GB2082328A - Continuously Monitoring a Subject for Pulmonary Oxygen Toxicity - Google Patents

Abstract

A device for continuously monitoring a subject for pulmonary oxygen toxicity comprises a force transducer 10 whose ends are attached to a band which passes around the subject's chest whereby the expansion and contraction of the chest wall during respiration results in a roughly sinusoidal output signal have maximum and minimum excursions FH and FL, respectively. A micro-processor (12) is programmed firstly to establish the time period within which the signal changes between two predetermined trigger values [FL+0.4 (FH-FL)] and [FL+0.9 (FH-FL)] on both the inspirational ( DELTA ti) and expirational ( DELTA te) parts of the signal and then to establish the ratio <IMAGE> of the latter two time periods. The operation is repeated on a plurality of successive breaths and the average value of the ratio is calculated and displayed (at 16) to indicate the oxygen toxicity. <IMAGE>

Description

SPECIFICATION Device for Continuously Monitoring a Subject for Pulmonary Oxygen Toxicity The present invention is concerned with a device for continuously monitoring a subject for pulmonary oxygen toxicity.

The understanding of the present invention is assisted by the following explanation of the background and context within which the invention has been made.

Although oxygen is essential to life, it can prove toxic when administered in excessive amounts.

The mechanism remains essentially unknown although many conflicting theories have been proposed. In man, the toxicity can take one of two forms depending upon the nature of the exposure.

For continuous exposures to oxygen partial pressures in excess of 2 to 3 atmospheres, the presenting symptoms are essentially neurologic in nature and appear relatively quickly; the higher the partial pressure the sooner they are observed.

The most serious and characteristic symptom of this form of oxygen poisoning is a convulsion, i.e.

repeated seizures virtually indistinguishable from a grand mal epileptic episode.

When breathing lower oxygen partial pressures continuously, symptoms take much longer to become manifest but then involve the lungs rather than the nervous system. This pulmonary oxygen toxicity is characterised by a sub-sternal pain in the chest, sometimes accompanied by coughing and developing into a general difficulty in breathing. It can be reversed by returning the subject to a sub-toxic level of oxygen but, if a toxic level is maintained, the respiratory distress is progressive and can ultimately result in death.

From a more physiological viewpoint, the excessive oxygen causes fluid to collect in the lungs (pulmonary oedema) adversely affecting their mechanical properties and widening the blood-air barrier across which oxygen molecules need to pass in order to effect adequate oxygenation of the blood. Hence, paradoxically if the subject dies, he does so from lack of oxygen to the brain or other vital organs (hypoxia). This hypoxia can be alleviated immediately by further raising the oxygen partial pressure to provide a higher gradient for oxygen diffusion in the lungs.

However such remission is only temporary since the further elevation of the oxygen level also accelerates the rate of formation of pulmonary oedema with a supra-proportional increase in the barrier to blood oxygenation. This also increases the barrier to the elimination of carbon dioxide.

Hence it is easy to enter a vicious cycle in the administration of elevated oxygen partial pressures in which death from hypoxia is the inevitable end. At least, it is a trap to be avoided before the subject becomes dependent upon elevation of the oxygen partial pressure (to2) for adequate oxygenation of his blood. In clinical situations, this tends to occur where a patient has a particular disorder which lowers his oxygen tolerance as characterised by his oxygen threshold. This is the P02 above which pulmonary oxygen toxicity will eventually start and then get progressively worse and below which it will recede.This threshold can be just under 0.5 Atm for some patients with lung diseases, while even fit, healthy young men will develop symptoms of pulmonary oxygen toxicity if allowed to breath pure oxygen at normal barometric pressure for 24 hours.

It can be seen from the above description that it is highly desirable to avoid pulmonary oxygen toxicity. However, to do so simply by keeping below the threshold P02 would be most inefficient and deprive the subject of the many benefits which elevated oxygen can provide for particular disorders on the short term. It is therefore highly desirable to detect the early onset of pulmonary oxygen toxicity and certainly to detect it before it has progressed to the stage at which the subject is dependent upon the elevated oxygen diffusion gradient across his lungs for providing his basic metabolic oxygen requirements.

