Malignant hyperthermia is a potentially lethal inherited metabolic disorder that may be triggered by inhalational anaesthetic agents. A large body of literature is dedicated to the anaesthetic management of the malignant hyperthermia susceptible patient and the complete avoidance of inhalational anaesthetic agents and suxamethonium is recommended by all. Potential alternatives include regional anaesthesia or total intravenous anaesthesia with non-triggering agents. Although there is no safe low level of inhalational agent known in malignant hyperthermia susceptible humans, levels less than 5 parts per million (0.0005%) are known to be well tolerated for malignant hyperthermia susceptible swine. Although many departments have a dedicated vapour-free anaesthesia workstation, others need to prepare workstations previously exposed to inhalational agents for malignant hyperthermia susceptible patients. A number of publications have examined the preparation of other anaesthesia workstations: the Siemens Kion, the Drager Fabius GS, the Drager Fabius Tiro, the Drager Primus and the DatexOhmeda.1–6.
To prepare any anaesthetic machine for a patient susceptible to malignant hyperthermia, the vaporisers should be removed from the operating theatre completely. The disposable tubing, reservoir bag and CO2 absorber canister should be removed and all replaced with new disposables. This should be followed by a flush of the machine at a high fresh gas flow (at least 10 l min−1) for variable periods. High fresh gas flows enhance the clearance of inhalational anaesthetics by increasing turbulence within the system and maintain a high concentration gradient between the fresh gas flow and components within the circuit. Reducing the fresh gas flow from 10 to 3 l min−1 will result in a rebound increase in inhalational anaesthetic due to on-going elution from rubber or silicone, and due to the redistribution from dead space areas within the breathing system. Therefore, it is recommended that the FGF should be continued at a high rate throughout the operation. Volatile anaesthetics are soluble in rubber and plastic to varying extents and the period of flushing may be shortened significantly by substituting the anaesthetic ventilator diaphragm with a new or autoclaved diaphragm (Fig. 2). The Quick Emergence Device: (QED, Anecare Laboratories, Salt Lake City, Utah, USA.) is a charcoal scrubber filter which may be incorporated into the breathing system and has been demonstrated to significantly reduce the time taken for the anaesthetic concentration to fall below 5 parts per million in the Drager Fabius (Tiro model).6
Our objectives were to study the clearance of inhalational anaesthetic agents in machines present within our department. The Fabius CE is the standard anaesthesia workstation in our anaesthesia induction rooms and is shown here with the Miran sapphire analyser (Thermo Electron Corporation, Waltham, Massachusetts, USA) incorporated (Fig. 1).
The Drager Fabius MRI may be identical to either the CE model or the Tiro model apart from being designed to work within magnetic fields. The Drager Fabius Tiro and GS are identical machines and use the same compact breathing system (COSY 2.6). The Drager Fabius CE is a less sophisticated machine and uses the older compact breathing system (COSY 2.5).
The Drager Zeus anaesthesia boasts a number of innovative features in the delivery of inhalational anaesthetic agents, ventilation and low-flow anaesthesia. The direct injection of volatile anaesthetic allows for what Drager terms as ‘target controlled anaesthesia’; ultralow fresh gas flows may be used while still affording rapid changes in the concentration of volatile anaesthetic delivered. The anaesthetic agent is mixed with fresh gas flow out of circle in the ‘fresh gas mode’ and within the circle in other modes. The Zeus machine has no ventilator diaphragm, but has a ‘Turbovent’ which acts as a pressure source to drive positive pressure ventilation. The heat emanating from the turbovent also serves to heat the gas within the system.
Ethical approval was not sought as this was an equipment study and did not involve patients. Six Fabius CE anaesthesia workstations and three Zeus workstations were studied in total. Three Drager Primus workstations were used as controls, washout times were equal to those previously published.2 The Miran XL sapphire analyser (Thermo Electron Corporation) was integrated into the inspiratory limb of the circle in the manner specified by previous authors2 (Fig. 1). It has an accuracy of 5% and a sensitivity of 0.1 parts per million. The analyser was calibrated before each experiment outside of the theatre environment where the concentration of sevoflurane was 0 parts per million.
