The Reflector: a new method for saving anaesthetic vapours

L. Perhag1,*, P. Reinstrup1, R. Thomasson2 and O. Werner1

1Department of Anaesthesia, University Hospital of Lund, S-22185 Lund, Sweden. 2Department of Inorganic Chemistry 2, University of Lund, PO Box 124, S-22100 Lund, Sweden

Accepted for publication: April 11, 2000


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Anaesthesia systems that minimize the use of volatile anaesthetics to reduce cost and pollution are of interest. Closed circuit anaesthesia is the ideal solution, but requires continuous adjustment of fresh gas flow and composition and thus is demanding in routine practice. We describe an alternative system, the Reflector system, which is open in regard to oxygen, nitrogen and N2O, and semiclosed in regard to volatile anaesthetics. The Reflector system is a circle system with a carbon dioxide absorber and an automatic vapour delivery device placed in the inspiratory limb of the circle. A zeolite filter, the Reflector, is placed between the ventilator and the circle. The Reflector functions as a molecular sieve, preventing the volatile anaesthetic from leaving the circle. Isoflurane consumption using the Reflector system in bench tests and an animal study was compared with that of an open system. In bench tests consumption was reduced by 79% and 82%, at a respiratory frequency of 10 and 20 min–1, respectively. The corresponding mean figures from the animal experiment were 65% and 77%.

Br J Anaesth 2000; 85: 482–6

Keywords: anaesthetic techniques, closed circuit; anaesthetic techniques, inhalation; anaesthetics volatile, isoflurane


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
An anaesthesia circuit with carbon dioxide absorption was described in 1915 by Jackson.1 In 1926, Waters2 developed the to-and-fro absorption method, and the introduction of cyclopropane in 1933 led to widespread use of closed circuit anaesthesia.

With the clinical introduction of halothane in 1956 by Raventos,3 low-flow and closed systems were used, but halothane vaporizers were imprecise at low gas flows, and lack of knowledge about the pharmacokinetics of halothane resulted in continued use of open/semiclosed systems.

The introduction of more ideal,4 but also more expensive, inhalational anaesthetics and an increasing awareness of atmospheric pollution5 have increased the interest in low-flow and closed circuit methods. These methods preserve humidity and body heat6, save money7 and reduce atmospheric pollution.

The greater use of gas monitors to measure inspired oxygen concentration8 and other gases of the circuit has increased the safety of closed circuit anaesthesia. However, because they require continuous adjustments of gas delivery to the circuit,9 these systems have not gained much popularity.

One solution could be a system that was open in regard to oxygen, nitrogen and N2O delivery, but closed in regard to volatile anaesthetics. Such a system has been created by introducing a molecular sieve (Reflector),10 a semipermeable filter that keeps the volatile anaesthetics on the patient side of the anaesthetic circuit but lets most other gases pass through. We evaluated such a Reflector system in bench tests and in an animal study.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Set-up and experimental procedure
An anaesthesia circuit with one-way valves and a carbon dioxide absorber (Monosorb; Siemens Elema AB, Solna, Sweden) was used. The tubing was made of polyethylene (Hytrel). The Reflector was placed between a ventilator (Servo 900 C; Siemens Elema AB) and the circuit. A heat–moisture exchanger (Siemens Humidifier 152; Siemens Elema AB) was placed at the tracheal tube connector.

A computerized automatic vapour delivery device (AVDD)11 was placed in the inspiratory limb of the circuit, distal to the absorber. The AVDD delivered isoflurane at a partial pressure close to 40 kPa, the carrier gas being sucked from the circuit. During each inspiration, the AVDD continuously adjusted the addition of isoflurane vapour into the circuit to keep a set inspiratory isoflurane concentration of 1.5%. To allow continuous assessment of the isoflurane consumption, the reservoir of the AVDD was placed on an electronic balance (G1200; Ohaus, NJ, USA).