Various tables and formulae have been proposed in attempts to help the physician in prescribing the oxygen exposure, but these tend to suffer from three major drawbacks. The first is the very wide variation in individual susceptibility to oxygen poisoning; the second revolves around the very practical point of knowing the oxygen partial pressure actually administered to the subject. The oxygen tent commonly used in the hospital ward may contain up to 50% air even some time after ventilation with pure oxygen, while the bearded diver breathing pure oxygen from an oral-nasal mask may actually derive 40% of his inhaled gas from the surrounding environment. Thirdly, there is the question if how to estimate the cumulative effect of elevated oxygen when it has been administered intermittently or at various levels. Two attempts have been proposed for this purpose.In the one published by the U.S. Navy, units (Units of Pulmonary Toxicity Dose or UPTDs) are added every minute to give a total reflecting the total pulmonary insult. The other is based upon the principle of superposition and continuously updates an index (the Cumulative Oxygen Toxicity index), again reflecting the cumulative insult to the lungs. However neither of these is anywhere near as meaningful as actually monitoring the subject himself and detecting pulmonary oxygen poisoning or even quantifying it if it has already progressed beyond the threshold stages.

The development of a direct oxygen toxicity monitor would be of great benefit in the clinical use of elevated oxygen partial pressures with particular relevance to the following situations:~ 1. The newborn where as oxygen-enriched atmosphere can help the baby survive a temporary oxygen insufficiency. It is most tempting for a nurse to given oxygen to a "blue" baby when it responds by turning the pink colour which one tends to regard as helathy for a newborn infant. However, if this is continued to the point where the toxic nature of the elevated oxygen has produced pulmonary oedema to the extent where a sub-toxic P02 may no longer provide the basic oxygen needs, then the baby may have already entered the vicious cycle described above.Pharmacological means of alleviating pulmonary oedema and breaking the cycle are not always successful and many deaths can be attributed to over-prescription of oxygen. It is therefore highly desirable to be able to detect the onset of oxygen toxicity in the newborn and so enable the physician to gain the maximum advantage to be derived from elevated oxygen without fear of entering the vicious cycle.

2. Other patients to whom elevated oxygen would be a benefit and those in whom it is difficult to predict the onset of pulmonary oxygen toxicity on account of the interaction with other medication.

3. Patients in whom it is most desirable to monitor pulmonary oedema whether induced by oxygen a disease process such as pneumonia, the medication or any combination or interaction of these factors.

4. Divers - a somewhat different group since these are essentially fit, healthy young men in the prime of life.

With particular regard to the latter category, when a diver descends into the ocean and stays on the bottom for a time which exceeds a limit determined by the depth, then he could experience any of a wide variety of symptoms (known collectively as decompression sickness) if he were to return directly to the surface. When the combination of depth and bottom time exceed those limits, then it is necessary to invoke gradual decompression in returning safely to the surface.

The formulation of the decompression procedure for each dive is complex and controversial. The procedure to be followed is termed a "diving table" and prescribes a series of times to be spent at various "stops", i.e. at certain depth intervals between the bottom and the surface. Modern tables do not only specify the times and depths but the particular mixture of gases to be breathed at those stops.

There are also procedures known as "treatment tables" which are used if the diving table proved inadequate and the diver needs to be recompressed in order to reduce the volume of the bubbles causing the trouble. He is then held at pressure for a sequence of stops and/or periods of breathing pure oxygen or an oxygen-enriched gas mix in order to dissolve the recompressed bubbles.

There is unanimous agreement that elevating the oxygen partial pressure greatly accelerates the treatment of limb "bends" (Type I decompression sickness) which are the presenting symptoms in 85 to 95% of cases.

Substitution of oxygen for inert gas during the initial decompression also accelerates the whole decompression procedure so that less time is needed in returning a diver from his working pressure to normal atmospheric pressure big advantage commercially. Hence a diver may experience many hours of breathing an oxygen partial pressure far above normal and often fluctuating widely as he changes depth and switches breathing mixes. If he then develops symptoms of decompression sickness, he will be recompressed and probably exposed to much more oxygen as therapy, so he becomes a prime target for pulmonary oxygen toxicity.

