Karon Dawkins, MD

Electroconvulsive therapy (ECT) has been practiced since 1938 and is   one of the best-studied treatment modalities. Because of its long   history‚ it has been the comparator for a number of subsequent   therapies‚ including pharmacological and other somatic interventions.   Even with ECT‚ however‚ the response rate for treatment-refractory   patients is sobering‚ and the treatment is not without risks and adverse   effects.

 Still‚ it can be effective when other interventions have   failed.

Treatment-refractory major depression is the primary indication for   ECT‚ although ECT may have particular utility in psychotic depression‚   catatonia‚ and suicidality as well.2 The American Psychiatric   Association (APA) guidelines also list an indication for schizophrenia.

The Case Vignette illustrates a fairly typical case for ECT referral‚ and the patient is a reasonable candidate.

John‚ 55 years old‚ is referred for ECT evaluation. He has a long   history of recurrent major depression (including in childhood)‚ with   onset of the current episode 6 months earlier. He presents with   depressed mood‚ anxiety‚ middle insomnia‚ social isolation‚ marked   fatigue‚ significant anhedonia‚ and profound loss of interest and   motivation. His concentration‚ attention‚ and short-term memory are also   negatively affected. His appetite is normal and his weight is stable.   He denies suicidal ideation and past suicide attempts but acknowledges   thinking that his life is pointless. He denies psychotic or manic   symptoms. He is having problems at work and there is significant marital   discord at home.

John first sought treatment in his early 30s. He has taken a number   of antidepressants of different classes‚ some of which worked partially‚   until they lost their effectiveness or he had unacceptable adverse   effects. He desperately wants symptom relief‚ as soon as possible‚ but   is very concerned about cognitive adverse effects of ECT.

To suppress the electrically induced tonic-clonic motor seizure   activity‚ and thus minimize musculoskeletal complications‚ paralytic   agents are administered. Because conscious paralysis would be very   distressful‚ patients receive general anesthesia. ECT entails a series   of treatments‚ typically 6 to 12‚ to achieve response or remission.   Cognitive adverse effects are the major concern‚ especially for patients   and their families.

 Even with full remission‚ relapse rates remain   high‚ particularly for the nonpsychotic medication–resistant   population.

 Generally‚ once symptom remission is achieved‚ ECT is   frequently discontinued. However‚ some patients get continuation (to   prevent relapse) or maintenance (recurrence prophylaxis) ECT.

The decision to pursue ECT is based on a number of factors‚ including   “  diagnosis‚ type and severity of symptoms‚ treatment history‚   consideration of the anticipated risks and benefits of ECT and   alternative treatment options‚ and patient preference.”3 The principal   diagnostic indication is unipolar or bipolar depression. “  Policies and procedures should be developed to ensure proper informed   consent‚ including when‚ how‚ and from whom consent is to be obtained   and the nature and scope of information to be provided. . . . Informed   consent should be obtained from the patient except when the patient   lacks capacity to consent.”

There is evidence to suggest that a significant number of patients   referred for ECT have not had adequate medication trials.

 The   response rate for patients who are not medication-resistant is higher‚   up to 90% in one multisite study (log-linear analysis x2 = 6.82‚ df = 1‚   P = .009)‚ with site differences in the rates of patients classified as   medication-resistant‚ patient age‚ Hamilton depression scores‚ and   total number of ECT sessions.

The global rating of reliability did not differ among patients who   were classified as medication-resistant and having inadequate   pharmacotherapy before ECT.

 The FDA efficacy review estimated the   overall treatment effect to be 78%.10 For treatment-refractory patients‚   the ECT response rate may be about 50% to 60%.

Personality disorders comorbid with major depression are common‚ with   poorer outcomes with psychotherapy or pharmacotherapy. Patients with   these comorbid conditions are likely to be referred for ECT. One study   examined 139 patients with a primary diagnosis of unipolar major   depression and scores of at least 20 on the 24-item Hamilton Depression   Rating Scale; personality disorders were assessed with a structured   interview. Outcomes were compared in patients with no personality   disorder‚ patients with borderline personality disorder‚ and patients   with personality disorders other than borderline personality disorder.   Only 22% of patients with borderline personality disorder met criteria   for remission‚ compared with 56% of patients with personality disorders   other than borderline personality disorder and 70% of patients with no   personality disorder.

