the pressure exposure. Animals that suffered from any severe symptoms were euthanized, and postmortem examination was carried out to determine the exact cause of the symptoms.
Series A: 15-min prebreathe at surface (0.1 MPa) followed by simulated submarine escape to 240 msw (2.5 MPa). Forty animals were included in this group; of these, 12 breathed oxygen (100 kPa inspired PO2) via a hyperbaric treatment hood (Sea-Long) for 15 min before they were placed in the escape chamber (O2PB group). Twenty-eight animals made up the control group (NPB). These animals were exposed to identical dive profiles and prebreathe protocol as the O2PB animals except that they breathed air before the escape. In all cases, the transfer time between taking the animals from the breathing hoods, moving them into the chamber, closing the pressure door, and starting the escape profile compression was not longer than 50 s.
Once inside the chamber, the animals were left free to move around and breathed ambient chamber air. The compression phase of the escape profile lasted for 24 s with 4 s at depth. The decompression speed was 2.75 msw/s (0.0275 MPa/s). Directly after escape, all animals were moved from the chamber to a holding pen. To monitor the health of the animals, breathing frequency and end-tidal oxygen and carbon dioxide concentrations were measured during the first 30 min after escape. To do so, a mask placed over the animal's snout was linked to a fast response CO2/O2 Servomex analyzer, which was in turn connected to a computer running the data collection program Labview (National Instruments).
Precordial Doppler monitoring was carried out every 5 min during the first 30 min, every 15th min between 30 and 120 min, and then every 1 h for 6 h or until an individual was free of any Doppler-detectable circulating bubbles. The amount of bubbles was scored using the Kisman-Masurel (KM) three-digit code, which is commonly used in this field of research. This method converts the code into a 12-point scale (19) and finally into to a numerical scale, where a "+" value adds 0.33 and a "–" value subtracts 0.33; for example, a score of III+ would be equal to 3.33. The KM scores were also converted to bubble severity estimates (Z-scores), also derived by Kisman et al. (18, 19) to give a more linear representation of the amount of circulating VGE, thus giving a better visual representation of the data.
Series B: saturation at 10 msw (0.2 MPa) for 6 h with 15-min prebreathe before simulated submarine escape to 240 msw (2.5 MPa) while breathing 60% O2-40% N2. The animals were introduced to the saturation chamber and kept at 0.2 MPa in air (0.042 MPa inspired O2) for 6 h before the simulated escape. One hour before the escape, they were moved, at pressure, from the saturation chamber to the escape chamber, put in restraints, and fitted with oronasal breathing masks. Twelve of the animals were given 15 min of oxygen (0.2 MPa) breathing just before the escape (Sat + O2PB), 8 were given a carbogen mixture (Sat + CPB) consisting of 97.5% O2 and 2.5% CO2 (equivalent to PCO2 of 5 kPa at 0.2 MPa) for 15 min before escape, and 10 remained on air until the start of the compression (Sat + NPB). During the escape phases, all of the animals breathed a 60% O2-40% N2 mixture via their masks supplied with gas from the hood-inflation system. As in previous experiments with escape from saturation (5), the compression phase lasted 30 s with 4 s at depth, and the decompression speed was 2.75 msw/s (0.0275 MPa/s). Directly after escape, all animals were returned to air breathing, moved to the holding pen, and then monitored as for series A (see above).
Statistics. For series A, which had only two treatment groups, namely O2PB and NPB, a Mann-Whitney U-test was used to test for statistical significance. For series B, comparisons between the three experimental groups regarding maximum bubble scores and time with circulating bubbles were made with the Kruskal-Wallis test (P < 0.05 significant). The Mann-Whitney U-test was used as a post hoc test, with the level of significance divided by the number of comparisons (n = 3) to compensate for repeated testing (P < 0.017 regarded as significant). Evaluations of both KM score and Z-score data were made.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
III (Z 6.65) was significantly reduced (P = 0.0034), from 28 min in the control group to 5 min in the O2PB group (Table 1). The median time to a KM or Z-score of zero (no bubbles) was significantly reduced (P = 0.0004) from 105 min in the control group to 38 min in the O2PB group. Figure 1A describes the time profile of bubble evolution using Z-scores.
|
|
There were three cases of fatal DCI: two in the Sat + O2PB group and one in the Sat + NPB group. Postmortem examination indicated that two fatalities were caused by pulmonary barotrauma, and one may have been caused by severe bloating and subsequent compression and ischemia of the small intestine. Thus supersaturation and subsequent bubble evolution may not have caused any of these incidences. Another animal in the Sat + CPB group suffered from a spinal CNS DCS incident on surfacing, losing control of the hind legs. After hyperbaric therapy, there was still no improvement; therefore, the animal was humanely destroyed.