Sometimes a diver's chest will already be hurting so badly from the effects of oxygen poisoning that he will not report a "bend" for fear of being subjected to a treatment which will involve more oxygen and further aggravation of his soreness. Failure to report decompression sickness is serious since delay in treatment is more likely to leas to permanent injury.

Although reducing the pain and general discomfort to the diver caused by oxygen toxicitwy is reason enough to try to detect and avoid it, there is one other justification. This concerns the recent finding that a lung poisoned with excessive exposure to oxygen is no longer an effective filter for the many bubbles found in the blood returned to the heart via the veins. These bubbles are not responsible for limb bends and give rise to no symptoms if filtered out of the circulation by the lungs. However, if the lung is no longer capable of trapping them, then they become potentially very dangerous. These bubbles are now in arterial blood and may seriously impede blood flow to parts of the brain or the heart if, by chance, they are in the blood flowing to those organs.These can cause some of the less common yet much more dangerous forms of "neurologic" decompression sickness since these are more likely to result in permanent disability.

Hence it is particularly desirable to have a noninvasive means of detecting the onset of pulmonary oxygen toxicity which can be used in the diving bell or in the deck decompression chamber where the men are decompressing. A direct monitor has the further advantage that it can allow for the possible potentialing action of decompression whereby the asymptomatic bubbles trapped by the lung can cause oedema to augment that already induced directly by the elevated oxygen.

Many studies have been directed towards developing a direct means of monitoring pulmonary oxygen toxicity in view of its importance~ in diving and clinical work in general. All manner of pulmonary function tests have been investigated and attempts made to correlate the implied changes in lung mechanics with oxygen history. The most encouraging indication was thought to be afforded by measuring the vital capacity, but this offers a poor correlation with subjective feeling and, in any case, has two major disadvantages. Firstly it is a measurement based upon a subjective effort in forcibly breathing out as hard as possible and one may not get a true reading if there is any holding back and this can be a painful process when the lungs have been poisoned with oxygen.Secondly, vital capacity cannot be measured on animals and therefore compared with the true indications of oxygen toxicity. These include lung weight, pulmonary oedema and the various mechanical properties of the excised lungs or their microscopic examination afforded by sacrificing animals at various stages of oxygen exposure.

The approach which led to the present invention was instigated by the results of a published investigation of lung insults caused by small amounts of air pollutants such as chlorine and nitric oxide - i.e. by gases generally recognised to be toxic to the lung. In that study based upon rabbits, the authors had shown a good correlation between the pathology of the excised lungs and the ratio (VE/Vl) of the rate of mid-expiratory flow (VE) to the rate of midinspiratory flow (Vl). Mid-inspiratory flow is measured as the rate of flow, or gradient of the lung volume vs. time curve at the mid-point of inspiration; while mid-expiratory flow refers to the corresponding point for expiration.

This approach seemed promising since it involves a non-invasive technique which could be applied to animals and, therefore, tested much more rigorously than otherwise. The fact that (VE/VI) offered such a good correlation with lung weight, and the gross pathology induced by air pollutants resulted in the initiation of further research to see whether the same index, i.e.

(VwV1) also offered a good reflection of lung damage induced by excessive oxygen.

Upon investigation, it was shown that (VE/K/I) offered just as good an index of lung damage induced by excessive oxygen exposure as it did when induced by small amounts of chlorine or other noxious gases. This was most encouraging from a research standpoint, but the original method of measurement was useless from a commercial viewpoint. In the original study of pullutants, air flows had been directly measured using a pneumotachograph or had been calculated from the gradient of the lung-volume vs. time curve (Fig. 1) which has been recorded using a capacitance plesythmograph. Both of these instruments are ideal for the research laboratory but are totally unsuitable for use in practical diving or even in other clinical situations.

