Chapter 5. Adverse Effects

5.1. Medical Complications

Precise rates of mortality attributable to ECT are difficult to determine due to methodological issues intrinsic to studies of medical mortality, such as uncertainty as to cause of death, time frame for linking death to ECT, and variability in reporting requirements. The mortality attributed to ECT is estimated to be approximately the same as that associated with minor surgery (McCabe 1985 Warner et al. 1993; Brand et al. 1994; Badrinath et al. 1995: Hall et al. 1997). Published estimates from large and diverse patient series over several decades report up to 4 deaths per 100,000 treatments (Heshe and Roeder, 1976; Fink, 1979; Weiner 1979; Babigian and Guttmacher, 1984; Crowe, 1984; Kramer, 1985: Abrams 1997b; Reid et al. 1998). Despite the frequent use of ECT in patients with significant medical complications and in the elderly (Sackeim 1993, 1998; Weiner et al. in press), rates of mortality appear to have decreased in recent years. A reasonable current estimate is that the rate of ECT-related mortality is 1 per 10,000 patients. This rate may be higher in patients with severe medical conditions. The rate of significant morbidity and mortality is believed to be lower with ECT than with treatment with some types of antidepressant medication (e.g., tricyclics) (Sackeim 1998). There is also evidence from longitudinal follow-up studies that mortality rates following hospitalization are lower among depressed patients who received ECT than patients who received alternative forms of treatment or no treatment (Avery and Winokur, 1976; Philibert et al. 1995)

When mortality occurs with ECT, it typically happens immediately following the seizure or during the postictal recovery period. Cardiovascular complications are the leading cause of death and of significant morbidity (Pitts 1982; Burke et al. 1987; Welch and Drop 1989; Zielinski et al. 1993; Rice et al. 1994). Despite the short-lived increases in cerebral blood flow and intracranial pressure, cerebrovascular complications are notably rare (Hsiao et al. 1987). Given the high rate of cardiac arrhythmias in the immediate postictal period, the majority of which are benign and resolve spontaneously, ECG should be monitored during and immediately following the procedure (see Section 11.8) and patients should not be taken to the recovery area until there is resolution of significant arrhythmias. Vital signs (pulse, systolic and diastolic pressure) should be stable prior to the patient's leaving the recovery area (Section 11.10). Patients with pre-existing cardiac illness are at greater risk for post-ECT cardiac complications (Prudic et al. 1987; Zielinski et al. 1993; Rice et al. 1994). Indeed, there is evidence that the type of pre-existing cardiac disease predicts the type of complication that may be encountered following ECT. For example, ventricular arrythmias are more common in patients with pre-existing ventricular abnormalities than in patients with ischemic heart disease (Zielinski et al. 1993). Management of cardiac complications is discussed in Chapter 11.

Two other possible sources of morbidity are prolonged seizures and tardive seizures (Weiner et al. 1980a). Management of prolonged seizures is described in Section 11.9. Failure to terminate seizures within a period of 3 to 5 minutes may increase postictal confusion and amnesia. Inadequate oxygenation during prolonged seizures increases the risk of hypoxia and cerebral dysfunction, as well as cardiovascular complications. In animal studies, seizure activity that is sustained for periods exceeding 30-60 minutes, regardless of steps taken to maintain appropriate levels of blood gases, is associated with an increased risk of structural brain damage and cardiovascular and cardiopulmonary complications (Meldrum et al. 1974; Ingvar 1986; Meldrum 1986; Siesjo et al. 1986; O'Connell et al. 1988; Devanand et al. 1994).

Prolonged seizures and status epilepticus may be more likely in patients receiving medications that lower seizure threshold or interfere with seizure termination (e.g. theophylline, even at therapeutic levels) (Peters et al. 1984; Devanand et al. 1988a; Abrams, 1997a), in patients receiving concomitant lithium therapy (Weiner et al. 1980b), in patients with pre-existing electrolyte imbalance (Finlayson et al. 1989), and with the repeated induction of seizures within the same treatment session (e.g., multiple monitored ECT) (Strain-and Bidder 1971, Maletzky 1981).

There has been concern as to whether the rate of spontaneous seizures is increased following the course of ECT (Assael et al. 1967; Devinsky and Duchowny 1983). The evidence indicates, however, that such events are extremely rare and probably do not differ from population base rates (Blackwood et al. 1980; Small et al. 1981). There are no data concerning rates of tardive seizures, i.e., seizures that occur following termination of the ECT-induced seizure, but experience indicates that these are also rare events. As noted in Section 11.9, prolonged or tardive seizures occurring during the immediate postictal period are often not accompanied by motor manifestations, underscoring the need for EEG seizure monitoring (Rao et al. 1993). Nonconvulsive status epilepticus may also occur in the interictal period, with an abrupt onset of delirium, unresponsiveness, and/or agitation as distinguishing clinical features (Grogan et al. 1995). Cessation of EEG abnormalities and improved cognitive function following short-acting anticonvulsant treatment (e.g. intravenous lorazepam or diazepam) may prove diagnostic (Weiner and Krystal, 1993).

Prolonged postictal apnea is a rare event that occurs primarily in patients with a pseudocholinesterase deficiency resulting in slow metabolism of succinylcholine (Packman et al. 1978). Maintaining adequate oxygenation is critical in instances of prolonged apnea, which will usually resolve spontaneously within 30 to 60 minutes. When prolonged apnea is encountered, it is helpful to obtain a dibuciane number assay or a pseudocholinesterase level prior to the next treatment in order to establish etiology. At subsequent treatments, either a very low dose of succinylcholine may be used or a non-depolarizing muscle relaxant, such as atracurium, may be substituted (Hickey et al. 1987; Hicks, 1987; Stack et al. 1988; Kramer and Afrasiabi 1991; Lui et al. 1993).

To some extent, medical adverse events can be anticipated. Whenever possible, the risks of such events should be minimized by optimization of the patient's medical condition prior to ECT and/or modifications in ECT procedures. Patients with preexisting cardiac illness, compromised pulmonary status, a history of CNS insult, or medical complications following prior courses of anesthesia or ECT are especially likely to be at increased risk (Weiner and Coffey 1988; Zieliniski et al. 1993). ECT psychiatrists should review the medical work-up and history of prospective ECT patients (see Chapter 6). Specialist consultations or additional laboratory studies may be called for, as well as changes in medication regimens. In spite of careful pre-ECT evaluation, medical complications may arise which have not been anticipated. ECT facilities should be staffed with personnel prepared to manage potential clinical emergencies and should be equipped accordingly (see Chapters 9 and 10). Examples of these events include cardiovascular complications (such as cardiac arrest, arrythmias, ischemia, hyper- and hypotension), prolonged apnea, and prolonged or tardive seizures and status epilepticus.

Major adverse events that occur during or soon after the ECT course should be documented in the patient's medical record. The steps taken to manage the event, including specialist consultation, use of additional procedures, and administration of medications, should likewise be documented. As cardiovascular complications are the most likely source of significant adverse events and are seen most frequently in the immediate post-ECT period, the treatment team should be capable of managing the major classes of cardiovascular complications. A set of predetermined procedures for dealing with instances of prolonged or tardive seizures and status epilepticus are helpful.