Of the two animals that apparently succumbed to pulmonary barotrauma and arterial gas embolism, one had to be humanely destroyed immediately on surfacing (Sat + O2PB group). Consequently, no data were collected from this animal; therefore, these data were not included in the results of the study. However, the other animals that suffered from DCI and DCS were not destroyed immediately, and some data (up to 5 data points) were collected. These results were included in the study.
The median peak bubble score was significantly reduced in the goats given oxygen before the escape compared with the Sat + NPB group (KM III+ vs. IV–, respectively; P < 0.001; see Table 1 for Z-scores). There was no significant reduction of peak KM or Z-score in the Sat + CPB group (Table 1). The time to bubble scores of <III (Z-score <6.65) was also significantly shortened in the O2PB group compared with the NPB control group (Table 1). With the conservative significance levels used, although this period in the Sat + CPB group was also much shorter, it did not reach statistical significance (P = 0.0173) when compared vs. the control. The time to a median KM or Z-score of zero (no bubbles) was reduced from >420 min in the NPB group to 360 min in the CPB and to 240 min in the O2PB groups (Table 1). Figure 1B describes time profile vs. bubble evolution for the three treatments investigated in series B, again using Z-scores.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
None of the animals that were exposed to the simulated escape profile without any prior hyperbaric exposure (series A) showed any signs or symptoms of DCS or oxygen toxicity on surfacing. Although the peak bubble score did not differ between the control and the O2PB groups, the animals that had breathed oxygen for 15 min before escape did have a lower total amount of bubbles detected postescape (Table 1 and Fig. 1A), particularly in the 30 min immediately after surfacing when most of the VGE are thought to originate from the CNS and other fast tissues. Median KM Doppler scores for the O2PB animals fell below KM III after only 5 min. Thus oxygen breathing for a period as short as 15 min appears to be efficacious in helping to reduce the initial bubble load after submarine escape.
With respect to the higher risk profile in series B, in which animals were exposed to 0.2 MPa for 6 h before escape, the peak bubble scores were significantly reduced by the 15-min period of oxygen breathing just before escape (Table 1 and Fig. 1B). A reduction of median scores from a KM of IV– to a KM of III+ may not seem to be of any major physiological significance; however, bubble-scoring scales are highly nonlinear (11, 19). KM three-digit codes are produced by an operator listening to a Doppler signal, most commonly from a precordial probe position, then grading the audio information for three qualities. First, the maximum frequency or number of bubbles present in one cardiac period is assessed. Next, the percentage of heart beats to contain that number of bubbles is considered. Finally, the amplitude or loudness of the bubble sound compared with that of the heart sound is graded; all grades are made on a scale of 1–4. Once these primary data have been collected, they can be transformed into the KM scale (see METHODS), into Z-scores, or into other evaluation grades such as bubble numbers (11).
Figure 1 describes the data using the bubble severity estimate scale (Z-scores) derived by Kisman et al. (19). In the KM scale, there are large differences in bubble numbers between the highest ranks (grades III and IV) (11); these differences are more clearly illustrated with the physiological severity scores (Z-scores). The correspondence between Z-scores and KM code was derived from a mathematical equation by Kisman et al., who tried to define a "physiological severity" based again on the frequency and percentage of bubbles containing cardiac cycles and amplitude of the bubble signals. In contrast to the highly nonlinear KM scores, the Z-scores appear to be more directly related to bubble counts and thus the amount of free gas in the bloodstream. Although this numeric inference of physiological severity was never published, other workers have since referenced and used this method in evaluating decompression risk (18). Because the Z-scoring system is directly related to KM codes, we considered that this system was preferable to the use of other scales, such as bubble numbers (11).
Calculations with a decompression model (14) calibrated against bubble numbers (11) made before these experiments predicted that 15 min of oxygen breathing would cause a large reduction in the peak level of bubbles. In the present study, in series A, Z-scores were reduced by 31%; in series B, scores were reduced from 18% (transient peak at 10 min) to 48% after 15 min of oxygen breathing.
