The effect of changes of lung characteristics on the functioning of automatic ventilators*

Transcription

1 VOL 17 NO 3 ANBSTHESA JULY 1962 The effect of changes of lung characteristics on the functioning of automatic ventilators* To study the effect of changes of lung characteristics on the functioning of automatic ventilators it is first necessary to understand the functioning of the various types of ventilator when connected to a normal lung. A set of concepts, with which it is possible to give a systematic account of the functioning of a ventilator, has already been expoundedl; but certain of these concepts are not very well understood. Therefore some of them will be considered here in rather more detail. GENERAL An automatic ventilator has to do four things: 1 nflate the patient s lungs 2 Change over from inspiration to expiration 3 Deflate the lungs - or permit passive expiration 4 Change over from expiration to inspiration t seems that the change-over, or cycling, processes are better understood than the processes of inflation or deflation. For instance, it is well-known that the change-over from inspiration to expiration can be either time-cycled, pressure-cycled or volume-cycled ; that is to say the process of inflation is stopped either after some preset time has elapsed, or as soon as some preset pressure is reached at the mouth (or at some other point), or when some preset volume has been delivered to the patient. n the same way the change-over from expiration to inspiration can be time-, pressure- or volume-cycled; in addition there is here the possibility of patient-cycling. What seems to be much less well understood is what goes on during the inspiratory phase. n some *Paper read at the Scientific Meeting of the Faculty of Anaesthetists of The Royal College of Surgeons of England, 12th May

2 ANBSTHESA 301 ventilators there may be equal confusion about what goes on during the expiratory phase, but in so many of them this is just a matter of allowing the patient s lungs to deflate passively to atmospheric pressure that it need not be considered further here. Therefore attention can be concentrated upon the inspiratory phase. During the inspiratory phase there are four factors to be considered : 1 Flowrate of gas into the lungs 2 Volume of gas in the lungs 3 Pressure in the alveoli 4 Pressure at the mouth For a complete understanding, the way in which these four factors vary from moment to moment throughout the inspiratory phase must be determined. These four factors are all inter-related in a rigorous cause-and-effect manner and the inter-relationships are determined entirely by the characteristics of the patient s lungs. Therefore only one of these four factors is free to be independently determined by the ventilator. n fact it is always either the flow into the patient s lungs or the pressure at the patient s mouth which is determined by the ventilator. Once one or other of these two factors has been determined by the ventilator the other three follow from the effect of the one factor on the characteristics of the patient s lungs. f it is the flowrate into the lungs which is determined by the ventilator then the ventilator may be called a flow generator; if it is the pressure applied to the mouth which is determined by the ventilator then it may be called a pressure generator. t seems to be possible to allot all ventilators to these two groups. FLOW GENERATORS: MECHANSMS The discussion will be restricted to the case of a constant-flow generator. A ventilator which operates as a constant-flow generator produces a flowrate into the patient s lungs which is constant throughout the inspiratory phase and furthermore, which is entirely determined by the ventilator and is uninfluenced by the patient s lung characteristics. There may be a knob on the ventilator to adjust the flowrate but once this is set, no matter what the lung characteristics, this constant flow will be driven into the patient for as long as the inspiratory phase lasts, that is until it is terminated by the cycling mechanism - unless some pressure-limit comes into operation earlier due, perhaps, to the opening of a safety valve. Once the ventilator has determined the flowrate into the lungs the other three factors listed above follow according to the lung characteristics.

3 302 ANESTHESA A number of mechanisms operate as constant-flow generators. n one the patient s lungs are connected more or less directly to an air compressor driven by a powerful constant-speed motor. (This system is found in the Aintree, Blease and Fazakerley* ventilators). This compressor goes on pumping out the same steady flow of gas throughout the inspiratory phase no matter how high the back pressure from the patient may rise. Another arrangement which operates as a constant-flow generator is shown in FG. 1. The reducing valve, on the top of the compressed- L 5-60 p.s.i cm RE S STANCE J. TO PATENT m S G R d cm H20 FG. 1 A constant-flow generator gas cylinder, delivers a steady pressure somewhere between 5 and 60 lb/in2 (between 300 and 4000cm H20). This is connected through a high resistance to the patient. The pressure in the patient s lungs rises during inflation from 0 to perhaps 30cm H2O - but this is a very small change compared to the 300 to 4000cm H20 at the reducing valve. Therefore the pressure difference across the high resistance is substantially constant throughout the phase and therefore the flow is also constant. t may be adjusted to any value up to say 100 litres/min by altering the magnitude of the high resistance, but once it is set the flow stays constant throughout every inspiratory phase. This system is found in the Pneumotron and Williams ventilators and also in some modes of operation of the Bird. *Descriptions of all the ventilators mentioned in this paper (apart from the Cyclator and Takaoka) will be found in Automatic Ventilation of the Lungs*

