Inhalation induction with sevoflurane for electroconvulsive therapy: a case series

AIMS AND METHOD Until 1999, the most common drug for induction of anaesthesia for 
electroconvulsive therapy (ECT) in the UK was methohexital. Cessation of 
production left few choices of induction agent. Increased use of sevoflurane 
in short procedures suggests that it might be suitable as a sole agent for 
anaesthesia in ECT. We therefore induced anaesthesia in five consecutive 
patients undergoing ECT to assess sevoflurane’s potential for this use. 
We recorded vital signs, needle phobia, face-mask toleration, duration of 
induction, seizure duration and recovery variables. RESULTS Anaesthesia was successfully induced in all patients without difficulty. 
One patient experienced mild hypoxia following the seizure. Seizure duration 
ranged from 24 to 72 seconds. Recovery times to eye opening and times to ‘
ward’ fitness were acceptable. CLINICAL IMPLICATIONS Sevoflurane may be a suitable alternative for induction of anaesthesia for 
ECT.

Inhalation induction with sevoflurane for electroconvulsive therapy: a case series AIMS AND METHOD Until1999, the most common drug for induction of anaesthesia for electroconvulsive therapy (ECT) in the UK was methohexital. Cessation of production left few choices of induction agent. Increased use of sevoflurane in short procedures suggests that it might be suitable as a sole agent for anaesthesia in ECT. We therefore induced anaesthesia in five consecutive patients undergoing ECT to assess sevoflurane's potential for this use. We recorded vital signs, needle phobia, face-mask toleration, duration of induction, seizure duration and recovery variables.

RESULTS
Anaesthesia was successfully induced in all patients without difficulty. One patient experienced mild hypoxia following the seizure. Seizure duration ranged from 24 to 72 seconds. Recovery times to eye opening and times to 'ward' fitness were acceptable.

CLINICAL IMPLICATIONS
Sevoflurane may be a suitable alternative for induction of anaesthesia for ECT.
The production of methohexital ceased in September 1999, ending 40 years of its use in anaesthesia for electroconvulsive therapy (ECT), first described by Friedman (1959). In December 1999, the Special Committee on Electroconvulsive Therapy of the Royal College of Psychiatrists issued guidance on the preferred induction agent for ECT (Freeman, 1999). These guidelines suggested propofol as the next best choice, but concluded by stating that methohexital remained the drug of choice. The requirements for an anaesthetic agent for ECT are difficult to realise: the drug must not only possess induction characteristics suited for repeated use, including rapid recovery to street fitness with no confusion following therapy, but also have a minimal effect on seizure duration.
Sevoflurane, a methyl-propyl ether first used in 1971 in the USA, has been increasingly available since 1997 in the UK. It is non-irritant, has a pleasant ethereal smell, and, with its low blood gas solubility of 0.6, induces anaesthesia rapidly. It also appears to have excitatory effect on the electroencephalogram (EEG) during inhalational induction (Yli-Hankala et al, 1999;Sato et al, 2002). It would appear to be a suitable agent for repeated induction of anaesthesia for short procedures, including ECT. To date there has been little study of sevoflurane in ECT. Ishikawa et al (2001) described a case of a pregnant woman receiving sevoflurane for maintenance of anaesthesia during her course of ECT. At the time of the series described here (1999), the use of sevoflurane for inhalational induction of ECT had not been presented. We describe an open observational study on a consecutive series of patients presenting for ECT.

Method
After ethical committee approval, five consecutive patients (four male, one female: age range 33-52 years) were recruited; all had had between one and ten ECT treatments with propofol as the induction agent. Informed consent was obtained from all participants. We excluded patients with a history suggestive of malignant hyperpyrexia or those with severe respiratory or cardiac disease. Patients received no premedication, but did receive all current psychiatric medication on the day of treatment. In the treatment room, all patients had monitoring applied and received pre-oxygenation through a face mask and Bain circuit.
At time zero, the anaesthetist turned the vaporiser to 8% (maximum) and an assistant cannulated a vein in the hand or arm. The time to loss of eyelash reflex and verbal contact, occurrence of adverse airway events and any purposive movements were recorded. Once the anaesthetist was satisfied with anaesthetic depth, succinylcholine was administered at a dose commensurate with the size of the patient or the same as that given during previously successful ECT therapy. The electrical stimulus was delivered bilaterally using an Ectonus 5CB-62 (Ectron Ltd, Letchworth, Herts, UK) and the duration of motor activity was recorded in an isolated foot. Sevoflurane administration was stopped when the stimulus was delivered. The patient was monitored, and ventilation assisted, until spontaneous respiration returned; the patient was placed in the recovery position and transferred to the recovery area. Time to eye opening after cessation of sevoflurane administration, times to follow simple commands, and time to street fitness were recorded. Recovery staff and patients also completed a simple questionnaire, which measured patient satisfaction and any confusion or excessive drowsiness in the recovery period.