Unlike the O2PB treatment in series B, the CPB treatment did not reduce the initial bubble scores compared with the results from the NPB group (Fig. 1B, Table 1). The reason that carbon dioxide was added to the prebreathe mixture was to overcome the vasoconstrictive effects of the high PO2 (1, 4) and thus allow a faster washout of nitrogen. The fact that the only two convulsive incidents that occurred in the present study were both in the CPB group suggests that the CNS of the animals in this group were exposed to higher PO2 levels due to the vasodilation effect of the added carbon dioxide. However, unlike prebreathing with oxygen, carbogen gas did not reduce the initial bubble score. There are at least two possible reasons for this result. Bubble formation and bubble growth are enhanced not only by high gas pressures but also by high gas concentrations. Carbon dioxide, because of its large solubility, may act as a large gas reservoir for bubble formation. In altitude decompressions, this appears to be well established (e.g., Ref. 28), whereas the data for hyperbaric decompressions are not as clear cut.
In a report on submarine escape in goats, Seddon (26) describes a number of experiments in which disabled submarine conditions were simulated by letting the animals breathe carbon dioxide at 2.5 kPa for 23 h, followed by an increase to 4.5 kPa for another 1 h. Direct decompression to surface or via submarine escapes resulted in increased bubble scores and in one series a larger incidence of DCS, compared with similar trials with pure air. However, in no case was there a statistically significant difference between the trials with increased carbon dioxide and air.
In caisson workers exposed to 6-h shifts at roughly 3 atm (0.3 MPa), an increase in the ambient carbon dioxide level in the decompression lock (1.8–2.3% CO2 when reaching surface) appeared to increase the incidence of DCS. Reduction of the carbon dioxide levels below 0.9% significantly reduced the incidence of DCS from 3% to <1% (22).
However, Bell et al. (2) showed that exposure to 2 kPa CO2 during a stay at 1.7 and 1.8 bar (0.17–0.18 MPa) in a nitrox atmosphere actually reduced the incidence and number of circulating bubbles compared with similar dives without carbon dioxide. Also, Masurel and Guillerm (24), using miniature pigs with implanted Doppler probes, did not find an effect on bubble scores with 5 and 7 kPa inspired carbon dioxide during 6-h-long, 20-m-deep air dives. It should be noted that, in the latter experiments, carbon dioxide was not present during the decompression and surfacing.
The results presented by Seddon (26), using the same model as the present study, would seem to support the notion that the higher bubble score in the carbogen group was caused by an enhancing effect of bubble growth due to an increased amount of dissolved carbon dioxide. However, the previous study involved 24-h carbon dioxide exposures, long enough to affect the slow carbon dioxide stores. The present study would only affect the fast stores. One should also bear in mind the particular ventilatory carbon dioxide exchange imposed on subjects performing submarine escapes (21) that would probably decrease the difference in carbon dioxide storage between the animals breathing carbogen and oxygen or air.
It has been shown in both humans (21) and goats (16) that fast compressions during submarine escapes cause an expiratory hypoventilation resulting in hypercapnia. Conversely, during the rapid decompression, the gas expansion causes an alveolar hyperventilation that brings about a hypocapnia. On surfacing, the alveolar PCO2 has been found to be 3.5 kPa in humans (21). Thus, after the submarine escape profiles, alveolar carbon dioxide is actually lower than normal when the supersaturation in the tissues will be at its largest. Because of the short duration of the escape profiles, there may still have been an increased amount of carbon dioxide left in the tissues after the carbogen breathing experiments; however, considering the time from the end of the prebreathe until surfacing from the escapes, around 1 to 3 half-times would have elapsed as the faster, soft tissue CO2 stores have time constants between 1.6 and 2.5 min (12).
Thus, considering all of these studies, the authors would tend to believe that the second explanation for the difference in results between the oxygen and carbogen prebreathe is the most likely: that is, that the increased oxygen supply to the fast tissues (mainly the CNS) during the prebreathe period exceeded the metabolic demands of the tissue. The washed out nitrogen would then be replaced by oxygen, which, at least for a time, would act as an inert gas with bubble-forming potential (5). This explanation is supported by findings from a human study carried out by Lambertsen et al. (20). Oxygen concentration was measured in internal jugular venous blood in four subjects who breathed either oxygen or a 2% carbogen mixture at 3.5 atm (7 kPa CO2). The carbogen breathing caused a very large increase in venous jugular PO2 (from <100 to 1,000 Torr), indicating oxygen loading of the cerebral tissues (20).
Oxygen bubbles or those containing very high oxygen concentrations are likely to shrink more quickly than nitrogen bubbles once excess oxygen starts to be metabolized. This theory is in agreement with the observation that there was a tendency for the bubbles in the CPB group to disappear faster than the bubbles in the NPB group, although a similar effect would also have been seen in carbon dioxide-rich bubbles due to the high solubility of carbon dioxide.