5.2. Systemic Side Effects

Headache is a common side effect of ECT and is observed in as many as 45% of patients during and shortly following the postictal recovery period (Devanand et al. 1995; Freeman and Kendell 1980; Gomez 1975; Sackeim et al. 1987d: Tubi et al. 1993; Weiner et al. 1994). However, the precise incidence of postECT headache is difficult to determine due to methodological issues such as the high baseline (preECT) occurrence of headache in patients with depression, the potential effects of concurrent medication or medication withdrawal, and differences between studies in the assessment of headache. PostECT headache appears to be particularly common in younger patients (Devanand et al. 1995) and especially in children and adolescents (Rey and Walter 1997; Walter and Rey 1997) It is not known whether pre-existing headache syndromes (e.g., migraine) increase the risk of postECT headache, but ECT may exacerbate a previous headache condition (Weiner et al. 1994). The occurrence of postECT headache does not appear to be related to stimulus electrode placement (at least bifrontotemporal vs. right unilateral) (Fleminger et al. 1970; Sackeim et al. 1987d; Tubi et al. 1993; Devanand et al. 1995), stimulus dosage (Devanand et al. 1995), or therapeutic response to ECT (Sackeim et al. 1987d; Devanand et al. 1995).

In most patients the postECT headache is mild (Freeman and Kendell 1980; Sackeim et al. 1987d), although a sizable minority will report severe pain associated with nausea and vomiting. Typically the headache is frontal in location and has a throbbing character.

The etiology of postECT headache is not known. Its throbbing character suggests a similarity with vascular headache, and ECT may be associated with a temporary change in quality of headache from muscle-contraction type to vascular type (Weiner et al. 1994; Weinstein 1993). Indeed, ECT upregulates 5-HT2 receptors and 5-HT2 receptor sensitization has been associated with development of vascular headache (Weiner et al. 1994). Other suggested mechanisms include electrically induced temporalis muscle spasm or acute increase in blood pressure and cerebral blood flow (Abrams 1997a; Weiner et al. 1994).

Treatment of postECT headache is symptomatic. Aspirin, acetaminophen, or non-steroidal anti-inflammatory drugs (NSAIDs) typically are highly effective, particularly if given promptly after the onset of pain. Sumatriptan, a serotonin 5HTID receptor agonist, has also been effective at doses of 6 mg subcutaneously (DeBattista and Mueller 1995) or 25 - 100 mg orally (Fantz et al. in press). Some patients will require more potent analgesics (e.g. codeine), although narcotics may contribute to associated nausea. Most patients also benefit from bed rest in a quiet, darkened environment.

PostECT headache may occur after any ECT treatment in a course, irrespective of its occurrence at any prior treatment. Patients who experience frequent postECT headache may benefit from prophylactic treatment, such as aspirin, acetaminophen, or NSAIDs given as soon as possible after ECT, or even immediately prior to the ECT treatment. Subcutaneous sumatriptan 6 mg given several minutes prior to ECT was also found to provide effective prophylaxis in a patient with severe, refractory postECT headache (DeBattista and Mueller 1995).

Estimates of the prevalence of nausea following ECT vary from 1.4% - 23% of patients (Gomez 1975; Sackeim et al. 1987d), but the occurrence is difficult to quantify because of methodological issues noted above for headache. Nausea may occur secondary to headache or its treatment with narcotics, particularly in patients with vascular-type headache. It may also occur independently either as a side effect of anesthesia or via other unknown mechanisms. When nausea accompanies headache, the primary treatment should focus on the relief of headache as outlined above. PostECT nausea is otherwise typically well controlled with dopamine-blocking agents, such as phenothiazine derivatives (e.g. prochlorperazine and others), butyrophenones (haloperidol, droperidol), trimethabenzamide, or metoclopramide. If nausea is severe or accompanied by vomiting these agents should be administered parenterally or by suppository. All of these agents have the potential to cause hypotension and motoric side effects, and may lower seizure threshold. If nausea does not respond to these treatments or if side effects are problematic, the serotonin 5HT3 receptor antagonists ondansetron or dolasetron may be useful alternatives. These medications may be given in single intravenous doses of 4 mg and 12.5 mg respectively, several minutes before or after ECT. The greater expense of these medications and their lack of proven superiority over traditional anti-emetics in the setting, of ECT may limit their routine use. If problematic nausea routinely follows the use of a particular anesthetic, an alternative anesthetic may be considered.

5.3 ). Treatment Emergent Mania

As with pharmacological antidepressant treatments, a small minority of depressed patients or patients in mixed affective states switch into hypomania or mania during the ECT course (Devanand et al. 1988b; Andrade et al. 1988b, 1990; Angst et al. 1992; Devanand et al. 1992). In some patients, the severity of manic symptoms may worsen with further ECT treatments. In such cases, it is important to distinguish treatment emergent manic symptoms from delirium with euphoria (Devanand et al. 1988b). There are a number of phenomenological similarities between the two conditions. However, in delirium with euphoria patients are typically confused and have pronounced memory disturbance. The confusion or disorientation should be continuously present and evident from the period immediately following the treatment. In contrast, hypomanic or manic symptomatology may occur in the context of a clear sensorium. Therefore, evaluating cognitive status may be particularly helpful in distinguishing between these states. In addition, states of delirium with euphoric are often characterized by a giddiness in mood or "carefree" disposition. Classical features of hypomania, such as racing thoughts, hypersexuality, irritability, etc. may be absent. In cases of delirium with euphoria an increase in the time between treatments, a decrease in the stimulus intensity, or a change to unilateral from bilateral electrode placement may lead to resolution of the condition.

There is no established strategy on how to manage emergent manic symptoms during the ECT course. Some practitioners continue ECT to treat both the mania and any residual depressive symptomatology. Other practitioners postpone further ECT and observe the patient's course. At times, manic symptomatology will remit spontaneously without further intervention. Should the mania persist, or the patient relapse back into depression, reinstitution of ECT may be considered. Yet other practitioners terminate the ECT course and start pharmacotherapy, often with lithium carbonate or other mood stabilizer, to treat emergent manic symptomatology.

5.4. Objective Cognitive Side Effects

The cognitive side effects produced by ECT have been the subject of intense investigation (Squire 1986; Sackeim 1992; McElhiney et al. 1995) and are the major complications limiting its use. ECT psychiatrists should be familiar with the nature and variability of cognitive side effects, and this information should be conveyed during the consent process (see Chapter 8).

The cognitive side effects of ECT have four essential features. First, the nature and severity of cognitive alterations rapidly change with time from last treatment. The most severe cognitive side effects are observed in the postictal period. Immediately following seizure induction, patients experience a variable, but usually brief, period of disorientation, with impairments in attention, praxis, and memory (Sackeim 1986). These deficits recede at variable rates over time. Consequently, the magnitude of deficits observed during the course of ECT will be a function, in part, of the time of assessment relative to the last treatment and the number of treatments received (Daniel and Crovitz, 1983a; Squire et al. 1985).

