Abstract
The enduring debate about the relative methods of bilateral versus unilateral ECT has led to a wealth of scientific literature, but also to idiosyncratic clinical practice that does not always serve patients optimally. [1]
In carefully selected, thoughtfully administered cases, electroconvulsive therapy (ECT) remains an effective medical intervention. Using strict remitter criteria, 95% of those with psychotic depression and 83% with nonpsychotic depression responded with an overall mean of 87%. They responded following a mean of 7.8 bitemporal ECT treatments administered at 50% above empirically derived, seizure thresholds in a prospective, multicentre, trial described under the acronym, CORE (Consortium for Research in ECT) [2]. However, and perhaps as a result of this recent and rigorously determined result, an earlier debate is being revisited – whether to use unilateral or bilateral ECT. Our aim is to explore the recent literature surrounding these two electrode placements and to describe a 2 year experience with one of these, namely right unilateral ECT administered at six times seizure threshold (6 × RUL ECT) in an Australian setting.
Efficacy has not, and has never been, the problem of ECT. ECT remains, indisputably, the single most efficacious treatment of serious depression. The problem with ECT has been, and remains, the need to diminish cognitive side-effects. [3]
This quotation captures the driving force behind the use of brief pulse, square wave, constant current devices and the unilateral and bifrontal electrode placements [4]. However a problem has arisen and that is the disappointing results from unilateral ECT. Sackeim has preferred the unilateral electrode placement, arguing for its efficacy while sparing cognitive side-effects. However, using his own data, unilateral ECT administered at 2.5 × seizure threshold has resulted in a 47% response rate [5]. Similar results have been found in more recent work by the same group from Colombia despite their continued careful and rigorous methodology. In their 2001 article, 46% either didn't respond or became nonremitters within 4–8 days. Of the 159 completers eligible to proceed for randomization into the continuation pharmacotherapy arm of the trial, 84% of patients who received placebo relapsed [6]. As Abrams points out, these results are substantially poorer than earlier studies [4]. The point is made more starkly by comparing the response rates in two landmark studies with near equivalent methodologies and identical remission criteria. In the CORE study, 253 patients with unipolar depression received bitemporal ECT and 87% responded. In the Columbia study, 290 unipolar patients received 2.9 × RUL and only 54.8% responded. A similar disappointment was noted in an Australian setting [8]. Using 2.5 × right unilateral ECT, only 31% of predominantly unipolar patients responded. In addition they received a mean total of 9.4 treatments. Abrams [4] provides a useful summarizing table of unilateral studies in order of increasing mean stimulus dosage (Table 1). Unlike bitemporal ECT, unilateral ECT is especially dosesensitive. That is, for the unilateral electrode placement, the higher the dosage, the better the outcome. This was so whether outcome was measured by percentage improvement or response rate. Further, and as indicated by Abrams’ inclusion of a dotted line, ‘… no study in the 195 mC or lower range obtained greater than 65% efficacy and no study in the 378 mC or greater range obtained less than 65% efficacy.’ He concluded that high stimulus dosing is quite effective for unilateral ECT, with better response rates and fewer treatments than lower dose ECT, the mean number of treatments being 6.8 and 8.9, respectively. The debate shifts therefore to comparing high dose unilateral ECT with that of bilateral ECT.
Brief-pulse unilateral ECT response by stimulus dose†
In a double blind, prospective study, 80 patients who met Research Diagnostic Criteria for major depressive disorder were randomly allocated to a number of electrode placements and dosages. For our purposes, these included 6 × RUL patients receiving a mean dose of 441 mC, and to bitemporal ECT at 1.5 × the seizure threshold, patients receiving a mean dose of 244 mC. Sackeim et al. argued comparable response rates but with significantly fewer cognitive side-effects using 6 × RUL ECT [9]. This study was important. For the first time it showed the possibility that unilateral ECT was capable of matching the effectiveness of bilateral (bitemporal) ECT. As such it received considerable scrutiny, most cogently through the accompanying commentary [10]. Abrams pointed out a number of problems not the least of which was the disappointingly low 65% response rate. He also noted that the most severely ill patients had been excluded and preferentially assigned to bitemporal ECT from the outset, that those patients who had failed to respond to unilateral ECT responded when switched to the bitemporal placement and that the choice of bitemporal ECT at 150% above seizure threshold was considerably higher than, for example, the CORE group, thus biasing cognitive side-effects against the bitemporal electrode placements. The debate turned again; if 6 × seizure threshold is insufficient, might it become so by further increasing the dosage? The answer is contained in the article by McCall et al. [11] of which Sackeim is a coauthor. They showed that as dosage increased through 12 × seizure threshold, improvement also increased. However, this occurred at the expense of cognitive sideeffects which rose at a rate greater than the rate of clinical improvement.