The experimental setup comprised an exposure or priming period of 2 h, a washout period and finally a period of low fresh gas flow at 3 or 6 l min−1.
During this time the machine, its circle system and an additional reservoir bag used as an artificial lung were primed for 2 h. The inspired sevoflurane was set at 2%, the FGF at 2 l min−1, the tidal volume at 500 ml min−1, ventilation rate at 15 min−1 and the I:E ratio set at 1 : 2.
Following this period of exposure, the sevoflurane was switched off, the vaporiser removed, the disposable tubing of the circle system was changed and a new CO2 canister, reservoir bag and disposable tubing attached (Fig. 2). The Miran Sapphire XL Ambient Gas Analyser was added to the inspiratory limb of the circle and its internal pump disabled to allow the ventilator to circulate gas through the machine passively. Ventilation was continued at a fresh gas flow of 10 l min−1, tidal volume 500 ml min−1 and rate of 15 min−1; this was followed by an exponential decrease in sevoflurane concentration.
Period of low flow
When the sevoflurane was measured at less than 5 parts per million, the FGF was reduced to either 3 l min−1 (Drager Fabius CE) or 6 l min−1 (Drager Zeus). As expected, each machine developed a rebound increase in sevoflurane concentration, reflecting elution from plastic, rubber or dead space within the breathing system. Measurement continued for a period of 1 h and the concentration of sevoflurane continued to decrease following a second peak at the lower fresh gas flow. The Drager Zeus failed to work in the semiclosed configuration with the Miran gas analyser when the FGF was reduced to 3 l min−1; therefore, a fresh gas flow of 6 l min−1 was employed during the period of low flow.
Drager Fabius CE
The protocol was as follows:
- Washout profile at a FGF of 10 l min−1, disposables changed and vaporisers removed.
- Washout profile at a FGF of 10 l min−1, disposables changed as in (1) and a new ventilator diaphragm and non-disposable ventilator tube (Fig. 3).
- Washout profile at a FGF of 10 l min−1, disposables changed as in (1) with an autoclaved diaphragm and non-disposable ventilator tube.
The protocol was as follows:
- Washout profile at a FGF of 10 l min−1, disposables changed and vaporisers removed.
- Washout profile at a FGF of 18 l min−1, disposables changed and vaporisers removed.
Descriptive statistics were used to determine mean (SD) values. Analysis using Kruskal–Wallis tests to determine whether washout times were significantly different between groups was undertaken. Results were deemed significant at the 5% level. The P values were corrected for multiple comparisons. Statistical analysis was performed using Stata (version 10, Stata Corp., College Station, Texas, USA).
Drager Fabius CE
The results of the experiments conducted on the Drager Fabius CE are presented on Table 1. The longest washout period seen was 196 min and the mean time taken for the anaesthetic agent to reach 5 parts per million was 141 min. Reductions in early washout periods were evident in section (2) of the ‘Experimental protocol’. Initial washout period decreased to an average of 14 min with a maximum of 18 min when the diaphragm and non-disposable ventilator tube were substituted for new parts never previously exposed (P = 0.017). Again when the diaphragm and non-disposable ventilator tube were removed and autoclaved in accordance with manufacturer's recommendations, the early washout periods decreased to an average of 22 min, with a maximum of 36 min (P = 0.031). No significant difference was seen when the effect of new components (2) was compared with the autoclaved components (3) (P = 0.345). When the fresh gas flow was reduced to 3 l min−1 in each of the experiments, there were, as expected, increases in measured sevoflurane within the circuits. Although the peaks reached in all of the circuits exceeded 5 parts per million, there was a progressive decline in the concentration as ventilation was continued over the following hour.