4 ANAESTHESA 303 Another mechanism which comes very near to being a constant-flow generator is a highly efficient injector such as is found in the Cyclator and also, though perhaps to a lesser extent, in the Aga and Drager ventilators. PRESSURE GENERATORS : MECHANSMS The discussion will be restricted to the case of a constant-pressure generator. Here it is the pressure at the mouth which is determined (and held constant) by the ventilator while the flowrate of gas and the volume and pressure in the alveoli result from the effect of this pressure on the patient s lung characteristics. (For other types of pressure generator and for an account of what happens when the pressure is held constant at some point separated from the mouth by a substantial resistance see Mushin et ap). One form of constant-pressure generator is shown in FG. 2. The weight provides a constant force acting on the constant crosssectional area of the bellows. Force per unit area is pressure so the pressure within the bellows is constant no matter whether the outlet is occluded and no flow occurs or whether the outlet is wide open and the air comes out in a rush. n use the bellows are connected to the patient s lungs FG. 2 A constant-pressure generator so that the lungs are inflated until the pressure $thin them equals the pressure in the bellows - unless the inspiratory phase is stopped beforehand. This system is found in the RPR and Radcliffe ventilators. Another form of constant-pressure generator is the second-stage reducing valve found in the Bennett ventilator. THE EFFECT OF THE VENTLATOR OUTPUT ON THE LUNGS Constant -30 w generators FG. 3 shows in graphical form the constant flow produced by a constant-flow generator and the resultant variation in the volume and pressures in the lungs. Numerical values have been indicated for the sake of illustration.

5 304 ANESTHESA At the start of the inspiratory phase the flow into the lungs rises suddenly to some value (30 litres/min in FG. 3) and stays there throughout the phase. Therefore (assuming there areno leaks in the system) the volume in the lungs increases steadily - at 30 litres/min=+ litre/sec and therefore reaches 500ml at the end of 1 sec. The compliance of the lungs may be assumed constant throughout the phase and therefore the steady increase in volume is accompanied by a steady increase in the pressure in the alveoli (PALV). f the total compliance of a normal anasthetised patient is taken to be 50ml/cm H20 then the steady increase in volume to 500ml entails a steady increase in alveolar pressure to locm H2O. The airway resistance is constant throughout the inspiratory phase and therefore the constant flow produces a steady pressure difference (PKPALv) across it. f the total airway resistance (including connectors) for a normal intubated patient is taken to be 6cm HzO per litre/sec the constant flow of 3 litre/sec will produce a constant pressure difference across it, i.e. between the mouth and the alveoli (PM- PaLv), of 3cm H2O. The pressure at the mouth is simply the sum of the pressure in the alveoli and the pressure difference between the mouth and the alveoli: it rises suddenly to 3cm H20 due to the pressure difference across the airway resistance and then rises steadily through a further locm H20 in 1 sec due to the steady rise in pressure in the alveoli. FLOW 30 ( /m in) 0 TME (sec) 0' 1 FG. 3 Flow, volume and pressure curves in the lungs for a constant-flow generator (30l/min) connected to a patient with a compliance of 50 ml/cm H2O and an airway resistance of 6 cm H2O per 1 /sec. PAL" =pressure in the alveoli, P,=pressure at the mouth. Constant-pressure generators A constant-pressure generator applies some pressure to the mouth at the start of the inspiratory phase and holds it constant throughout the phase. Suppose the generated pressure is 12cm H20 : PMoUTH= 12cm HzO