What then may be concluded from this recent review of unilateral and bitemporal electrode placements? At least for the unipolar depressive disorders, bitemporal ECT remains the ‘gold standard’ offering as it has the capability of delivering a mean response rate of 87%. Although comparable data is not available, there appears to be no disagreement on bitemporal ECT also being the treatment of choice for severely ill patients suffering from a number of other psychiatric conditions [12]. Fortunately, and due to careful study, the literature is now clear on the use of unilateral ECT: it is dose sensitive. If it is to be used, it is to be used at substantially higher doses than previously considered. Grampians Psychiatric Services (GPS) has been using right unilateral ECT at doses 6 × seizure threshold over the past 2 years. We present a retrospective account of that experience.
Method
Grampians Psychiatric Services (GPS) is a publically funded, Area Mental Health Service (AMHS) that covers 48 000 2 km in Victoria, Australia with a catchment population of 203 000. Electroconvulsive therapy is administered thrice weekly using a Thymatron DGx device. At the discretion of the treating psychiatrist, one of three ECT options are selected: bitemporal; bifrontal; or D'elia 6 × RUL. Anaesthesia is produced by thiopentone and succinylcholine at a dose of 1–5 mg/kg and 0.5–1 mg/kg, respectively. Seizure threshold determinations, until an effective seizure is produced, are made by empirical titration at the first treatment session. Seizure adequacy requires an electroencephalograph (EEG) duration ≥ 30 s plus a motor (using the cuff method) seizure ≥ 25 s plus a postictal suppression index (PSI) ≥ 74% plus an EEG showing morphological regularity and high amplitude slow waves. Deterioration in any one of these four parameters prompts an increase in stimulus dosage. Following ECT, all patients receive continuation pharmacotherapy as clinically determined by their treating psychiatrist.
The records of all patients who received 6 × RUL ECT between July 2000 and June 2002 were identified and examined. Demographic details, seizure thresholds, mean electrical dosage received, seizure duration and postictal suppression indices were noted. The number of patients who required either a further increase in electrical dosage or change to a bilateral placement were also recorded. Response and cognitive side-effects were clinically assessed by the treating psychiatrist. An unequivocal response was defined as complete remission of symptoms and discharge. In addition, total number of treatments received, number and percentage of patients relapsing and time to relapse were noted.
Results
Results of the demographic details, seizure parameters and treatment response are shown in Table 2. Results will be described for the total of 21 patients followed by age-based programmes. Small numbers precluded meaningful statistical analysis.
Profile of patients receiving 6×RUL ECT
Demographic details
One hundred and twenty-three ECT patients received ECT at GPS between July 2000 and June 2002, 66 from the adult and 57 from the psychogeriatric (PGAT) programme. Of these, 21 (17%) received 6 × RUL ECT. Six (9%) were adult and 15 (25%) were from PGAT. The mean ages were 43 and 77 years, respectively. Eighteen (86%) of the 21 patients had previously been admitted with a mean number of 3.4 psychiatric admissions. Eighteen patients (86%) had previously received ECT, four adults and 16PGAT patients. Of the latter group, 13 (62%) had multiple medical problems that involved the cardiovascular, respiratory and/or endocrine systems. Of the 21 patients, 12 (57.2%) had a DSM-IV diagnosis of recurrent major depression; four (19%) had bipolar affective disorder, depressive phase; three (14.3%) had bipolar effective disorder, manic phase; and two (9.5%) had with a diagnosis of ‘psychosis not otherwise specified’.
Seizure parameters
The mean seizure threshold across all patients was 64.8 mC. It was higher in the elderly (50.4 mC for adults, 70.6 mC for PGAT) and for male patients (72.0: 61.2 mC for male: female). At this threshold, the mean EEG seizure length was 62.5 s (68.4 s for adults, 59.6 s for PGAT; 27 s for male although on the basis of one patient only, 74.9 s for female); the mean motor seizure was 40.2 s (28 s for adults, 42.2 s for PGAT; 28.8: 45.4 s for male: female) and the mean PSI was 73.4% (63.6% for adults, 76.1% for PGAT; 67.4: 76.8% for male: female). At treatment number 2, all patients received energy at 6 × seizure threshold with a mean energy dose of 388.8 mC (302.4 mC for adults and 423.6 mC for PGAT; 432: 369.2 for male: female). This resulted in a mean EEG seizure of 51.2 s (45.3 s for adult, 53.6 s for PGAT; 45.3: 54.2 s for male: female), a mean motor seizure of 31.2 s (29.2 s for adults, 32.5 s for PGAT; 28.6: 32.5 s for male: female) and a mean PSI of 85% (89% adult, 84% PGAT; 89: 83% for male: female). In order to sustain predetermined criteria for seizure adequacy, energy was increased in 15 (71%) patients (two adult and 13 PGAT; five male and 10 female). On average this occurred at treatment number 5 (range 3–11). The final mean seizure parameters included a seizure energy of 684 mC (479 mC for adults, 766 mC for PGAT; 618: 716 mC for male: female). This resulted in an EEG seizure of 45 s (41.5 s for adults, 46.3 s for PGAT; 42.3: 46.2 s for male: female), a motor seizure of 28 s (16 s for adults, 34 s for PGAT; 25: 29 s for male: female) and a PSI of 83% (82% for adults, 83% for PGAT; 83% for both male and female).
