Why is this? Several factors likely explain this rather dramatic difference. First, the patients given ECT in the early decades of its use most likely suffered with a diverse range of dysphoric states (e.g., patients with “neurotic” depression, dysthymia, or personality disorders), while modern research has been limited to relatively homogeneous samples of patients with major depression defined according to strict research criteria.3 When you try to correlate a putative predictive variable such as presence vs. absence of melancholia with an outcome variable such as reduction in depression ratings, the less variability there is on the predictive variable, the less strong the correlation will be.
Another possible factor accounting for the lack of predictability is a broadened concept of melancholia. According to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV),4 it is possible to have melancholic depression without having weight loss, psychomotor retardation, or excessive guilt—3 signs classically thought to be inherent in the melancholic syndrome. If melancholia is defined more narrowly, to include requirements for weight loss and psychomotor change, then it is probably more likely to correlate with ECT response.
A final potential reason for lack of predictability is the method of ascertaining melancholia. In modern samples, usually the Hamilton Rating Scale for Depression is used.5 Such a scale is administered in a brief interview by a research clinician not familiar with any other aspects of the patient’s mental status. In contrast, in some early studies,2 melancholic signs were ascertained by complete psychiatric history and evaluations as done in clinical practice. Such methods are likely to yield more reliable data on weight loss, guilty ideations, psychomotor activity, and other signs than is a 15-to 20-minute interview conducted by a research technician.
Along these lines, Hickie et al,6 utilizing a thorough evaluation of psychomotor activity before ECT, found that psychomotor retardation robustly predicted positive ECT response. Their scale utilized numerous items assessing agitation and retardation, and required a longer period of assessment than did the Hamilton scale; the latter scale has only 2 items for psychomotor activity, each one a global assessment of agitation or retardation.
So do you consider ECT for patients with melancholic features? Yes. Patients with classic melancholic features such as weight loss, pronounced guilt, and especially psychomotor retardation stand an excellent chance of substantial relief with a course of ECT. Additionally, patients with nonmelancholic depression have good response rates with ECT.
WHAT TO TELL YOUR ECT PATIENTS ABOUT MEMORY DISTURBANCE
Patients and their families are frequently concerned about the effects of ECT on memory. The 3 types of memory disturbances to discuss are:
- Post-treatment confusion and disorientation. This state usually lasts from a few minutes to several hours or, in the case of some elderly patients after receiving numerous treatments, several days. This state is always reversible.
- Anterograde amnesia. This is the inability to recall newly learned information during and up to a few weeks after the course of treatments. During this time, any information given to the ECT patient may not be remembered. Important strategies are to write down instructions and make sure that family members are informed of the need to repeat things and monitor the patient if an outpatient. Fortunately, antero-grade amnesia is also reversible.
- Retrograde amnesia. This refers to the forgetting of personal life events and general knowledge about the world. Usually, the events and knowledge that are “wiped out” by the treatments are those from up to a few months before the treatments begin to about a month after the treatments are done. Even more remote memories may be forgotten as well. Unlike the other types of ECT-induced memory impairment, retrograde amnesia may be permanent.
Catatonic features It has been known for decades that catatonic features, regardless of etiology, respond robustly and often quickly to a course of ECT.1,7 But in recent years, the literature has documented the high rates of efficacy of benzodiazepines, usually lorazepam, in the initial treatment of catatonic signs such as mutism, stupor, waxy flexibility, posturing, stereotypies, and rigidity.
Bush et al8 treated 21 acutely catatonic patients, who were so diagnosed according to a standardized catatonia rating scale, with parenteral and oral lorazepam at doses up to 8 mg/d. Sixteen responded dramatically, usually within a day or so. Four of the lorazepam nonresponders were given ECT with excellent results, not only for the catatonic signs but also for other underlying psychopathological features.
Ungvari et al9 treated 18 catatonic patients with either lorazepam or diazepam; all patients had some degree of improvement after several days, but 9 of the 18 exhibited insufficient response. For these, ECT was administered with excellent resolution of the psychopathology, including catatonia.