Anaesthesia, 1990, Volume 45, pages 855-858 6PPARATUS
Mechanical ventilation during low-flow anaesthesia Experience with an alternative to the bag-in-bottle L. BERNTMAN, H. H. LUTTROPP
AND
0. WERNER
Summary
Clinical experience with low-flow anaesthesia during controlled ventilation of the lungs is described. The anaesthesia circle is separated by a corrugated hose that serves as a large deadspace. This open connexion has no bellows or overflow valve and therefore the risk of mechanical dysfunction is small. No mixing of circle and ventilator gas occurs during normal operation. Major decreases in the oxygen concentration in the system are unlikely even if the fresh gasflow is interrupted or significant leaks from the circle occur because 100% oxygen is delivered by the ventilator. A hose volume larger than 1650 ml prevented gas mixing at tidal volumes of 380-11 70 ml. There was no system-related mishap in over 600 patients, who comprised about 40% of the neurosurgical patients anaesthetised during that period. The cost of isoflurane was reduced to about 33% of that incurred during previous periods.
Key words Anaesthetic techniques; closed system. Ventilation:artificial.
Low-flow anaesthesia (LFA) has recently regained popularity. One reason for this is the reduced consumption of isoflurane, and another that LFA requires gastight systems, which reduce gas pollution in the operating room. Finally, LFA helps to reduce emissions of halogenated hydrocarbons and nitrous oxide into the atmosphere. Mechanical control of ventilation with LFA is usually achieved by a bag-in-bottle arrangement, i.e. the ventilator gas is separated from the patient circle by bellows inside an airtight container. The function of this device is simple, but critically dependent on proper function of the bellows and the circle's overflow valve. We used an old method'J as an alternative whereby the ventilator is connected to the anaesthesia circle via a large deadspace that prevents ventilator gas from entering the circle. This is a safe and simple method that combines the facilities of a powerful ventilator with the advantages of the circle system. This paper presents the principles of operation, bench experiments, trials in pigs, evaluation in clinical practice, and economy.
Materials and metbods Principle of operation
The ventilator (Servo 900 C, Siemens-Elema, Solna, Sweden), delivers oxygen into a large deadspace that consists of a corrugated polyethylene (Hytrel) hose with 2.2
litres internal volume. The other end of the hose is connected to a conventional anaesthesia circle with a COz absorber (Monosorb, Siemens-Elema, Solna, Sweden), one-way valves and fresh gas inlet (Fig. 1). The ventilator pushes a tidal volume of oxygen into the deadspace and the same volume of anaesthetic gas mixture is expelled from the other end into the circle. The gas column in the hose oscillates back during expiration: a tidal volume of exhaled gas is collected in the circle end of the deadspace, while one tidal volume of oxygen, mixed with some excess gas from the circle leaves via the ventilator. The pressure generated by the ventilator is transmitted unchanged to the circle since the hose is wide (3.3 cm internal diameter). In essence, the setting on the ventilator (minute volume, frequency etc.) will determine the pattern of ventilation, while the gas composition in the circle is determined by the vaporizer and flowmeter settings connected to the circle's fresh gas inlet. The different ventilatory modes and other functions of the ventilator can be used without restriction. Ventilation volumes, frequency etc. are monitored on the display of the ventilator. However, there is some discrepancy between the inspiratory and expiratory volumes as measured in the ventilator and volumes inhaled and exhaled by the patient. Thus, the fresh gas flow that enters the circle during the inspiratory phase of the ventilator, i.e. when the expiratory valve is closed, will increase the inhaled volume, while fresh gas entering the circle during expiration will increase the flow through the expiratory
L. Berntman, MD, PhD, Consultant, H.H. Luttropp, MD, Registrar, 0. Werner, MD, PhD, Consultant, Department of Anaesthesia, University Hospital, S-221 85 Lund, Sweden. Accepted 22 February 1990. 0003-2409/90/100855 + 04 $03.00/0
@ 1990 The Association of Anaesthetists of Gt Britain and Ireland
855
L. Berntman, H.H. Luttropp and 0 . Werner
856
1,.