Both techniques are expensive and require trained technicians to operate them. In addition, the high voltages associated with the capacitance plesythmograph present an explosion hazard if used close to oxygen, while the pneumotachograph is difficult to calibrate under pressure and is rather bulky. Moreover, its use would require modification to the plumbing of the oxygen supply or overboard dump.

The aim of the present invention is therefore to provide a device which is simple and safe to use, relatively inexpensive to manufacture, but which is capable of analysing the subject's breathing to compare predetermined aspects of the expiratory and inspiratory rates and produce a meaningful displayed result representative of the prevailing pulmonary oxygen toxicity.

According to the present invention there is provided a device for continuously monitoring a subject for pulmonary oxygen toxicity comprising a transducer for fitting to the subject to provide an electrical signal representative of the cyclic expansion and contraction of the subject's chest during respiration, means for operating on said signal to establish the time period within which said signal changes between predetermined trigger values on both an inspirational and an expirational part of the signal, and means for establishing the ratio of the latter time periods to provide an indication of said oxygen toxicity.

Preferably, said inspirational and expirational parts of the signal are disposed on the adjacent rising and falling portions of the signal corresponding to a given single breath, the trigger values having been established from the maximum and minimum levels reached by a preceding breath. Advantageously, the latter breath is the immediately preceding breath.

Advantageously, this operation is repeated on a plurality of successive breaths, say ten, and the average value of the ratio is calculated and displayed. Alternatively, before display, those individual ratio values which differ from the average by more than a predetermined amount, say +15%, are rejected and a new average calculated and displayed from the remaining values.

It has been found to be most appropriate to select the trigger values for a given breath at levels corresponding respectively to 40% and 90% of the maximum excursion of the transducer signal on the previous breath, although it is not intended that the invention should be limited strictly to the use of these limits.

The invention is described further hereinafter, by way of example, with reference to the accompanying drawings, in which:~ Fig. 1 is a block circuit diagram of one embodiment of an oxygen toxicity measuring device in accordance with the present invention which includes a transducer which responds to the respiration of a subject under test: and Fig. 2 is a diagrammatic illustration of a typical output curve obtained from the transducer in the apparatus of Fig. 1.

The device of Fig. 1 includes a transducer arrangement 10 which is indicated highly diagrammatically in the drawing but which in practice comprises a band which is placed around the subject's chest and whose two ends are connected by way of a force transducer, such as a strain gauge transducer. The expansion and contraction of the chest wall during respiration, i.e. changes in circumference of the chest, then result in a roughly sinusoidal output signal as indicated in Fig. 2 of the drawings.

As explained above, previous approaches to the present problem have involved the manual measurement of rates of inspiration and expiration. The latter have been obtained by direct measurement of the slopes of the relevant parts of the sinusoidal curve at the mid-inspiratory and mid-expiratory points respectively. For the present purposes, the use of mid-respiratory and midexpiratory points has had to be abandoned for two principal reasons~ 1. When oxygen poisoning is advanced, it is difficult to decide where expiration ends and the next inspiration begins, so that it is hard to select mid-points.

2. While it is easy for a technician to draw lines on chart paper and calculate a ratio, this is too cumbersome an operation, too wasteful in manpower, too prone to error in unskilled hands and represents too long a delay if one if cautious and averages the last ten breaths.

A computer could be used to analyse the signal and produce the average ratio value, but this is much too costly and just not feasible in the practical situations where it would be needed.

Instead, the present device involves the use of a new ratio as follows:~ With reference to Fig. 2, the maximum excursion (FH~FL) in the output from the force transducer for one breath is determined by taking the difference between the highest (FH) and lowest (FL) values of the tranducer signal.

'Triggers' are then set to time the next breath between 40 and 90% of the tidal fluctuation during inspiration and, in the reverse sense, during expiration - i.e. between

and

The time intervals for both inspiration (At1) and expiration (Ate) are then measured between those limits, time being of course a parameter which can be measured with immense precision.

The ratio (Ati)/(Ate) otherwise depicted as r, is then automatically calculated and displayed.

