Clinical Effectiveness of Maintenance Electroconvulsive Therapy in Patients with Schizophrenia: A Retrospective Cohort Study

Abstract

Objective

To assess the clinical effectiveness and cognitive effects of maintenance electroconvulsive therapy (mECT) in patients with schizophrenia or schizoaffective disorder and explore factors associated with both outcomes.

Methods

In this retrospective cohort study, we examined clinical records of 47 patients with a DSM-IV diagnosis of schizophrenia or schizoaffective disorder treated with mECT at an academic mental health hospital between April 2010 and July 2016. Sixty-two mECT courses were reviewed. We assessed clinical effectiveness and cognitive effects as well as factors associated with response to treatment, including psychiatric diagnosis, concomitant pharmacological treatment and previous treatment response.

Results

mECT was able to maintain clinical response in 48 (77%) treatment courses. Significant cognitive adverse effects were reported in 7 (11%) of the courses. Use of antipsychotic, antidepressant or benzodiazepine medications, psychiatric disorder and gender were not associated with response.

Conclusion

This study shows meaningful clinical effectiveness and good tolerability of mECT in patients with resistant schizophrenia over extended periods of time.

1. Introduction

Schizophrenia is a serious psychiatric illness that affects around 1% of the population.¹ Antipsychotics represent the mainstay of treatment, although up to 25% of patients do not respond to first line antipsychotics and up to 83% of these patients do not respond to second-line antipsychotics.² Only clozapine has shown established efficacy after two failed trials, but response can be incomplete: 40–70% of individuals do not respond to clozapine. Furthermore, some patients decline clozapine or are unable to tolerate it secondary to side effects, the need for bloodwork monitoring and the risk of agranulocytosis.³

For patients who do not respond or have partial response to clozapine and those who cannot take clozapine, ECT has demonstrated efficacy^4–8 in the setting of partial antipsychotic response. For those who respond to ECT augmentation, there remains a challenge in how to maintain this response. Evidence indicates that maintenance ECT (mECT) may prove effective in maintaining the improvement achieved during acute treatment although literature specific to this topic is limited to small sample sizes and relatively short treatment duration.⁹ Prospective studies are limited. One trial randomized 51 treatment resistant schizophrenia patients who remitted on acute ECT + flupenthixol to one of three treatment arms and followed them over 6 months within a single-blind design: flupenthixol alone, ECT alone, or combination therapy. ECT was administered weekly × 1 month, then biweekly × 5 months. Over 90% of those assigned to only flupenthixol or ECT relapsed, in contrast to 40% for combination treatment.¹⁰ Another prospective study by this group evaluated the combination of bilateral ECT and flupenthixol in 21 individuals with refractory schizophrenia over 1 year, reporting no relapses as well as notable improvement both clinically (70%) and in terms of functioning (91%), as measured by the Brief Psychiatric Rating Scale (BPRS) and Global Assessment of Functioning (GAF).¹¹ Another study, in 62 schizophrenia patients who responded to acute ECT, observed significantly lower relapse rates and longer relapse-free survival in those that received mECT combined with risperidone compared to risperidone only.¹²

To address this gap in knowledge, the present investigation sought to describe the effectiveness and tolerability in a relatively large sample of patients with schizophrenia or schizoaffective disorder that received mECT. We examined clinical response and cognitive adverse effects as well as possible predictors of each. We sought to include cognitive side effects as it has not well described in patients with schizophrenia receiving ECT compared to patients with depression who receive ECT and its position as one of the more common and troublesome side effects associated with ECT.¹³

2. Methods

2.1. Study Design & Subjects

This study was conducted at the Centre for Addiction and Mental Health (CAMH), an academic tertiary care psychiatric hospital in Toronto, Canada. We completed a retrospective chart review including all patients treated in the hospital’s ECT treatment program from April 2010 to July 2016. Any patient with a DSM-IV diagnosis of schizophrenia or schizoaffective disorder who received at least one mECT treatment was included in this study.