Pop-off volve
Fresh gas A
thing system supplied 50:50 N 2 0 / 0 2with isoflurane directly to the pig. Isoflurane concentration and duration of anaesthesia were the same as above. Isoflurane consumption was assessed by weighing the vaporizer on a precision balance (Galaxy 1200-S0, Ohaus, USA). The experiments were done in random order and separated by a one-hour washout phase.
absorber
Ventilator
Fig. 1. The low-flow system. The pop-off valve is functionally closed (35 cm H,O) during mechanical ventilation.
flowmeter of the ventilator, although no corresponding volume is exhaled. With the standard setting of the respiratory cycle, i.e. inspiration time 25% and pause time lo%, 35% of the fresh gas flow (FGF) will be added to the inspired minute ventilation and the rest will bypass the patient but be monitored by the ventilator. Bench tests Experiments were undertaken to evaluate the deadspace volume necessary to prevent mixing of ventilator gas with that of the circle. The fresh gas flow was set to one litre/ minute of 5050 N,O/O,. The circle was connected to a lung model (a 4-litre anaesthetic bag). Oxygen concentration in the circle was measured with an oxygen analyser (Hudson Ventronics 5590, Temecula, California, USA) placed in the expiratory limbs. The ventilator delivered 10 litres/minute of 0,. Tidal volume ( VT) was varied from 380 to 1170 ml by changing the ventilator frequency. The hose volume (VD) was stepwise increased from 550 to 2200 ml and the reading on the oxygen analyser was noted. The compressible volumes were measured at different VD by obstructing the Y-piece, then noting the expiratory at a plateau pressure of 2.0 kPa. volume (h)
Tests in pigs The tests compared the low-flow system’s isoflurane requirements with those of a non-rebreathing system. Four pigs (1 8.5-25 kg) were studied. Anaesthesia was induced with azaperon (Stresnil) and etomidate chloride (Hypnodil) and the tracheas of the pigs were intubated. The minute ventilation (4-6 litres/minute at a respiratory rate of 10 minute) was adjusted to give an end-tidal carbon dioxide concentration of approximatively 4.5%, as measured at the tracheal tube on a Normocap analyser (Datex, Finland). Gas sampled by the analyser was returned to the circle. Fresh gas flow during LFA was one litre/minute of 5050 N 2 0 / 0 2 with isoflurane. End-tidal isoflurane concentration (Servo Gas Monitor, Siemens, Sweden) was kept as 1.5%. Isoflurane anaesthesia was maintained for 2 hours. The ventilator in the non-rebrea-
Table 1.
VD (ml) 550 1100 1650 2200
Vc (ml/kPa) 0.22 0.26 0.29
0.33
Clinical use of the system The system was used in routine anaesthesia in over 600 adult patients scheduled for neurosurgical procedures that lasted up to 14 hours. Patients were unselected except that those with known severe lung dysfunction were excluded. Anaesthesia was induced by thiopentone and maintained by N20, fentanyl and 0.5-1% isoflurane. The circle was connected to the ventilator via the corrugated hose after tracheal intubation. The pop-off valve in the circle was set at 3.5 kPa; this was closed functionally since airway pressures normally do not reach this level. The fresh gas flow was 1.5 litres/minute of oxygen and 3 litres/minute of nitrous oxide with isoflurane as needed during the initial 2&30 minutes that were spent in the induction room. The higher flow was used for denitrogenation and to allow for the patients’ initial high uptake of nitrous oxide. The flowmeters were set to 0.5 litres/minute each of 0, and N,O for the duration of anaesthesia after transfer into the operating theatre. Respiratory monitoring consisted of expired minute volume and airway pressures as displayed by the ventilator, and O2and CO, concentrations sampled from the tracheal tube into a Normocap (Datex)’ which withdrew 150 ml minute from the circle. The sampled gas was diverted to the scavenging system, after passing the analyser, instead of being delivered back to the circle. Isoflurane concentration was measured (Servo Gas Monitor, Simens, Sweden) at the tracheal tube in most patients and the sampled gas returned to the circle. Anaesthesia gases were discontinued after surgery, and oxygen flow via the fresh gas inlet was increased to 6 litres/minute. The circle was separated from the corrugated hose when the patient was ready for spontaneous breathing and instead reconnected to an anaesthesia bag.