6 ANESTHESA 305 Suppose also that the total (intubated) airway resistance is as before: R = 6cm H20 per litre/sec At the start of inspiration, there is no pressure in the alveoli and therefore the whole of the 12cm H20 pressure at the mouth is dropped across the airway resistance. Therefore since the airway resistance is 6cm H2O for 1 litre/sec a pressure drop of 12cm H20 will produce a flow of 2 litres/sec: nitial flow into lungs = 2 litres/sec Therefore nitial volume increase in lungs = 200ml in 1 / 10 sec f the compliance is 50ml/cm H20, as before, a volume increase of 200ml will involve a pressure increase of 4cm H20. Therefore nitial pressure increase in alveoli = 4cm H20 in 1/10 sec. But as the pressure in the alveoli increases the pressure difference between the mouth and the alveoli decreases (because the mouth pressure is held constant by the ventilator). t is this pressure difference which provides the force driving gas into the lungs; therefore as it decreases, so the flow into the lungs decreases and the rate at which the volume in the lungs increases becomes less and the rate at which the pressure in the alveoli increases becomes less : however the alveolar pressure does go on increasing, but the closer it gets to the mouth pressure the slower it increases (FG. 4). From a mathematical point of view equilibrium will occur only after an infinite time; but from a clinical point of view there would be no advantage (and possibly some disadvantage) in prolonging the inspiratory phase to get the last few per cent of pressure or volume into the lungs. t is therefore important to know how long it takes to achieve 'near equilibrium' - say 90 to 99 % equilibrium. There are only three independent factors within the system (of ventilator plus patient) on which the time to reach a given degree of equilibrium could possibly depend : 1 Compliance 2 Resistance 3 Generated pressure VOLUME / rnl FG. 4 Pressure, flow and volume curves in the lungs for a constant-pressure generator (1 2 cm H2O) connected to a patient with a compliance of 50 ml/cm H20 and an airway resistance of 6 cm H20 per l/sec.

7 306 ANRSTHESA All the other variables within the system are derived from the effect of the generated pressure on the compliance and resistance of the patient's lungs. n fact the generated pressure has no influence on the time taken to reach a given degree of equilibrium so that only the compliance and resistance can be involved*. How can these parameters affect the time for equilibration? Consider the following calculation in which compliance and resistance are given the same values as before. C = 0.05 litre/cm H20, litre - - 0'05cm H20 ' R = 6cm H20 per L/sec, 6 cm H20 sec - litre litre cm H20 sec Therefore C x R = 0.05 x 6 X 9 cm H20 litre = 0.3 sec. The product of the compliance and airway resistance of a normal, ansesthetised, intubated patient is 0.3 sec. t seems odd that a lung should have a time, but the significance of this time can be seen from the following calculation. Assume that, as in the previous calculation for a constant-pressure generator: P,,,, = 12cm HzO nitial pressure increase in lungs = 4cm H20 in 1/10 sec. Now C x R = 0.3 = 3/10 sec. Therefore if initial rate of pressure increase in lungs were maintained for 3/10 sec the alveolar pressure would reach 3 x 4 = 12cm H20. Therefore this time of 3/10 sec, the time which is equal to the product of the compliance and resistance of the patient's lungs, is the time in which the inflation of the lungs wouzd be complete if the initial rate of inflation were maintained. n fact, as has already been shown, the initial rate of inflation is not maintained and therefore inflation is not complete at the end of this time. *n a more accurate analysis it would be necessary to take account of the way in which compliance wries with the volume in the lungs and resistance varies with flowrate so that the size of the generated pressure would have some indirect influence on the time taken to achieve a given degree of equilibrium. For the present discussion however the compliance and resistance may be regarded as independent of volume and flow, and hence of generated pressure, which then indeed has no influence on the time to achieve any given degree of equilibration.