The ventilator delivered a gas mixture of 40% oxygen in nitrogen with a volume-controlled ventilatory pattern and constant inspiratory flow. The inspiratory time of the ventilatory cycle was 33% and the set end-expiratory pressure was set to 5 cm H2O.

Side-stream gas analysers (Siemens 120; Siemens Elema AB), recirculating the sampled gas to the circuit, were used. Inspired, end-tidal and reflected isoflurane concentrations and inspired oxygen concentration were measured.

Measurements were also taken at the expiratory port of the ventilator to detect ‘isoflurane breakthrough’ (lower detection limit=0.1%), i.e. leakage of isoflurane on the ventilator side of the Reflector. The set-up of the Reflector system is shown in Figure 1.



View larger version (13K):
[in this window]
[in a new window]
 
Fig 1 Set-up of the Reflector system. AVDD, automatic vapour device. HME, heat–moisture exchanger. The ‘subject’ is the connection to model lung or animal. Numerals indicate gas sampling points.

 
The reflector
The Reflector consisted of a cylindrical plastic container, with a diameter of 70 mm, in which a 30 mm (RS30) or 50 mm (RS50) deep adsorption bed was arranged between two steel grids. A spreader was placed on each side to create laminar flow through the adsorption bed.

The pressure gradient over the RS30 was 1.2 and 4.6 cm H2O at 0.5 and 1.0 litres s–1, respectively, and that over the RS50 was 3.0 and 11 cm H2O at 0.5 and 1.0 litres s–1, respectively. The dead space of the plastic container, with the zeolite volume subtracted, was identical in both filters.

Pellets (110 or 185 ml) of ultrastable zeolite Y, a synthetic substance, were used as adsorption material. The pellets were about 1 mm in diamater and were prepared using about 15% by mass of a binder material (Al2O3).

Ultrastable zeolites are crystalline microporous silicates built up from corner-sharing [SiO4]4–-tetrahedral forming an indefinite framework structure, characterized by the presence of cavities and channels of well-defined shape and size (3–10 Å). Each type of zeolite has a unique framework structure, denoted by a three-letter code. Ultrastable zeolite Y is of the Faujasite (FAU) structural type, characterized by channels in three dimensions, with an aperture of 7.4 Å, connecting to microvoid cages ({alpha}(II) cages) within the structure. Ultrastable zeolites have high thermal stability (up to 1200°C) and are chemically inert; for example, they are insoluble in acids (except HF) and stable to alkali below pH 10.

Bench tests
Experiments were undertaken to compare isoflurane consumption by RS30 and RS50 with that by a non-rebreathing system. The different components of isoflurane consumption using the two Reflector systems were further studied. The reduction in isoflurane consumption by the added deadspace caused by the geometry of the Reflectors was evaluated using a dummy filter. This filter had a deadspace identical to that of RS30 and RS50 but was filled with glass spheres of 1 mm diameter. With connections all filters had a deadspace of 125 ml. To assess the amount of isoflurane lost due to ‘isoflurane breakthrough’, a Douglas bag, made of aluminium foil, was placed at the expiratory port of the ventilator. The amount of isoflurane trapped in the Reflector was also assessed by weighing the filter before and after each experiment

In bench tests the anaesthesia circle was connected to a lung model consisting of a U-shaped vessel filled with water. The end-expiratory volume of the model lung was 2.5 litres. The ventilator delivered 10 litres min–1 with a respiratory frequency of 10 and 20 min–1. The inspiratory concentration of isoflurane was set to 1.5%.

In control experiments, the Y-piece of the ventilator tubing was connected directly to the model lung, creating an open system. The AVDD was placed in the inspiratory limb.

Animal experiments
After approval from the local Animal Studies Committee, two pigs (34 and 28 kg; one female and one male) were studied. The pigs were premedicated with 15 mg midazolam i.m. 30 min before anaesthesia. Anaesthesia was induced with ketamine i.v. 1.5 mg kg–1 and maintained with a continuous infusion of 7 mg kg–1 h–1. Continuous infusion of pancuronium 0.35–0.50 mg kg–1 h–1 preserved muscle relaxation.