Second, the methods used in ECT administration profoundly impact on the nature and magnitude of cognitive deficits. For example, the methods of ECT administration will strongly determine the percentage of patients that develop delirium, characterized by continuous disorientation (Miller et al. 1986; Daniel and Crovitz 1986; Sackeim et al. 1986, 1993). In general, as described in Table 1, bilateral electrode placement, sine wave stimulation, high electrical dosage relative to seizure threshold, closely spaced treatments, larger numbers of treatments, and high dosage of barbiturate anesthetic agents are each independently associated with more intense cognitive side effects compared to right unilateral electrode placement, brief pulse waveform, lower electrical intensity, more widely spaced treatments, fewer treatments, and lower dosage of barbiturate anesthesia (Miller et al. 1985; Sackeim et al. 1986; Weiner et al. 1986b: Sackeim et al. 1993; Lerer et al. 1995; McElhiney et al. 1995). Optimization of these parameters can minimize short-term cognitive side effects and likely reduce the magnitude of long-term changes (Sobin et al. 1995). In patients who develop severe cognitive side effects, such as delirium (Summers et al. 1979; Miller et al. 1986; Mulsant et al. 1991), the attending physician and ECT psychiatrist should review and adjust the treatment technique being used, such as switching to unilateral ECT, lowering the electrical dosage administered, and/or increasing the time interval between treatments, and decrease the dosage or discontinue any medications being administered that may exacerbate cognitive side effects.

Third, patients vary considerably in the extent and severity of cognitive side effects following ECT. There is limited information about the factors that contribute to these individual differences. There is evidence that among depressed patients without known neurological disease or insult, the extent of preECT global cognitive impairment, i.e., mini-Mental State Exam (MMSE) scores, predicts the magnitude of retrograde amnesia for autobiographical information at long-term follow-up. While ECT typically results in improvement in global cognitive status in these patients, as a function of symptomatic response, nonetheless, these same patients may have greater persistent amnesia for personal memories (Sobin et al. 1995). Similarly, there is evidence that the duration of disorientation immediately following the ECT treatment is independently predictive of the magnitude of retrograde amnesia for autobiographical information. Patients who require prolonged periods to recover orientation may be at greater risk for more profound and persistent retrograde amnesia (Sobin et al. 1995). Patients with pre-existing neurological disease or insult (e.g., Parkinson's disease, stroke) may also be at increased risk for ECT-induced delirium and memory deficits (Figiel et al. 1991). Magnetic resonance imaging (MRI) findings of basal ganglia lesions and severe white matter hyperintensities have also been linked to the development of an ECT-induced delirium (Figiel et al. 1990). Some medications may exacerbate ECT-induced cognitive side effects. These include lithium carbonate (Small et al. 1980; Weiner et al. 1980b), and medications with marked anticholinergic, properties, particularly in elderly patients.

Fourth, ECT results in highly characteristic cognitive changes. Across diagnostic groups, prior to receiving ECT, many patients have deficits in attention and concentration that limit their capacity information (Byrne 1977; Pogue-Geile and Oltmanns, 1980; Cornblatt et al. 1981; Sackeim and Steif, 1988). For example, patients with severe psychopathology often have deficient recall of information that was just presented to them (immediate memory). In depressed patients, these deficits are most marked for unstructured material that requires effortful processing in order to impose organization (Weingartner and Silberman 1984; Roy-Byrne et al. 1986). However, such patients are considerably less likely to have deficits in retaining the new information that they do learn (delayed memory) (Cronholm and Ottosson 1961; Sternberg and Jarvik 1976; Steif et al. 1986). With symptomatic response following ECT, the deficits in attention and concentration usually resolve. Consequently, measures of immediate memory are either unchanged or improved within a few days of ECT termination (Cronholm and Ottosson, 1961; Steif et al. 1986; Weiner et al. 1986b; Rossi et al. 1990; Sackeim et al. 1993). Since attention and concentration are essential to many aspects of cognitive function, it is not surprising that shortly following completion of the ECT course improvement may be observed in a wide variety of neuropsychological domains, including global cognitive status (Sackeim et al. 1991; Sobin et al. 1995) and measures of general intelligence (IQ) (Huston and Strother 1948; Stieper et al 1951; Squire et al. 1975; Malloy et al. 1981; Sackeim et al. 1992). There is no evidence that ECT results in impairments of executive functions (e.g., the capacity to shift mental sets), abstract reasoning, creativity, semantic memory, implicit memory, or skill acquisition or retention (Weeks et al. 1980; Frith et al. 1983; Squire et al. 1984; Taylor and Abrams 1985; Jones et al. 1988).

Against this background of unchanged or improved neuropsychological performance, ECT selectively results in anterograde and retrograde amnesia. The anterograde amnesia is characterized by rapid forgetting of newly-learned information (Cronholm and Ottosson 1961; Squire 1986; Steif et al. 1986; Weiner et al. 1986b; Frith et al. 1987; Sackeim et al. 1993). As noted, compared to preECT baseline, a few days following ECT patients may recall more items in a list that was just presented. However, recall after a delay will often be impaired (Korin et al. 1956; Cronholm and Ottosson 1961; Cronholm and Molander 1964; Squire and Miller 1974; Steif et al. 1986; Weiner et al. Squire and Chace 1975; d'Elia 1976; Robertson and Inglis 1978, 1986b; Calev et al. 1989b; Sackeim et al. 1993). The extent and persistence of this rapid forgetting of newly learned information varies among patients and should be taken into account when making recommendations regarding the postECT convalescence period. Until there is substantial resolution of the anterograde amnesia, returning to work, making important financial or personal decisions, or driving may be restricted. The anterograde amnesia rapidly resolves following the termination of ECT. Indeed, no study has documented anterograde amnestic effects of ECT more than a few weeks following the ECT course (Strain et al. 1968; Bidder et al. 1970; Heshe et al. 1978; Jackson, 1978; Fraser and Glass, 1980; Weeks et al. 1980; Gangadhar et al. 1982; Frith et al. 1983; Weiner et al. 1986b; Sackeim et al. 1993). It is unlikely that ECT has any long-term effect on the capacity to learn and retain new information.

Following ECT, patients also display retrograde amnesia. Deficits in the recall of both personal (autobiographical) and public information are usually evident, and the deficits are typically greatest for events that occurred temporally closest to the treatment (Janis, 1950; Cronholm and Molander 1961; Strain et al. 1968; Squire 1975; Squire et al. 1975, 1976, 1981; Weeks et al. 1980; Sackeim et al. 1986; Wiener et al 1986b; Sackeim et al 1993; McElhiney et al. 1995). The magnitude of the retrograde amnesia is greatest immediately following the treatment. A few days following the ECT course, memory for events in the remote past is usually intact, but there may be difficulty in recalling events that transpired several months to years prior to ECT. The retrograde amnesia over this time span is rarely complete. Rather, patients have gaps or spottiness in their memories of personal and public events. Recent evidence suggests that the retrograde amnesia is typically greater for public information (knowledge of events in the world) as compared to personal information (autobiographic details of the patient's life) (Lisanby et al. in press). The emotional valence of autobiographical events, i.e., memories of pleasant or distressful events, is not related to their likelihood of being forgotten (McElhiney et al. 1995).