8 ANESTHESA 307 This process of gradual approach towards equilibrium in which the rate of approach diminishes in proportion to the nearness to which the limit is approached is not peculiar to the process of inflating lungs but occurs in such diverse fields as the charging of an electrical condenser through a resistor or the decay of a radioactive substance. n these fields of knowledge the characteristic time derived above (the time in which the change would be complete if the initial rate of change were maintained) is given the name time constant and is usually represented by the Greek letter T (tau). Thus at the end of a period of time T (equal to the product of the compliance and resistance of the lungs) the lungs would be fully inflated if the initial rate of inflation were maintained. n fact the rate of inflation falls off as the alveolar pressure rises (FG. 5) and at the end PRESSURE MOUTH FG. 5 ncrease of alveolar pressure with time (in units of time constant, T) for a r - constant-pressure generator. The very thin broken line shows that the alveolar / pressure would reach the mouth pressure after one time constant if the initial rate of increase were maintained. CT T+ of one time constant the change is only about 2/3 complete - or 1/3 incomplete. n the period of the second time constant the same thing happens again (in a sense) : the pressure difference between mouth and alveoli is reduced to approximately 1/3 of what it was at the beginning of the period - and therefore to 1/3 of 1/3 = 1/9 of what it was at the start of the inspiratory phase. At the end of three time constants the remaining pressure difference is about 1/3 of 1/3 of 1/3 = 1/27 of what it was at the start of the inspiratory phase. n fact, at the end of a period of one time constant, the fraction of change which is incomplete is not exactly 1/3 but where e is the base 1 of natural logarithms and is equal to = 37 %. Therefore at the end of one time constant the change is 37 % incomplete and 63 % 1 1 complete. At the end of two time constants it is- x -= 13 % incomplete and 87 % complete. At the end of three time constants it is 5 % incomplete and 95 % complete. These results are summarised and extended in TABLE 1. They are true no matter what the lung character-

9 308 ANESTHESA Table 1 Exponential ncrease of Alveolar Pressure in a Lung Connected to a Constant-pressure Generator. Time is given in units of T where T is the time constant of the lungs (the product of compliance and airway resistance). Time T 2i 3T 4T istics may be, in the sense that, for instance, inflation will always be 95 complete in a time which is three times as great as that obtained by multiplying the compliance and airway resistance together. For the normal, anasthetised, intubated patient, for whom the time constant is 0.3 sec, inflation will be 95% complete in 3 x 0.3 = 0.9 sec. Thus in these circumstances there will be little point in prolonging the inspiratory phase beyond 1 sec*. The foregoing has established the patterns of flow, volume and pressure produced in normal lungs by constant-flow generators and constant-pressure generators. t is now possible to go on to consider how these patterns are modified by changes in lung characteristics. THE EFFECT OF CHANGES OF LUNG CHARACTERSTCS Constant-pressure generator FGURE 6 shows the alveolar pressure and volume curves, plotted against time, for a constant-pressure generator connected to a normal lung (continuous curves in all three pairs of graphs) and to lungs in which the compliance has been halved, the airway resistance doubled, and in which both changes have occurred simultaneously (broken curves). The blobs on the curves mark off intervals of one time constant. When the compliance is halved (perhaps due to the surgeon leaning on the chest) the most obvious effect is that the final steady volume is halved. But halving the compliance also halves the product of compliance and resistance - the time constant. Therefore it takes only half as long to reach any given degree of equilibration between the alveolar pressure and the mouth pressure. This can be seen by studying the blobs on the curve in FG. 6: at the end of one time constant (the first blob) both the continuous and the broken curves have reached 63 % of the applied pressure but the time constant is only half as long for *f apparatus resistance is in series with the airway resistance leading to an increased time constant, or if a constant-flow generator is used, then prolonging the inspiratory phase beyond 1 sec will indeed lead to further inflation of the lungs.

10 PALV tr ""if- ANESTHESA 309 Rx2 FG. 6 The pressure and volume in the alveoli plotted against time for a constant-pressure generator. The continuous curves are for a normal lung and the broken curves apply when compliance and airway resistance deviate from normal in the manner indicated. The blobs on the curves mark off intervals equal to one time constant. the broken line as for the continuous line because of the halved compliance. Although the final steady volume in the lungs is halved because of the halved compliance, this lower volume is approached more quickly because of the halved time constant. f instead of the compliance being halved the resistance is doubled (perhaps due to an obstruction or the accumulation of secretions) the ultimate pressure or volume in the alveoli is unaltered but because the time constant is doubled (by the doubling of the resistance) it takes twice as long to reach any given degree of equilibration. f the compliance is halved and the resistance is doubled at the same time (one main bronchus occluded perhaps) then the time constant (the product of the two) is unaltered. Therefore the alveolar pressure rises in exactly the same way as before. The final steady volume in the