Results Bench-tests
Figure 2 shows the oxygen concentration in the circle at different tidal ( VT) and hose ( VD) volumes. Oxygen concentration was unaffected by changes in VT when VD was 1650 or 2200 ml. When VD was reduced to 1100 ml significant mixing occurred at tidal volumes of or above 765 ml. The results at a VD of 550 ml suggested significant mixing with oxygen from the ventilator at all tidal volumes tested. The calculated compressible volumes of the system, at different values of VD are given in Table 1. Tests in pigs
Isoflurane consumption during 2 hours of anaesthesia was 0.4 (SD 0.1) (g/kg)/hour when a low-flow technique was
used. In the non-rebreathing system the isoflurane consumption was 1.9 (SD 0.1) (g/kg)/hour.
Mechanical ventilation during low--owanaesthesia
5
0
857
Ioo1 90
1100 ml
1650 ml 2200 rnl
50
OQ
4bO
6AO
8kO
10’00 1260
Tidal volume (ml)
Fig. 2. Oxygen concentration (mean of duplicate experiments) in the circle at different tidal and hose volumes ( VD). Fresh gas flow to the circle was one litre/minute of 50% oxygen in nitrous oxide. The ventilator minute volume was 10 litres. Mixing of ventilator oxygen with circle gas is indicated by an increase in the measured oxygen concentration.
Clinical experience
The low-flow system has been used since August 1987 without any system-related incidents. From September 1987 low-flow anaesthesia was used in 60% of elective, daytime neuroanaesthesia cases. Figure 3 shows the cost for isoflurane per each 6 month-period from 1985 to 1989. Approximately 62000 SEK (about E5600) was spent on isoflurane during the first 6 months of 1987 while the corresponding sum for 1988 was 26000 SEK (approximately &2300), in spite of a 15% increase in total anaesthesia time at the unit.
Discussion Closed-system and low-flow anaesthesia have been used for many years, but many anaesthetists prefer non-rebreathing systems because of their greater simplicity. The development of in-circle vapour and O,/CO, analysers with alarms have increased the patient safety during LFA and permit anaesthetists to attend to other duties during the procedure. An advantage of the present system is that moving parts such as bellows and a special overflow valve are not needed. This minimises the risk of mechanical dysfunction. Cleaning the system is simple. A possible disadvantage is that the absence of bellows that move may make the system intuitively less easy to understand than a bag-in-bottle ventilator. During steady-state anaesthesia expired minute ventilation ( VE) at the ventilator will be: the inspired minute ventilation ($I) set at the ventilator + fresh gas flow+xygen consumption-gas consumed by the gas analysers. If the patient consumes 250 ml 0, per minute, the analyser samples 150 ml/minute and fresh gas flow is one litre/ minute the measured h will be 0.6 litres/minute larger than the set $I. The exact oxygen concentration in the circle will be a function of the composition and flow of fresh gas, the oxygen consumption and anaesthetic gas ~ p t a k eand ~ . ~the amount of gas removed via gas analysers. We usually obtain an oxygen concentration of 4 1 4 3 % in the circle at steady state when the flowmeters are set at 0.5
Fig. 3. Cost of isoflurane for the first ( I ) and second (2) &month period, 198688; the total expense is given for 1985.
