var fontSizeTitle='Font size'; var bigger='Bigger'; var reset='Reset'; var smaller='Smaller'; var biggerTitle='Increase size'; var resetTitle='Revert styles to default'; var smallerTitle='Decrease size';

ECT - How good, and for how long?

ECT - How good, and for how long?

ECT - How good, and for how long?

Date Created: 04/04/1999 Last Modified: 21/05/2001 Last Checked: 02/05/2002

ECT is without doubt a safe and effective treatment for serious depression.


What sort of depression responds to ECT? For how long do patients stay well afterwards?

The idea that particular 'endogenous' symptoms predict a specific and differential response to ECT is not sustained by the current evidence. It seems that with major, or 'endogenous', depression the same percentage of patients respond irrespective of the 'sub-type' of depression they have. ie the presence of retardation or melancholia, agitation or psychotic features does not increase the chances of response.

Sackeim's group conclude:-- "The findings cast doubt on the utility of these depression subtypes in predicting ECT response. ECT is a viable treatment option for patients with major depression regardless of the presence or absence of psychosis, retardation or agitation".

There seems to be a widely held myth that ECT is an 'ultimate measure'. Whether clinicians consciously think that or not they tend to behave as if it were true; so if a patient is still ill after ECT they often give up on drug treatment.

Many patients who fail to respond to ECT will subsequently achieve full remission if given one of the following regimes:--

  1. tranylcypromine (with lithium augmentation if needed)
  2. clomipramine (or amitriptyline), with lithium augmentation if needed)
  3. sertraline plus either nortriptyline, desipramine, or reboxetine (with lithium augmentation if needed)
  4. bupropion, by itself, and in combinations with either a selective serotonin reuptake inhibit or tranylcypromine
  5. high dose venlafaxine (with lithium augmentation if needed).

NB The above are examples from a range of possible regimes.

If there are no adequate psychiatric back-up services a competent physician may be able to supervise some of these regimes satisfactorily.

Without good drug treatment after ECT a great majority of patients will relapse within 1-3 months.

The practice of continuing the same drug which failed to improve the patient before ECT, as treatment after ECT, is common. It is both puzzling and illogical that so many specialists do this. It makes me wonder whether the side effects of ECT are not only on the patients memory, but also on the critical faculties of the clinician giving it.

Sackeim's group have published several papers in 2000/1 (below) that highlight some fundamental points.

  1. Overall rate of response to ECT was 60% (176 of 290) but 17 patients did not even maintain improvement long enough to be classed as "remitters". Of those who completed an ECT course 55% achieved illness remission by their predefined criteria (ie 159 out of 290 patients).
  2. Intention to treat (ITT) analysis for comparison with drug trials would be 316 patients started treatment, 159 achieved remission, giving a 50% success rate.

Bipolar Disorder (both type 1 and 2) had fewer ECTs and responded more rapidly than unipolar. The final proportion of responders was the same for unipolar and bipolar groups.

They were left with 84 patients who remitted with ECT and who further consented for a medication continuation study. There were three groups; nortriptyline alone (started n=27 / completed n=25), nortriptyline and lithium (n=28 / 23), and placebo (n=29 / 25). Note their figures are for completers of the 24 weeks follow up, an Intention to treat analysis would give worse figures.

  1. Those who received no drug treatment post ECT had a relapse rate of 84% (up to 24 weeks); 50% of these relapses were within the first 4 weeks post ECT; 70% within the first three months.
  2. Those who received nortriptyline post ECT had a relapse rate of 60% and for nortriptyline + lithium, 40 per cent. Eight (out of the total of 23 completers- there were 5 dropouts) relapsed with nortriptyline + lithium within 5 weeks post ECT; indeed only 1 of the 23 on nortriptyline and lithium relapsed after this 5 week post ECT cutoff. This suggests to me that had these patients been given better drug treatment sooner then less of them would have relapsed. After all we know, both from trials and good old clinical experience, that antidepressants take 4-6 weeks to really start working.

They stratified cases for randomisation for the continuation study by medication resistance prior to ECT. This was because adequacy of previously administered treatment predicts a lower ECT response rate. In other words, the 'better' the treatment already given the less likely ECT is to work. With the benefit of the retrospectoscope this may seem blindingly obvious and confirms clinical impressions. It is thus even more important that someone has produced more formal evidence supporting it.

The distinction between bipolar and unipolar had no predictive value in determining final ECT response rate.

So what are the lessons? In what way might we reasonably modify our practice?

  1. These studies illustrate what I think is a telling indictment of the level of therapeutic knowledge and practice in psychiatry: the standard of pharmacological treatment, not only prior to referral for ECT, but also after it, will benefit from a critical reassessment.
  2. The remission rate (on an Intention to treat basis) of at least 50% makes it an effective treatment by any standard. One might remember that the, usually lower, success rate of drugs is typically defined by the lesser standard of a 50% improvement in scores on a rating scale (eg HAM-D scores) which does not equate with true remission.
  3. Although ECT is a good treatment for severe depression there is no reason to lessen therapeutic resolve if it fails; patients may subsequently respond to appropriate regimes designed for refractory cases.
  4. Without adequate drug treatment ECT is but a temporary and all too transient remedy for the patients suffering and illness.
  5. Auditing of ECT practice demonstrates that even after many years of use and education it is often performed sub-optimally; I suggest those who have trouble using drugs to the best advantage are going to be more challenged still using, as they inevitably must, the combination.
  6. Until further evidence accrues one would have to go with the presumption that combined drug treatment (of one type or another) is advisable before / with a course of ECT in order to lessen the high incidence of early relapse.


1. Continuation pharmacotherapy in the prevention of relapse following electroconvulsive therapy: a randomized controlled trial. JAMA, 2001. 285(10): p. 1299-307.
2. ECT in bipolar and unipolar depression: differences in speed of response. Bipolar Disord, 2001. 3(2): p. 95-104.
3. A prospective, randomized, double-blind comparison of bilateral and right unilateral electroconvulsive therapy at different stimulus intensities. Arch Gen Psychiatry, 2000. 57(5): p. 425-34.
4 Who responds to electroconvulsive therapy? A comparison of effective and ineffective forms of treatment. British Journal of Psychiatry, 1996. 169(3): p. 322-8.