Data for this study was obtained from referrals to the ECT service with the referring psychiatrist providing information on diagnosis, indication, and medications. Referral indications for mECT included: failed pharmacotherapy, prior good response to ECT, suicidality, failed continuation maintenance pharmacotherapy, intolerance of adequate pharmacotherapy, and patient preference. Subjects included in this study were patients identified on the referral form as suffering from “schizophrenia” or “schizoaffective disorder”. The information provided by the referring physician was confirmed by the ECT psychiatrist.

If a patient was found to lack capacity to consent to mECT treatment due to their mental illness by the treating physician, then a substitute decision maker (SDM) was appointed. If the patient chooses they can contest this finding at the Consent and Capacity Board (CCB), which is an independent government panel composed of a psychiatrist, lawyer and community member. If the finding of incapacity is upheld, or if the patient does not challenge the finding of incapacity then informed consent is provided by an SDM whose identity is defined based on provincial legislation and follows a hierarchy with individuals appointed by the courts at the top of the hierarchy, followed by family members, and lastly government-employed healthcare trustees. The study was approved by the research ethics board at CAMH.

2.2. ECT Technique

The ECT machine used for this study was a MECTA spECTrum 5000Q. A 1.0 millisecond (ms) pulse width with 800 mA was used for bilateral (BL) stimulation. For the right unilateral (RUL) setting, the pulse width was set to an ultrabrief pulse width of 0.3–0.37 ms in more recent treatment courses, while the remainder received RUL with a pulse width of 1.0 msec. The choice of electrode placement was determined by the consultant ECT psychiatrist during the acute course, based on variables such as risk of cognitive side effects, need for rapid response, and prior treatment history. In general, BL electrode placement was recommended based on the prevailing reports in the literature.¹⁴ The frequency of mECT treatments was determined by the consultant ECT psychiatrist based on clinical assessment in which the maintenance of response was balanced with tolerability. Standard practice is to provide treatment once weekly for one month, then biweekly for two months then monthly for three months; however, maintenance treatments were provided more frequently if symptoms worsened. The stimulus titration method was used to determine the seizure threshold. After threshold was achieved the stimulus intensity was set at 1.5 times the seizure threshold for BL and at 6 times for RUL. The anesthetic agent used was methohexital 0.75–1.0 mg/kg IV and the muscle relaxant was succinylcholine 0.5–0.75 mg/kg IV.

2.3. Assessment of Treatment Response

2.3.1. Chart Review Technique

All patients who received at least one mECT treatment were eligible for the assessment of treatment response. ECT treatment response was assessed by chart review using a 4-point scale, previously described,^4,15,16 to estimate the clinician global impression improvement, and we have referred to this as the clinical note CGI (CN-CGI). Note that in the acute ECT retrospective study,⁴ response was assessed compared to pre-treatment baseline. In the present study of mECT, the main treatment effect is the preservation of the effect from the acute treatment (which is the purpose of mECT). Still, the patient’s condition before the acute ECT course was used as the baseline for the assessments here as well.

The CN-CGI scale was determined as follows:

1. Excellent:

The patient chart demonstrated dramatic benefit from ECT treatment and that benefit is well preserved. Examples of this level of response include: rapid discharge during maintenance treatment, reduction in need for medications, resolution of target symptoms, and statements such as “dramatic response” or “greatly improved”.

2. Good:

The patient chart indicated that the patient responded and keeps well. Examples of this level of response include: significant reduction in severity of target symptoms, and statements such as “responded well”, or “good response”.

3. Moderate:

The patient chart indicated that the patient had some amount of benefit. Examples of this level of response include: slight-moderate reduction in severity of target symptoms, and statements such as “improved somewhat”, or “partial response”.

4. Poor:

The patient chart indicated that the patient had minimal to no benefit or rapid recurrence of symptoms during mECT. Examples of this level of response include treatment discontinuation after few sessions due to mECT ineffectiveness or statements regarding exacerbation.