litres/minute of N 2 0 and 0.5 litres/minute of 0,. Some modifications were introduced after the initial evaluation of the system, for example the gas from the CO, monitor is now returned to the circle and the oxygen concentration in the circle is adjusted to about 35% by reducing the oxygen supply. A hose volume ( VD) of 1650 ml prevented entry of ventilator oxygen into the circle, using a clinically relevant range of adult tidal volumes. We elected to use 2200 ml in order to increase the margins. A low-flow system should be gas tight for proper function, but leaks may nevertheless occur e.g. around the cuff of the tracheal tube. The result of a leak in the present system varies with the leak volume (VL). If VL is smaller than the excess gas in the circle (in our case approximately 0.6 litres/minute, see above), measured h will decrease slightly, but there will be no change in oxygen concentration in the circle. If VL is larger, 0, from the Ventilator will enter the circle and cause an increase in the oxygen concentration. A disconnexion anywhere in the system will immediately be recognised as a decrease in measured h. The pressure in the system is controlled by both the setting of the pop-off valve and that of the high-pressure limit on the ventilator. We usually set the high-pressure limit below that of the pop-off valve, and it could be argued that we could as well close the pop-off valve entirely when switching from manual to mechanical ventilation (Fig. 1). Interruption of fresh gas supply is unlikely to cause hypoxia, because under these conditions the oxygen and N,O uptake by the patient will cause a net flow of oxygen from the ventilator via the hose towards the anaesthesia circle. Oxygen concentration will soon increase above the previous level (Fig. 4) after an initial decrease. However, interruption of only the oxygen while the nitrous oxide flow remains will cause hypoxia if this accident is not detected by the 0, analyser. The system has the vaporizer outside the circle. With this arrangement the inhaled vapour concentration may be considerably lower than that set on the vaporizer. The difference is dependent on F G F , ventilation and the patient’s uptake of vapour. The concentration can and should be monitored by an analyser, sampling gas from the tracheal tube, to minimise the risk of intra-operative awareness. This is particularly important when an inhalant-only
L. Berntman, H.H. Luttropp and 0. Werner
858
II20 4Ol
c
2
c C 0) 0 C 0 0
0
1
2
I
I
1
1
I
3
4
5
6
7
Hours
4
Minutes
Fig. 4. Changes in O2concentration in the circle after turning off (at arrow) the fresh gas flow. Four trials in the same patient.
anaesthetic is given. Our patients all had a N,O-fentanyl background and in most patients the isoflurane concentration was monitored. The low-flow system was used in 60% of elective daytime neurosurgical operations, which comprise about 40% of the total case load. The cost of isoflurane in the neuroanaesthesia service was reduced to 33% of that when only non-rebreathing systems were used, in spite of a 15% increase in anaesthesia time. The relatively larger reduction in cost than in the number of anesthetics with open systems is because of a preferential use of LFA with isoflurane for long operations. A calculation of costs for maintenance anaesthesia (Fig. 5) (1% isoflurane, 0, and N,O) with open and closed systems is shown in Figure 5. LFA not only affects the cost of isoflurane, but also reduces the consumption of N,O (cost 0.7 p/litre) and increases that of 0, (cost 0.1 p/litre) which is used to drive the ventilator. In addition LFA requires a CO, absorber for E9. The two systems break even after 30 minutes, after which time the cost is about 15% of that of a non-rebreathing system. The calculations do not include costs for induction or reversal drugs and the comparison refers to maintenance cost after the break even
Fig. 5. Calculated cost of maintenance anaesthesia with 1% isoflurane in either the low-flow (0.5 litres OJO.5 litres N,O+ 12 litres 0, per minute, see text) or a non-rebreathing system (5 litres 0,/7 litres N,O per minute).
point. The isoflurane consumption (and cost) in our animal experiments in LFA was 22% of that using the open system. A retrospective study in humans6 compared total drug costs for relatively short procedures (about 2.5 hours) and showed that the cost for LFA was almost 50% of that with an open system. Gas costs become more and more important during longer procedures.
References 1. VOSS JV. The adaption of ventilators for anaesthesia, with particular reference to paediatric anaesthesia. South African Medical Journal 1967; 41: 1079. 2. ADAMSAP, HENVILLE JD. A new generation of anaesthetic ventilators. Anaesthesia 1977; 32: 34-40. 3. BENGTSON JP, SONANDER H, STENQVIST 0. Oxygen analyzers in anaesthesia: performance in a simulated clinical environment. Acta Anaesthesiologica Scandinavica 1986; 30: 656-9. 4. SEVERINGHAUS JW. The rate of uptake of nitrous oxide in man. Journal of Clinical Investigation 1954; 33: 1183-9. 5. VIRTUE R, SHERRILL DL, SWANSON GD. Uptake of nitrous oxide by man. Canadian Anaesthetists' Society Journal 1982; 2 9 42k7. 6. BENGTSON JP, SONANDER H, STENQVIST 0. Comparison of costs of different anaesthetic techniques. Acta Anaesthesiologica Scandinavica 1988; 3 2 33-5.