11 3 10 ANESTHESA lungs is halved (by the halved compliance) but this lower volume is approached at the same rate as that at which the full volume is approached in the normal lung - because the time constant is the same. The foregoing explains the changes in the volume and pressure patterns produced by a constant-pressure generator when the lung characteristics change in certain ways; but the clinician is more interested in the resultant changes in tidal volume and ventilation. To determine these it is necessary to take account not only of what sort of generator the ventilator is working as, but also what sort of cycling mechanism is in operation. For instance if the constant-pressure generator is time-cycled (e.g. the Radcliffe) and the patient s compliance is halved, either with or without a doubling of airway resistance, (the top two sets of curves in FG. 6) then the tidal volume will certainly be reduced, probably down to half what it was before. f the change-over from expiration to inspiration is similarly time-cycled then the duration of the respiratory cycle, and hence the respiratory rate, will be fixed. Therefore the total ventilation will be reduced in the same proportion as the tidal volume. f the change in lung characteristics consists only of an increase of airway resistance (lower set of curves in FG. 6) then the effect on tidal volume and ventilation depends largely on the length of the inspiratory phase set by the time-cycling mechanism. f this is long (beyond the end of the lower graph in FG. 6) then the volume in the lungs will have near enough reached its equilibrium value, even with the doubled time constant consequent upon the doubled airway resistance. Therefore the change in airway resistance has little or no effect on the tidal volume. However, if the time-cycling mechanism is set to give rather a short inspiratory phase - say one-third of the way along the time axis of the lower set of graphs in FG. 6 (that is two normal time constants or one double time constant) - then doubling the airway resistance results in a substantial reduction in tidal volume. f instead of being time-cycled the constant-pressure generator is volume-cycled (as in the French RP R ventilator) then it is clear from FG. 6 that any decrease in compliance or increase in airway resistance will result in the ventilator taking a longer time to deliver the preset tidal volume. Therefore, unless there is a compensating decrease in expiratory time the rate and hence the total ventilation will be decreased. ndeed, if the cycling mechanism is set to operate at a volume which is greater than half the final equilibrium volume for a normal lung, and if the compliance is then halved (with or without an increase in airway resistance) the cycling volume will never be achieved and the rate, and therefore the total ventilation, will drop to zero.

12 FLOW - -- ANiESTHESA 311 palv! / E 0..<.. Rx2 0 p 1 / / / k /'..!..... Cx+ Rx2 FG. 7 Flow, volume and pressure curves in the lungs for a constant-flow generator when connected to a normal lung (continuous lines) or when connected to lungs with the indicated changes from normality (broken lines) of compliance and airway resistance. Constant-Jow generator FGURE 7 shows the flow, volume and pressure patterns produced by a constant-flow generator when connected to a normal lung (the continuous lines in all three sets of graphs - which are the same as shown in FG. 3) and also when connected to lungs in which the compliance has been halved, or the airway resistance doubled, or both changes have occurred simultaneously (the broken lines in the three successive sets of graphs). Changes of lung characteristics can have no effect on the fl ow from the ventilator, nor on the pattern of volume increase in the lungs

13 312 ANRSTHESA since, assuming there are no leaks, the volume is simply the integral of the flow into the lungs. However, when the compliance is halved (broken lines in the left-hand set of graphs in FG. 7) the resultant pressure rise in the alveoli is doubled. Therefore the pressure at the mouth, after a normal initial step (due to the pressure difference across the normal airway resistance), increases at twice the normal rate due to the halved compliance. When the airway resistance is doubled (centre set of graphs) there is no change in the alveolar pressure curve but the pressure difference between the mouth and alveoli (due to the constant flow through the doubled airway resistance) is doubled. Therefore the pressure at the mouth has a doubled initial step but then rises at a normal rate. When the compliance is halved and the airway resistance doubled at the same time (broken lines in the right-hand set of graphs in FG. 7) the two effects occur simultaneously. Therefore the pressure at the mouth shows both a doubled initial step and a doubled subsequent rate of rise. As with the constant-pressure generator it is necessary to take account of the cycling mechanisms to determine the effect of these changes in waveform, consequent upon changes in lung characteristics, on the tidal volume and ventilation. f a constant-flow generator is time-cycled (e.g. the Pneumotron) then the constant flow for a preset time will deliver a fixed tidal volume despite any changes in lung characteristics. Assuming there is no change in the expiratory time, the rate will stay the same and so the total ventilation wil be unaltered. The same will also be true of a constant-flow generator which is volume-cycled (e.g. the Greer and Donald) since the constant flow will take a fixed time to deliver the preset tidal volume and, providing the expiratory time is fixed, the rate and total ventilation will again be independent of changes in lung characteristics. f a constant-flow generator is pressure-cycled the situation is very different. Suppose that when connected to a normal lung the cycling pressure is set so that the inspiratory phase lasts just as long as the graphs in FG. 7 so that the tidal volume delivered is that indicated by the ends of the volume curves. n that case the cycling pressure must be that indicated by the dotted lines in FG. 7. Therefore when the compliance is reduced or the airway resistance increased or when both occur simultaneously the cycling pressure is reached much earlier and therefore the tidal volume is decreased. To take a specific instance : if the compliance is halved the inspiratory time is halved and the tidal volume is halved. f the expiratory time is fixed (as in the Aintree, Blease, Cyclator and Fazakerley) then the complete respiratory cycle will be only slightly shortened and the halving of the tidal volume will