Results
All participants were rated as class II on the American Society of Anesthesiologists classification (Saklad, 1941), and their body mass indices ranged from 22.6 kg/m 2 to 30.4 kg/m 2 . Montgomery & -sberg Depression Rating Scale (Montgomery & -sberg, 1979) scores ranged from 15 to 35, and patients' current medication included zopiclone, clozapine, venlafaxine, paroxetine and chlorpromazine. No patient had needle phobia. Skull impedance ranged from 215 O to 351 O in four patients, and was unrecorded in one. Charge delivered ranged from 235 mC to 676 mC. Face mask tolerance was good in four patients, and acceptable in one.
Duration of induction was acceptable in all patients (range 79-268 seconds) and there was no adverse event. Motor seizure duration ranged from 24 to 72 seconds. No additional drug was given to any patient either during or after the procedure. Four patients maintained acceptable oxygen saturation during the entire procedure, but one patient's saturation level fell to 94%. All patients increased their heart rate and systolic blood pressure during the ECT procedure. Recovery in all patients was rapid and uncomplicated with no postoperative nausea or vomiting. Two patients took a longer time to respond to verbal commands than the other three, but their times to street fitness were similar (range 12-21minutes). Reasons for this are unclear. All patients completed the questionnaire. The most important finding was that four of the five patients liked the method of induction, even though two liked it less than an intravenous induction. Two of the five felt that induction was slower, and two (the patients with the fastest induction times) thought it was faster; four of the five felt that something was different, and recalled both the cannula and face mask; and three recalled the smell of the volatile. Interestingly, one patient could not say what had induced anaesthesia. Sensations of restlessness and confusion in recovery were reported by three patients and corroborated by recovery staff.

Discussion
Methohexital, a methylated oxybarbiturate with rapid onset of action and smooth induction characteristics, has been the drug of choice for induction of anaesthesia for ECT for many years. The drug's ability in some cases to produce abnormal spike discharges in some patients with epilepsy but a normal EEG might have been influential in its selection (Harel et al, 1975). It is more likely, however, that its popularity was due to its recovery profile being superior to that of thiopental, although in this it does not compare with propofol (Pollard et al, 2003).
The introduction of propofol predictably led to a number of studies comparing propofol with methohexital, propofol with thiopental, and propofol with both methohexital and etomidate. Studies examining the potential effects of anaesthetic agents on the therapeutic effect of ECT often assume that duration of motor seizure, or cumulative seizure duration of a course of ECT, provides a main indicator of beneficial effect (Maletzky, 1978;American Psychiatric Association, 1990), although the duration of the motor seizure is not now felt to be essential (Fear et al, 1994). Although the current ECT Handbook suggests that a seizure of at least 15 seconds peripherally or 25 seconds on EEG is likely to produce a consistently beneficial effect (Scott & Lock, 1995), there is no current substantive evidence for a minimum effective seizure length. In fact, since the change in seizure threshold (possibly due to the release of endogenous opioid anticonvulsants), particularly with unilateral ECT, over a treatment course, may be more relevant to outcome (Sackeim, 1999).
It should be noted that measurable seizure quality indices that relate to how generalised a seizure is, rather than to its length, such as mean integrated amplitude and postictal suppression (how quickly the EEG flattens after a seizure), are not affected by anaesthetic agent choice. There is also no difference in outcome (Fear et al, 1994;Kirkby et al, 1995), at least between propofol and methohexital.
In studies comparing induction agents for ECT in terms of seizure duration, etomidate appears to have the least effect both on motor and EEG seizures, followed by methohexital, thiopental and finally propofol, but the reduction in seizure duration experienced with propofol and methohexital is dose-dependent, whereas that of etomidate is not (Boey & Lai, 1990;Avramov et al, 1995;Kirkby et al, 1995;Geretsegger et al, 1998). The effects of different doses, concentrations and methods of application of sevoflurane on seizure duration require detailed investigation. In terms of cardiovascular stability, propofol has been shown to be most effective at blunting the haemodynamic response to ECT and etomidate the least (Boey & Lai, 1990;Avramov et al, 1995), but since recent work suggests an intriguing link between a vigorous cardiovascular response to ECT and an earlier antidepressant effect, this might be beneficial in some individuals (Saravanan et al, 2002). Etomidate, however, is associated with high rates of postoperative nausea and vomiting, and in repeated doses has a dangerous adrenocortical suppressive effect (Watt & Ledingham, 1984); its use is therefore discouraged. Thiopental has an undesirable recovery profile, and may lead to prolonged sedation and confusion, especially among the elderly, and is considered unsuitable (Freeman, 1999).
Inhalational induction of anaesthesia with sevoflurane has been shown to be acceptable for day case surgery and may be performed either by stepwise increase in concentration of inhaled anaesthetic agent or by inhalation of high concentrations from the start. Induction is slower than with the intravenous agents, but emergence is more rapid and apnoea less common; however, the increased rate of postoperative nausea and vomiting may reduce these perceived advantages in some population groups (Pollard et al, 2003). The effect of sevoflurane on the electrophysiology of the brain has been studied, with conflicting results. Several workers have recorded epileptiform EEG tracings accompanied in some cases by visible motor manifestations (Yli-Hankala et al, 1999;Sato et al, 2002). The ability of sevoflurane to modulate neuroexcitability was shown in one study on patients with epilepsy to be dose-dependent, manifested most at near burst suppression doses of 1.5 minimum alveolar concentration (MAC). During induction with sevoflurane inhaled concentrations of up to 4 MAC are used, and although concentrations at the effect site will be well below this -usually 1-2 MAC -inhalation of sevoflurane before ECT could affect seizure threshold and possibly duration. Comparing our results with those of Calarge et al (2003) shows some interesting differences, particularly induction times (generally shorter in our series), seizure duration (longer in our series) and lack of nausea in our patients. This may reflect the male skew or the number of smokers in our group (males and smokers suffer less postoperative nausea and vomiting) and possibly the technique used to induce and maintain anaesthesia.
In conclusion, although recognising the small size of the study and hence the lack of evidence as to the safety of the technique for repeated use, we present additional preliminary findings on the efficacy of sevoflurane in a new area of application. The high level of acceptability of the method to patients, the rapid recovery with low numbers of adverse effects and the fact that this agent is indicated for repeated inductions for surgery suggest that sevoflurane might be a useful alternative to current intravenous induction agents, particularly if the potential exists for reducing the charge delivered if seizure duration is prolonged.