Mechanical ventilation during low-flow anaesthesia Experience with an alternative to the bag-in-bottle L. BERNTMAN, H. H. LUTTROPP
AND
0. WERNER
Summary
Clinical experience with low-flow anaesthesia during controlled ventilation of the lungs is described. The anaesthesia circle is separated by a corrugated hose that serves as a large deadspace. This open connexion has no bellows or overflow valve and therefore the risk of mechanical dysfunction is small. No mixing of circle and ventilator gas occurs during normal operation. Major decreases in the oxygen concentration in the system are unlikely even if the fresh gasflow is interrupted or significant leaks from the circle occur because 100% oxygen is delivered by the ventilator. A hose volume larger than 1650 ml prevented gas mixing at tidal volumes of 380-11 70 ml. There was no system-related mishap in over 600 patients, who comprised about 40% of the neurosurgical patients anaesthetised during that period. The cost of isoflurane was reduced to about 33% of that incurred during previous periods.
Key words Anaesthetic techniques; closed system. Ventilation:artificial.
Low-flow anaesthesia (LFA) has recently regained popularity. One reason for this is the reduced consumption of isoflurane, and another that LFA requires gastight systems, which reduce gas pollution in the operating room. Finally, LFA helps to reduce emissions of halogenated hydrocarbons and nitrous oxide into the atmosphere. Mechanical control of ventilation with LFA is usually achieved by a bag-in-bottle arrangement, i.e. the ventilator gas is separated from the patient circle by bellows inside an airtight container. The function of this device is simple, but critically dependent on proper function of the bellows and the circle's overflow valve. We used an old method'J as an alternative whereby the ventilator is connected to the anaesthesia circle via a large deadspace that prevents ventilator gas from entering the circle. This is a safe and simple method that combines the facilities of a powerful ventilator with the advantages of the circle system. This paper presents the principles of operation, bench experiments, trials in pigs, evaluation in clinical practice, and economy.
Materials and metbods Principle of operation
The ventilator (Servo 900 C, Siemens-Elema, Solna, Sweden), delivers oxygen into a large deadspace that consists of a corrugated polyethylene (Hytrel) hose with 2.2
litres internal volume. The other end of the hose is connected to a conventional anaesthesia circle with a COz absorber (Monosorb, Siemens-Elema, Solna, Sweden), one-way valves and fresh gas inlet (Fig. 1). The ventilator pushes a tidal volume of oxygen into the deadspace and the same volume of anaesthetic gas mixture is expelled from the other end into the circle. The gas column in the hose oscillates back during expiration: a tidal volume of exhaled gas is collected in the circle end of the deadspace, while one tidal volume of oxygen, mixed with some excess gas from the circle leaves via the ventilator. The pressure generated by the ventilator is transmitted unchanged to the circle since the hose is wide (3.3 cm internal diameter). In essence, the setting on the ventilator (minute volume, frequency etc.) will determine the pattern of ventilation, while the gas composition in the circle is determined by the vaporizer and flowmeter settings connected to the circle's fresh gas inlet. The different ventilatory modes and other functions of the ventilator can be used without restriction. Ventilation volumes, frequency etc. are monitored on the display of the ventilator. However, there is some discrepancy between the inspiratory and expiratory volumes as measured in the ventilator and volumes inhaled and exhaled by the patient. Thus, the fresh gas flow that enters the circle during the inspiratory phase of the ventilator, i.e. when the expiratory valve is closed, will increase the inhaled volume, while fresh gas entering the circle during expiration will increase the flow through the expiratory
L. Berntman, MD, PhD, Consultant, H.H. Luttropp, MD, Registrar, 0. Werner, MD, PhD, Consultant, Department of Anaesthesia, University Hospital, S-221 85 Lund, Sweden. Accepted 22 February 1990. 0003-2409/90/100855 + 04 $03.00/0
@ 1990 The Association of Anaesthetists of Gt Britain and Ireland
855
L. Berntman, H.H. Luttropp and 0 . Werner
856
1,.