A tracheotomy was performed; control of the airway was achieved by a cuffed tube. A pressure transducer attached to a catheter placed in the carotid artery monitored arterial pressure. A central venous cannula was placed in the jugular vein and an i.v. infusion (glucose 25 g l–1 with 70 mmol Na+, 45 mmol Cl and 25 mmol acetate per litre) was given at a rate of 6 ml kg–1 h–1.

In the animal experiments only the RS30 was tested. The pigs were ventilated at a respiratory rate of 10 and 20 min–1 and tidal volume (Vt) was adjusted to keep end-tidal carbon dioxide within 4.0–4.6 kPa. End-tidal carbon dioxide was measured with an in-line carbon dioxide analyser (Novametrix 7000, Wallingford, USA). A heat–moisture exchanger was placed at the tracheal tube. Inspiratory concentration of isoflurane was set to 1.5% throughout each experiment.

Each phase of the experiments, using the Reflector or the open system in a varied order, lasted 90 min. The zeolite material in the Reflector, when used, was renewed before a new experiment was started. Between each phase there was a 90 min washout period without a filter in the system. During the washout periods the end-expiratory isoflurane concentration decreased rapidly and no isoflurane was detectable after 30 min.

Calculation of measurement error
To verify the validity of isoflurane consumption in the open system, isoflurane expenditure was calculated (Tables 1 and 3) using 1.496 x 103 g litre–1 as isoflurane density at 25°C and 184.5 as molecular weight.12


View this table:
[in this window]
[in a new window]
 
Table 1 The gravimetrically measured consumption at 1.5% inspired concentration of isoflurane. RR=respiratory rate
 
The quantity of isoflurane used may differ from the calculated quantity because of measurement errors in the weight, the gas analyser and inaccuracy of the delivered volume from the ventilator. The measurement errors of the balance, gas analyser and ventilator are ±2%, ±5% and ±5% respectively, giving a combined measurement error of 7%.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Bench tests
The calculated isoflurane consumption over 90 min using a ventilator flow of 10 litres min–1 and an inspiratory isoflurane concentration of 1.5% is 103.8 g. The gravimetrically measured isoflurane consumption values for the open system at a respiratory frequency of 10 s–1 and 20 s–1 (Vt 1000 and 500 ml) were 100.7 and 102.6 g, respectively.

Compared with the open system, the isoflurane consumption was reduced by 12% and 25% at a respiratory frequency of 10 and 20 min–1, respectively, when using the Reflector system with the dummy filter. Compared with the open system, the RS30 reduced the consumption of isoflurane by 79% and 82%, respectively. The corresponding figures when using the RS50 were 80% and 82%. The absolute consumption is listed in Table 1.

Using the RS30, ‘breakthrough’, or tracings >=0.1% of isoflurane on the ventilator side of the Reflector, was detected after 30 min at a respiratory frequency of 10 s–1 (Vt 1000 ml) and after 44 min at a respiratory frequency of 20 s–1 (Vt 500 ml). No such breakthrough was detected with the RS50. The amount of isoflurane retrieved from the Douglas bag after 90 min sampling of scavenger gas and the gravimetrically measured weight gain of the filters are listed in Table 2.


View this table:
[in this window]
[in a new window]
 
Table 2 Isoflurane consumption displayed in its different components using the Reflector system. RR=respiratory rate
 
Animal tests
In the reflector experiments (RS30), the inspired oxygen concentration was 37–40%. The preset inspired isoflurane concentration of 1.5 % was reached after 4 ± 2 min (mean ± SD) and kept at this level (±0.1%) by the AVDD. Within 10 min an end-tidal isoflurane concentration of at least 1.2% was reached. Reflection of isoflurane was detected within 1 min and rapidly reached a plateau of 1.1% within 15 min (Figure 2). Using the RS30, at a respiratory frequency of 10 or 20 min–1, isoflurane consumption was reduced by 61–78% of that of an open system (Table 3).



