Abstract
Cardiac dysrhythmias are frequent complications of electroconvulsive therapy (ECT) [1, 2]. In the majority of cases, they are transient and benign, and rarely lead to the termination of ECT, even in elderly patients with pre-existing cardiovascular morbidity.
Asystole, a potentially lethal complication, is defined as a period of electrical silence on electrocardiogram (ECG) lasting at least 10 s [3]. Asystole is a well-known, but probably underreported, adverse cardiac effect of ECT [4]. We report here a Chinese patient who developed asystole after a subconvulsive stimulus during ECT.
Case history
Mr A., a 65-year-old, 63-kg Chinese man with a 2-year history of depression and good past medical health, was admitted to our psychiatric unit for ECT, following lack of response to adequate trials with various antidepressants. During his previous depressive episode, he had responded well to a course of ECT without adverse effects. A titration procedure to determine seizure threshold was not undertaken at that time.
Mr A. presented now with low mood, anhedonia, poor appetite and constipation. A pervasive sense of hopelessness led him to attempt suicide by hanging prior to this admission. His medication regimen included sertraline 150 mg/day, and piracetam 7.2 g/day. A 12-lead ECG before ECT revealed sinus rhythm with no abnormalities. In the first ECT session, anaesthesia was induced by 1.97 mg/kg of intravenous thiopentone followed by 0.58 mg/kg of suxamethonium for muscle relaxation. A constant current, bi-directional, brief impulse electrical stimulus was delivered by a MECTA SR-2 machine (Mecta, Lake Oswego, OR, USA) using bifrontotemporal electrode placement. A single-channel electroencephalogram (EEG) built into the ECT machine measured seizure activity. According to the standard stimulus dosing schedule [5], the following stimulus parameters were chosen: pulse width, 1.0 ms; frequency, 40 Hs; duration, 1.25 s and current, 0.8 A; the stimulus charge was 80 mC. No seizure activity was detected either by EEG or inspection using the cuff method. The ECG was monitored by a separate cardiac monitor and also on one of the channels of the ECT machine. Immediately after the stimulus, Mr A. developed a 18-s asystole, followed by bradycardia of 40 beats per minute for 10 s. The bradycardia resolved spontaneously before therapeutic intervention was effected. No further electrical stimulation was attempted in this session.
Mr A. was transferred to a cardiology unit for observation. Investigations included 24-h ECG and blood pressure monitoring, blood electrolytes and cardiac enzymes. No abnormalities were detected. Mr A. subsequently received six sessions of ECT with the addition of 0.6 mg of intravenous atropine premedication before anaesthesia. The stimulus charge ranged from 192 to 576 mC. Seizure activity was detected both by EEG monitoring and observation of peripheral muscle activity. The duration of seizures measured by EEG ranged from 15 to 46 s, while the duration of peripheral seizures lasted 15–30 s. No further episode of asystole, bradycardia or any other complication occurred and Mr A.'s depression responded well to ECT.
Discussion
Brief periods of asystole lasting a few s are frequently encountered in the course of ECT, particularly immediately following the stimulus, when a vagus-mediated parasympathetic effect results in bradycardia and asystole. Clement [6] found that 22% of patients had asystole during ECT lasting 1–5 s, despite premedication with 1 mg subcutaneous atropine 30 min before ECT. These brief episodes of asystole are without consequence. However, since asystole longer than 10 s is a potentially dangerous event, 10 s was chosen as a cut-off point to distinguish between inconsequential and clinically significant asystole [3]. Data about the frequency of asystole are conflicting. Most papers refer to cardiac arrest as a rare complication of ECT [3,7–11]. A yet unpublished, retrospective survey of 2424 sessions of ECT delivered in the past 8 years in our adult and geriatric unit found that stimulus dosing was used in 929 sessions. No asystole occurred except the one reported here. In a prospective, controlled study on the safety of ECT in 40 depressed patients with compromized cardiac status (mean age: 68.9 ± 7.2 years), three cases of sinus bradycardia and one case of a 13-s asystole were detected [12]. Another recent prospective study [4] found asystole lasting 5 s or longer in 40.1% of 364 ECT sessions delivered to 38 elderly patients who had received no atropine premedication. Eight of the 38 patients experienced asystole of 10 s or more during ECT immediately following the stimulus. Younger age and higher succinylcholine dose predicted asystole while, surprisingly, in patients with pre-existing ECG block and abnormal rhythm, significantly less asystole was detected. In Burd and Kettl's study [4], it was found that sex, ECT stimulus parameters, electrode placement, seizure duration, cardiac drugs and other medication given during ECT (e.g. methohexital, caffeine, esmolol, nicardipine) did not predict asystole. These results need to be confirmed on a larger sample.