Pop-off volve
Fresh gas A
thing system supplied 50:50 N 2 0 / 0 2with isoflurane directly to the pig. Isoflurane concentration and duration of anaesthesia were the same as above. Isoflurane consumption was assessed by weighing the vaporizer on a precision balance (Galaxy 1200-S0, Ohaus, USA). The experiments were done in random order and separated by a one-hour washout phase.
absorber
Ventilator
Fig. 1. The low-flow system. The pop-off valve is functionally closed (35 cm H,O) during mechanical ventilation.
flowmeter of the ventilator, although no corresponding volume is exhaled. With the standard setting of the respiratory cycle, i.e. inspiration time 25% and pause time lo%, 35% of the fresh gas flow (FGF) will be added to the inspired minute ventilation and the rest will bypass the patient but be monitored by the ventilator. Bench tests Experiments were undertaken to evaluate the deadspace volume necessary to prevent mixing of ventilator gas with that of the circle. The fresh gas flow was set to one litre/ minute of 5050 N,O/O,. The circle was connected to a lung model (a 4-litre anaesthetic bag). Oxygen concentration in the circle was measured with an oxygen analyser (Hudson Ventronics 5590, Temecula, California, USA) placed in the expiratory limbs. The ventilator delivered 10 litres/minute of 0,. Tidal volume ( VT) was varied from 380 to 1170 ml by changing the ventilator frequency. The hose volume (VD) was stepwise increased from 550 to 2200 ml and the reading on the oxygen analyser was noted. The compressible volumes were measured at different VD by obstructing the Y-piece, then noting the expiratory at a plateau pressure of 2.0 kPa. volume (h)
Tests in pigs The tests compared the low-flow system’s isoflurane requirements with those of a non-rebreathing system. Four pigs (1 8.5-25 kg) were studied. Anaesthesia was induced with azaperon (Stresnil) and etomidate chloride (Hypnodil) and the tracheas of the pigs were intubated. The minute ventilation (4-6 litres/minute at a respiratory rate of 10 minute) was adjusted to give an end-tidal carbon dioxide concentration of approximatively 4.5%, as measured at the tracheal tube on a Normocap analyser (Datex, Finland). Gas sampled by the analyser was returned to the circle. Fresh gas flow during LFA was one litre/minute of 5050 N 2 0 / 0 2 with isoflurane. End-tidal isoflurane concentration (Servo Gas Monitor, Siemens, Sweden) was kept as 1.5%. Isoflurane anaesthesia was maintained for 2 hours. The ventilator in the non-rebrea-
Table 1.
VD (ml) 550 1100 1650 2200
Vc (ml/kPa) 0.22 0.26 0.29
0.33
Clinical use of the system The system was used in routine anaesthesia in over 600 adult patients scheduled for neurosurgical procedures that lasted up to 14 hours. Patients were unselected except that those with known severe lung dysfunction were excluded. Anaesthesia was induced by thiopentone and maintained by N20, fentanyl and 0.5-1% isoflurane. The circle was connected to the ventilator via the corrugated hose after tracheal intubation. The pop-off valve in the circle was set at 3.5 kPa; this was closed functionally since airway pressures normally do not reach this level. The fresh gas flow was 1.5 litres/minute of oxygen and 3 litres/minute of nitrous oxide with isoflurane as needed during the initial 2&30 minutes that were spent in the induction room. The higher flow was used for denitrogenation and to allow for the patients’ initial high uptake of nitrous oxide. The flowmeters were set to 0.5 litres/minute each of 0, and N,O for the duration of anaesthesia after transfer into the operating theatre. Respiratory monitoring consisted of expired minute volume and airway pressures as displayed by the ventilator, and O2and CO, concentrations sampled from the tracheal tube into a Normocap (Datex)’ which withdrew 150 ml minute from the circle. The sampled gas was diverted to the scavenging system, after passing the analyser, instead of being delivered back to the circle. Isoflurane concentration was measured (Servo Gas Monitor, Simens, Sweden) at the tracheal tube in most patients and the sampled gas returned to the circle. Anaesthesia gases were discontinued after surgery, and oxygen flow via the fresh gas inlet was increased to 6 litres/minute. The circle was separated from the corrugated hose when the patient was ready for spontaneous breathing and instead reconnected to an anaesthesia bag.
