Electroconvulsive Therapy: Side Effects
The risk of death with ECT is very low: around 1 per 10,000 patients. This rate is comparable to the rate that would be expected from a series of brief anesthetic procedures alone. Most deaths occur in high-risk cases, which are therefore associated with a greater mortality rate. The most common cause of death with ECT is cardiovascular decompensation, and other causes include prolonged apnea, status epilepticus, and cerebral herniation (e.g., in unrecognized cases of brain tumor).
The systemic morbidity associated with ECT is related to the known physiological effects of induced seizures in a setting of general anesthesia and muscular relaxation. Both sympathetic and parasympathetic discharges occur and produce transient disturbances in cardiac rate and rhythm and elevated blood pressure. Although they are most often benign, these changes are rarely associated with more pathological arrhythmias or cardiac ischemia. Cardiovascular complications are most likely to occur in patients whose cardiac status is already severely compromised. Risks from these complications are increased in such persons, and pharmacological prophylaxis may be indicated (American Psychiatric Association Committee on ECT 2000; Zielinski et al. 1993). Prolonged apnea and status epilepticus are rare events, although resources to manage their occurrence should be available. Bone fractures and other musculoskeletal injuries are prevented by adequate muscular relaxation, whereas injuries to teeth or soft tissues of the mouth are minimized by use of an oral protective device during the application of the electrical stimulus. More common systemic side effects include headache, muscle pain, and nausea, all of which are usually mild and can be managed supportively, with attention to appropriate pharmacological prophylaxis, if indicated (American Psychiatric Association Committee on ECT 2000).
Amnesia is the most commonly discussed side effect of ECT. Many patients receiving ECT experience difficulties with memory function that increase over a course of treatments and begin to diminish as soon as the ECT is stopped. However, it should also be noted that less often memory function improves following ECT. This tends to occur in persons who have depression-related cognitive deficits or even frank pseudodementia, both of which are typically relieved by the ECT. Memory effects are more prominent and last longer in patients with preexisting cerebral disease and in those who receive bilateral electrode placement or have a large number of treatments in an index ECT series.
When memory deficits occur with ECT, they are of two types. The most prominent of these is retrograde amnesia, that is, difficulty in remembering information learned before the ECT course. The severity of this deficit, when present, is greatest for more recent memories, particularly those occurring during the months before the ECT. The persistence of retrograde amnesia after ECT has been a cause of concern. Based on results of objective testing, these losses disappear over a period of weeks to months, except for some material covering more recent memories, particularly those during the period of treatment with ECT and the preceding months. The specific percentage of patients who display persistent retrograde amnesia on objective testing following ECT is unknown, but patient surveys suggest that this may occur in a sizable minority.
The second type of memory deficit with ECT is anterograde amnesia, that is, difficulty in retaining newly learned information. This problem is most severe during the ECT course but typically disappears within a matter of days to weeks following completion of treatment. The likelihood, severity, and persistence of ECT-induced amnesia are influenced by the type of ECT administered, particularly stimulus electrode placement and stimulus dosing. If amnesia during the ECT course is severe, it can be managed by increasing the interval between treatments, changing to a different ECT type (if applicable), or, if necessary, aborting the treatment course altogether. There is an extensive history of attempts to ameliorate ECT-induced amnesia pharmacologically (e.g., by using nootropics, hormones, stimulants, and peptides), although the findings have thus far not been replicable (Krueger et al. 1992).
Despite the findings described above, complaints of more extensive or severe, long-lasting memory deficits have also been reported (American Psychiatric Association Committee on ECT 2000; Freeman and Kendell 1986; Sackeim 1992). The incidence of such complaints is not precisely known but is much lower than that observed with those experiencing the more typical effects described above. The etiology of these complaints is unclear, and in many cases either the patients perform normally on objective memory testing (although it is difficult to test for retrograde amnesia); the complaint has nonphysiological characteristics, such as a retrograde amnesia without a temporal gradient; or other etiological factors are involved (e.g., medication effects). Although some of these complaints may be biologically based, psychological factors may also be present, including an increased sensitivity to the phenomenon of normal forgetting (following the experience of a transient organic amnesic effect), the development of a conversion-type syndrome (analogous to the finding that hysterical seizures are most common in patients with epilepsy), the presence or recurrence of residual depressive symptoms, and the effects of secondary gains.
A period of postictal confusion lasting from 10 minutes to more than 1 hour after the induced seizure (depending on factors such as ECT type) is common. The occurrence of interictal delirium, however, is much less common. It is most likely to appear in patients with preexisting cerebral impairment (including Parkinson’s disease).
Structural Brain Changes
Available evidence indicates that structural brain changes do not occur with ECT (Coffey et al. 1991; Devanand et al. 1994). This evidence includes contemporary experimental studies of metabolic, neurochemical, and neuropathological changes associated with both electrically and chemically induced seizures, as well as prospective long-term follow-up of brain magnetic resonance imaging in humans. This work has been important, given the existence of considerable misinformation, based in part on very early, methodologically deficient studies.
Revision date: July 6, 2011
Last revised: by Jorge P. Ribeiro, MD