Since this technique is effective using one breath to set the limits (FL and FH) for timing the next, there can be a variation due to random changes in (FH#FL), even during spontaneous breathing. Hence it is desirable to display a value which is the average of say the last ten breaths and to discard any value differing by more than say 10 to 15% from the mean. This has the further advantage of tending to eliminate any breath which was not spontaneous, i.e. when the subject sighed, coughed, sneezed, talked or was consciously controlling his respiration.

As indicated in Fig. 1, the apparatus for performing this technique includes a microprocessor 12 which is coupled to the output of the force transducer 10 by an analogue to digital converter 14 which converts the analogue signal produced by the transducer 10 into a digital signal capable of being handled by the microprocessor 12. The output of the microprocessor 12 drives a 4 digit display 16 via a driver amplifier 18, the microprocessor 12 and transducer 10 having a common power supply unit 20. The microprocessor also has an external clock timer 22.

The correct operation of the apparatus can be checked by the provision of a test signal supplied to the analogue to digital converter 14 from a signal generator 24 and by the monitoring of the transducer output on a suitable indicator 26 via an operational amplifier 26.

The microprocessor is programmed to perform the following operations on the signal provided by the transducer.

1. For a given positive gradient (inspiration) of one breathing cycle a predetermined number of points are measured on the curve, digitized and stored in a memory or store. The number of points selected depends on the accuracy required.

2. When the curve has completed its positive gradient between FL and FH (the latter being the turning points of the curve), the stored points are processed and trigger points calculated for:-

and

3. These trigger points are then used on the positive and negative gradients of the next breathing cycle to measure the time interval for the increase Ati between points (a) and (b) on the positive gradient of that cycle and to measure the time interval for the decrease Ate between points (b) and (a) on the negative gradient of that cycle.

4. The ratio

is then calculated for that cycle, and the result stored in the memory.

5. The values of FL and FH are meanwhile obtained for the second cycle and used to repeat steps 3 and 4 for the next cycle to obtain r2 in the memory.

6. The operation is repeated until the values (r1, r2 ... r1 o) are present in the memory.

7. The average FAv is calculated using the values from step 6.

8. The average value FAv is compared with each of the ten values (rl, r2 . . . r10) obtained from step 6 and rejecting any of the latter values which differ from the average FAv by more than, +15%. The latter limits are variable to suit particular situations.

9. The average ratio is then re-calculated using the remaining acceptable values.

10. The result of step 9 is displayed on the display 16, a typical value being 0.8 to 0.9.

11. Steps 1 to 10 are repeated for as long as the apparatus is in operation.

The result is that the physician or diver supervisor has an immediate indication of the state of the person's lungs without the need to draw blood or to use any invasive techniques.

However, although simple to operate and cheap to install, the value of the device is only as good as the r ratio is meaningful.

The chosen ratio has the following features:~ 1. As a ratio, many potential errors in monitoring r cancel, out, since they apply almost equally to inspiration and expiration. For instance, a cheap transducer can be used since any effect of temperature does not change the ratio significantly, although it can change Ati and Ate individually.

2. r is almost independent of respiration rate with a value of about 0.8 under normal conditions. Hence r is meaningful whatever the state of rest or exercise of the subject.

3. When the lung is poisoned with oxygen, the ratio increases with Ati getting larger and Ate, if anything, decreasing.

4. r shows a continuous change with oxygen exposure and can therefore be used as a quantitative index of the degree of poisoning as opposed to just an indication of inset. For instance, r has been found to increase from 0.8 to 1.8 in a rabbit exposed to 1.5 Atm of 02 for 8 hours, while, close to death, the value rises to above 3.0.

5. The wide range of r values cited above is far in excess of normal variation (+10%) and therefore represents quite a sensitive index.

6. The ratio has the considerable advantage that it can be determined on experimental animals and therefore evaluated much more rigorously than could be permitted on man.

7. The r value determined immediately before an animal is sacrificed correlates well with~ (a) weight of the excised lung, (b) its mechanical properties, e.g. compliance, (c) general pathological findings.

8. On man, the technique is non-invasive, involves no pain and could be used by anyone, without special skills.

9. In the newborn, the chest-band could be of very soft rubber to ensure no interference with the baby's normal breathing.