As time from ECT increases, there is usually substantial reduction in the extent of retrograde amnesia. Older memories are more likely to be recovered. The time course for this shrinkage of retrograde amnesia is often more gradual than that for the resolution of anterograde amnesia. In many patients the recovery from retrograde amnesia will be incomplete, and there is evidence that ECT can result in persistent or permanent memory loss (Squire et al. 1981; Weiner et al. 1986b; McElhiney et al. 1995; Sobin et al. 1995). Owing to a combination of anterograde and retrograde effects, many patients may manifest persistent loss of memory for some events that transpired in the interval starting several months before and extending to several weeks following the ECT course. There are individual differences, however, and, uncommonly, some patients may experience persistent amnesia that extends back several years prior to ECT. Profound and persistent retrograde amnesia may be more likely in patients with pre-existing neurological impairment and patients who receive large numbers of treatments, using methods that accentuate acute cognitive side effects (e.g., sine wave stimulation, bilateral electrode placement, high electrical stimulus intensity).

To determine the occurrence and severity of cognitive changes during and following the ECT course, orientation and memory functions should be assessed prior to initiation of ECT and throughout the course of treatment (see Chapter 12 for details).

5.5. Adverse Subjective Reactions

Negative subjective reactions to the experience of receiving ECT should be considered adverse side effects (Sackeim 1992). Prior to ECT, patients often report apprehension; rarely, some patients develop intense fear of the procedure during the ECT course (Fox 1993). Family members are also frequently apprehensive about the effects of the treatment. As part of the consent process prior to the start of ECT, patients and family members should be given the opportunity to express their concerns and questions to the attending physician and/or members of the ECT treatment team (see Chapter 8). Since much of the apprehension may be based on lack of information, it is often helpful to provide patients and family members with an information sheet describing basic facts about ECT (see Chapter 8). This material should be supplemental to the consent form. It is also useful to make available video material on ECT. Addressing the concerns and educational needs of patients and family members should be a process that continues throughout the course. In centers that regularly conduct ECT, it has been found useful to have ongoing group sessions led by a member of the treatment team, for patients receiving ECT and/or their significant others. Such group sessions, including prospective and recently treated patients and their families, may engender mutual support among these individuals and can serve as a forum for education about ECT.

Shortly following ECT, the great majority of patients report that their cognitive function is improved relative to their pre-ECT baseline (Cronholm and Ottosson 1963b; Shellenberger et al 1982; Frith et al 1983; Pettinati and Rosenberg 1984; Weiner et al 1986b; Mattes et al 1990; Calev et al 1991; Sackeim et al. 1993 ); Coleman et al 1996). Indeed, recent research has shown that two months following completion of ECT the memory self-ratings of former patients are markedly improved relative to their pre-ECT baseline and indistinguishable from healthy controls (Coleman et al. 1996). In patients who have received ECT, memory self-ratings show little association with the results of objective neuropsychological testing (Cronholm and Ottosson 1963b; Frith et al 1983; Squire and Slater 1983; Weiner et al 1986b; Squire and Zouzounis 1988; Calev et al 1991a; Coleman et al 1996). Likewise, in healthy and neurological samples, subjective memory assessments have generally shown weak or no association with objective neuropsychological measures (Bennett-Levy and Powell 1980; Broadbent et al. 1982; Rabbitt 1982; Larrabee and Levin 1986; Sackeim and Stem 1997). In contrast, strong associations are observed between mood state and memory self-ratings among patients who have received ECT, as well as other populations (Stieper et al. 1951; Frith et al 1983; Pettinati and Rosenberg 1984; Weiner et al. 1986b; Mattes et al 1990; Coleman et al. 1996). In essence, patients who benefit the most from ECT in terms of symptomatic response typically report the greatest improvement in subjective evaluations of memory.

A small minority of patients treated with ECT later report that they have suffered devastating consequences (Freeman and Kendell 1980, 1986). Patients may indicate that have dense amnesia extending far back into the past for events of personal significance and/or that broad aspects of cognitive function are impaired such that they are no longer able to engage in former occupations. The rarity of these subjective reports of profound cognitive deficits makes determination of their absolute base rates difficult. Multiple factors likely contribute to these perceptions by former patients.

First, in some patients self-reports of profound ECT-induced deficits may be accurate. As noted, as with any medical intervention, there are individual differences in the magnitude and persistence of ECT's cognitive effects. In rare cases, ECT may result in a more dense and persistent retrograde amnesia that extends back to years prior to the treatment.

Second, some of the psychiatric conditions treated with ECT result in cognitive deterioration as part of their natural history. This may be particularly likely in young patients in their first psychotic episode (Wyatt 1991, 1995), and in older patients where ECT may unmask a dementing process. While in such cases, cognitive deterioration would have occurred inevitably, the experience of transient short-term side effects with ECT may sensitize patients to attribute the persistent changes to the treatment (Squire 1986; Sackeim 1992).

Third, as noted above, subjective evaluations of cognitive function typically show poor association with objective measurement and strong association with measures of psychopathology (Coleman et al. 1996). Only one study recruited patients with long-term complaints about effects of ECT and compared them to two control groups (Freeman et al. 1980). Objective neuropsychological differences among the groups were slight, but there were marked differences in assessments of psychopathology and medication status. Patients who reported persistent deficits due to ECT were less likely to have benefited from the treatment, and were more likely to be presently symptomatic and receiving psychotropic treatment (Freeman et al. 1980; Frith et al. 1983).

Recommendations

5. 1. General

a) Physicians administering ECT should be aware of the principal adverse effects which may accompany its use.

b) The type, likelihood, and persistence of adverse effects should be considered on a case-by-case basis in the decision to recommend ECT and in the informed consent process (see Chapter 8).

c) Efforts should be made to minimize adverse effects by optimization of the patient's medical condition prior to treatment, appropriate modifications in ECT technique, and the use of adjunctive medications (see also Section 4.1).

5.1.1. Cardiovascular Complications

a) The electrocardiogram (ECG) and vital signs (blood pressure, pulse, and respiration) should be monitored during each ECT treatment to detect cardiac arrythmias and hypertension (see Section 11.8).

b) The ECT treatment team should be prepared to manage the cardiovascular complications known to be associated with ECT. Personnel, supplies, and equipment necessary to perform such a task should be readily available (see Chapters 9 and 10).

5.1.2. Prolonged Seizures

Each facility should have policies outlining the steps to be taken to terminate prolonged seizures and status epilepticus (see Section 11.9.4).

5.1.3 Prolonged Apnea

Resources for maintaining an airway for an extended period, including intubation, should be available in the treatment room (see Chapters 9 and 10).

Systemic Side Effects

Headache and nausea are the most common systemic side effects of ECT. Systemic side effects should be identified and symptomatic treatment considered.

5.3 Treatment Emergent Mania

Instances in which patients switch from depressive or affectively mixed states into hypomania or mania during a course of ECT should be identified, and a determination to continue or suspend further treatment with ECT.

5.4. Cognitive Dysfunction

a) Orientation and memory function should be assessed prior to ECT and periodically throughout the ECT course to detect and monitor the presence of ECT-related cognitive dysfunction (see Section 12.2.1 for details). This assessment should attend to patient self-reports of memory difficulty.

b) Based on the assessment of the severity of cognitive side effects, the physician administering ECT should take appropriate action. The contributions of medications, ECT technique, and spacing of treatments should be reviewed. Potential treatment modifications include changing from bilateral to right unilateral electrode placement, decreasing the intensity of electrical stimulation, increasing the time interval between treatments, and/or altering the dosage of medications, or, if necessary, terminating the treatment course.