View larger version (19K):
[in this window]
[in a new window]
 
Fig 2 Tracing of mean inspired, end-tidal and reflected isoflurane concentrations in animal experiments using the Reflector system (RS30).

 

View this table:
[in this window]
[in a new window]
 
Table 3 Calculated and gravimetrically measured isoflurane consumption in animal studies using the Reflector (RS30) and open systems after 30, 60 and 90 min. Vt=tidal volume (mean±SD); RR=respiratory rate. * indicates breakthrough of isoflurane through the filter which happened after 75 min (pig 1 weighed 34 kg and pig 2 28 kg)
 
In one pig, at a respiratory frequency of 10 min–1, ‘isoflurane breakthrough’ was detected after 75 min.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
High gas flow administered to an anaesthesia circuit with carbon dioxide absorption is used as a standard technique by many anaesthetists. The inspired concentration of gases is close to the concentrations in the fresh gas delivered into the circuit. The major disadvantages of this approach are its expense and the loss of heat and humidity.

The Reflector system was designed as a simple high-flow system which is closed with regard to volatile anaesthetics, in this case isoflurane. This is achieved by inserting a zeolite filter at the interface between the anaesthetic circuit and the Y-piece from the ventilator (Figure 1). The ventilator provides the circle with fresh gas with each tidal volume. Because about 25% of carbon dioxide is reflected by the molecular sieve,13 an anaesthesia circuit with a carbon dioxide absorber was used.

We compared the amount of isoflurane consumed using the open system with calculated values. The discrepancy between the calculated consumption of isoflurane and the amount used in the open system can be explained by measurement errors in the components, as consumption of an open system depends only on the delivered volume, if inspired volatile concentration is constant. Thus, the combined error of isoflurane consumption depends on the accuracy of the volume delivered by the ventilator, the inspired isoflurane concentration measured by the gas analyser and the accuracy of the balance. This error was calculated as 7%, and all the measured values were within this level.

Assessment of the Reflector system in bench tests was done by comparing a zeolite filter with a dummy filter, using an open system as control. The reduction of isoflurane consumption when changing from the open system to the Reflector system, but using a dummy filter filled with glass spheres, was caused by the added deadspace of this filter when situated between the ventilator and the circle and was more pronounced at high ventilatory frequencies equal to low tidal volumes.

Zeolite can reflect the expired isoflurane and thereby retain the volatile anaesthetic in the circle system. The microporous arrangement of the zeolite gives rise to a large internal surface area (1000 m2 g–1) and microvoid space (0.5 cm3 g–1). Because of the connectivity of the tetrahedral building units, all oxygen atoms in the three-dimensional framework are interconnected and shared between two silicon atoms: there are no terminal hydroxyl groups. This makes the (internal and external) surfaces of the zeolite hydrophobic. Together with the size of the micropores, this gives ultrastable zeolite Y a high affinity and a high adsorption capacity for non-polar substances, e.g. volatile anaesthetics, and a low affinity for polar substances, e.g. water. The size and shape of the molecules entering the pore system, i.e. the aperture of the channels, restrict the adsorption. The amount adsorbed depends on several factors, including the partial pressure, molecular weight, polarity and polarizability of the adsorbate and on the temperature.

We found that both RS30 and RS50 reduced isoflurane consumption by about 80% compared with the open system. Our results support the findings of Thomasson, Luttrop and Werner10 using a similar filter but a different set-up. However, in a truly closed system the amount of isoflurane used should only be equivalent to the volume of the circle and test lung, saturated to 1.5%, in this set-up approximately 0.6 g of isoflurane.

Isoflurane use during Reflector experiments was greater than that in a closed system because of ‘isoflurane breakthrough’, the passage of volatile anaesthetics through the Reflector and absorption of vapour in the molecular seive of the Reflector. Breakthrough was more pronounced at a large tidal volume while absorption was independent of respiratory frequency.

