To maximize the benefits of ECT and minimize the risks, it is essential that
the patient's illness be correctly diagnosed, that ECT be administered only
for appropriate indications, and that the risks and adverse effects be
weighed against the risks of alternative treatments. Risks and adverse
effects of ECT can be divided into two categories: (1) Those medical
complications that can be substantially reduced by the use of appropriately
trained staff, best equipment, and best methods of administration and (2)
those side effects, such as spotty but persistent memory loss and transient
posttreatment confusion, that can be expected even when an optimal treatment
approach is used. In this report, we will focus on the risks still present
with adequate treatment techniques.

In the early days of ECT, mortality was a significant problem. The commonly
quoted overall mortality rate in the first few decades was 0.1 percent or 1
per 1,000. Over the years, safer methods of administration have been
developed, including the use of short-acting anesthetics, muscle relaxants,
and adequate oxygenation. Present mortality is very low. In the least
favorable recent series reported, there were 2.9 deaths per 10,000 patients,
another series, 4.5 deaths per 100,000 treatments were reported. Overall, the
risk is not different from that associated with the use of short-acting
barbiturate anesthetics. The risk of death from anesthesia, although very
small, is present and should be considered when evaluating the setting for
performing ECT. 

In the past, up to 40 percent of patients suffered from various
complications, the most common being vertebral compression fractures. With
present techniques, these risks have been virtually eliminated. In one recent
study of almost 25,000 treatments, a complication rate of 1 per 1,300 to
1,400 treatments was found. These included laryngospasm, circulatory
insufficiency, tooth damage, vertebral compression fractures, status
epilepticus, peripheral nerve palsy, skin burns, and prolonged apnea.

During the few minutes following the stimulus, profound and potentially
dangerous systemic changes occur. First, there may be transient hypotension
from bradycardia caused by central vagal stimulation. This may be followed by
sinus tachycardia and also sympathetic hyperactivity that leads to a rise in
blood pressure, a response that may be more severe in patients with essential
hypertension. Intracranial pressure increases during the seizure.
Additionally, cardiac arrhythmias during this time are not uncommon (but
usually subside without sequelae). Thus, certain patient groups that would be
adversely affected by these manifestations are at increased risk.

There are two categories of central nervous system effects: The immediate
consequences of the ECT seizure and the more enduring effects, both of which
are affected by the treatment course. Immediately after awakening from the
treatment, the patient experiences confusion, transient memory loss, and
headache. The time it takes to recover clear consciousness, which may be from
minutes to several hours, varies depending on individual differences in
response, the type of ECT administered, the spacing and number of treatments
given, and the age of the patient. 

The severity of this acute confusional state is greatest after bilateral sine
wave treatment and least when nondominant unilateral pulsed ECT is
administered. Severity also appears to be increased by longer seizure
duration, close spacing of the treatments, increasing dosage of electrical
stimulation, and each additional treatment.

Depressive disorders are characterized by cognitive deficits that may be
difficult to differentiate from those due to ECT. It is, however, well
established that ECT produces memory deficits. Deficits in memory function,
which have been demonstrated objectively and repeatedly, persist after the
termination of a normal course of ECT. Severity of the deficit is related to
the number of treatments, type of electrode placement, and nature of the
electric stimulus. Greater deficit occurs from bilateral than from unilateral
placement. Sine wave current has been found to impair memory more than pulsed

The ability to learn and retain new information is adversely affected for a
time following the administration of ECT; several weeks after its
termination, however, this ability typically returns to normal. There is also
objective evidence, based on neuropsychological testing, of loss of memory
for a few weeks surrounding the treatment; such objective tests have not
firmly established persistent or permanent deficits for a more extensive
period, particularly for unilateral ECT. However, research conducted as long
as 3 years after treatment has found that many patients report that their
memory was not as good as it was prior to the treatment. They report
particular difficulties for events that occurred on average 6 months before
ECT (retrograde amnesia) and on average 2 months after the treatment
(anterograde amnesia). Because there is also a wide difference in individual
perception of the memory deficit, the subjective loss can be extremely
distressing to some and of little concern to others.

There are other possible adverse effects from ECT. Some patients perceive ECT
as a terrifying experience; some regard it as an abusive invasion of personal
autonomy; some experience a sense of shame because of the social stigma they
associate with ECT; and some report extreme distress from persistent memory
deficits. The panel heard eloquent testimony of these attitudes from former
patients who had been treated with ECT. It is clear, however, that these
attitudes are not shared by all ECT patients. The panel also heard moving
testimony from former patients who regarded ECT as a wholly beneficial and
lifesaving experience. There are insufficient systematic studies to permit
any definitive assessment of the prevalence of these various perceptions
among ECT patients.

Numerous ECT studies have been conducted with animal models. Many of these
suffer from methodological shortcomings. In studies that have been controlled
for fixation artifacts, hypoxia, and other methodological problems, neuronal
cell death has not been detected. Cerebral vasospasm and alterations in
capillary permeability are of short duration and of insufficient magnitude to
lead to neuronal cell death. The precise mechanism of the anterograde and
retrograde memory deficit has not been established; it may represent
alterations in neuronal function that are not detectable with present
methods. Computerized axial tomography (CAT) studies of patients who have had
ECT are very preliminary and open to alternative interpretations. Definitive
studies of in vivo brain metabolism with positron emission tomography (PET)
and studies of tissue changes detectable by magnetic resonance imaging (MRI)
remain to be done.


Electroconvulsive Therapy. NIH Consens Statement 1985 Jun 10-12;5(11):1-23.