It has been suggested [3] that asystole during ECT is usually the result of multiple factors including the administration of a beta-receptor blocking agent as a premedication [8–11], bilateral electrode placement, which produces a greater cholinergic surge than unilateral ECT [3, 10], the use of thiopentone as an anaesthetic [13] and higher doses of succinylcholine [14, 15]. These factors, singly or in combination [3], increase the risk of cardiac dysrhythmia, bradycardia and asystole.
The optimal balance between the efficacy and sideeffects of ECT can be achieved by applying a moderately suprathreshold electrical stimulus [16, 17]. Since the seizure threshold shows up to 40-fold interindividual variations [16], stimulus dosing at the commencement of ECT is highly recommended by recent practice guidelines [5]. The main disadvantage of stimulus titration is the application of a subconvulsive stimulus, which decreases the heart rate [18] and increases the risk of asystole [3, 8, 19]. A subconvulsive stimulus is thought to cause asystole via unopposed stimulation of parasympathetic system, followed by an immediate upsurge of a peripheral sympathetic response associated with the seizure [8, 12, 18, 19].
A further disadvantage of this procedure is protraction of the anaesthesia by an average of 30 s [20]. In most cases, however, the benefits of seizure threshold titration far outweigh its risks. As the risk of bradycardia or asystole may increase during the titration of seizure threshold in patients with significant cardiac disease, some authors (e.g. [21]) suggest that stimulus dosing should be avoided in such cases.
To date, three incidents of asystole following subconvulsive stimuli have been reported [3, 8, 19]. All three patients had either a history of myocardial infarction [8, 19] or conduction defect [3], which might have contributed to their asystole, independent of the subconvulsive stimulus. Intravenous propranolol premedication was given in two cases [8, 9] without concomitant atropine or glycopyrrolate. The present report is unusual in that it describes asystole following a subconvulsive stimulus in a patient who had no prior cardiovascular abnormality and received no beta-adrenergic blocking agent prior to ECT.
Premedication with anticholinergic drugs in general and before stimulus dosing in particular, is a controversial issue because it raises the rate pressure product and hence the workload of the myocardium during ECT [22]. In addition, atropine predisposes to cardiac dysrhythmias and also increases the regurgitation of gastric content [5]. For these reasons, the Royal College of Psychiatrists guidelines do not recommended routine anticholinergic premedication even during stimulus dosing except for patients with pre-existing bradycardia [5]. In contrast, in his influential textbook, Abrams [21] asserts that anticholinergic drugs should always be used in stimulus dosing in order to prevent bradycardia and asystole. This approach has been adopted by some [12, 16], but not all [17], experts. Atropine is also recommended when a beta-adrenergic blocking agent is used as a premedication [9]. Glycopyrrolate could be an alternative vagolytic agent to atropine, although glycopyrrolate is a less potent anticholinergic than atropine [23].
Our experience shows that, with proper precautions, asystole does not necessitate the suspension of further ECT sessions [3]. In cases of pre-existing conduction defect or the need for beta-blockade, intravenous atropine should be considered as premedication. Inducing anaesthesia with methohexital, avoiding excessive amounts of succinylcholine and subconvulsive stimulus, and employing unilateral electrode placement may further lessen the likelihood of asystole in susceptible cases.
The role of concurrent medications is important when considering cardiac side-effects. Our patient received both sertraline and piracetam. Sertraline given at a therapeutic dose (150 mg/day) has not been demonstrated to have cardiac effects. Piracetam [(2-oxopyrrolidin1yl)- acetamide] is a nootropic drug which has been widely used for more than 20 years for a variety of psychiatric disorders associated with cognitive dysfunction [24], including the prevention or attenuation of ECT-induced memory disturbances [25]. In several previous trials, piracetam exhibited no cardiovascular effects even in doses up to 45 g/day over 7 years [26]. On the basis of these data, it is very unlikely that piracetam contributed to the development of asystole in our patient.
The ethnicity of our patient is another intriguing aspect of this case report in the context of the substantial and fast-growing literature on ethnicity and psychopharmacology [27]. Despite the extensive use of ECT in Asia [28], there is a paucity of ethno-specific data on ECT. Stimulus parameters, premedication effects and the efficacy of ECT have been rarely studied in Oriental psychiatric patients [22, 29, 30] and deserve further research.