Findings from the study indicate that the diagnosis of borderline   person­ality disorder is a predictor of poor response to ECT (x2 =   11.63‚ df = 1‚ P = .001) and has potentially significant implications   for the selection of candidates for ECT.

 Still‚ the APA guidelines do   not suggest a prohibition of ECT in patients with comorbid borderline   personality disorder. There is a similar stance for patients with   comorbid dysthymia‚ because ECT is rarely used for dysthymia alone.

ECT can be rapidly effective in mania and catatonia‚ and it can be an   early consideration for depressions with psychotic features. Although   antipsychotics continue to be the first-line intervention‚ ECT can be   considered in patients with medication-resistant schizophrenia. However‚   the probability of significant improvement may be low. There are   special considerations for children and adolescents (additional   evaluation from a child and adolescent psychiatrist is recommended) and   for pregnant women (procedural considerations and additional monitoring   are needed).

Unfortunately‚ some practitioners see ECT as a treatment of last   resort‚ whatever the presenting diagnosis. This kind of thinking on the   part of referring psychiatrists and patients is discouraged. There are   relative contraindications for ECT‚ but few absolute ones. Many patients   present with comorbid psychiatric disorders for which ECT is not   indicated. There are comorbid medical conditions that also affect the   risk to benefit ratio.

Electrically induced seizures can be associated with transient   cardiovascular changes (asytole‚ bradycardia‚ tachycardia‚ hypertension‚   increased myocardial oxygen demand)‚ increased intracranial pressure‚   increased intraocular pressure‚ release of catecholamines‚ and other   effects that can exacerbate preexisting conditions. These conditions   require optimization of their management and/or procedural changes to   minimize morbidity and mortality.

Precautions must be taken in patients with unstable or severe   cardiovascular conditions (eg‚ recent myocardial infarction‚ unstable   angina‚ con­gestive heart failure‚ uncontrolled hypertension)‚ aneurysm‚   increased intracranial pressure‚ recent cerebral infarction‚ or   pulmonary conditions (eg‚ chronic obstructive pulmonary disease‚ asthma‚   pneumonia)‚ and in patients who fall into the American Society of   Anesthesiologists rating level 4 or 5.3

The mortality rate has fallen over the decades of ECT practice.12 The   number of deaths resulting from ECT is estimated to be 1 per 10‚000   patients‚ or 1 per 80‚000 treatments.

 The Veterans Affairs (VA)   National Center for Patient Safety database estimate is less than 1   death per 73‚400 treatments (or less than 1 per 14‚000 patients). No   associated deaths were reported in any VA hospital between 1999 and   2010.

 In California from 1984 to 1994 and in Texas from 1993 to 1998‚   there were less than 2 deaths per 100‚000 treatments.

The mainstays for the reduction of adverse effects include electrode   placement‚ stimulus energy‚ waveform‚ and frequency of treatment (

The two best-studied placements are right unilateral (RUL) and   bilateral (BL‚ also known as bitemporal and bifrontotemporal)   placements. RUL ECT has fewer cognitive adverse effects than BL ECT;   however‚ unless it is administered well above seizure threshold‚ RUL ECT   is less effective (double-blind randomized trial‚ P = .054).

   Suprathreshold RUL ECT can approach the efficacy of BL ECT. In one   study‚ the response rates of both were 65%.9 In another study‚ in which   patients were randomized to titrated moderate suprathreshold or fixed   high-dose RUL ECT‚ the response rates were 39% and 67%‚ respectively (x2   = 5.6‚ P = .02).