The results from experiments conducted on the Drager Zeus are presented in Table 2. The mean early washout phase at a fresh gas flow of 10 l min−1 was 85 min. When the fresh gas flow was increased to 18 l min−1, this early washout period decreased to 16 min. When the fresh gas flow was decreased to 6 l min−1 in the period of low flow, the peak increase in sevoflurane was much higher in the machines flushed at 18 l min−1 for 16 min when compared with those flushed for 10 l min−1 for a mean of 85 min (70 versus 27 parts per million, data not presented).
The Zeus failed to work in the semiclosed configuration with the Miran gas analyser when the FGF was reduced to 3 l min−1 at the settings (i.e. tidal volume 500 ml, rate 15, I: E ratio 1 : 2). The reasons for this are two-fold. At a FGF of 3 l min−1, the leak in the system as a result of the gas analyser was such that it could not function in the semiclosed configuration. It repeatedly requested that the operator increase the fresh gas flow. The second reason was the manufacturers guidelines suggest that the machine will not work in a closed configuration if reservoir bags which are too distensible or, in technical language, reservoir bags that cross the 1 Mbar threshold are used. The reservoir bag that was employed when changing to the clean system was the standard 2-l capacity bag.
The results from the Fabius CE experiments may be compared with results previously published by authors who have examined other members of the Fabius family: the GS and the Tiro. Although there are similarities, there are significant between study differences. In our study the ventilator settings used and the way in which the Miran gas analyser was incorporated into the breathing system were almost identical to Whitty et al. and based on previous work by the same authors. Whitty et al. analysed the washout of isoflurane 1.5% in the Fabius GS and found the time to reach 5 parts per million was 10 min longer than our washout of sevoflurane (141 versus 151 min) and, following autoclaving, the time to 5 parts per million was 20 min longer (22 versus 42 min).5 Isoflurane has been shown to be more soluble in the rubber and plastic components of a breathing system than sevoflurane; therefore, it is likely to elute at a slower rate, thus explaining the delayed washout demonstrated by Whitty et al.1 Despite this, Prinzhausen et al.2 found that isoflurane 1.5% was washed out more rapidly than sevoflurane 2.5% in the Drager Primus model. We believe the most important reason for the delayed washout of isoflurane seen by Whitty et al. may relate to the arrangement of the compact breathing system. Although the volume of compact breathing systems of the Fabius CE, the Tiro and the GS models are identical (2.8 l), the relative positioning of the reservoir bag is different (Fig. 4a and b). The reservoir bag in the CE model is situated close to the inspiratory valve: the COSY 2.5 arrangement. The Tiro and GS models tested by others uses the COSY 2.6 arrangement in which the reservoir bag is situated the on the expiratory side of the circle, the other side of the CO2 absorber.5,6 As fresh gas flow is decoupled and diverted during inspiration, proximity to the inspiratory side of the circle would ensure the reservoir bag is filled predominantly with fresh gas rather than added from the expiratory side of the circle as recycled gas. Although one group have described the bag position here as being in a ‘cul de sac’, this arrangement is likely to be more efficient than the COSY 2.6 arrangement in terms of anaesthetic agent clearance.8 The COSY within the Fabius CE predisposed the system to occasional large leaks termed the ‘peak puff’ leak. This may occur where low pulmonary compliance may generate brief periods of high flow on exhalation. These high flows or ‘peak puffs’ are preferentially vented through the scavenger relief valve rather than dissipated within the CO2 absorber.9 The COSY 2.6 version in the Tiro and GS models may absorb large flows intermittently and prevent this leak and, therefore, this COSY version has the potential to allow closed circuit anaesthesia.
One initial surprising finding was the large disparity in clearance rates (between 77 and 196 min, Table 1) in the initial experimental washout phase. The disparity disappeared when the ventilator diaphragm and non-disposable ventilator tube were changed or autoclaved. We know rubber and plastic will absorb all inhalational anaesthetic agents to varying extents. It has been shown by us and other investigators that the rubber ventilator diaphragm plays a significant role as a ‘sink’ in which volatile anaesthetic builds up over time and is released slowly.