Treatment response was defined as a CN-CGI score of 1 or 2, and treatment non-response was defined as a CN-CGI score of 3 or 4.

2.3.2. Validation of Technique

While attending psychiatrists were encouraged to complete the CGI-I after an mECT treatment course, this information was available only for a subset of the treatment courses (n=11). Therefore, we used this subset to validate the CN-CGI technique. CN-CGI was measured by JR, and when there was a disagreement with the CGI-I scale, an experienced ECT psychiatrist (MI) reviewed the treatment response, and this score was used.

2.4. Cognitive Impairment

The subset of 11 treatment courses with treatment response CGI-I available also had a similar scale rating the level of transient cognitive impairment following ECT treatments. This was rated as ‘none’, ‘mild’, ‘moderate’, or ‘severe’ cognitive impairment by the treating clinician. Similar to treatment response, a chart review of transient cognitive impairment was done based on the same 4-point scale. ‘Non’ or ‘mild’ were considered to not be clinically significant, while ‘moderate’ or ‘severe’ were considered to be clinically significant for the purposes of statistical analysis.

2.5. Statistical Analysis

We compared treatment response and cognitive impairment rates by various clinical and treatment features. The differences in these features between responders and non-responders were calculated to determine their potential association with treatment outcome. Barnard’s test was used to compare rates of categorical data and logistic regression was used for numerical data. All statistical tests were two-tailed with significance level set at 0.05 and were performed in R (version 3.2.2, R Development Core Team).

3. Results

3.1. Demographics and ECT Details

This study included 47 patients who received 62 maintenance ECT courses. All patients were fully or partially resistant to at least one antipsychotic medication prior to being referred to ECT. In most courses patients received initial maintenance treatment as inpatients (73%), lacked capacity to provide consent for treatment (66%), received bilateral electrode placement (90%), and received antipsychotic medications (85%) during mECT. In 32% of the courses, patients were treated with benzodiazepines (at doses ≤ 2mg of lorazepam or equivalent). The average maintenance course consisted of 112 treatments delivered over 763 days and most individuals could not be spaced beyond biweekly treatments with many requiring weekly mECT treatments due to symptomatic worsening (Table 1). Ninety percent of the courses were bilateral. This proportion was very closely maintained in the responder and non-responder groups as well as in the cognitive impairment and non-impairment groups so all courses were analyzed together.

Table 1:

Patients Characteristics

Demographics and clinical characteristics:	(62 courses)
Mean age (SD, range)	54, (13,24–80)
Female	33 (53%)
Incapable to consent	41(66%)
Schizophrenia	27(44%)
SzA-Depression	23(37%)
SzA-Manic	12(19%)
Medications:
Antipsychotic	53(85%)
Clozapine	19(31%)
Antidepressant	14(23%)
Bz ≥ 2mg of lorazepam / equivalent	20(32%)
Treatment setting:
Bilateral electrode placement	57(90%)
RUL -> BL	1(2%)
BL -> RUL	0
Mean Tx number per course (SD, median)	112 (85,106)
Mean course duration (days, SD, median)	763 (771, 383)

SzA-Schizoaffective Disorder, RUL-right unilateral, BL-bilateral, Tx-treatment

3.2. Treatment Response

Based on the CN-CGI, 77% of the maintenance courses resulted in a marked response (i.e. the good or excellent response to the acute ECT course was preserved throughout the mECT course). All 11 courses for which the CGI was available, as assessed by the treating clinician, were classified as treatment responders (i.e. CGI of 1 or 2), though one of the 11 courses was assessed by the independent reviewer as a non-responder (CN-CGI of 3). Table 2 summarizes the response frequencies to mECT divided by various potential predicting factors. No significant differences were found between responder status and any of the following factors: gender, capacity to provide consent for treatment, use of antipsychotic, benzodiazepine or antidepressant medication or diagnosis of schizoaffective disorder. 66% of treatment courses lasted over 6 months and 85% lasted over 3 months. We examined factors that might be associated with the longer courses and found older age to be significantly associated with courses that last over 6, 3 and 1 month (all p values < 0.05). Diagnosis of schizoaffective disorder-manic was also significantly associated with courses that last 6 months and longer.