Increasing the amount of zeolite in the filter, from the RS30 to the RS50, did not increase efficacy because each millilitre of zeolite added to the absorption bed traps about 0.1 g of isoflurane when saturated. Therefore, increasing the amount of zeolite to prevent loss through the filter, was in this case, offset by the amount of isoflurane trapped in the filter. Lengthy anaesthesia will be more cost effective, as will the reuse of a loaded filter.

The difference between isoflurane use in the open pig experiments (Table 3) and that in the open bench tests (Table 1) was caused by differences in tidal volume, which were necessary to keep end-tidal carbon dioxide within normal limits.

In the animal experiment, isoflurane consumption by the open system and the RS30 was not quite comparable because the tidal volumes used were higher in the RS30. One reason was that the volume of this system was 2 litres greater, which increased the compressible volume. Futhermore, end-tidal carbon dioxide was only kept within the limit of 4.0–4.6 kPa.

The calculated uptake of isoflurane by the pigs in our experiments was 8–9 g.14 The reduction in concentration of isoflurane in the expired air, caused by uptake of isoflurane by the pig, may ease the burden on the zeolite filter (RS30) as suggested by the fact that ‘isoflurane breakthrough’ was always found in the bench tests but only detected in one phase of the animal experiments. Uptake by the pigs was similar to the amount of isoflurane lost through the filter (RS30) in bench tests, thus explaining the small difference between consumption in the bench tests and animal experiments.

In conclusion, by using a ventilator attached to the circle system with a zeolite reflector placed in between, isoflurane consumption is reduced by approximately 80% in a bench test and approximately 70% in an animal study.


    Footnotes
 
* Corresponding author. Present address: Department of Anaesthesia, Ystad Hospital, S-27182 Ystad, Sweden Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
1 Jackson DE. A new method for the production of general analgesia and anaesthesia with a description of the apparatus used. J Lab Clin Med 1915; 1: 1–12

2 Waters RM. Advantages and technique of carbon dioxide filtration with inhalation anaesthesia. Anesth Analg 1926; 5: 2–4

3 Raventos J. The action of fluothane—a new volatile anaesthetic. Br J Pharmacol 1956; 11: 394–409

4 Jones, RM. Desflurane and sevoflurane: inhalation anaesthetic for this decade. Br J Anaesth 1990; 65: 527–36[ISI][Medline]

5 Rowland FS. Chlorofluorocarbons and the depletion of stratospheric ozone. Am Sci 1989; 77: 36–45[ISI]

6 Bengtson JP, Sonander H, Stenqvist O. Preservation of humidity and heat of respiratory gases during Anaesthesia. Acta Anaesthesiol Scand 1987; 31: 127–31[Medline]

7 Cotter SM, Petros AJ, Dore CJ, Barber ND, White DC. Low flow anaesthesia—practice, cost implications and acceptability. Anaesthesia 1991; 46: 1009–12[ISI][Medline]

8 Wilson RS, Laver MJ. Oxygen analysis: advances in methodology. Anesthesiology 1972; 37: 112–16[ISI][Medline]

9 Virtue RW. Minimal-flow nitrous oxide anesthesia. Anesthesiology 1974; 40: 196–98[ISI][Medline]

10 Thomasson R, Luttropp HH, Werner O. A reflection filter for isoflurane and other anaesthetic vapours. Eur J Anaesthesiol 1989; 6: 89–94[ISI][Medline]

11 Perhag L, Larsson A, Luttropp HH, Persson J, Werner O. Method to deliver a preset inspired concentration of anesthetic regardless of degree of rebreathing. Anesthesiology 1991; 75: A1010

12 Stevens WC, Kingston HGG. Inhalation anesthesia. In: Barasch PG, Cullen BF, Stoelting RK, eds. Clinical Anesthesia, 2nd edn. Philadelphia: JB Lippincott, 1992; 439–65

13 Thomasson R. Zeolites, Curvature and Anesthesia. Thesis, 1991. University of Lund, Sweden.

14 Lowe HJ, Ernst EA. The Quantitative Practice of Anesthesia—Use of Closed Circuit. Baltimore; Williams and Wilkins, 1981; 16, 67–98