ECT practitioners must often choose between maximizing effectiveness   and minimizing adverse effects. This prompts some practitioners to use   BL placement‚ frequently at high stimulus energy‚ almost exclusively.   Others start with RUL and switch to BL placement if there is a lack of   response‚ generally after 4 to 6 treatments.19 The APA guidelines   suggest reassessing the indication for continued ECT and making a change   in implementation after 6 to 10 treatments if the patient has slow or   minimal clinical improvement. These changes could include switching   electrode placement from RUL to BL‚ increasing stimulus energy‚ and   augmentation strategies.

Semkovska and colleagues20 performed a meta-analysis (with an effort   to control for publication bias) of 39 studies (1415 patients) to   clarify the advantages of RUL relative to BL‚ stimulus intensity‚ and   time interval between final assessment and cognitive assessment. They   found that cog­nitive differences associated with electrode placement   and stimulus intensity were limited to the first 3 days after unilateral   treatment: effect sizes ranged from –1.10 to –0.21 (95% confidence   interval [CI]‚ –1.53 to –0.67‚ –0.40 to 0.01‚ respectively).

Within this time frame‚ there were fewer cognitive disturbances   compared with BL placement and more compared with higher stimulus   intensity. After this subacute period‚ there were no significant   differences between RUL and BL and no remaining significant relationship   between stimulus intensity and cognitive performance after RUL ECT. The   time interval after final treatment was predictive of continuous   improvement in anterograde episodic memory‚ autobiographical mem­ory‚   and executive functioning.

Left unilateral ECT is considered an alternative when RUL is   contraindicated‚ typically when there is a skull defect. It is also a   consideration for left-handed patients‚ in an effort to avoid   stimulation of language centers. However‚ the lateralization of language   centers for an estimated 70% of this group of patients is similar to   that for right-handed patients; 15% have bilateral representation; and   only 15% are right dominant for language. Handedness may also be an   unreliable predictor. While most practitioners use RUL placement   regardless of handedness‚ an alternative strategy is to switch   laterality over the initial sessions to gauge which is associated with   fewer cognitive adverse effects.

Bifrontal ECT has been investigated as a way to maximize response   without increasing cognitive adverse effects (by avoiding stimulation   over the temporal region) and directing treatment toward circuits in the   frontal cortex that may play an important role in depression.

 bifrontal placement was initially sidelined   because benefits were intermediate; the electrodes were too close   together and there were skin burns secondary to shunting. Interest was   renewed in the 1990s with a number of published clinical trials and   apparent widespread use.

One study showed bifrontal ECT had the best antidepressant (P <   .01) and cognitive (P < .05) outcomes and required the fewest   treatments‚ compared with BL and RUL ECT. However‚ the differential   cognitive effects were not evident 3 months after the last ECT.

 randomized 48 patients to either BL or   bifrontal ECT (stimulus titration at first treatment‚ subsequent   treatments 1.5 times seizure threshold). All 24 patients in the   bifrontal group and 23 of the 24 patients in the BL group met remission   criteria (Hamilton Depression scale less than 10‚ first Hamilton   Depression scale less than 3‚ or Clinical Global Impression improvement   score less than 3) by the 12th treatment. The BL group had a significant   decrease in the mini-mental state score that was considered to be   clinically significant‚ albeit small (P = .03).

The researchers attributed efficacy and improved cognitive   performance to avoidance of stimulation over the temporal areas while   retaining bilateral stimulation. They also speculated that bifrontal   placement may have the advantage of fewer dental complications‚ since   the masseter muscles may receive less direct stimulation.

In a double-blind controlled trial‚ Eschweiler and colleagues

   randomized 92 patients to either 6 RUL (2.5 times seizure threshold) or 6   BL (1.5 times seizure threshold) sessions. While there was no   difference in the response rate and only a minor cognitive difference   between the two groups‚ only 12 of the 46 patients (26%) in each group   responded (odds ratio tested by Fisher’s exact test and used to   calculate 95% CI; CI‚ 18-36; odds ratio‚ 0.35 to 2.8). The reasons for   this low response rate were unclear‚ although this may have been a   function of selection bias of a particularly treatment-resistant   population late in the course of their illness.