Table 2:

Factors associated with treatment response

Factor	Response	No response	Barnard’s test
Total (N=62)	48 (77%)	14 (23%)
Demographic / Clinical:
Female	28	5	0.65
Mean age (SD)	55 (12)	49 (13)	0.16*
Capable to Consent to ECT	17	4	0.74
Cognitive impairment present	4	3	0.21
Failed-Pharmacotherapy	18	8	0.21
Prior good response to ECT	38	10	0.6
Schizophrenia Diagnosis	22	5	0.6
SzA-Depression	17	6	0.68
SzA-Mania	9	3	0.89
Bilateral electrode placement	43	13	0.92
Inpatient	36	9	0.47
Medications:
Antipsychotic	41	12	1
Clozapine	15	4	0.88
Bz (≥ 2mg of lorazepam or equiv)	15	5	0.83
Antidepressant	11	3	0.88
Mood Stabilizer	10	2	0.69

^*logistic regression; SzA-Schizoaffective disorder; Bz-Benzodiazepine

3.3. Cognitive Impairment

Based on the clinical note cognitive impairment scale, significant cognitive adverse effects of ECT were evident in 11% of the courses. Based on the attending psychiatrist score (available for 11 of the treatment courses), significant cognitive adverse effects were reported in 9% of the courses. Table 3 summarizes the cognitive impairment frequencies divided by various factors. No significant differences in clinical or treatment factors were found to be associated with presence of cognitive adverse effects (defined by clinical note score ≥ 3).

Table 3:

Factors associated with cognitive impairment

Factor	Cognitive impairment	No Cognitive impairment	Barnard’s test
Total (N=62)	7 (11%)	55 (89%)
Demographic / Clinical:
Female	3	30	0.62
Mean age (SD)	51 (11)	54 (13)	0.642*
Capable to Consent to ECT	2	19	0.79
ECT Treatment Response	4	44	0.2
Failed-Pharmacotherapy	4	22	0.43
Prior good response to ECT	4	44	0.2
Schizophrenia Diagnosis	2	25	0.52
SzA-Depression	3	20	0.79
SzA-Mania	2	10	0.61
Bilateral electrode placement	6	50	0.92
Inpatient	5	40	0.93
Medications:
Antipsychotic	6	47	1
Clozapine	2	17	0.95
Bz (≥ 2mg of lorazepam or equiv)	2	18	0.87
Antidepressant	0	14	0.14
Mood Stabilizer	2	10	0.61

^*logistic regression; SzA-Schizoaffective disorder; Bz-Benzodiazepine

4. Discussion

The present investigation, found in a sample of 47 patients who received 62 maintenance ECT courses, a response rate of 77%. Rates of significant transient cognitive side effects, approximated 10% of patients. Notably, this was the case despite a mean number of treatments exceeding 100 over a period greater than 2 years. No relationship was found for either response or cognitive side effects with respect to age, gender, capactiy, diagnosis, or concomitant psychotropic medications (i.e. antipsychotic, benzodiazepines, antidepressants). A relationship was found between older age and longer mECT duration, implying greater effectiveness in older patients.