Table 1. Treatment factors that may increase or decrease the severity of adverse cognitive side effects

5.1. Medical Complications

Precise rates of mortality attributable to ECT are difficult to determine due to methodological issues intrinsic to studies of medical mortality, such as uncertainty as to cause of death, time frame for linking death to ECT, and variability in reporting requirements. The mortality attributed to ECT is estimated to be approximately the same as that associated with minor surgery (McCabe 1985 Warner et al. 1993; Brand et al. 1994; Badrinath et al. 1995: Hall et al. 1997). Published estimates from large and diverse patient series over several decades report up to 4 deaths per 100,000 treatments (Heshe and Roeder, 1976; Fink, 1979; Weiner 1979; Babigian and Guttmacher, 1984; Crowe, 1984; Kramer, 1985: Abrams 1997b; Reid et al. 1998). Despite the frequent use of ECT in patients with significant medical complications and in the elderly (Sackeim 1993, 1998; Weiner et al. in press), rates of mortality appear to have decreased in recent years. A reasonable current estimate is that the rate of ECT-related mortality is 1 per 10,000 patients. This rate may be higher in patients with severe medical conditions. The rate of significant morbidity and mortality is believed to be lower with ECT than with treatment with some types of antidepressant medication (e.g., tricyclics) (Sackeim 1998). There is also evidence from longitudinal follow-up studies that mortality rates following hospitalization are lower among depressed patients who received ECT than patients who received alternative forms of treatment or no treatment (Avery and Winokur, 1976; Philibert et al. 1995)

When mortality occurs with ECT, it typically happens immediately following the seizure or during the postictal recovery period. Cardiovascular complications are the leading cause of death and of significant morbidity (Pitts 1982; Burke et al. 1987; Welch and Drop 1989; Zielinski et al. 1993; Rice et al. 1994). Despite the short-lived increases in cerebral blood flow and intracranial pressure, cerebrovascular complications are notably rare (Hsiao et al. 1987). Given the high rate of cardiac arrhythmias in the immediate postictal period, the majority of which are benign and resolve spontaneously, ECG should be monitored during and immediately following the procedure (see Section 11.8) and patients should not be taken to the recovery area until there is resolution of significant arrhythmias. Vital signs (pulse, systolic and diastolic pressure) should be stable prior to the patient's leaving the recovery area (Section 11.10). Patients with pre-existing cardiac illness are at greater risk for post-ECT cardiac complications (Prudic et al. 1987; Zielinski et al. 1993; Rice et al. 1994). Indeed, there is evidence that the type of pre-existing cardiac disease predicts the type of complication that may be encountered following ECT. For example, ventricular arrythmias are more common in patients with pre-existing ventricular abnormalities than in patients with ischemic heart disease (Zielinski et al. 1993). Management of cardiac complications is discussed in Chapter 11.

Two other possible sources of morbidity are prolonged seizures and tardive seizures (Weiner et al. 1980a). Management of prolonged seizures is described in Section 11.9. Failure to terminate seizures within a period of 3 to 5 minutes may increase postictal confusion and amnesia. Inadequate oxygenation during prolonged seizures increases the risk of hypoxia and cerebral dysfunction, as well as cardiovascular complications. In animal studies, seizure activity that is sustained for periods exceeding 30-60 minutes, regardless of steps taken to maintain appropriate levels of blood gases, is associated with an increased risk of structural brain damage and cardiovascular and cardiopulmonary complications (Meldrum et al. 1974; Ingvar 1986; Meldrum 1986; Siesjo et al. 1986; O'Connell et al. 1988; Devanand et al. 1994).

Prolonged seizures and status epilepticus may be more likely in patients receiving medications that lower seizure threshold or interfere with seizure termination (e.g. theophylline, even at therapeutic levels) (Peters et al. 1984; Devanand et al. 1988a; Abrams, 1997a), in patients receiving concomitant lithium therapy (Weiner et al. 1980b), in patients with pre-existing electrolyte imbalance (Finlayson et al. 1989), and with the repeated induction of seizures within the same treatment session (e.g., multiple monitored ECT) (Strain-and Bidder 1971, Maletzky 1981).

There has been concern as to whether the rate of spontaneous seizures is increased following the course of ECT (Assael et al. 1967; Devinsky and Duchowny 1983). The evidence indicates, however, that such events are extremely rare and probably do not differ from population base rates (Blackwood et al. 1980; Small et al. 1981). There are no data concerning rates of tardive seizures, i.e., seizures that occur following termination of the ECT-induced seizure, but experience indicates that these are also rare events. As noted in Section 11.9, prolonged or tardive seizures occurring during the immediate postictal period are often not accompanied by motor manifestations, underscoring the need for EEG seizure monitoring (Rao et al. 1993). Nonconvulsive status epilepticus may also occur in the interictal period, with an abrupt onset of delirium, unresponsiveness, and/or agitation as distinguishing clinical features (Grogan et al. 1995). Cessation of EEG abnormalities and improved cognitive function following short-acting anticonvulsant treatment (e.g. intravenous lorazepam or diazepam) may prove diagnostic (Weiner and Krystal, 1993).

Prolonged postictal apnea is a rare event that occurs primarily in patients with a pseudocholinesterase deficiency resulting in slow metabolism of succinylcholine (Packman et al. 1978). Maintaining adequate oxygenation is critical in instances of prolonged apnea, which will usually resolve spontaneously within 30 to 60 minutes. When prolonged apnea is encountered, it is helpful to obtain a dibuciane number assay or a pseudocholinesterase level prior to the next treatment in order to establish etiology. At subsequent treatments, either a very low dose of succinylcholine may be used or a non-depolarizing muscle relaxant, such as atracurium, may be substituted (Hickey et al. 1987; Hicks, 1987; Stack et al. 1988; Kramer and Afrasiabi 1991; Lui et al. 1993).

To some extent, medical adverse events can be anticipated. Whenever possible, the risks of such events should be minimized by optimization of the patient's medical condition prior to ECT and/or modifications in ECT procedures. Patients with preexisting cardiac illness, compromised pulmonary status, a history of CNS insult, or medical complications following prior courses of anesthesia or ECT are especially likely to be at increased risk (Weiner and Coffey 1988; Zieliniski et al. 1993). ECT psychiatrists should review the medical work-up and history of prospective ECT patients (see Chapter 6). Specialist consultations or additional laboratory studies may be called for, as well as changes in medication regimens. In spite of careful pre-ECT evaluation, medical complications may arise which have not been anticipated. ECT facilities should be staffed with personnel prepared to manage potential clinical emergencies and should be equipped accordingly (see Chapters 9 and 10). Examples of these events include cardiovascular complications (such as cardiac arrest, arrythmias, ischemia, hyper- and hypotension), prolonged apnea, and prolonged or tardive seizures and status epilepticus.

Major adverse events that occur during or soon after the ECT course should be documented in the patient's medical record. The steps taken to manage the event, including specialist consultation, use of additional procedures, and administration of medications, should likewise be documented. As cardiovascular complications are the most likely source of significant adverse events and are seen most frequently in the immediate post-ECT period, the treatment team should be capable of managing the major classes of cardiovascular complications. A set of predetermined procedures for dealing with instances of prolonged or tardive seizures and status epilepticus are helpful.