14 ANBSTHESA 313 result in a large reduction of total ventilation. However, if when the compliance is halved, it is found that the halved tidal volume is not only inflated in half the time, but also expired in half the time (as it will be if the ventilator is a pressure-cycled constant-flow generator in expiration as well as in inspiration e.g. approximately, the Aga, Drager Poliomat and Takaoka ventilators) then, although the tidal volume is halved the rate is doubled and the total ventilation is unchanged. SUMMARY AND CONCLUSONS This paper does not give a comprehensive account of the functioning of all types of ventilator in all circumstances but illustrates a method of analysis by describing the most important aspects of the problem in detail. Thus only the inspiratory phase has been considered and even within that phase the many types of non-constant-flow and nonconstant-pressure generators have been omitted. Likewise no mention has been made of the effects of pressure and volume limits or of various other devices which modify the functioning of some ventilators. A more comprehensive, but much more concise, account will be found in Automatic Ventilation of the Lungsl. The present treatment is intended to give a fuller and more leisurely account of some of the fundamental aspects of the problem in the hope that it will lead to a greater understanding which can then be applied to other circumstances not covered here. CONSTANT-FLOW However, it is convenient to sum- GENERATOR marize the results that have been TME CYC. o o 0 TDAL NSF! TOTAL VOL. TME VENT, + + +or0 derived in this account and this is O O O done in FG. 8. An important con- CONST~-PRESSURE clusion which can be drawn from GENERATOR this Figure is that it is dangerous to. ; ; ;,. ; : 0 4 try to relate the responses of a t +or0 ventilator to changes in lung char- FG. 8 The effect of decreased compliacteristics to a single aspect of the ance or increased resistance on the performance of various types of ventilator s functioning. For inventilator (assuming no leaks from stance, a volume-cycled ventilator the system). is traditionally regarded as immune from changes in lung characteristics; it is true that the tidal volume is held constant by this means, but if the ventilator operates as a pressure-generator then the inspiratory time is increased and, unless there is a compensating decrease in expiratory time, the rate and therefore the total ventilation, are decreased. Similarly a pressure-cycled ventilator is traditionally regarded as particularly susceptible to

15 314 ANESTHESA changes in lung characteristics; but if it operates as a flow generator in both phases (and is pressure-cycled in both phases) any change in tidal volume resulting from a change in lung characteristics is precisely compensated for by a change in rate so that the total ventilation is held constant. t may be concluded that the only way to determine how a given type of ventilator will perform in the face of changes in lung characteristics is either to try it experimentally or else to work systematically through a complete functional analysis of its operation in the manner outlined above. Acknowledgements t is a pleasure to acknowledge the encouragement of Professor William W. Mushin firstly to develop this method of analysis of the functioning of automatic ventilators and secondly to develop what hope is a clear exposition of it. Reference 1MUSHN, W. W., RENDELL-BAKER, L., and THOMPSON, P. W. (1959). Automatic Ventilation of the Lungs (with chapters on the Physics of Automatic Ventilators by W. W. Mapleson). Oxford. Blackwell Scientific Publications.