Theoretically, the Drager Zeus, which has a turbine rather than a piston-driven ventilator diaphragm, should have a short time constant. This was not the case however, and this machine was a relatively poor performer in this regard when compared with other ventilators. The breathing system in the Zeus is relatively large at 4.5 l.7 Although the early washout period was markedly reduced by increasing the FGF to 18 l min−1, this may reflect increased turbulence and a greater dilution of sevoflurane rather than true clearance from the breathing system. During the period of low flow in the Zeus protocol (6 l min−1), the peak concentration seen was observed to be markedly higher in the 18 l min−1 washout than following the 10 l min−1 washout (70 versus 25 parts per million, data not presented). Therefore, it is reasonable to recommend the longer preparation time at 10 l min−1 for the Zeus.
This study has a number of limitations. As the Miran Sapphire analyser was on loan, there was a limited period during which we could perform our analysis. Second, it is very difficult to draw accurate conclusions from just three Drager Zeus machines. Third, sevoflurane was chosen for this study, as it is the predominant inhalation anaesthetic in use and it would have been interesting to have studied desflurane also. Previous authors have examined the partition coefficients of inhalation agents in rubber and plastic components of breathing systems. Solubility, from the most soluble to the least soluble, is halothane>isoflurane>sevoflurane>desflurane.1 This may not reflect how well desflurane is cleared from breathing systems as other physiochemical properties are also very important in determining clearance from breathing systems. Therefore, the results presented could not be extrapolated to include the preparation of an anaesthesia machine in which desflurane has been used.
In conclusion, we recommend that when preparing the Drager Fabius CE for patients with suspected malignant hyperthermia, remove the vaporiser from the operating theatre and change the disposable ventilator tubing and the CO2 absorber for new components. Exchange the ventilator diaphragm and non-disposable tube with either a new or autoclaved diaphragm and non-disposable ventilator tube and flush the machine at 10 l min−1 for a minimum of 36 min [the longest time taken in experimental parts (2) and (3) in the ‘Experimental protocol’ section of ‘Drager Fabius CE’]. When preparing the Drager Zeus for the malignant hyperthermia susceptible patient, remove the vaporiser from the operating theatre, replace the disposable ventilator tubing, reservoir bag and CO2 absorber and flush at a FGF of 10 l min−1 for a minimum of 90 min. In both the Fabius CE and Zeus, continue at a fresh gas flow of 10 l min−1 throughout the operation. We would caution against adopting these suggestions in machines in which desflurane has been in use.
This work was supported by Drager Ireland who supplied the Miran Sapphire XL gas analyser, and also by the Department of Anaesthesia, Beaumont Hospital who supplied the disposables. No other sources of funding or conflicts of interest apply.
The authors thank Dr Joe Lee for assistance with the study.
1. Kim TW, Nemergut ME. Preparation of modern anesthesia workstations for malignant hyperthermia
-susceptible patients: a review of past and present practice. Anesthesiology
2. Prinzhausen H, Crawford M, O’Rourke J, Petros G. Preparation of the Drager Primus anesthetic machine for malignant hyperthermia
-susceptible patients. Can J Anaesth
3. Petroz GC, Lerman J. Preparation of the Siemens KION anesthetic machine for patients susceptible to malignant hyperthermia
4. Beebe JJ, Sessler DI. Preparation of anaesthesia machines for patients susceptible to malignant hyperthermia
5. Whitty RJ, Wong GK, Petroz GC, et al. Preparation of the Drager Fabius GS workstation for malignant hyperthermia
-susceptible patients. Can J Anaesth
6. Gunter JB, Ball J, Than Win S. Preparation of the Drager Fabius anesthesia machine for the malignant-hyperthermia susceptible patient. Anesth Analg
7. Technical Specifications, Training Zeus: Breathing System PowerPoint: http//http://www.dragermedical.co.uk
. [Accessed 15 November 2011]
8. Dosch MP, Loeb RG, Brainerd TL, et al. Time to a 90% change in gas concentration: a comparison of three semi-closed anesthesia breathing systems. Anesth Analg
9. Philips JH. New anesthesia delivery systems. http//www.etherweb.bwh.harvard.edu
. [Accessed 2 January 2012]