5.2. Systemic Side Effects

Headache is a common side effect of ECT and is observed in as many as 45% of patients during and shortly following the postictal recovery period (Devanand et al. 1995; Freeman and Kendell 1980; Gomez 1975; Sackeim et al. 1987d: Tubi et al. 1993; Weiner et al. 1994). However, the precise incidence of postECT headache is difficult to determine due to methodological issues such as the high baseline (preECT) occurrence of headache in patients with depression, the potential effects of concurrent medication or medication withdrawal, and differences between studies in the assessment of headache. PostECT headache appears to be particularly common in younger patients (Devanand et al. 1995) and especially in children and adolescents (Rey and Walter 1997; Walter and Rey 1997) It is not known whether pre-existing headache syndromes (e.g., migraine) increase the risk of postECT headache, but ECT may exacerbate a previous headache condition (Weiner et al. 1994). The occurrence of postECT headache does not appear to be related to stimulus electrode placement (at least bifrontotemporal vs. right unilateral) (Fleminger et al. 1970; Sackeim et al. 1987d; Tubi et al. 1993; Devanand et al. 1995), stimulus dosage (Devanand et al. 1995), or therapeutic response to ECT (Sackeim et al. 1987d; Devanand et al. 1995).

In most patients the postECT headache is mild (Freeman and Kendell 1980; Sackeim et al. 1987d), although a sizable minority will report severe pain associated with nausea and vomiting. Typically the headache is frontal in location and has a throbbing character.

The etiology of postECT headache is not known. Its throbbing character suggests a similarity with vascular headache, and ECT may be associated with a temporary change in quality of headache from muscle-contraction type to vascular type (Weiner et al. 1994; Weinstein 1993). Indeed, ECT upregulates 5-HT2 receptors and 5-HT2 receptor sensitization has been associated with development of vascular headache (Weiner et al. 1994). Other suggested mechanisms include electrically induced temporalis muscle spasm or acute increase in blood pressure and cerebral blood flow (Abrams 1997a; Weiner et al. 1994).

Treatment of postECT headache is symptomatic. Aspirin, acetaminophen, or non-steroidal anti-inflammatory drugs (NSAIDs) typically are highly effective, particularly if given promptly after the onset of pain. Sumatriptan, a serotonin 5HTID receptor agonist, has also been effective at doses of 6 mg subcutaneously (DeBattista and Mueller 1995) or 25 - 100 mg orally (Fantz et al. in press). Some patients will require more potent analgesics (e.g. codeine), although narcotics may contribute to associated nausea. Most patients also benefit from bed rest in a quiet, darkened environment.

Post-ECT headache may occur after any ECT treatment in a course, irrespective of its occurrence at any prior treatment. Patients who experience frequent post-ECT headache may benefit from prophylactic treatment, such as aspirin, acetaminophen, or NSAIDs given as soon as possible after ECT, or even immediately prior to the ECT treatment. Subcutaneous sumatriptan 6 mg given several minutes prior to ECT was also found to provide effective prophylaxis in a patient with severe, refractory postECT headache (DeBattista and Mueller 1995).

Estimates of the prevalence of nausea following ECT vary from 1.4% - 23% of patients (Gomez 1975; Sackeim et al. 1987d), but the occurrence is difficult to quantify because of methodological issues noted above for headache. Nausea may occur secondary to headache or its treatment with narcotics, particularly in patients with vascular-type headache. It may also occur independently either as a side effect of anesthesia or via other unknown mechanisms. When nausea accompanies headache, the primary treatment should focus on the relief of headache as outlined above. PostECT nausea is otherwise typically well controlled with dopamine-blocking agents, such as phenothiazine derivatives (e.g. prochlorperazine and others), butyrophenones (haloperidol, droperidol), trimethabenzamide, or metoclopramide. If nausea is severe or accompanied by vomiting these agents should be administered parenterally or by suppository. All of these agents have the potential to cause hypotension and motoric side effects, and may lower seizure threshold. If nausea does not respond to these treatments or if side effects are problematic, the serotonin 5HT3 receptor antagonists ondansetron or dolasetron may be useful alternatives. These medications may be given in single intravenous doses of 4 mg and 12.5 mg respectively, several minutes before or after ECT. The greater expense of these medications and their lack of proven superiority over traditional anti-emetics in the setting, of ECT may limit their routine use. If problematic nausea routinely follows the use of a particular anesthetic, an alternative anesthetic may be considered.

5.3 ). Treatment Emergent Mania

As with pharmacological antidepressant treatments, a small minority of depressed patients or patients in mixed affective states switch into hypomania or mania during the ECT course (Devanand et al. 1988b; Andrade et al. 1988b, 1990; Angst et al. 1992; Devanand et al. 1992). In some patients, the severity of manic symptoms may worsen with further ECT treatments. In such cases, it is important to distinguish treatment emergent manic symptoms from delirium with euphoria (Devanand et al. 1988b). There are a number of phenomenological similarities between the two conditions. However, in delirium with euphoria patients are typically confused and have pronounced memory disturbance. The confusion or disorientation should be continuously present and evident from the period immediately following the treatment. In contrast, hypomanic or manic symptomatology may occur in the context of a clear sensorium. Therefore, evaluating cognitive status may be particularly helpful in distinguishing between these states. In addition, states of delirium with euphoric are often characterized by a giddiness in mood or "carefree" disposition. Classical features of hypomania, such as racing thoughts, hypersexuality, irritability, etc. may be absent. In cases of delirium with euphoria an increase in the time between treatments, a decrease in the stimulus intensity, or a change to unilateral from bilateral electrode placement may lead to resolution of the condition.

There is no established strategy on how to manage emergent manic symptoms during the ECT course. Some practitioners continue ECT to treat both the mania and any residual depressive symptomatology. Other practitioners postpone further ECT and observe the patient's course. At times, manic symptomatology will remit spontaneously without further intervention. Should the mania persist, or the patient relapse back into depression, reinstitution of ECT may be considered. Yet other practitioners terminate the ECT course and start pharmacotherapy, often with lithium carbonate or other mood stabilizer, to treat emergent manic symptomatology.

5.4. Objective Cognitive Side Effects

The cognitive side effects produced by ECT have been the subject of intense investigation (Squire 1986; Sackeim 1992; McElhiney et al. 1995) and are the major complications limiting its use. ECT psychiatrists should be familiar with the nature and variability of cognitive side effects, and this information should be conveyed during the consent process (see Chapter 8).

The cognitive side effects of ECT have four essential features. First, the nature and severity of cognitive alterations rapidly change with time from last treatment. The most severe cognitive side effects are observed in the postictal period. Immediately following seizure induction, patients experience a variable, but usually brief, period of disorientation, with impairments in attention, praxis, and memory (Sackeim 1986). These deficits recede at variable rates over time. Consequently, the magnitude of deficits observed during the course of ECT will be a function, in part, of the time of assessment relative to the last treatment and the number of treatments received (Daniel and Crovitz, 1983a; Squire et al. 1985).