Results Bench-tests
Figure 2 shows the oxygen concentration in the circle at different tidal ( VT) and hose ( VD) volumes. Oxygen concentration was unaffected by changes in VT when VD was 1650 or 2200 ml. When VD was reduced to 1100 ml significant mixing occurred at tidal volumes of or above 765 ml. The results at a VD of 550 ml suggested significant mixing with oxygen from the ventilator at all tidal volumes tested. The calculated compressible volumes of the system, at different values of VD are given in Table 1. Tests in pigs
Isoflurane consumption during 2 hours of anaesthesia was 0.4 (SD 0.1) (g/kg)/hour when a low-flow technique was
used. In the non-rebreathing system the isoflurane consumption was 1.9 (SD 0.1) (g/kg)/hour.
Mechanical ventilation during low--owanaesthesia
5
0
857
Ioo1 90
1100 ml
1650 ml 2200 rnl
50
OQ
4bO
6AO
8kO
10’00 1260
Tidal volume (ml)
Fig. 2. Oxygen concentration (mean of duplicate experiments) in the circle at different tidal and hose volumes ( VD). Fresh gas flow to the circle was one litre/minute of 50% oxygen in nitrous oxide. The ventilator minute volume was 10 litres. Mixing of ventilator oxygen with circle gas is indicated by an increase in the measured oxygen concentration.
Clinical experience
The low-flow system has been used since August 1987 without any system-related incidents. From September 1987 low-flow anaesthesia was used in 60% of elective, daytime neuroanaesthesia cases. Figure 3 shows the cost for isoflurane per each 6 month-period from 1985 to 1989. Approximately 62000 SEK (about E5600) was spent on isoflurane during the first 6 months of 1987 while the corresponding sum for 1988 was 26000 SEK (approximately &2300), in spite of a 15% increase in total anaesthesia time at the unit.
Discussion Closed-system and low-flow anaesthesia have been used for many years, but many anaesthetists prefer non-rebreathing systems because of their greater simplicity. The development of in-circle vapour and O,/CO, analysers with alarms have increased the patient safety during LFA and permit anaesthetists to attend to other duties during the procedure. An advantage of the present system is that moving parts such as bellows and a special overflow valve are not needed. This minimises the risk of mechanical dysfunction. Cleaning the system is simple. A possible disadvantage is that the absence of bellows that move may make the system intuitively less easy to understand than a bag-in-bottle ventilator. During steady-state anaesthesia expired minute ventilation ( VE) at the ventilator will be: the inspired minute ventilation ($I) set at the ventilator + fresh gas flow+xygen consumption-gas consumed by the gas analysers. If the patient consumes 250 ml 0, per minute, the analyser samples 150 ml/minute and fresh gas flow is one litre/ minute the measured h will be 0.6 litres/minute larger than the set $I. The exact oxygen concentration in the circle will be a function of the composition and flow of fresh gas, the oxygen consumption and anaesthetic gas ~ p t a k eand ~ . ~the amount of gas removed via gas analysers. We usually obtain an oxygen concentration of 4 1 4 3 % in the circle at steady state when the flowmeters are set at 0.5
Fig. 3. Cost of isoflurane for the first ( I ) and second (2) &month period, 198688; the total expense is given for 1985.