 report the results of a large‚ double-blind‚   controlled trial of 230 patients who were randomized to bifrontal   placement at 1.5 times seizure threshold‚ BL at 1.5 times seizure   threshold‚ or RUL at 6 times seizure threshold. Antidepressant outcomes   were comparable‚ with remission rates of 55% for RUL (95% CI‚ 43-66)‚   61% for bifrontal (95% CI‚ 50-71)‚ and 64% for BL (95% CI‚ 53-75). BL   placement resulted in a faster reduction of symptoms‚ but there were few   cognitive differences among the 3 treatment arms. The researchers note   that BL placement could be used preferentially in urgent clinical   situations. RUL placement can be effective at high stimulus intensity   and could be the initial choice if there is concern about retrograde   amnesia. Bifrontal placement did not have a cognitive advantage‚ nor was   RUL consistently better than BL placement on cognitive measures.

One potential advantage of bifrontal ECT is a lower risk of   bradycardia and asystole compared with BL and standard pulse width RUL   ECT. Bifrontal ECT could be considered for patients with arrhythmia.   Ultrabrief pulse width RUL ECT is associated with less bradycardia (P =   .004) and asystole (P = .001) than standard pulse width RUL ECT‚   although the data were collected in the context of clinical service‚ via   printed ECGs‚ and were not systematically controlled.

Left anterior‚ right temporal (LART‚ or left frontal‚ right temporal)   placement seeks to decrease cognitive adverse effects by minimizing the   impact on the temporal/hippocampal areas. This placement is less well   studied‚ although in small open and double-blind pilot trials‚ LART had   comparable efficacy and fewer cognitive adverse effects.

 It is not   known how widely this modality is used. Some practitioners use it as an   intermediate step between RUL and BL ECT.

It is well established that the stimulus energy must be above seizure   threshold (suprathreshold) to maximize treatment efficacy. Seizure   threshold-the minimum amount of energy to elicit a generalized   seizure-is highly variable‚ up to 40-fold in some estimates. Men‚ the   elderly‚ and BL placement tend to have a higher seizure threshold.   Seizure threshold is also inconstant because ECT has an anticonvulsant   effect that raises seizure threshold (paired t‚ P < .001).

   Treatment protocols vary from titration procedures to fixed-dose   strategies because there are different schools of thought about how much   above seizure threshold a stim­ulus should be.

Some practitioners‚ during the first treatment session‚ determine   seizure threshold by delivering increasingly larger stimuli until a   generalized seizure of adequate length (at least 15 seconds as measured   by motor activity or electroencephalogram) is obtained. There are a   number of methodologies by which this stimulus titration is done.   Electrode placement‚ sex‚ age‚ dosages of anesthetics‚ and concomitant   medications can be used to predict seizure threshold.

 However‚ seizure   threshold typically increases over the series. Proxies such as seizure   length‚ seizure quality‚ and clinical response are signs that   stimulation is no longer suprathreshold and stimulus intensity should be   increased.

Seizure threshold increases with age; thus‚ a method that sets the   stimulus energy based on the patient’s age can be used.

 Because this   might cause patients to be stimulated at higher intensity than   necessary‚ some practitioners use the “  half age” method.

 The APA guidelines recommend that unilateral   treatments be moderately to markedly suprathreshold (2.5 to 6 times   seizure threshold) and that BL treatments be moderately suprathreshold   (1.5 to 2.5 times seizure threshold) because the efficacy of BL   placement is less sensitive to suprathreshold dosing than is unilateral   placement.

There have been calls for higher-output machines in the United States   to achieve the dosages required for effective unilateral ECT.

   However‚ whatever the placement‚ higher-energy stimulation is associated   with increased cognitive adverse effects. Manipulation of pulse width   may allow for more efficiency in inducing seizures‚ with fewer cognitive   adverse effects.

The first ECT apparatus used sine waves‚ which were inefficient in   inducing seizures and were associated with significant cognitive adverse   effects.

 Brief pulse has fewer cognitive adverse effects‚ without a   reduction in efficacy. As a result‚ the APA guidelines state that the   continued use of sine wave stimulation is not justified.