10. In the diving situation, the device might only be used towards the end of decompression or during recompression therapy when the diver has nothing else to do, so he would not wear it in the water and it would not hamper the job in hand. Alternatively, the device could be worn permanently with the electronics sealed into a water-tight environment. In a further possibility for use in this situation, the signals from the transducer could be transmitted by radio link to a microprocessor on shore or on a support boat.

It should be noted that the particular limits of 0.4 and 0.9 chosen in the trigger points

and

are capable of variation, provided that the resulting timing periods At, and Ate correspond to approximately 0.5 (FH~FL)

Claims (11)

Claims

1. A device for continuously monitoring a subject for pulmonary oxygen toxicity comprising a transducer for fitting to the subject to provide an electrical signal representative of the cyclic expansion and contraction of the subject's chest during respiration, means for operating on said signal to establish the time period within which said signal changes between predetermined trigger values on both an inspirational and an expirational part of the signal, and means for establishing the ratio of the latter time periods to provide an indication of said oxygen toxicity.

2. A device as claimed in claim 1, in which said inspirational and expirational parts of the signal are disposed on the adjacent rising and falling portions of the signal corresponding to a given single breath, the trigger values having been established from the maximum and minimum levels reached by a preceding breath.

3. A device as claimed in claim 2, in which said preceding breath is the immediately preceding breath.

4. A device as claimed in claim 1, 2 or 3, in which the operation is repeated on a plurality of successive breaths and the average value of the ratio is calculated.

5. A device as claimed in claim 4, wherein any individual ratios which differ from the average by more than a predetermined amount are rejected and a new average calculated from the remaining values.

6. A device as claimed in any of claims 1 to 5, wherein the timing periods (At and Ate) on the inspirational and expirational parts, respectively, are each chosen to correspond to approximately 0.5 (FH~FL), wherein FH and FL are the maximum and minimum levels reached by a preceding breath.

7. A device as claimed in claim 6, wherein the trigger values are set at levels corresponding respectively to approximately 40% and 90% of the maximum excursion of the transducer on said previous breath.

8. A device as claimed in any of claims 1 to 7, wherein the transducer is a force transducer whose ends are coupled by a band adapted to be worn around a subject's chest whereby the expansion and contraction of the chest wall during respiration results in a roughly sinusondal output signal.

9. A device as claimed in any of claims 1 to 8, wherein said means for operating on the transducer signal and deriving said ratio comprises a microprocessor.

10. A device as claimed in claim 9, wherein the microprocessor is programmed to perform the following operations on the signal provided by the transducer:~ (i) for a given positive gradient (inspiration) of one breathing cycle a predetermined number of points are measured on the signal curve, digitized and stored in a memory or store, the number of points selected depending on the accuracy required;; (ii) When the curve has completed its positive gradient between FL and FH (the latter being the turning points of the curve), the stored points are processed and trigger points calculated force

and

(iii) These trigger points are then used on the positive and negative gradients of the next breathing cycle to measure the time interval for the increase At between points (a) and (b) on the positive gradient of that cycle and to measure the time interval for the decrease Ate between points (b) and (a) on the negative gradient of that cycle; (iv) The ratio

is then calculated for that cycle, and the result stored in the memory; (v) The values of FL and FH are meanwhile obtained for the second cycle and used to repeat steps 3 and 4 for the next cycle to obtain r2 in the memory;; (vi) This operation is repeated until n values ......... Fn) are present in the memory; (vii) The average FAv is calculated using the values from step 6; (viii) The average value FAv is compared with each of the n values (r1, r2 . . . Fn) obtained from step 6 and any of the latter values which differ from the average FAv by more than +15% are rejected, the latter limits being variable to suit particular situations; (ix) The average ratio is then re-calculated using the remaining acceptable values; (x) The result of step 9 is displayed on a display; and (xi) Steps 1 to 10 are repeated for as long as' the apparatus is in operation.

11. A device for continuously monitoring a subject for pulmonary toxicity, substantially as hereinbefore described with reference to the accompanying drawings.