Second, the methods used in ECT administration profoundly impact on the nature and magnitude of cognitive deficits. For example, the methods of ECT administration will strongly determine the percentage of patients that develop delirium, characterized by continuous disorientation (Miller et al. 1986; Daniel and Crovitz 1986; Sackeim et al. 1986, 1993). In general, as described in Table 1, bilateral electrode placement, sine wave stimulation, high electrical dosage relative to seizure threshold, closely spaced treatments, larger numbers of treatments, and high dosage of barbiturate anesthetic agents are each independently associated with more intense cognitive side effects compared to right unilateral electrode placement, brief pulse waveform, lower electrical intensity, more widely spaced treatments, fewer treatments, and lower dosage of barbiturate anesthesia (Miller et al. 1985; Sackeim et al. 1986; Weiner et al. 1986b: Sackeim et al. 1993; Lerer et al. 1995; McElhiney et al. 1995). Optimization of these parameters can minimize short-term cognitive side effects and likely reduce the magnitude of long-term changes (Sobin et al. 1995). In patients who develop severe cognitive side effects, such as delirium (Summers et al. 1979; Miller et al. 1986; Mulsant et al. 1991), the attending physician and ECT psychiatrist should review and adjust the treatment technique being used, such as switching to unilateral ECT, lowering the electrical dosage administered, and/or increasing the time interval between treatments, and decrease the dosage or discontinue any medications being administered that may exacerbate cognitive side effects.

Third, patients vary considerably in the extent and severity of cognitive side effects following ECT. There is limited information about the factors that contribute to these individual differences. There is evidence that among depressed patients without known neurological disease or insult, the extent of preECT global cognitive impairment, i.e., mini-Mental State Exam (MMSE) scores, predicts the magnitude of retrograde amnesia for autobiographical information at long-term follow-up. While ECT typically results in improvement in global cognitive status in these patients, as a function of symptomatic response, nonetheless, these same patients may have greater persistent amnesia for personal memories (Sobin et al. 1995). Similarly, there is evidence that the duration of disorientation immediately following the ECT treatment is independently predictive of the magnitude of retrograde amnesia for autobiographical information. Patients who require prolonged periods to recover orientation may be at greater risk for more profound and persistent retrograde amnesia (Sobin et al. 1995). Patients with pre-existing neurological disease or insult (e.g., Parkinson's disease, stroke) may also be at increased risk for ECT-induced delirium and memory deficits (Figiel et al. 1991). Magnetic resonance imaging (MRI) findings of basal ganglia lesions and severe white matter hyperintensities have also been linked to the development of an ECT-induced delirium (Figiel et al. 1990). Some medications may exacerbate ECT-induced cognitive side effects. These include lithium carbonate (Small et al. 1980; Weiner et al. 1980b), and medications with marked anticholinergic, properties, particularly in elderly patients.

Fourth, ECT results in highly characteristic cognitive changes. Across diagnostic groups, prior to receiving ECT, many patients have deficits in attention and concentration that limit their capacity information (Byrne 1977; Pogue-Geile and Oltmanns, 1980; Cornblatt et al. 1981; Sackeim and Steif, 1988). For example, patients with severe psychopathology often have deficient recall of information that was just presented to them (immediate memory). In depressed patients, these deficits are most marked for unstructured material that requires effortful processing in order to impose organization (Weingartner and Silberman 1984; Roy-Byrne et al. 1986). However, such patients are considerably less likely to have deficits in retaining the new information that they do learn (delayed memory) (Cronholm and Ottosson 1961; Sternberg and Jarvik 1976; Steif et al. 1986). With symptomatic response following ECT, the deficits in attention and concentration usually resolve. Consequently, measures of immediate memory are either unchanged or improved within a few days of ECT termination (Cronholm and Ottosson, 1961; Steif et al. 1986; Weiner et al. 1986b; Rossi et al. 1990; Sackeim et al. 1993). Since attention and concentration are essential to many aspects of cognitive function, it is not surprising that shortly following completion of the ECT course improvement may be observed in a wide variety of neuropsychological domains, including global cognitive status (Sackeim et al. 1991; Sobin et al. 1995) and measures of general intelligence (IQ) (Huston and Strother 1948; Stieper et al 1951; Squire et al. 1975; Malloy et al. 1981; Sackeim et al. 1992). There is no evidence that ECT results in impairments of executive functions (e.g., the capacity to shift mental sets), abstract reasoning, creativity, semantic memory, implicit memory, or skill acquisition or retention (Weeks et al. 1980; Frith et al. 1983; Squire et al. 1984; Taylor and Abrams 1985; Jones et al. 1988).

Against this background of unchanged or improved neuropsychological performance, ECT selectively results in anterograde and retrograde amnesia. The anterograde amnesia is characterized by rapid forgetting of newly-learned information (Cronholm and Ottosson 1961; Squire 1986; Steif et al. 1986; Weiner et al. 1986b; Frith et al. 1987; Sackeim et al. 1993). As noted, compared to preECT baseline, a few days following ECT patients may recall more items in a list that was just presented. However, recall after a delay will often be impaired (Korin et al. 1956; Cronholm and Ottosson 1961; Cronholm and Molander 1964; Squire and Miller 1974; Steif et al. 1986; Weiner et al. Squire and Chace 1975; d'Elia 1976; Robertson and Inglis 1978, 1986b; Calev et al. 1989b; Sackeim et al. 1993). The extent and persistence of this rapid forgetting of newly learned information varies among patients and should be taken into account when making recommendations regarding the postECT convalescence period. Until there is substantial resolution of the anterograde amnesia, returning to work, making important financial or personal decisions, or driving may be restricted. The anterograde amnesia rapidly resolves following the termination of ECT. Indeed, no study has documented anterograde amnestic effects of ECT more than a few weeks following the ECT course (Strain et al. 1968; Bidder et al. 1970; Heshe et al. 1978; Jackson, 1978; Fraser and Glass, 1980; Weeks et al. 1980; Gangadhar et al. 1982; Frith et al. 1983; Weiner et al. 1986b; Sackeim et al. 1993). It is unlikely that ECT has any long-term effect on the capacity to learn and retain new information.

Following ECT, patients also display retrograde amnesia. Deficits in the recall of both personal (autobiographical) and public information are usually evident, and the deficits are typically greatest for events that occurred temporally closest to the treatment (Janis, 1950; Cronholm and Molander 1961; Strain et al. 1968; Squire 1975; Squire et al. 1975, 1976, 1981; Weeks et al. 1980; Sackeim et al. 1986; Wiener et al 1986b; Sackeim et al 1993; McElhiney et al. 1995). The magnitude of the retrograde amnesia is greatest immediately following the treatment. A few days following the ECT course, memory for events in the remote past is usually intact, but there may be difficulty in recalling events that transpired several months to years prior to ECT. The retrograde amnesia over this time span is rarely complete. Rather, patients have gaps or spottiness in their memories of personal and public events. Recent evidence suggests that the retrograde amnesia is typically greater for public information (knowledge of events in the world) as compared to personal information (autobiographic details of the patient's life) (Lisanby et al. in press). The emotional valence of autobiographical events, i.e., memories of pleasant or distressful events, is not related to their likelihood of being forgotten (McElhiney et al. 1995).

As time from ECT increases, there is usually substantial reduction in the extent of retrograde amnesia. Older memories are more likely to be recovered. The time course for this shrinkage of retrograde amnesia is often more gradual than that for the resolution of anterograde amnesia. In many patients the recovery from retrograde amnesia will be incomplete, and there is evidence that ECT can result in persistent or permanent memory loss (Squire et al. 1981; Weiner et al. 1986b; McElhiney et al. 1995; Sobin et al. 1995). Owing to a combination of anterograde and retrograde effects, many patients may manifest persistent loss of memory for some events that transpired in the interval starting several months before and extending to several weeks following the ECT course. There are individual differences, however, and, uncommonly, some patients may experience persistent amnesia that extends back several years prior to ECT. Profound and persistent retrograde amnesia may be more likely in patients with pre-existing neurological impairment and patients who receive large numbers of treatments, using methods that accentuate acute cognitive side effects (e.g., sine wave stimulation, bilateral electrode placement, high electrical stimulus intensity).