 However‚ the   ideal pulse width continues to be studied.

The chronaxie‚ or optimal width for neuronal polarization‚ is   estimated to be 0.1 to 0.2 millisecond.

 Until recently‚ most   devices had a pulse width range of 0.5 to 2 milliseconds. Ultrabrief   pulse width of less than 0.5 millisecond had been considered earlier in   ECT history but was not pursued because of concerns about efficacy. Now   there is resurgent interest.

Using a double-masked study‚ Sackeim and colleagues

 sought to   determine whether ultrabrief pulse and RUL placement would minimize   cognitive adverse effects but not at the expense of efficacy. Patients   were randomized to RUL ECT at 6 times seizure threshold or BL ECT at 2.5   times seizure threshold‚ using brief pulse (1.5 milliseconds) or   ultrabrief pulse (0.3 millisecond). The remission rate was 73% for   ultrabrief pulse RUL ECT‚ 65% for standard brief pulse BL ECT‚ 59% for   standard brief pulse RUL ECT‚ and 35% for ultrabrief pulse BL ECT (all P   < .05 after covariate adjustment). The ultrabrief pulse RUL group   had less severe cognitive adverse effects (P < .001).

The researchers concluded that ultrabrief stimulus markedly reduces   cognitive adverse effects (acute‚ short-term‚ and long-term)‚ and   efficacy can be preserved with markedly suprathreshold RUL ECT-efficacy   comparable to high-dose BL ECT. There was speculation that the efficacy   of BL ECT may have been improved if stimulus intensity had been greater   than 2.5 times the initial seizure threshold‚ although cognitive adverse   effects may have also increased.

 compared 0.25-millisecond   ultrabrief pulse with 0.5-millisecond brief pulse in BL ECT. In their   study‚ where response was defined as 50% or greater decrease in Hamilton   depression score and re­mission as a score of less than 7‚ 73.6% of   ultrabrief pulse recipients were responders (42.1% remitters) and 75.6%   of brief pulse recipients were responders (45.6% remitters); no   significant difference was found between the two groups (P = .947). The   researchers attributed different outcomes to study design (retrospective   vs prospective)‚ patient population (unipolar depression only vs both   bipolar and unipolar depression‚ their sample perhaps more severely   depressed)‚ prior history of ECT (their sample was ECT-naive)‚ and the   differences in the comparator pulse widths (0.25 and 0.5 millisecond vs   0.3 and 1.5 milliseconds).

 compared ultrabrief (0.3 millisecond)   bifrontal ECT at 1.5 times seizure threshold with RUL ECT at 6 times   seizure threshold. Of the 64 patients‚ 78% were responders (n = 50)‚ and   65% (n = 42) were remitters. There was no significant difference in   response and remission between the two groups. Not only was there no   deterioration in neuropsychological measures (attention‚ executive   function/working memory‚ anterograde episodic memory‚ episodic   autobiographical memory‚ and subjective memory)‚ but patients assessed   their memory as improved and there were no significant differences   between groups.

The authors caution that ultrabrief ECT may require more sessions   (and therefore may produce more cognitive adverse effects)‚ that   perceived memory improvement may be a function of improved depression‚   and that only 72.8% of their sample completed the post-ECT battery of   neuropsychological tests.

 compared ultrabrief pulse (0.3 millisecond) at 6   times seizure threshold with brief pulse (1.0 millisecond) at 5 times   seizure threshold in RUL placement in a study limited by its   naturalistic‚ nonrandomized design. The remission rate was lower than   that found in the Sackeim40 study‚ but the difference was not   significant. If those patients who completed treatment with RUL ECT only   (not switched to BL placement) are considered‚ the response/remission   rates were 97%/61% for the ultrabrief pulse RUL group and 79%/57% for   the brief pulse RUL group.

Articles

With ECT‚ the response rate for treatment-refractory patients is sobering‚ and the treatment is not without risks and adverse effects.

Karon Dawkins, MD

Articles

Refinements in ECT Techniques