These findings align with the limited existing evidence. Other retrospective studies evaluating mECT have also reported effectiveness in schizophrenia,^17–20 and reviews of this same topic have drawn a similar conclusion.^13,21–24 To date, however, there exist only a small number of prospective studies.^10–12

A previously published investigation, in a small sample, also failed to identify predictors of response with mECT when examining age, gender, ECT indication, number of ECT sessions and series, outcome, use of mECT as well as antipsychotics, duration of illness and most recent exacerbation.¹⁸ Approaching the subject from a different perspective, a recent report investigated what factors may be associated with risk of relapse in individuals with schizophrenia who responded to ECT acutely (N=43). All were treated with bilateral ECT and antipsychotics were maintained post-ECT. The median relapse free period was 21.5 months. The number of acute ECT sessions was associated with increased risk of relapse (hazard ratio: 1.159, p=0.033). In contrast, use of concomitant mood stabilizers was linked to a lower risk of relapse (hazard ratio: 0.257, p=0.047).²⁵ Though we failed to find an association between benzodiazepine use and treatment response in this study, given that previous evidence has suggested benzodiazepines may impair treatment response in ECT for depression,²⁶ these medications should likely be discontinued, minimized or reversed with flumazenil for mECT in individuals with schizophrenia or schizoaffective disorder.

An important clinical finding in this study is that we failed to find a significant difference in clinical outcomes between individuals with schizophrenia compared to schizoaffective disorder in terms of response or transient cognitive impairment; however, a significantly larger proportion of individuals with schizoaffective disorder and a recent manic episode had a mECT course lasting greater than 6 months. This suggests that mECT may be beneficial for both schizophrenia and schizoaffective disorder, though the strength of this conclusion is limited by the small sample size. Furthermore, though only six individuals were treated with RUL electrode placement, five of these individuals had a treatment response to mECT, which suggests that clinicians may consider this electrode placement for individuals with psychotic illnesses as it may be associated with reduced cognitive side effect burden.²⁷ However, severity and acuity of illness may necessitate rapid improvement that is more likely to be achieved with a bitemporal electrode placement.

ECT is considered relatively safe,^28–30 although mECT rightfully raises unique concerns related to longer-term safety issues, in particular cognitive side effects. While systematic reviews of this topic are lacking, those reports that have taken into account safety issues suggest mECT is no less safe than acute ECT.^{13,18,22,24,31,32}. Another study compared 10 patients with schizophrenia undergoing mECT to 10 matched patients who had never been treated with ECT. A comprehensive neuropsychological battery failed to show any significant difference between the two groups.³³ Our study contributes to the safety data regarding long-term mECT. Further, reports have suggested cognitive improvement related to successful treatment of the underlying psychiatric condition, as well as gains on measures of functional recovery.^22,34,35 Interestingly, the relationship we found between older age and longer mECT duration appears in-line with a recent study that found that older age was also found to be associated with good response to ECT in depressed patients.³⁶ Another finding emerging from the present study is the relatively high treatment frequency needed to maintain effectiveness. Only 31% of treatment courses consisted of biweekly or less frequent treatment schedule. This frequency is higher compared to mECT in affective disorders and aligns with the conclusions of a recent review on mECT in schizophrenia.⁹

Limitations impacting the present data include the retrospective approach that resulted in variability in some aspects of the methodology. More comprehensive and detailed outcome measures and scales such as the brief psychiatric rating scale,³⁷ the positive and negative syndrome scale³⁸ and the Montreal cognitive assessment scale³⁹ to define response and assess cognitive side effects throughout are needed to better delineate predictors or variables associated with favorable or unfavorable response. In addition, measures of quality of life, disability and functioning would be informative.

Limitations notwithstanding, this study adds to the growing body of evidence indicating that both acute and maintenance ECT can be useful in schizophrenia and schizoaffective disorder, in particular for refractory patients. The fact that in two thirds of the cases, maintenance courses lasted over 6 months and in about half the cases, courses lasted over a year suggests that patients receiving this treatment can have a sustained period of wellness. Indeed, mECT use can fill a gap in treatment needs, in that other options to manage treatment resistance in schizophrenia are lacking.^6,40–42 Given the public health burden of schizophrenia, there is a pressing need to enhance our knowledge regarding how ECT might best be utilized in this population.