To determine the occurrence and severity of cognitive changes during and following the ECT course, orientation and memory functions should be assessed prior to initiation of ECT and throughout the course of treatment (see Chapter 12 for details).

5.5. Adverse Subjective Reactions

Negative subjective reactions to the experience of receiving ECT should be considered adverse side effects (Sackeim 1992). Prior to ECT, patients often report apprehension; rarely, some patients develop intense fear of the procedure during the ECT course (Fox 1993). Family members are also frequently apprehensive about the effects of the treatment. As part of the consent process prior to the start of ECT, patients and family members should be given the opportunity to express their concerns and questions to the attending physician and/or members of the ECT treatment team (see Chapter 8). Since much of the apprehension may be based on lack of information, it is often helpful to provide patients and family members with an information sheet describing basic facts about ECT (see Chapter 8). This material should be supplemental to the consent form. It is also useful to make available video material on ECT. Addressing the concerns and educational needs of patients and family members should be a process that continues throughout the course. In centers that regularly conduct ECT, it has been found useful to have ongoing group sessions led by a member of the treatment team, for patients receiving ECT and/or their significant others. Such group sessions, including prospective and recently treated patients and their families, may engender mutual support among these individuals and can serve as a forum for education about ECT.

Shortly following ECT, the great majority of patients report that their cognitive function is improved relative to their pre-ECT baseline (Cronholm and Ottosson 1963b; Shellenberger et al 1982; Frith et al 1983; Pettinati and Rosenberg 1984; Weiner et al 1986b; Mattes et al 1990; Calev et al 1991; Sackeim et al. 1993 ); Coleman et al 1996). Indeed, recent research has shown that two months following completion of ECT the memory self-ratings of former patients are markedly improved relative to their pre-ECT baseline and indistinguishable from healthy controls (Coleman et al. 1996). In patients who have received ECT, memory self-ratings show little association with the results of objective neuropsychological testing (Cronholm and Ottosson 1963b; Frith et al 1983; Squire and Slater 1983; Weiner et al 1986b; Squire and Zouzounis 1988; Calev et al 1991a; Coleman et al 1996). Likewise, in healthy and neurological samples, subjective memory assessments have generally shown weak or no association with objective neuropsychological measures (Bennett-Levy and Powell 1980; Broadbent et al. 1982; Rabbitt 1982; Larrabee and Levin 1986; Sackeim and Stem 1997). In contrast, strong associations are observed between mood state and memory self-ratings among patients who have received ECT, as well as other populations (Stieper et al. 1951; Frith et al 1983; Pettinati and Rosenberg 1984; Weiner et al. 1986b; Mattes et al 1990; Coleman et al. 1996). In essence, patients who benefit the most from ECT in terms of symptomatic response typically report the greatest improvement in subjective evaluations of memory.

A small minority of patients treated with ECT later report that they have suffered devastating consequences (Freeman and Kendell 1980, 1986). Patients may indicate that have dense amnesia extending far back into the past for events of personal significance and/or that broad aspects of cognitive function are impaired such that they are no longer able to engage in former occupations. The rarity of these subjective reports of profound cognitive deficits makes determination of their absolute base rates difficult. Multiple factors likely contribute to these perceptions by former patients.

First, in some patients self-reports of profound ECT-induced deficits may be accurate. As noted, as with any medical intervention, there are individual differences in the magnitude and persistence of ECT's cognitive effects. In rare cases, ECT may result in a more dense and persistent retrograde amnesia that extends back to years prior to the treatment.

Second, some of the psychiatric conditions treated with ECT result in cognitive deterioration as part of their natural history. This may be particularly likely in young patients in their first psychotic episode (Wyatt 1991, 1995), and in older patients where ECT may unmask a dementing process. While in such cases, cognitive deterioration would have occurred inevitably, the experience of transient short-term side effects with ECT may sensitize patients to attribute the persistent changes to the treatment (Squire 1986; Sackeim 1992).

Third, as noted above, subjective evaluations of cognitive function typically show poor association with objective measurement and strong association with measures of psychopathology (Coleman et al. 1996). Only one study recruited patients with long-term complaints about effects of ECT and compared them to two control groups (Freeman et al. 1980). Objective neuropsychological differences among the groups were slight, but there were marked differences in assessments of psychopathology and medication status. Patients who reported persistent deficits due to ECT were less likely to have benefited from the treatment, and were more likely to be presently symptomatic and receiving psychotropic treatment (Freeman et al. 1980; Frith et al. 1983).

Recommendations

5. 1. General

a) Physicians administering ECT should be aware of the principal adverse effects which may accompany its use.

b) The type, likelihood, and persistence of adverse effects should be considered on a case-by-case basis in the decision to recommend ECT and in the informed consent process (see Chapter 8).

c) Efforts should be made to minimize adverse effects by optimization of the patient's medical condition prior to treatment, appropriate modifications in ECT technique, and the use of adjunctive medications (see also Section 4.1).

5.1.1. Cardiovascular Complications

a) The electrocardiogram (ECG) and vital signs (blood pressure, pulse, and respiration) should be monitored during each ECT treatment to detect cardiac arrythmias and hypertension (see Section 11.8).

b) The ECT treatment team should be prepared to manage the cardiovascular complications known to be associated with ECT. Personnel, supplies, and equipment necessary to perform such a task should be readily available (see Chapters 9 and 10).

5.1.2. Prolonged Seizures

Each facility should have policies outlining the steps to be taken to terminate prolonged seizures and status epilepticus (see Section 11.9.4).

5.1.3 Prolonged Apnea

Resources for maintaining an airway for an extended period, including intubation, should be available in the treatment room (see Chapters 9 and 10).

Systemic Side Effects

Headache and nausea are the most common systemic side effects of ECT. Systemic side effects should be identified and symptomatic treatment considered.

5.3 Treatment Emergent Mania

Instances in which patients switch from depressive or affectively mixed states into hypomania or mania during a course of ECT should be identified, and a determination to continue or suspend further treatment with ECT.

5.4. Cognitive Dysfunction

a) Orientation and memory function should be assessed prior to ECT and periodically throughout the ECT course to detect and monitor the presence of ECT-related cognitive dysfunction (see Section 12.2.1 for details). This assessment should attend to patient self-reports of memory difficulty.

b) Based on the assessment of the severity of cognitive side effects, the physician administering ECT should take appropriate action. The contributions of medications, ECT technique, and spacing of treatments should be reviewed. Potential treatment modifications include changing from bilateral to right unilateral electrode placement, decreasing the intensity of electrical stimulation, increasing the time interval between treatments, and/or altering the dosage of medications, or, if necessary, terminating the treatment course.

Table 1. Treatment factors that may increase or decrease the severity of adverse cognitive side effects

next: Chapter 6. Pre-ECT Evaluation
~ all Shocked! ECT articles
~ depression library articles
~ all articles on depression