DCI. There was no incidence of limb DCS (bends) in the present study. It has previously been reported that, after 6 h of air breathing at 0.2 MPa, direct decompression to surface caused limb bends in 3 of 24 goats (17). Direct decompression from saturation at 0.18 MPa on air did not cause any bends in a similar group of animals. However, when an escape from 2.5 MPa was added onto the profile, limb bends were again quite frequently observed (17). Therefore, a dive profile of 6 h at 0.2 MPa during air breathing followed by a submarine escape from 2.5 MPa might be expected to cause a large number of limb bends. Because no limb bends were observed in either the O2PB or in NPB animals, the previous finding that hyperoxic (60% O2-40% N2) gas appears to protect against late-occurring bends (5) is further supported here.
Only one case of supersaturation-induced CNS DCS was observed: a spinal bend in one of the CPB animals. Although the present study showed that the initial amount of detectable circulating bubbles decreased, which is an indirect indication that oxygen prebreathing protects against fast tissue DCS, the number of DCS incidents in this trial is too low to determine the effectiveness of the prebreathe method.
However, bubble scores have been shown to vary with severity of dive profile in the goat model (26). Although it is difficult to directly correlate DCS with KM Doppler score, the risk of DCS increases if a subject has the highest Doppler score (KM IV). In a longitudinal study on the goat model, comparisons were made between animals that had experienced DCS and those that did not. In the former category, 16 of 17 animals had KM scores of IV, whereas, in the latter category, only 7 of 13 animals had scored KM of IV (P < 0.05 by Fisher's exact test) (6).
In the present study, there were three DCI fatalities, apparently not caused by supersaturation, but instead initiated by pulmonary barotrauma or severe bloating. A certain number of cases of barotrauma are always expected during submarine escapes, and the amount of oxygen in either a prebreathe gas or that supplied during the escape will not affect the incidence of barotrauma. Therefore, these incidents must be excluded when evaluating any benefits of oxygen breathing.
Oxygen toxicity. Oxygen toxicity is always of concern when such deep pressure profiles are performed and when augmenting inspired PO2 levels for the purpose of nitrogen washout. However, as shown previously (5), escapes can be performed safely from 25 ata (2.5 MPa) during breathing of hyperoxic gas without causing oxygen convulsions. This is despite the fact that the maximum inspired PO2 level at the deepest part of the escape is 1.5 MPa. It has been shown in another animal model that the minimum latency period before the onset of oxygen toxicity convulsions is 4 min, even at pressures up to 3 MPa (8). However, in the present study, one group of animals was also exposed to 15 min of oxygen breathing at 0.2 MPa before the escape, but this increased oxygen exposure was also apparently not severe enough to cause acute oxygen toxicity during these escapes.
When carbon dioxide was added to the prebreathe gas, two animals suffered from the effects of acute oxygen toxicity. Previous studies on humans have shown that carbon dioxide is a potent instigator of oxygen toxicity (20). As discussed previously, inspired carbon dioxide reverses oxygen-induced vasoconstriction and increases cerebral oxygenation (20), an action that we wished to capitalize on in an attempt to increase nitrogen washout from the fast tissues. However, the present findings suggest that the carbon dioxide fraction increased the delivery of oxygen to the CNS to a point that susceptibility to oxygen toxicity was reached for two animals. It should be noted that both of the animals that convulsed during the ascent phase survived the rapid decompression. Obviously, in these animals, the convulsions did not induce laryngospasm or closing of the glottis.
In a disabled submarine scenario, it is conceivable that the survivors will be exposed to high PO2 levels for a considerable period of time before a submarine escape can occur. Before hyperoxic gas is introduced into the escape systems, more research needs to be carried out to establish the oxygen toxicity limits of this particular exposure, i.e., long periods with slightly increased PO2 levels terminating with a short period of extremely high PO2 levels during the escape phase. It is also highly likely that the carbon dioxide level in the atmosphere will be increased in a disabled submarine. Also, the effects of long periods of slightly increased carbon dioxide on subsequent oxygen ventilation will have to be assessed.
In conclusion, oxygen breathing before submarine escape reduces initial bubble scores and increases the speed of disappearance of bubbles from the circulation. However, the effectiveness of oxygen prebreathing in reducing DCS, especially central nervous DCS, needs to be further investigated.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Arieli, E. Boaron, and A. Abramovich Combined effect of denucleation and denitrogenation on the risk of decompression sickness in rats J Appl Physiol, April 1, 2009; 106(4): 1453 - 1458. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