Treatment response
Eighty percent of patients had an unequivocal clinical response (100% for adults, 71% for PGAT; 83% for males and 79% for females). Including the first titration session, this took a mean number of 7.0 treatments (6.8 for adults and 7.1 for PGAT). An 85 year-old man who had a modest response to treatment became medically unwell as a result of which further ECT was cancelled. One 80 year-old woman who only had a modest response after six treatments was changed, and responded to two bitemporal treatments. Four patients from the PGAT programme had cognitive side-effects recorded in their clinical file. For one, continuation ECT was stopped, one was changed from a thrice to twice weekly regimen, one had a pre-existing vascular dementia and one had developed a recurring urinary tract infection at the time of documentation. Eleven (52%) patients (five adult and six from PGAT) relapsed within a mean of 6.3 months. The range was 2–13 months. One 73 yearold woman, who had become confused during an index course of 6 × RUL, responded to a second course 7 months later. She has remained well for 12 months. Five patients were subsequently treated with, and responded to, bilateral ECT (4 × bitemporal, 1 × bifrontal). The remaining five patients were managed pharmacologically. Three PGAT patients (two male, one female) died from medical illnesses, 16, 18 and 12 months after their last ECT treatment.
Discussion
6 × RUL ECT is capable of matching the responsivity of bitemporal ECT while sparing cognition. This review was a 2 year examination of the first half of that claim in a publicly funded service. Seventeen percent of patients who underwent ECT between July 2000 and June 2002 received ECT via this electrode placement. This reflected the history of this service, namely a preference for bitemporal ECT. Of the 21 patients, seven were male and 15 female. Six were from the adult programme and 15 from PGAT with mean ages of 43 and 77, respectively. Importantly, and in contrast to the literature, 6 × RUL ECT was administered across a range of diagnostic disorders. The most common was recurrent major depression (57.2%), followed by bipolar affective disorder, depressive phase (19%), bipolar affective disorder, manic phase (14.3%) and psychotic disorder NOS (9.5%). The sample also reflected a severity of illness with 86% having been admitted previously, 62% of psychogeriatric patients having one or more medical problems and 86% of all patients having received ECT previously. As such they represent a population from the clinical field.
Abrams suggests that a seizure energy at or above 378 mC is necessary to achieve a response rate at or above 65% for unilateral electrode placement. The mean initial stimulus energy administered to our patients was 388.8 mC. Eighty percent of patients who received RUL ECT at this level had an unequivocal response – remission of symptoms and discharge. This occurred with a mean of 7.0 treatments which included the first titration session. Importantly, the responding group included two patients with mania and two with psychosis not otherwise specified. Details of these patients are described elsewhere [13]. In addition, both relapses and cognitive side-effects were also noted.
Our study has significant limitations. It was retrospective, diagnoses were assigned clinically, no rating scales were used and therefore response to treatment and cognitive side-effects were assessed by the judgement of the treating psychiatrists. Subtle impairments in cognition would have been missed.
What then may be concluded from this review? It would seem that 6 × RUL ECT administered at or above 388.8 mC energy is effective. In particular, it was effective for 21 patients who might have attended any public psychiatric service between 2000 and 2002. Obvious cognitive side-effects were noted in four patients (21%) as was a relapse rate of 52% of patients despite continuation pharmacotherapy. The results, namely unequivocal responsivity in 80% of patients, must be viewed cautiously. The debate will continue concerning the most useful placement. However, a useful approach may be to remain focused on the response rate. The secondary question then follows: how was the outcome achieved? The answer is likely to include both technique and patient selection. We support the literature that advocates RUL ECT should only be administered an effective dosage. As to patient selection, there is a body of evidence to suggest that a response rate above 80% with single figure treatment numbers should now be routine. If the technique is unable to achieve this result, then either that technique or the patients selected, need to change.
Acknowledgements
We thank staff at Grampians Psychiatric Services, the medical records staff at Ballarat Base Hospital and Queen Elizabeth Centre, Lesley Hopkins, Norma Worswick and Gemma Willcocks.