litres/minute of N 2 0 and 0.5 litres/minute of 0,. Some modifications were introduced after the initial evaluation of the system, for example the gas from the CO, monitor is now returned to the circle and the oxygen concentration in the circle is adjusted to about 35% by reducing the oxygen supply. A hose volume ( VD) of 1650 ml prevented entry of ventilator oxygen into the circle, using a clinically relevant range of adult tidal volumes. We elected to use 2200 ml in order to increase the margins. A low-flow system should be gas tight for proper function, but leaks may nevertheless occur e.g. around the cuff of the tracheal tube. The result of a leak in the present system varies with the leak volume (VL). If VL is smaller than the excess gas in the circle (in our case approximately 0.6 litres/minute, see above), measured h will decrease slightly, but there will be no change in oxygen concentration in the circle. If VL is larger, 0, from the Ventilator will enter the circle and cause an increase in the oxygen concentration. A disconnexion anywhere in the system will immediately be recognised as a decrease in measured h. The pressure in the system is controlled by both the setting of the pop-off valve and that of the high-pressure limit on the ventilator. We usually set the high-pressure limit below that of the pop-off valve, and it could be argued that we could as well close the pop-off valve entirely when switching from manual to mechanical ventilation (Fig. 1). Interruption of fresh gas supply is unlikely to cause hypoxia, because under these conditions the oxygen and N,O uptake by the patient will cause a net flow of oxygen from the ventilator via the hose towards the anaesthesia circle. Oxygen concentration will soon increase above the previous level (Fig. 4) after an initial decrease. However, interruption of only the oxygen while the nitrous oxide flow remains will cause hypoxia if this accident is not detected by the 0, analyser. The system has the vaporizer outside the circle. With this arrangement the inhaled vapour concentration may be considerably lower than that set on the vaporizer. The difference is dependent on F G F , ventilation and the patient’s uptake of vapour. The concentration can and should be monitored by an analyser, sampling gas from the tracheal tube, to minimise the risk of intra-operative awareness. This is particularly important when an inhalant-only
L. Berntman, H.H. Luttropp and 0. Werner
858
II20 4Ol
c
2
c C 0) 0 C 0 0
0
1
2
I
I
1
1
I
3
4
5
6
7
Hours
4
Minutes
Fig. 4. Changes in O2concentration in the circle after turning off (at arrow) the fresh gas flow. Four trials in the same patient.
anaesthetic is given. Our patients all had a N,O-fentanyl background and in most patients the isoflurane concentration was monitored. The low-flow system was used in 60% of elective daytime neurosurgical operations, which comprise about 40% of the total case load. The cost of isoflurane in the neuroanaesthesia service was reduced to 33% of that when only non-rebreathing systems were used, in spite of a 15% increase in anaesthesia time. The relatively larger reduction in cost than in the number of anesthetics with open systems is because of a preferential use of LFA with isoflurane for long operations. A calculation of costs for maintenance anaesthesia (Fig. 5) (1% isoflurane, 0, and N,O) with open and closed systems is shown in Figure 5. LFA not only affects the cost of isoflurane, but also reduces the consumption of N,O (cost 0.7 p/litre) and increases that of 0, (cost 0.1 p/litre) which is used to drive the ventilator. In addition LFA requires a CO, absorber for E9. The two systems break even after 30 minutes, after which time the cost is about 15% of that of a non-rebreathing system. The calculations do not include costs for induction or reversal drugs and the comparison refers to maintenance cost after the break even
Fig. 5. Calculated cost of maintenance anaesthesia with 1% isoflurane in either the low-flow (0.5 litres OJO.5 litres N,O+ 12 litres 0, per minute, see text) or a non-rebreathing system (5 litres 0,/7 litres N,O per minute).
point. The isoflurane consumption (and cost) in our animal experiments in LFA was 22% of that using the open system. A retrospective study in humans6 compared total drug costs for relatively short procedures (about 2.5 hours) and showed that the cost for LFA was almost 50% of that with an open system. Gas costs become more and more important during longer procedures.
References 1. VOSS JV. The adaption of ventilators for anaesthesia, with particular reference to paediatric anaesthesia. South African Medical Journal 1967; 41: 1079. 2. ADAMSAP, HENVILLE JD. A new generation of anaesthetic ventilators. Anaesthesia 1977; 32: 34-40. 3. BENGTSON JP, SONANDER H, STENQVIST 0. Oxygen analyzers in anaesthesia: performance in a simulated clinical environment. Acta Anaesthesiologica Scandinavica 1986; 30: 656-9. 4. SEVERINGHAUS JW. The rate of uptake of nitrous oxide in man. Journal of Clinical Investigation 1954; 33: 1183-9. 5. VIRTUE R, SHERRILL DL, SWANSON GD. Uptake of nitrous oxide by man. Canadian Anaesthetists' Society Journal 1982; 2 9 42k7. 6. BENGTSON JP, SONANDER H, STENQVIST 0. Comparison of costs of different anaesthetic techniques. Acta Anaesthesiologica Scandinavica 1988; 3 2 33-5.
